RESEARCH NEEDS IN BIOTECHNOLOGY
AND THE ENVIRONMENT
FINAL REPORT
November 1985
-------
RESEARCH NEEDS IN BIOTECHNOLOGY AND THE ENVIRONMENT
FINAL REPORT
Prepared for the
U. S. Environmental Protection Agency
Washington, D.C.
by the
Office of Public Sector Programs
American Association for the Advancement of Science
Washington, D.C.
November 1985
-------
PREFACE
This report presents the results of a workshop convened by the
American Association for the Advancement of Science and the U.S.
Environmental Protection Agency to examine research needs in
biotechnology and the environment. The workshop was held at Coolfont
Conference Center, Berkeley Springs, West Virginia, 29 April - 1 May
1984.
Four draft papers prepared by EPA staff — on environmental effects,
health effects, monitoring and quality assurance and control
technologies — served to focus discussions at the workshop. Two of
these papers, revised following the workshop, are included in this
report. In the two other cases, the workshop review groups, and AAAS
and EPA staff determined that supplementary papers by outside
individuals were warranted in order to better treat the issues
considered at the workshop. These two papers, commissioned separately
following the workshop, have also been incorporated into this report.
This report, therefore, represents a blend of the workshop discussions
and papers and additional work performed subsequently.
Many people contributed to the workshop and to this report and
deserve thanks for their contributions. Among them are, of course, the
paper authors, and other workshop participants, who gave generously of
their time and whose names are listed in an appendix at the end of this
report, and, especially, the workshop chairman, Gilbert S. Omenn. EPA
officials, led by John R. Fowle III and Morris A. Levin, were most
supportive and helpful in arranging and conducting the workshop and in
preparing this report. We also appreciate the assistance of the
consultants who prepared the two commissioned papers following the
workshop, Marvin Rogul and David Glaser and his colleagues. Finally,
many current and former AAAS staff members contributed their time and
energy to the project in its various stages, including Jill H. Pace,
Barbara Dworsky, and, especially, Mary I. Haddock.
Albert H. Teich
Head, Office of Public
Sector Programs
Washington, D.C.
November 1985
-------
CONTENTS
Preface
PART ONE: WORKSHOP SUMMARY
Page
A. Introduction and Background 1
B. Key Concerns and Recommendations 5
C. Environmental Effects 8
D. Health Effects 10
E. Monitoring and Quality Assurance 12
F. Control Technologies 13
PART TWO: WORKSHOP PAPERS
Tab
"Research Plan for Test Methods Development for Risk A
Assessment of Novel Microbes Released into Terrestrial
and Aquatic Ecosystems"
Authors: Ramon Seidler, Oregon State University
and A1 Bourquin, EPA
"Biotechnology Health Risk Assessment Research Plan" B
Authors: Marvin Rogul, The Rogul Group, and
John R. Fowle III, and David Kleffman,
EPA
"Environmental Engineering Research Support Proposal" C
Authors: John Burckle and Albert D. Venosa,
EPA
"Monitoring Techniques for Genetically Engineered D
Microorganisms"
Author: David Glaser, £t al^ Harvard University
APPENDIX
Workshop Participants E
-------
PART ONE
WORKSHOP SUMMARY
Prepared by Gilbert S. Omenn, Workshop Chairman
and
Albert H. Teich, Project Director
A. Introduction and Background
Approximately 60 people, including outside peer reviewers and EPA and
AAAS staff, participated in the AAAS/EPA Biotechnology Workshop at Coolfont
Conference Center, Berkeley Springs, West Virginia, 29 April - 1 May 1984.
As noted by EPA Assistant Administrator for R&D, Bernard D. Goldstein in
the opening session, the charge to the workshop was to assess the
scientific needs and researchable problems facing the agency as it
prepares to deal constructively and responsibly with proposals that might
lead to release of genetically-altered organisms to the working or general
environment.
The project began with a workshop on 14-16 December 1983 at EPA.
Attendees at this meeting included over 50 representatives of EPA program
offices (including Pesticide Programs; Toxic Substances; Policy and
Resource Management; Drinking Water; Solid Waste; Water; and Air, Noise and
Radiation), the Office of Research and Development, and EPA field
laboratories. AAAS's role in this first meeting consisted of assisting EPA
with the meeting's logistical arrangements and suggesting means of
facilitating session deliberations. Most importantly, AAAS involvement in
the first workshop facilitated translating the results of that meeting into
the basis for the second workshop, as described below.
-------
2
Based on the deliberations at the first workshop, EPA laboratory and
office representatives prepared documents suggesting research approaches
which might be necessary to support EPA regulatory efforts in four major
areas, which generally paralleled the working groups from the first
workshop: health effects, environmental effects, monitoring and quality
assurance, and containment and control technologies. Taken as a whole, the
four papers constituted elements of a draft biotechnology research agenda.
The second workshop, held at the Coolfont Conference Center in
Berkeley Springs, West Virginia, 29 April to 1 May 1984, was convened to
subject these papers to an intensive peer review by researchers from
outside the agency and by EPA staff, to further define the agency's
research plans, needs and capabilities, evaluate them, and modify them to
produce an appropriate and feasible research agenda. Names of participants
in the Coolfont workshop may be found in the Appendix.
Workshop participants were selected to represent a range of
backgrounds, and affiliations. EPA laboratory representatives and
headquarters staff brought to the meeting their knowledge of relevant
research areas, their understanding of laboratory capabilities, and their
own scientific interests and expertise. They represented the interests of
those who will actually carry out the research planned at the workshop.
Program office staff represented the perspectives of the clients of
users of the research. They brought to the meeting their knowledge of
-------
3
EPA's regulatory mission, their understanding of its needs and priorities,
and their views of what needed to be learned in order to carry out the
agency's mission and address its priorities.
The outside experts, drawn from universities, government agencies and
industrial firms, lent to the deliberations their special expertise
regarding scientific aspects of the problems under discussion, their
knowledge of other research being conducted in these areas, and their
informed opinions regarding the feasibility of proposed efforts.
Participants were assigned to each of the four working groups, and
were sent draft documents in advance of the meeting. Each participant
brought to the workshop a written review of the paper closest to his or her
area of expertise. The workshop began with a plenary session at which
summaries of all the draft documents were presented. Brief scenarios of
possible EPA involvement in biotechnology, and a summary of the N1H
experience were also presented to stimulate discussions.
Most of the meeting was devoted to discussions in four workgroups
(panels) corresponding to the four draft papers prepared by EPA scientists:
(1) "Proposed Biotechnology Research Plan for Test Methods
Development for Risk Assessment of Novel Microbes Released
into Terrestrial and Aquatic Ecosystems" (environmental effects
group);
-------
n
(2) "Proposed Biotechnology Research Plan for Test Methods
Development for Risk Assessment of Health Effects Associated
with Biotechnology" (health effects group);
(3) "Proposed Biotechnology Research Plan for Monitoring Systems and
Quality Assurance" (monitoring and quality assurance group);
(H) "Proposed Biotechnology Research Plan for Environmental
Engineering and Technology" (control technologies group).
Plenary sessions were held at the beginning, middle and conclusion of
the workshop. The discussions throughout were lively and provocative,
going to the root of EPA's role, as well as providing a rigorous review of
the proposed data gathering and research projects.
Following the workshop, revised drafts of the papers were prepared and
circulated to all participants for comment. Reviews were also solicited
from a variety of other individuals who had not attended the workshops.
Following this review, it was determined that two of the revised papers
did not serve the purpose intended by the workshop organizers, and AAAS and
EPA staff agreed to commission new papers to better address the issues.
These replacement papers were submitted in draft form and reviewed by EPA
and AAAS staff subsequent to the workshop. A preliminary report was
prepared by AAAS immediately following the workshop, highlighting the main
ideas of the papers and identifying areas for EPA emphasis and attention.
Part one of the final report, contains a AAAS summary of the key
recommendations and concerns expressed at the workshop. Part two includes
revised versions of two of the workshop papers and the two replacement
papers: "Monitoring Techniques for Genetically Engineered Microorganisms,"
by David Glaser and colleagues of Harvard University, and "Biotechnology
-------
5
Health Risk Assessment Research Plan," by Marvin Rogul of The Rogul Group
and John R. Fowle III and David Kleffman of EPA.
B. Key Concerns and Recommendations
Workshop deliberations resulted in identification of a number of
areas of concern and recommendations for EPA research activities. No
attempt was made to force consensus among workshop participants, but
several items were considered high priority for EPA attention. These are
enumerated below, and are followed by highlights from the discussions of
each working group.
(1) EPA's primary emphasis in biotechnology research should be on
potential environmental and health effects of deliberate or
accidental release of genetically-altered organisms to the
environment. EPA's mandate establishes its lead-agency role in
this important area.
(2) Attempts at risk assessment should begin with well-selected
specific cases, including, especially, the applications being
developed within EPA to control certain pollutants or
contaminated sites. A major regulatory need, particularly for
OTS, is predictive risk assessment models for products of bio-
technology, analogous to the structure-activity relationship
models employed for predicting chemically induced effects. It
-------
6
is, however, premature to attempt development of a general pre-
dictive model for assessing the risks of release of genetically
altered organisms. Because of the vast number of biological
possibilities for biotechnology products (e.g., organisms,
vectors, gene sequences, products), it is not possible to pre-
dict potential effects without specific knowledge of a number of
important parameters. Thus, experience must be gained first on
a case by case basis.
(3) Much information pertinent to EPA's biotechnology activities may
already be available. The currently available literature should
be systematically reviewed, analyzed and used to focus and set
priorities for EPA's future efforts. This review will also help
foster complementary efforts, and avoid duplication of work
performed by other organizations. The scope of the search should
include published microbiological and public health information,
NIH/RAC sponsored risk assessment efforts, and reports on recent
and current molecular biological and ecological studies. The
literature should be continuously monitored and collaborative
efforts with organizations such as NIH/RAC should be encouraged
to keep EPA abreast of developments in the field.
(1) There is a clear need to enhance EPA's molecular biology
capabilities and its in-house expertise in biotechnology. A
strategy should be developed for gradual, long-term development
-------
7
of a viable research program capable of attracting and retaining
highly competent molecular biologists. Given the Agency's
undisputed lead role in environmental assessment, priority should
be given to enhancing expertise in this area. Expertise is
needed also in such related fields as monitoring of organisms
and their products in the environment, fermentation engineering,
modern molecular analysis, and health-oriented molecular biology.
In general, in order to attract and keep competent molecular
biologists, EPA should support some basic research, filling gaps
identified in more applied studies.
Close communication and collaborative efforts among EPA
scientists with biotechnology-related expertise should be
facilitated. Opportunities to discuss research problems and
results during weekly seminars and during informal daily contacts
foster good science. Spreading modest resources in this area
over a number of geographic locations will dilute the talent,
perhaps to the point of rendering it ineffective. Consideration
should be given to initiating the EPA biotechnology program at a
single laboratory site where all ORD office are represented and
providing strong management to coordinate activities between
offices. It should be noted, however, that EPA's precise per-
sonnel needs are inadequately defined at this time.
-------
8
(5) Outside experts should be identified and called upon to help EPA
develop its scientific capabilities in biotechnology, to identify
problem areas, to develop and revise research strategies, to peer
review research proposals, etc. All of EPA's research proposals
should be subjected to rigorous peer review. This workshop set
a good precedent.
(6) EPA should actively seek input from industry, public sector
interest groups, academia, and other federal agencies on its
proposed activities. In turn, EPA should demonstrate to these
others that the kinds of approaches and testing any proposed
guidelines might include are, indeed, feasible, by leading the
way in its own environmental control technology development
applications.
C. Environmental Effects
Five research areas were identified which require investigation in
order to estimate the actual and potential environmental effects of
genetically-altered organisms. The workshop regarded them all as high
priority and listed them in the order in which they felt the areas
should be addressed:
(1) Methods to identify and enumerate organisms are central to
research on environmental effects, as well as monitoring and
-------
9
health effects. Existing technologies for these functions
should be assessed with respect to their applicability to
environmental hazards.
(2) Existing methods for measuring survivability and growth para-
meters should be evaluated to determine where they are
effective and whether the kinds of data they yield can be
useful in hazard assessment.
(3) The impacts of chemical, physical and biological environ-
mental factors on survival and growth of organisms need to be
investigated.
(4) Genetic transfer. Questions in this area, particularly re-
lating to mechanisms, probability, and similarities to or
differences from naturally-occurring gene transfer are very im-
portant, but they must be recognized as targets for basic in-
vestigation, not simply methodology development.
(5) Hazard assessment. Development and validation of methods are
needed for studying effects (pathogenicity and infectivity) on
non-target organisms. Existing methods for studying disrup-
tion and perturbation of environmental processes need to be
evaluated as well, particularly with regard to their utility
for hazard assessment.
-------
10
The area of transport was judged to be of significantly lower
priority than any of the above topics. The panel believed that existing
transport models for various media should be applied first, before any
new experimental research is undertaken.
D. Health Effects
The panel felt that because of the large variety of possible
organisms and applications, it would not be feasible at this stage to
undertake development of a predictive model for health effects. It
recommended, instead, a case-by-case approach to estimation of potential
health hazards related to release to the environment of genetically-
altered organisms. This approach has several aspects (listed in order
of priority):
(1) Development of test protocols. Animal test protocols
for genetically-altered organisms should be integrated
with existing protocols in Subdivision-M of the Pest-
icide Assessment Guidelines (which already include
tests for microbials) and refined. This effort should
include, if possible, a scheme for categorizing
bacteria according to their degree of pathogenicity.
(Subdivision-M also needs to be revised to provide
rules for testing viral organisms, particularly those
that contain potentially oncogenic genes.)
-------
(2) Data-gathering and information management. Any
research effort needs access to a wide range of
relevant, up-to-date information. This involves
subscription to relevant computer data bases, use of
computers to catalog in-house experience, and
acquisition of first-rate library resources.
(3) Identification of least pathogenic bacteria. To
facilitate both industrial and in-house applications
of modified organisms, the agency should identify
organisms which are incapable of colonizing man and
which may be useful commercially, and should seek to
validate that they are indeed safe.
(U) Specific experiments. Among the experiments proposed
in the draft paper, those which address possible risks
of Bacillus thuringiensis and possible dissemination of
baculovirus DNA were considered to be most valuable and
feasible. The need to identify genetic markers other
than antibiotic resistance for tagging organisms (more
generally, the need to conceive and evaluate new genetic
markers) was also supported. Such markers would have
applications in ecological, health and monitoring efforts.
-------
12
E. Monitoring and Quality Assurance
Monitoring efforts are central to the conduct of R&D work within
EPA. The panel recognized that specific monitoring protocols will need
to be developed for individual experiments. In view of this, it felt
that the overall emphasis of the monitoring R&D program should be: (a)
to evaluate existing tests, (b) to develop new tests for identifying
and quantifying genetically-altered organisms in environmental
settings, (c) and, based on this experience, to provide guidance on
monitoring approaches. Four broad areas of investigation were suggested
(in order of priority):
(1) Molecular probes (DNA and RNA probes, immunological methods),
(2) Conventional microbiological methods,
(3) Sampling procedures, and
(4) Quality assurance.
Highest priority was placed on research on the application of
molecular methods to monitoring, specifically using DNA fingerprinting
methods with microbial populations. This approach possibly could be
coupled with the use of DNA and RNA probes to monitor the movement of
R-DNA in in situ microbial populations.
The panel also devoted attention to the problem of research needs
related to monitoring strategies. While specific strategies should be
considered on a case-by-case basis, there is need to compile existing
-------
13
information on strategies and to develop and test some general
guidelines for planned releases including such factors as site
selection, climatic conditions, sampling arrays, and sampling frequence
inside and outside of the release area. The use of simulants —
analogous organisms whose behavior and properties are well-characterized
— to test alternative strategies was recommended.
F. Control Technologies
The panel took as given that there is a potential hazard in the
release of modified organisms either accidentally or deliberately from
manufacturing processes or field trial applications. It focused on four
problems: (1) assessment of release and exposure; (2) methods for
minimizing release; (3) techniques to prevent worker exposure; and
(4) management of release in situations where it was needed. The
research recommendations fell into two major categories. One set or
recommendations relates to biologically-based manufacturing processes
in which releases can be either accidental or deliberate. These include
(in order of priority):
(1) Studies to assess the potential for release at various points
throughout industrial processes, including evaluations of
individual pieces of equipment;
(2) Assessment of the potential for worker exposure in manu-
facturing plants;
-------
11
(3) Evaluation of techniques for containment of specific pieces of
process equipment, should there be an accidental release;
(1) Development of monitoring needs and strategies in the context
of manufacturing processes;
(5) Assessment of alternative decontamination techniques;
(6) Evaluation of alternative worker protection equipment.
The second set of recommendations concerned deliberate release of
modified organisms in field applications. These include (again, in
order of priority):
(1) Studies to specify the characteristics of sites which make
suitable for various types of field tests;
(2) Evaluation of various alternative approaches to containment
materials to be used and interaction of micro-organisms with
those materials;
(3) Assessment of monitoring needs and strategies specific to
field testing situations;
-------
15
(1) Evaluation of alternative decontamination methods and
materials;
(5) Evaluation of alternative technologies for application of
genetically-altered organisms to the environment.
Coordination with other research efforts and the desirability of
recommendation to the manufacturing community of guidelines for
desirable properties of organisms were also endorsed. Finally, EPA was
urged to make its own applications of genetically-altered organisms for
control of environmental problems models for the larger research and
technical community.
-------
PART TWO
WORKSHOP PAPERS
-------
A
-------
Research Plan for
Test Methods Development for Risk Assessment of Novel
crobes Released Into Terrestrial and Aquatic Ecosystems
by
A1 Bourquin
Environmental Research Laboratory, EPA
Gulf Breeze, Florida
Ramon Seidler
Oregon State University
Corvallis, Oregon
Office of Environmental Processes and
Effects Research
Washington, D.C.
May 22, 1984
-------
TABLE OF CONTENTS
Page
I. INTRODUCTION 4
A. Goal 4
B. Risk Assessment 5
C. Two Major Approaches . 6
1. Data base development 6
2. Research 7
D. Short and Long Term Needs 7
E. Relevance to EPA Needs 8
II. NOVEL ORGANISMS 10
III. DEVELOPMENT OF TEST METHODS FOR THE DETECTION, IDENTIFICATION
AND ENUMERATION OF NOVEL ORGANISMS 12
A. Statement of Research Problems 12
B. Availability of Data Base 13
C. Approaches 13
1. Conventional techniques 15
2. Molecular techniques 15
D. Short Term Products 17
E. Long Term Products 17
IV. DEVELOPMENT OF TEST METHODS FOR ASSESSING FATE
OF NOVEL ORGANISMS 19
A. Statement of Research Problems 20
B. Availability of Data Base 20
C. Approaches 21
1. The microcosm approach 21
2. Rationale for selecting ecosystems 22
a. Terrestrial research 22
b. Aquatic research 23
D. Short Term Products 25
E. Long Term Products 26
V. DEVELOPMENT OF TEST METHODS FOR ASSESSING GENETIC STABILITY
OF NOVEL ORGANISMS 28
A. Statement of Research Problems 28
B. Availability of Data Base 29
C. Approaches 30
1. Naked plasmid DNA 30
2. Stability of DNA in novel organisms 31
3. Sources of cultures and plasmids 35
D. Short Term Products. 35
E. Long Term Products 36
2
-------
VI. DEVELOPMENT OF TEST METHODS FOR ASSESSING HAZARDS OF RELEASED
NOVEL ORGANISMS 38
A. Statement of Research Problems 38
B. Availability of Data Base .. 39
C. Approaches................... 39
D. Short Term Products 41
E. Long Term Products 42
VII. SUMMARY 44
VIII. ACKNOWLEDGEMENTS 46
IX. LITERATURE CITED 47
3
-------
I. INTRODUCTION
The Office of Pesticides and Toxic Substances (OPTS), under the authority
of the Toxic Substances Control Act (TSCA) and the Federal Insecticide, Fungicide,
Rodentlclde Act (FIFRA), will regulate parts of the biotechnology industry (1).
The major concern leading to this regulatory oversight is uncertainty over
human health and environmental effects of organisms released specifically to
the environment* This uncertainty comes from the unassessed potential hazard
of novel organisms in the environment* In this research plan a series of
experimental approaches are presented for developing methods which EPA can cite
for measuring risk assessments from the release of novel as well as naturally
occurring indigenous microbes* "Novel" microbes Include naturally occurring
microorganisms placed in environments where they are not native (nonlndigenous
or exotic), and microorganisms altered or manipulated by humans through techniques
of genetic engineering.
The dearth of scientific information on the potential risks from release
of novel organisms and the need for a scientific evaluation of risk assessment
point to a significant role for EPA's research program in the evolution of the
Agency's regulatory scheme for biotechnology.
A. Goal
This document has one major goal: the presentation of a research plan for
develoing test methods and other information for risk assessment from the
effects of novel organisms released to terrestrial and aquatic ecosystems.
Test methods need to be experimentally developed so that EPA can cite them in
requiring manufacturers to develop relevant data*
The guiding element throughtout this document is to meet OPTS needs for
testing methods aimed at identifying hazards and exposures and determining dose-
response relationships for novel microbes released to terrestrial and
4
-------
aquatic environments. This focus is justified on the basis of program office
needs (2), ORD's research priorities In biotechnology (3), and the expertise
of the Office of Environmental Processes and Effects Research (OEPER) in test
methods development and exposure assessment. It 16 clearly recognized that
microbes also become dispersed by air currents. However in the initial phases
of the risk assessment, containment constraints override concerns for methods
development involving air dispersal. Furthermore, it is known that air
pollution, as well as chemical and biological warfare defenses have provided
mathematical models for describing movement and diffusion of small particles
such as microbes. The models of dispersion are generally accepted as an
effective replacement of field monitoring for estimating particulate
concentrations downwind (65). Therefore, the risk assessment for air
dispersal of microbes may not require new methods development.
This document 16 a research plan and not a research proposal. Many
specific research proposals will be developed from this plan and a subsequent
workshop to define research to answer some of the more pressing needs of OPTS.
B. Risk Assessment
The basic elements of risk assessment guided the development of this
research plan. Risk, defined as a measure of the likelihood and severity of
harm (A), is generally assessed through three kinds of investigations: (1)
exposure assessment (determining conditions of exposure); (2) hazard
identification (attributing adverse effects to the hazard) including dose-
response assessment (relating exposure to effects); and (3) risk
characterization (estimating overall risk) (4,5).
Exposure assessments make several determinations: (a) the segments of
the environment exposed to the agent; (b) the intensity, frequency, and length
of exposure; and (c) the concentration and fate of the agent (4,5). Hazard
5
-------
identification, or the process of determining whether exposure to an agent
causes an adverse effect, often i6 a non-systematic Investigation (4,5). In
fact, a major difficulty, particularly with environmental effects, is in
developing tests that will identify the specific impact from the myriad
possibilities. Dose-response assessment often requires studies of possible
threshold effects and extrapolations from high to low dose, laboratory
research to field applications, and few species to many species (4,5). The
interdependence among these assessments of hazard (including dose-response)
and exposure, is apparent. Finally, risk characterization, the estimated
incidence of the adverse effect on a system in a given population, is based on
these assessments (5).
C. Two Major Approaches
This research plan presents two major approaches to accomplish the goal
of test method development for risk assessment of novel organisms in
terrestrial and aquatic ecosystems. For the purposes of this plan, aquatic
will include both freshwater and marine (including e6tuarine) ecosystems,
fully understanding that methodology will vary in many instances. Terrestrial
will Include habitats above (phyllosphere), on the surface, and within the
soil ecosystem.
1. Data base development
First, data base development 16 an important emphasis. The vast and
diverse body of ecological and pathological literature on many organisms
in many environments contains pertinent information on effects, exposures,
and methodologies. We propose to identify, collect, analyze, and evaluate
this information to use in test methods development. This objective Is
consistent with ORD's ranking of data base development as an important
priority item In the development of its biotechnology research strategy
(3).
6
-------
2. Research
The second approach for te6t methods development relies on laboratory
research to meet unique needs associated with risk assessment of novel
organisms and to complement the information and approaches obtained from
data base development. The terrestrial and aquatic microcosms developed
by OEPER scientists for the study of xenobiotic degradation are especially
well suited for cost effective containment needed for some of the more
advanced ("Tier II") 6tudies described in this plan.
D. Short and Long Term Needs
This plan contains both short and long term (one to two year and three-
plus years, respectively) research needs and is developed with the following
limits and assumptions. The research is planned for microbes, specifically
bacteria, fungi and viruses. This emohasis is derived (1) from the greater
likelihood that microbes will be exploited for TSCA- and FIFRA-covered
biotechnology purposes in the near future; (2) from the likelihood that
microbes would be more difficult than macroorganisms to contain in the
environment and, therefore, may be a greater hazard, and (3) from the OEPER
laboratory's mission for test method development and its current research
activities for microbial pest control agents (MPCA). Because of the
limitations on research resources and the easy availability of many bacteria
already used in gene cloning activities, bacteria have been chosen as the
primary model test organisms for early methods development in new aspects of
this research plan. An effort to develop the necessary test methods for
macroorganisms such as plants would require at the lea6t, resources similar in
magnitude to those listed in this document for microorganisms. As the test
:n*tt.ods develop and when new agents become candidates for possible release to
the environment, it is our full intention to Include species other than
7
-------
bacteria in the long term research components. We are veil aware of the
viruses and fungi which have been released to control various species of pests
(57,58). These viruses and fungi will be studied as part of the MPCA risk
assessment methods discussed In Section VI. Because the regulatory offices
will define the extent of their oversight of organisms released to the
environment, no attempt was made in thi6 document to rigorously define novel
organisms or genetic engineering. The research plan and test methods can be
adapted to whatever groups of organisms become subject to regulation.
E. Relevance to EPA Needs
This plan is presently designed in terms of developing citable
methodologies for OPTS regarding terrestrial and aquatic hazards to the
environment and will include environmental effects and exposure assessment
research. Initial emphasis is on blotic effects within aquatic, soil and
above-soil ecosystems. While direct focus on abiotic environmental effects is
not a high priority in this plan, consideration is given to the interaction of
blotic and abiotic factors.
This research plan was based on the acceptance of the following
assumptions: 1) certain information will be available to the regulatory
offices concerning the novel microbes. In particular, it is assumed that the
regulatory offices will have at least the following information: the identity
and source of the microbe, the phenotypic or physiological characteristics of
the organism type or at least those characteristics of the unaltered parent,
the nature of any genetic manipulation, unique attributes for which it was
developed, and intended use including sites, quantity, and manner of
dispersal. The eventual use of methods developed under this research plan
will be dependent on this information. 2) To conduct a risk assessment of a
novel organism OPTS is going to need information on the following (47):
8
-------
1. ability of the novel organism to survive;
2. ability of the organism to reproduce or persist;
3. ability to be transported to and establish in nlche6 other than those
intended;
A. ability to transfer traits to and from other members of its niche;
5. ability to cause adverse environmental effects
a. pathogenicity, including toxicity and infectlvity, to non
target species
b. disruption of environmental processes (e.g. nutrient cycling,
nitrogen fixation).
As an initial effort to coordinate the research effort among OEPER's
terrestrial and aquatic research groups, a workshop will be held in
Washington, D.C. in November 1984. The objective of this workshop will be to
assess the state-of-the-art in survival of genetically altered microorganisms
and the potential for genetic material transfer in natural environments,
including aquatic and terrestrial ecosystems.
As resources in biotechnology become available, scientists competent in
the fields of microbial ecology and microbial genetics will be assembled at
the workshop to discuss the research areas, to assess accurately the state-of-
the-art and to define precisely the approaches to be used by both in-house and
potential extramural cooperative researchers. This workshop would center on
defined needs and research strategy proposed in the Office of Exploratory
Research Biotechnology Workshop and will bring in scientists who have worked
in the pertinent scientific disciplines and help to fine-tune research
needs. This group would include scientists with whom we could have potential
cooperative agreements as well as representatives from other government
agencies. By including other agencies, we could Increase the potential
funding levels and avoid potential duplication of existing research efforts.
-------
II. NOVEL ORGANISMS
The use of novel organisms i6 growing at a rapid pace. They are being
used for a variety of purposes ranging from control of pest populations to the
manufacture of pharmaceuticals. Many of these organisms are simply species
that are being released into new environments, while others are genetically
engineered for specific purposes.
Human health effects research and containment strategies relating to the
release of genetically engineered E. coll strains have been supported by the
National Institutes of Health. The logic developed from that research very
early led to the development of specific physical containment and vector
requirements for cloning hazardous genes (48). In specific situations,
investigators are required to work with debilitated mutants of E. coli which
cannot colonize the natural environment. However, there is a paucity of
information available on the biotic environmental effects associated with
genetically engineered organisms of other species. This lack of information
is significant and has raised concern (6-8). First, strains designed for
purposeful release (including microbiological pe6t control agents, MPCAs)
will, by choice, not be debilitated since they will need to function
effectively when released. Second, these microbes will be comprised of
species very different, both ecologically and physiologically, from coli.
Furthermore, very little research ha6 been done to assess the hazards
resulting from the accidental release of genetically engineered organisms.
Genetic modifications have now been achieved in many classes of organisms
ranging from viruses and other microorganisms to higher plants and animals
(10). The intent in creating new combinations of genes within a single
organism is to provide humanity with new biological tools to benefit health,
welfare, and the environment (10-15). Progress in the genetic manipulation of
10
-------
these organisms has exceeded our knowledge of their fate and effect on natural
ecosystems (9,14). Before EPA can make decisions concerning release of such
organisms, information is needed to establish the potential hazard of
engineered microbes in the environment and the consequences of their
interactions with other indigenous microbes.
Genetically engineered microbes are being constructed to carry out a
plethora of new metabolic activities. Cultures are now able to metabolise new
combinations of organic pollutants, to mineralize metals, enhance oil
recovery, increase efficiency of nitrogen fixation in leguminous plants, and
make animal and human hormones (10,11,12,13,22,23). Very little has been done
to assess the ecological fate and effects of such engineered microbes in
aquatic, terrestrial and/or agricultural ecosystems (14).
It should be made very clear that human activities have already impacted
on the genetic constitution of microbes in the environment. Through the
overuse of antibiotics there has been a dramatic Increase in the incidence of
multiple antibiotic resistance expressed by DNA plasmids which can be
transferred into other bacterial species. Various Industrial processes
including dumping of pollutants and even chlorine disinfection of water
selected for strains that are resistant to metals, to antibiotics, and which
possess novel metabolic capacities such as the biodegradation of PCB's and
chlorinated aromatic pesticides. It is known that transfer of antibiotic
resistance, heavy metal resistance, and unique metabolic capacities by cell-
to-cell contact occurs in laboratory media, in certain institutions
(hospitals), in the animal and human gastrointestinal tract, and in habitats
simulating the natural environment (8,13,24,29-31,46). It is most unlikely
that the gene pools from novel organisms, which would be released by the
billions, will not influence other, new biotic and abiotic ecosystem
processes.
11
-------
III. DEVELOPMENT OF TEST METHODS FOR THE DETECTION, IDENTIFICATION. AND
ENUMERATION OF NOVEL ORGANISMS.
A. Statement of Research Problems
The requirement for a technique to Identify and enumerate the novel
organism following accidental or purposeful release is one of the most
fundamental prerequisites of the entire process of risk assessment of novel
microbes. Therefore, the development of one or more combinations of specific,
convenient, reliable, and sensitive tracer methodologies should be an early If
not the first consideration for research needs. Such methods need to be
experimentally developed so that EPA can plte them in requiring manufacturers
to develop relevant data. The development of these appropriate methodologies
is the principal objective of this methods research. EPA will not do actual
testing on individual novel microbes proposed for commercialization. A
desirable constraint of any detection/enumeration technique is that it be
widely applicable and suitable In technical application for detecting any
desired novel organism. The test methods that are developed should do little
or nothing to alter the mission of the microbe in its intended use of
application. The identification and enumeration procedures must discriminate
the specific novel strain from other organisms present in the environment and
also be capable of discriminating the novel organism from other strains of the
same 6pecies.
Fortunately, microbiologists have a battery of criteria and techniques
for fulfilling these needs. Two broad categories of methods are available.
These are: 1) "conventional" selective enrichment methods, and 2) "modern"
molecular approaches which rely on the fundamental techniques of genetic
engineering. As will become apparent below, there are advantages and
disadvantages of both methodologies, and In the final analyses, a combination
12
-------
approach would seem to best satisfy the fundamental criteria for
identification and enumeration.
B. Availability of Data Base
Several modern reference books provide selective media formulations for
the recovery of bacteria from their natural ecosystems (49-52). Published
references document the usefulness of tagged resistant bacteria as a sensitive
tool for monitoring their fate in natural ecosystems.
Applications of the DNA probe for detecting unique DNA sequences in
several types of bacteria have demonstrated the reliability and sensitivity of
this technique. There is no apparent need for a major effort involving data
base development.
C. Approaches
1. Conventional techniques
Microbiologists have relied upon the principles of selective
enrichment (enhancement) culture and selective/differential media for
nearly 100 years for the recovery of various metabolic types of organisms
from various ecosystems. Vith these techniques, the media and other
physical/chemical conditions of incubation are adjusted so as to promote
the growth of one or more related metabolic types of bacteria. For
example, if the novel organism metabolizes cellulose, a medium can be
formulated to contain this compound as a sole source of carbon for
growth. A broth enrichment medium would be used to detect novel organisms
present at counts below the detection limit required for direct plating
onto agar media (< 100 cells/gm of matter or ml of water). The broth
enrichment has the disadvantage of poor direct enumeration but would
provide a presence or absence result. In this example, agar media could
provide a direct enumeration of cellulose digestors. However, both
13
-------
enrichment media (broth and agar) have the disadvantage of not being
specific for the novel strain which was introduced to the environment.
Any cellulose degrader in the environment would grow on broth or agar.
Therefore it is mandatory to introduce into the novel organisms markers
which would Increase the specificity of recovery.
The uses of strains which are marked or tagged with a particular
resistance or metabolic activity has the advantage of being very specific
and allows direct enumeration. There are finite limitations to the kinds
of tags available. These tags specifically include resistances to
antibiotics, dyes, and heavy metals which can be incorporated into culture
media. The novel organism will grow If it is made resistant to these
agents. The use of multiple tags will alleviate the problem of a loss of
a single resistance marker due to a spontaneous mutation.
Initially, proven laboratory procedures will be used to select
spontaneous mutants in chromosomal DNA for the desired resistance. The
O Q
frequency of spontaneous resistance occurs about 1/10-10 cells. At
least two such resistance markers should be successively introduced into
the novel organism.
The offensiveness of releasing antibiotic resistant mutants can be
partially offset by selecting those antibiotics which are not of known
importance in treating diseases which the species of novel organism might
cause or by relying entirely upon mutants resistant to heavy metals or
dyes.
There are certain potential problems with the use of resistant
mutants. The effects are unpredictable and must be tested for each novel
organism. Potential problems in using organisms with resistance to
antibiotics, dyes, or heavy metals include: possible dependence upon the
14
-------
presence of the antibiotic for growth to occur, changes in growth rates,
mutability by the selective agent, and possible differences in ability to
persist in an ecosystem. Also, certain pesticides (arsenicals,
mercurials) may enhance metal resistance in Indigenous microbes and mask
any such resistance in the novel organism. All these possibilities will
have to be experimentally tested to verify that there are no such
changes. Specific experiments would involve comparing growth rates in
laboratory culture of the parent novel strain and the resistant mutant.
Persistence of the two strains will also be compared in simple axenic
ecosystems as well as in simple ecosystems containing the natural
microbial flora.
2. Molecular techniques
There are several molecular techniques which are useful for the
identification of organisms. It should be noted, however, that the
molecular techniques do not have the ability to enhance selective growth
nor do they provide a selective, differential milieu which can be provided
by conventional techniques. The molecular techniques will not be useful
if the novel organism has declined in numbers to the point where it cannot
be directly detected on a suitable agar plating medium (limits are < 100
cells/gm matter or ml of water).
A DNA gene probe, specific for the DNA base sequences present in the
novel organism, provides a powerful, specific, and reliable detection and
enumeration tool (54,55). The DNA probe oust reflect DNA sequences which
are unique to the novel organism. This can be accomplished by either
preparing the probe to reflect a novel plasmid or a smaller portion of a
custom DNA sequence Incorporated into a plasmid in the novel organism.
The probe can be radiolabeled and hybridized to colonies of the novel
15
-------
organism growing on Che surface of a nitrocellulose filter in contact with
an appropriate agar medium. By radloautography one can locate and count
colonies of the novel organisms present on the original filter surface.
This technique is relatively expensive and not as convenient as a
conventional direct selective plating medium used in conjunction with
tagged strains.
A second molecular technique involves the introduction of a plasmld
containing DNA sequences which will respond upon command to produce
(Induce) a visible chromogenlc substance by the novel organism. Examples
would Include the induction of an enzyme which causes a color change in an
ingredient in the culture medium or Induces the production of a pigment
such as the red-colored prodigiosln compound. The experimental details of
such a plasmld and its induction mechanism have recently been described
(56). The induction of the chromogenlc substance triggered only upon
application of a chemical or a temperature change would provide a means of
distinguishing the novel organism from any resident microbes which may
also possess the same metabolic capacity.
Once the novel organism is isolated and obtained in pure culture, its
plasmld DNA could be isolated and "fingerprinted" with specific types of
restriction enzymes (16,53). This will verify the identity of the unique
DNA fragments as those present in the novel organism prior to its
release. This fingerprinting procedure would serve as a verification
check that the selective culturing conditions were indeed specific for
recovering the unique novel organism of concern.
16
-------
D. Short Term Products
Early efforts should be directed at compiling selective and differential
media formulations from published modern literature appropriate to novel
organisms which are likely to be released. There are several recent
publications available to accomplish this task. Another early effort should
Involve inducing resistance markers in novel organisms and examining them for
growth rate effects and survival capabilities in "simple" ecosystems.
Persistence and growth rate studies of the resistant and parent novel
organisms should be compared. Appropriate quality assurance parameters should
be established for all identification and enumeration assays using EPA
recommended guidelines. Initial studies should be conducted in conjunction
with part IV, fate and transport, to examine novel organism recovery and
detection efficacies from suitable model ecosystems.
E. Long Term Products
Test the efficacy of detection procedures using molecular techniques,
with initial efforts devoted to DNA probes which are already available for
specific organisms. Test applicability of DNA probe for recovery and
detection of novel organisms specifically from terrestrial and aquatic
ecosystems. Evaluate how conventional and molecular techniques can be best
utilized and how procedures meet expectations of the quality assurance program
guidelines (validation of test methods). Prepare technical papers on
performance and validation of test methods.
17
-------
Time Lines: Detection, Identification and Enumeration Methods
4 mo
12
24
36
48
Document on guidelines,
protocols
Data base on selective
media for novel organisms
Resistance markers and
growth rate/persistence
"simple" ecosystems
Working model
microcosm
(Coordinate W/F-T)
Apply detection,
Recovery assays to
microcosm studies
f*— **—
00
*
Add organisms,
microcosm complexities
Interface v/Ecolog
Processes Hazards
Development,
Shakedown of Molecular techniques
Molecular marker variations.
Develop microcosms most
applicable to additional novel
organisms
Field
Validation
Methods
- >
Solid arrows Indicate
est(mated document
completion period (data base as technical report)
**OTS protocols and support document
-------
IV. DEVELOPMENT OF TEST METHODS FOR ASSESSING FATE OF NOVEL ORGANISMS
A. Statement of Research Problems
Test procedures must be designed for determining whether novel organisms
persist or multiply to high levels and therefore increase the risk of escaping
from the niche for which their use was intended. Experimental aquatic and
terrestrial ecosystems for studying persistence and fate will also provide
information on biological systems and ecological processes which become
exposed to or affected by the novel microbe. Such information will be
valuable in selecting relevant methods to deal with infectivity and
pathogenicity and ecosystem perturbation models in long-term research
evaluating exposure-risk phenomena appropriate for novel microbes (Section
VI).
Furthermore, persistence (monitoring population changes) and ecological
fate (niche and physical location) testing methodologies will be useful and
applicable in developing protocols for validating genetic expression in
complex ecosystems and for delineating biotic and abiotic factors which
contribute to survival or nonsurvival of the organism or its genetic traits.
The choice of the model ecosystem will be predicated on practical
criteria which can fulfill the following 6et of conditions. The model
ecosystem must be flexible in design and capabilities to accommodate the
intended use of a spectrum of novel organisms. It must be housed in such a
way as to prevent accidental release of novel organisms during and after the
test period. This requirement imposes size restrictions on the type of
ecosystem and the number of experimental units employed. For example, since
plants will be a biotic component in terrestrial ecosystems, regulation of
photoperiod would be mandatory. The test units should accommodate appropriate
physical and biotic trophic components i.e., a terrestrial unit should be
19
-------
capable of housing plants, invertebrates (chewing insects, earthworms, etc).
Similar concerns for the design of aquatic units are also apparent.
Furthermore, as part of the proposed experimental protocols, it is envisioned
that various procedures will be used to test, on a worst case basis, changes
which can prolong the survival or encourage growth of the novel organism.
Such perturbations might involve the addition of decaying plant and animal
matter in both the terrestrial and aquatic systems, as well as variations in
plant and invertebrate stocking densities. It is clear that a variety of
simple (flasks, mason jars, aquaria) and complex (microcosms) containment
systems are available for fulfilling the necessary test methods criteria.
B. Availability of Data Base
A vast amount of data base information is available on containment
systems used for assessing fate of chemicals. The EPA has been involved in
most of the significant contributions in this area (25-27) and OEPER
laboratories have developed and are using such systems for both aquatic and
terrestrial research. These containment ecosystems are adaptable to the
present needs. A review of the vast amount of literature should be made to
derive the most appropriate experimental designs for application to the
present tasks.
Studies conducted over some 60 years are available dealing with the
persistence of certain microbes In natural ecosystems. Most of this
literature deals with organisms associated with animals and plants (pathogens
and pollution indicator bacteria). Other, more contemporary information needs
to be summarized on persistence, cell densities, and fate dealing more
specifically with genera of organisms likely to be released including species
which are associated with the plant phyllospheres and root rhizosphere
(Pseudomonas. Xanthomonas. Erwlnla, Klebsiella, etc.) and those likely to be
20
-------
found in fresh and marine water systems (Pseudomonas, Vibrio t Witrosomonas,
Alcaligenes, Actineobacter. Aeromonas. etc.).
C. Approaches
1. The microcosm approach
Initially a series of method developments are envisioned for
monitoring survival of novel organisms in simple containment systems such
as flasks, mason jars, membrane filter chambers, etc. In later, more
complex risk assessments, and for purposes of containment, scientific
control, habitat simulation and convenience of experimental manipulations,
tests should be conducted in microcosms. The microcosm envisioned for
terrestrial research (43) uses a soil/plant/water ecosystem with water
runoff collected in an aquarium. Thus the fate of soil, plant and
invertebrate-associated microbes which are transported through the soil
into the aquatic ecosystem can also be readily monitored. The transport
and fate of the novel organism in fish can also be investigated. The
terrestrial/plant microcosms will be maintained to simulate growing
conditions and soil temperatures and moisture can be readily controlled.
As resources permit, seasonal conditions will be selected to allow plant
senescence and the accumulation of decaying animal and plant matter. Two
r
different kinds of microcosms are appropriate for studying the fate of
these organisms in aquatic environments. These include Intact sediment
cores and flow through chemostats. These devices will contain both micro
and macroorganisms appropriate for studying fate of novel organisms in
aquatic ecosystems.
21
-------
2. Rationale for selecting ecosystems
a. Terrestrial research
Attention will be focused on five separate kinds of
experimental approaches for assessing fate. All experimental
approaches will be conducted In a terrestrial model microcosm
which the EPA has developed. The five ecosystems to be
Investigated in experimental microcosms are:
1. Rhizoblum/legume soil ecosystem.
2. Root rhlzosphere of easily manipulated and cultivated plants
(radishes and potatoes).
3. Soil/plant ecosystem involving fate and effect of engineered
microorganisms capable of metabolizing pesticides.
A. Vegetables undergoing microbial decay simulating remains left la
fields following commercial harvest.
5. Plant leaf surfaces.
Since terrestrial ecosystems will be used, novel organism fate and
transport will be monitored not only through and on plants and soil, but
through other blotic components such as insects and earthworms and
mammals, (surfaces and Intestines) as well as decaying plant and animal
matter.
The research strategy has been built around a selection of those
unique natural terrestrial habitats which support high natural cell
densities in nature. For example. Inside an average legume nodule
7 9
approximately 10 -10 viable cells are found (32,33). Plant root
rhlzospheres contain approximately 10® or more bacteria per cm length of
root surface (34,35). The spray application onto soil and plants of
herbicide-metabolizing bacteria (11,12) will probably not initially
22
-------
achieve high cell densities unless there is significant regrowth of the
organism. However, it is envisioned that application of pesticides will
achieve two unique conditions. First, many of the indigenous soil
bacteria will be killed by the pesticide making both nutrients and niches
available for the resistant bioengineered or plasmld containing microbe.
Second, the vegetation will undergo natural decay and the remaining
indigenous soil microbes will greatly Increase In numbers. This will
allow growth of bacteria of many genera including Pseudomonas, Erwinia,
Klebsiella, Enterobacter, A1callgenes, Pectobacterlum, and others.
Representatives of these genera are involved in bloenglneering efforts and
can easily become candidates for release to the environment. Finally, It
Is common knowledge that the commercial harvest of vegetables leaves
behind hundreds of pounds of plant material per acre. As this material
decays, there will be a great Increase in selected portions of the
terrestrial microflora, especially those plant pathogenic soft rot
bacteria of the genera Erwinia, Pectobacterlum, and Pseudomonas.
Representatives of these genera and other Enterobacterlaceae are well
known for their promiscuity Involving plasmld transfer (10,14,16,31,34)
and could easily come into contact with the novel organism,
b. Aquatic research
Studies with aquatic systems may utilize "Eco-cores" which
consist of sediment cores taken from the bottom of lakes, streams and
estuaries (59-61). These cores are currently used to determine
degradation rates of xenobiotics by the organisms present in them.
These microorganisms, as well as mixed culture chemostats (12). can
be employed In studies designed for Investigating consequences of
novel organism release into aquatic ecosystems. Also, membrane
23
-------
chambers can be used to develop methods for studying persistence and
die-off kinetics of novel organisms in self-contained aquatic
ecosystems containing various levels of nutrients.
Much consideration has also been used in selecting those unique
aquatic ecosystems which support naturally high densities of
microbial populations. Ecosystems with high microbial densities are
those most likely to foster occurrences of gene flow between species
components. There are a number of ecosystems which support high
microbial densities that can be experimentally evaluated in the eco-
core microcosm. These include oertain aquatic sediments, Spartina
salt-marsh sediments, and various liquid/surface interfaces.
Certainly the evidence of genetic transfer in nature between
organisms appears to infer that cell-to-cell movement of genes
depends more on ecological intimacy than on evolutionary relatedness
and the proposed aquatic microcosms will be designed to provide the
opportunity for detecting gene transfer events.
There are additional justifications for selecting these ecosystems for
the study of fate and transport. In many cases it has already been
demonstrated that these environments contain microbes important in
agricultural processes (11,13,32-42). Microbes from these environments have
already been subjected to much research, often involving gene manipulation and
bioengineerlng. It is likely that bacteria from these environments will
someday be purposefully released to solve applied biological problems
associated with agriculture and pollution abatement (62). In 6ome cases, gene
transfer between plasmid containing strains of Escherichl coll and other
bacteria obtained from these environments has already been demonstrated
24
-------
(28). Finally, many of these ecosystems contain bacteria which can become
pathogens (or opportunistic pathogens) of humans, animals, and plants.
D» Short Term Products
The first short term milestone will involve the preparation of a document
which deals with designs of microcosms appropriate for the current task.
Similar documents already exist and this task should not require much
effort. In addition a document will be developed to summarize available
literature on survival of novel species which are prime candidates for release
to aquatic/terrestrial/plant environments. A suitable containment laboratory
will be designed and equipped with the experimental microcosms. Investigators
will establish which novel organisms to use in initial evaluation of protocols
for fate and transport; they will establish appropriate choices of higher
plants and invertebrates and the appropriate range for stocking densities.
Physical and environmental factors which contribute to transport and survival
of novel organisms will be measured in two or more of the proposed
ecosystems. However, initial studies will rely heavily on more simple and
inexpensive experimental ecosystems (soil-jar, aquaria, etc.) to evaluate
general physical/cheraical/nutrltional factors influencing survival and fate
of novel organisms. Conditions will be identified which influence organism
persistence or growth in simple systems and used for specific experimental
design considerations in developing the more complex microcosms. The
scientific importance and relevance of the microcosm approach for studying
fate and transport of novel organisms must be established. This will be
accomplished by publishing technical papers dealing with survival of novel
species which are prime candidates for release; by publishing reports on the
description and use of the microcosm with preliminary scientific data
establishing design specifications, containment, and Initial data. OEPER
25
-------
should sponsor a workshop program at a national scientific meeting on,
"Ecosystem approaches for studying novel organism fate and transport".
E. Long Term Products
The microcosm experimental approaches will be expanded to include other
ecosystems and also reflect the needs of new developments of novel organism
applications. Investigators will establish parameters necessary for
fulfilling quality assurance and data validation and publish additional
technical reports. OEPER will organize a workshop with environmental and
microbial ecologists to critique experimental protocols and discuss state of
the art alternatives for simulating natural ecosystem processes for fate and
transport studies in terrestrial and aquatic ecosystems. The proceedings of
this workshop will be published in the EPA Technology Series.
26
-------
Time Lines: Fate Methods
4 mo 8 12 24 36 48
Data bases:
Review of blotlc
microcosm models;
~
Survival times/habitats
of novel species
Establish flora, fauna and research protocols
&th
for working microcosm;
select specific novel
organisms
(Coordinate
Detection, Genetic Stability)
Measure Fate due to environmental factors
Biotic/ablotlc Influences on Fate in 2 or 3 ecosystems
- ^ - J* ** *
Interface: Ecolog. Process. Hazards
Workshop on Ecosystem
Approaches
—^ •
Workshop to
critique protocols
— — — -—a t
Add. organisms
focused activities
In 2 or 3 different ecosystems
Field Validation
Methods
solid arrows indicate estimated
document completion period (Data ba9e or technical report)
nrnf-nrnlfl and Bunoort document
-------
V. DEVELOPMENT OF TEST METHODS FOR ASSESSING GENETIC STABILITY OF NOVEL
ORGANISMS
A. Statement of Research Problems
Test procedures mist be developed to quantitatively measure transfer
frequencies of genetic material from and into novel organisms which enter the
environment. Suitable systems oust be established so that genetic traits of
novel organisms can be monitored for stability. The transfer of genes into
new species makes it exceedingly difficult to monitor fate and effects since
we must now monitor gene pools and new organisms. If genetic trait6 are
transferred into microbes which have ecosystem functions different from the
novel organism, new abiotic and biotic components of that system may be
adversely impacted.
It should be mentioned at the outset that assessment of genetic stability
will be largely influenced by results obtained in companion studies discussed
in this document. The level of concern or risk involving genetic stability or
lack thereof will, for example, Increase or decrease dependent upon results of
the persistence, fate and transport, and hazard assessment data. If, for
example, it is found that the population of specific novel organisms declines
rapidly (hours to days) and never regrows following entry into the
environment, it is more unlikely that concern will arise from its genetic
capabilities. If, on the other hand, an organism colonizes an ecosystem and
achieves sufficient cell densities to affect on other biotic processes, an
increased level of concern will arise over the fate of its novel genetic
information and the possible impacts or hazards that might result from it.
A major research challenge is formulating an all inclusive approach to
assess the ways genetic stability is to be measured. The approach must
consider the myriad of possible ways an organism can be bloengineered or
28
-------
otherwise genetically altered, the kind6 of species Involved, the ways in
which gene transfer can occur, and that commercial products may contain not
just one but rather a consortium of genetically and taxonooically undefined
organisms.
It is also recognized that phy6lochemical characteristics of different
types of soil, water, and other environs significantly affect the fate and
genetic stability of novel organisms. Consequently, considerable attention
should be focused on studying a broad spectrum of simple microcosms so that as
many physlcochemical conditions as possible can be examined.
The research will build upon existing information on microbial genetics.
DNA plasmid biology, and a synthesis of fundamental principles of microbial
ecology. A multitier approach is proposed, proceeding from simple, rapid
laboratory experiments to the habitat simulated microcosm analyses.
B. Availability of Data Base
Data base needs indicated earlier for persistence, fate, and specific
environmental components impacted by various 6pecies will be of use in the
present study of genetic stability. In addition, information must be gathered
on the published information concerning gene transfer capabilities of major
bacterial groups. This should include information on whether chromosomal DNA
can be transferred and a compilation of functions and types of plasmids which
have been found in the representative organisms. Similarly, this review
should include information on other types of DNA transfer mechanisms such a6
transformation and phage mediated transduction. A compilation of gene
transfer frequencies, the kinds of gene functions which are transferred and
the species which might serve as recipients of genetic information from novel
organisms will be made. Much of the previous research on these subjects is
limited to bacteria which are animal or plant pathogens, but extensive
29
-------
information is available for bacteria of the genera Escherichla, Pseudomonas.
Bacillus, Aclnetobacter, and Rhizoblum. Assembling this data base will be no
small task, and a number of scientific laboratories should become involved.
However, the work should be distributed to the reviewer with expertise on the
particular genus or group of bacteria.
C. Approaches
1. Naked plasmld DNA
In the first phase of genetic stability studies, the fate and effect
of naked DNA plasmids released to the microcosm will be investigated for
both maintenance of their original physical characteristics (molecular
weight, covalently-closed circular nature, nicks, and enzyme restriction
patterns) and for biological activities (ability of released plasmid to
transform competent suspensions of E. coli). The ability of resident
strains in the ecosystem to take up selected plasmid DNA molecules will be
investigated.
There is some debate as to whether experiments on the fate and effect
of naked DNA is important. Some scientists suggest that naked DNA will be
quickly inactivated by DNA hydrolyzing enzymes present in the natural
ecosystem. The experimental approaches which follow are intended to
carefully verify this possibility. In preliminary control experiments
plasmid DNA will be added to various simple ecosystems and its stability
will be evaluated by periodic extraction and testing for physical and
biological activities. If the DNA is not 6table, additional experiments
would not be conducted.
If the DNA is found to maintain physical integrity, the following
approaches will be used. The fate of plasmids of different molecular
weights and incompatability groups will be Investigated. The reviewed
30
-------
literature (14) as well as two research publications (44,45) have already
demonstrated the feasibility of such studies and have commented on the
importance of establishing the fate cf naked DNA in microbial
ecosystems. The relevance of conducting such experiments lies in learning
the biological fate of plasmid DNA released from lysed bacterial cells.
It is also important to know the fate and effect resulting from the
release of plasmid DNA Into specific niches which could be used to
genetically transform indigenous microbes to carry out agriculturally
relevant activities (increased legume host range, antibiosis in root
rhisospheres, etc.) or those organisms responsible for various degradation
activities in water and soil systems.
The fate of naked plasmid DNA will be investigated in:
A. Soil.
8. Plant associated habitats (plant root rhlzospheres, plant
phyllosphere, decaying plants, legume nodule environments).
C. Fresh and estuarine water and associated habitats.
The fate of naked DNA will be followed In Indigenous bacteria (see page
22) including selected species of plant pathogens (Pseudomonas syrlngae,
Agrobacterium).
2. Stability of plasmid DNA in novel organisms
In the second phase of studies the stability of the genetic material
carried in novel organisms will be Investigated with regard to its
transmisslbillty to indigenous bacteria.
in these initial Tier I tests, the novel organism will be grown to
high cell densities in laboratory media and combined in separate tubes,
with the bacteria it would be likely to contact in nature. Using routine
selective media, an examination will be made for recombinant organisms.
31
-------
i.e.. those which have received marker genes from the novel organism.
Calculations of transfer frequencies will be documented (recombinants per
donor cell). In appropriate cases, transfer of chromosomal DNA will also
be examined.
In Tier II studies, organisms exhibiting gene transfer in Tier I will
be placed into terrestrial and aquatic test systems as described in
Section III, fate and transport. A similar study will be conducted here,
with the purpose of examining for recombinants among the microbes present
in the ecosystem. Gene transfer frequencies will be documented.
The fate in various novel bacterial species of conjugatlve and
nonconjugatlve plasmids of different incompatibility groups will also be
investigated in both Tier I and II studies. These studies will employ
genetically characterized broad host range plasmids as well as "safe"
nonmobillzable plasmids. Bacterial 6pecies chosen as prospective donors
(potential novel organism candidates) are appropriate to the different
ecosystems under study:
A. The soil ecosystem (ex. Pseudomonas, Alcaligenes,
Klebsiella, Enterobacter, Rhizobiutn).
B. The aquatic ecosystem(s) (ex. Pseudomonas, Escherichia,
Klebsiella
C. The plant-associated habitats (ex. Pseudomonas,
Alcaligenes, Agrobacterlum, Cltrobacter, Enterobacter,
Erwlnla, Pectobacterium, Klebsiella).
D. The invertebrate and mammalian ecosystems (genera as above)
plus Escherichia coll.
Potential recipients of genetic material will consist of genera present in
the ecosystem intended to receive the novel organism.
32
-------
All mating combinations Investigated in microcosms will be Initially
tested in the laboratory for feasibility and frequency of occurrence.
Screening systems will also be tested for the detection of appropriate
biological markers in recombinant organisms. Plasmid containing strains
trill serve as donors while recipients, in this initial phase, will be
those naturally occurring strains isolated from the various ecosystem
models in which the novel organism will be released. In addition, well
characterized strains of E. coll will serve as control recipients.
In the microcosm analyses the occurrence of triparental matlngs will
be measured In addition to the direct plasmid transfer between two
participating cells. Triparental matlngs will provide an index of the
incidence of helper plasmids in the terrestrial microcosm flora and will
define the capabilities of natural helper plasmids to mobilize different
classes of plasmids present in the novel organisms added to the
ecosystems. This aspect of the study is especially important in assessing
the fate and effect of novel microbes and has not been Investigated
before. Plasmid containing donor cells added to ecosystems will be
constructed, as far as applicable, to contain various combinations of a
conjugatlve plasmid or a nonconjugative plasmid as well as those which are
or are not mobilizable, as well as strains which contain pairs of
different plasmid types.
Certain genetic elements are designated by the term, transposons
(16,17,20). Transposons or "hopping genes" have the ability to leave a
plasmid and enter the chromosomal DNA of the organism which contains it or
can be transferred into another bacterium if carried on a mobilizable
plasmid. Transposons are known to occur among microbes in nature.
Therefore, consideration should be given to the importance of transposons
33
-------
In transferring genes from a novel organism plasmid to its chromosome or
vice-versa. The subsequent transfer of these genes to other Indigenous
organisms could be enhanced by the occurrence of transposons and helper
plasmids in the natural microflora. Research elements should be
investigated for developing citable methods for evaluating the roles of
both transposons and helper plasmids in assessing the genetic stability of
released novel organisms.
In one group of the terrestrial in vivo mating trials, Rhizoblum
strains will be used which nodulate subclover or soybeans. Inoculations
of seeds will be conducted using two.Rhizoblum species or strains which
have the potential to occupy the same legume nodule (double infection;
32,33). In this manner the possible in vivo transmission of plasmids
within individual legume nodules can be measured.
In addition, specific strains of Alcaligenes and Pseudomonas will be
used as potential donors of plasmids which code for the metabolism of 2,4-
D biodegradation (13,46). These plasmids have already been shown to have
a broad host range and are moblllzable to a variety of bacterial species,
at least under laboratory conditions (13,40).
Screening and detection of recombinants will be made possible through
the use of genetically tagged donor strains. Recovery of recombinants
will be achieved by plating out specimens on selective media containing,
for example, antibiotics appropriate to the plasmid resistance markers,
and a carbon source not utilized by the donor strains. Screening and
detection of recombinants will be facilitated through the use of
appropriate markers Induced by transposon mutagenesis (17,20).
Presumptive in vivo plasmid transfer will always be verified. This
will Involve demonstrating the taxonomic or phenotyplc distinctiveness
34
-------
between donor and recombinant as well as actual physical demonstration of
the appropriate plasmid In recombinants. The latter will involve
characterization of plasmid molecular weight and enzyme restriction
patterns. Recombinants will also be examined to determine if they can in
turn serve as donor cells to transmit the plasmid which they have
received.
3. Sources of cultures and plasmlds
Bacterial cultures will be obtained from several sources. These
Include specific cultures, already bioenglneered, requested from
scientists who have described such strains in the scientific literature.
Cultures will also be obtained directly from the environment (plant root
rhizosphere, decaying vegetation, sediment, etc.) and from international
culture collections. Cultures Isolated from the environment will be
characterized taxonomlcally to at least the genus level. These strains
will serve as experimental donors and recipients of plasmlds representing
various types of conjugative and nonconjugatlve groups which may be
employed in the construction of various kinds of novel organisms.
Plasmlds are widely available from international culture collections,
biological supply companies, or other scientists, and can also be
constructed in the laboratory using basic bioengineerlng techniques.
D. Short Term Products
The first short term products will be the technical documents which
compile known genera which have gene transfer capabilities. It is envisioned
that information will be compiled by genus to include not only chromosome,
plasmid, naked DNA, and bacterial virus gene transfer mechanisms, but also the
precise methods under which DNA exchange was established (growth phase, cell
densities, temperature, etc.). Tabulated information on the habitats where
35
-------
the organism resides and whether literature Is published on natural ecosystems
where DNA transfer has been documented will also be obtained.
Initial research will be conducted in soil and leaf phyllosphere habitats
to investigate the stability of naked plasmld DNA.
Tier I test procedures for conjugation and plasmid DNA transfer will be
initiated. Initial protocols will use common bacteria which are likely
candidates for release (Klebsiella, Xanthomonas, Pseudomonas) and which are
known to exhibit gene transfer capabilities. Working protocols will be
established.
Quality assurance protocols will be.developed in parallel with research
developments.
E. Long Term Products
Technical reports will be published on methods developed including fate,
effects, and stability of naked plasmid DNA.
Tier I DNA transfer experiments will be continued using various types of
DNA plasmlds. The host range of plasmid transfers (donor-recipient
combinations) will be based on Intended uses of novel organisms and organisms
In habitats impacted. Experimental niches will be established and tested for
trlparental matlngs In terrestrial microcosms.
EPA will sponsor a workshop at a national professional meeting on,
"Prokaryotlc gene transfer in natural ecosystems: relevance to release of
novel organisms." Appropriate technical reports will be published and all
appropriate methodologies will be validated.
36
-------
Time Lines: Genetic Stability Hethoda
4 mo 8 12 24 36 48
Data base
gene transfer,
methods, organisms
habitats for gene
transfer
publ. review document
in primary lit.
Naked plasmid DNA
stability, fate
Conjugation
Tier I
u
Evaluate research,
continue if warranted;
**
study fate in various
ecosystems
Vary genera,
plasmlds
Workshop **
>
>
Genetic stabil. in microcosms
Tier II gene exchange studies;
Coordinate W/D,I,E, Fate atudles,
Ecolog. Processes Hazards
Consider 5 ecosystems: water, soli, aerial plant
- ~ —^
parts, legumes, etc; invertebrates/mammals
and in pathogens of these blotlc species
Field
Validation
Methods -
solid arrows indicate estimated
document completion period (Data base or technical report)
**OTS protocols and support document
-------
VI. DEVELOPMENT OF TEST METHODS FOR ASSESSING HAZARDS OF RELEASED NOVEL
ORGANISMS
A. Statement of Research Problems
The regulatory goal of permitting the release of only low-risk novel
organisms requires the development of test methods that will distinguish
between hazardous and non-hazardous organisms and test methods which allow a
determination of the nature of the hazards if they exist. This portion of the
research plan is organized around two major kinds of potential environmental
hazards from novel organisms released to the environment: 1) pathogenicity,
including toxicity and infectivity for various non-target organisms and 2)
other effects on environmental processes.
This approach to hazard identification is based on the assumption that a
variety of test methods will be developed applicable to a variety of intended-
use ecosystems. The selection of methods to which to subject a given organism
would be based on the knowledge of the organism's identity, source, unique
attributes for which the organism is released, genetic manipulations, and
intended use (site, concentration, application method). Portions of the
hazard identification methodology will be dependent upon exposure assessment
for determining populations exposed to the agent. The latter information will
be available from experimental data obtained from Section IV on fate and
transport.
There are two categories of research problems which reflect the two kinds
of environmental hazards. First, there is a need to develop test methods for
pathogenicity, including toxicity and infectivity of novel organisms released
in terrestrial and aquatic ecosystems; second is the need for test methods for
determining the nature, magnitude, and consequences, if any, of novel microbes
on natural ecosystem processes other than pathogenicity and gene stability.
38
-------
It Is anticipated that data requirements in subdivision M (63) would be useful
for screening novel organisms (MPCAs and non-MPCAs) for pathogenicity,
toxicity, etc. Simple and complex containment studies would be the approach
for assessing hazards to ecological processes.
B. Availability of Data Base
The Office of Pesticide Programs (OPP) has developed guidelines for
testing hazards to non-target organisms by microbial pesticides. These
guidelines provide detailed methods for testing pathogenicity, Including
toxicity, and lnfectlvlty, of microbial pesticides to a variety of organisms
from terrestrial habitats (63). In addition, a vast amount of Information
available In the plant, animal, and microbial pathology literature would
provide methods and data concerning other (non-pesticidal) microbes.
Guidelines for similar testing strategies involving engineered MPCAs are
predominantly the same as for non-engineered MPCAs (64).
On the other hand, OPP guidelines on ecosystem processes and expression
data requirements for microbial agents are relatively undeveloped and,
therefore, will be of very little value in this test methods development.
However, there exists a vast literature In plant, animal, and microbial
ecology and environmental microbiology which contains experimental approaches,
and test methodologies pertinent to ecosystem processes.
There is no similar data base for organisms that may be used in
industrial processes that are covered under the Office of Toxic Substances
Programs.
C. Approaches
Short-term research is directed at a) data base development and b) an
evaluation of initial experimental data obtained from fate, transport and
survival studies developed from Section IV.
39
-------
In the first approach, a search and evaluation will be conducted of
literature for information and methods for testing pathogenicity, toxicity,
and infectivlty, of novel microbes to various organisms in target ecosystems
and in ecosystems to which they are transported. A small workshop with
program office personnel and outside experts will be held to provide guidance
on naming likely novel microbes and identifying model(s) used for developing
test methods for measuring hazard(s) to target and non-target ecosystems. The
workshop should also assess applicability of OPP non-target hazard guidelines
to TSCA-covered novel organisms. Recent assessments of OPP guidelines should
be helpful in this applicability study. Goals of the workshop would also be
to identify the applicability of and/or changes in guideline tests for other
novel organisms.
The second approach, will be to develop methods of hazard assessment
based on microcosm experimental data obtained from fate and transport and
genetic stability studies. This data will provide Insights for formulating
models of risk assessment. Possible hazard risk assessment models currently
considered as candidates Include: a) involvement of aquatic invertebrates and
insects as vectors which transport novel agents to and from target and non-
target plant species including plant pathogenic viruses, nuclear-polyhedrosis
virus, bacteria, and fungi; b) application of a novel organism to a toxic
waste applied to soil or water and the toxic waste or its degradation product
Is mutagenic; c) use of a novel organism with Increased capacity to mineralize
matter that results in a soil pH change affecting plant growth; or reaches
significant numbers in terrestrial or aquatic systems and influences a natural
blogeochemical process (C, N, S cycles); d) niche displacement of microbes
associated with plants, insects, etc. by aggressive colonizing novel
bacteria. These approaches to hazard assessment applicable for target and
40
-------
non-target organisms will Include the spectrum of biological species
(microbes, plants, Invertebrates, birds, mammals).
Longer term research (laboratory, microcosm) In this area would be of
three kinds: (1) continue to revise and test the OPP hazard guidelines for
microbiological pest control agents (HPCAs); (2) fill gaps in ecological
process test methods which are apparent after completion of short-term data
base development and evaluation; (3) validate test methods.
D. Short Term Products
OEPER will conduct the workshop and prepare a report on the novel
organisms most likely to be used and their most likely route of release and
the ecosystems Impacted. A document will be prepared on data base development
and evaluation which contains methods and other pertinent Information from
literature searches.
OEPER will prepare a separate report evaluating applicability of OPP
guidelines to TSCA covered novel organisms containing recommendations and
modified guidelines with indications of gaps in information. A search and
evaluation will be conducted of the ecological literature for information and
methods for testing ecosystem expressions and processes. It will also be
Important to delineate discrete ecosystem types and discrete testable end-
point effects in the various ecosystem types. Guidance should be available
from the program office.
Five to 10 ecosystem processes will be Identified that are likely to be a
target of novel organism risk assessment. Then it should be possible to
delineate possible end-point effects within these ecosystems. Measurable end-
points should exist, for example, In the systems of xenoblotic degradation,
nitrogen, carbon, and mineral cycling; In Ritualistic relationships such as
Rhlzoblum-legume and mycorhlzzal associations; in competition In the
41
-------
rhlzosphere and phyllosphere of planes (changes in resident populations, niche
displacement, effects on trophic levels).
E. Long Term Products
Investigators will publish technical papers showing methods development
dealing with examples of hazards (pathogenicity/toxicity, ecological process
changes) and their measurement. Technical papers will also detail validation
of test methods. OEPER will prepare a set of suggested guidelines for OTS and
OPP.
A significant goal is to develop research systems in which to test for
effects on various endpoints as delineated in short-term research. We would
anticipate that the most likely approach will be with the use of microcosms.
However, the test system will be developed to gain information about specific
endpoints and not ecosystems in general. As methods proceed and develop,
there will be appropriate validation of all tests.
42
-------
Summary Time Lines, All Test Methods
Initial Workshop November 1984
Data base documents
All sections 6-10 mo
Established working model microcosms 12 mo
Detection, Identification,
Enumeration
OTS PROTOCOLS.
SUPPORT DOCUMENTS:
Conventional Techniques
Molecular Techniques
36 mo
40 mo
Fate
Microcosms
Blotlc/ablotlc factors
on Pate and Transport
Post workshop updated
document
24 mo
36 mo (Draft)
42 mo
Genetic Stability
Fate/Effect naked DNA
Fate/Effect plasmld DNA
Tier I
Tier II
34 mo
30 mo
36-40 mo
Assessing Hazards
MPCA
Insects, Avian
Mammals, Plants
28-30 mo
40-48 mo
Ecological Processes
36-48 mo
-------
VII. SUMMARY
The goal of this document is to present a research plan for developing
test methods for risk assessment involving the accidental or purposeful
release of novel organisms to the terrestrial or aquatic environment. The
document is not a research proposal but many detailed research proposals will
be developed from the plan.
The plan is developed on the basis of program needs, OEPER's research
priorities in biotechnology, and the expertise in test methods development in
our research group at various laboratories.
There are two major approaches to the plan: data base development and
research. The scope involves both short (1-2 years) and long term needs. The
methods development addressed regarding release of novel organisms are:
1. Detection, identification and enumeration;
2. persistence in the environment;
3. fate and transport to niches other than those Intended;
4. gene transfer to and from other microbes in the ecosystem; and
5. involvement in environmental effects
a) pathogenicity and acute and chronic toxicity to non-target
species
b) disruption of environmental processes.
44
-------
Time Lines: Methods for Assessing Hazards
MPCA (ongoing efforts, insects, avian)
0 4 12 Ik 36 48
Data base
Invertebrates, mammals, plants
Research measuring
Pathogenicity, toxicity, infectivlty:
Insects - — - !~**
Avian, crustacians •
Examine, develop tests and possible revisions
Mammals to OPP hazard guidelines (all_species) ^
Plants ...
Ecological Processes
Data base and workshop: n^vel organisms,
£ env. impacted, routes of release ^
Analyze, evaluate for methods development:
ecosystem expressions and processes -
Integrate with all areas of ongoing research including MPCA
Identify 5-10 y
ecosystem processes
Integrate with ongoing
microcosm research -
Test methods for measuring Impacts on ecological
processes
Plants, insects, crustacians, avian, mammals, cycles of
mineralization, niche displacement ..
Ify. ?>.
solid arrows indicate Field
estimated completion period Validation
Methods
**0TS protocols and support document
-------
VIII. ACKNOWLEDGEMENTS
This research plan was written by Ramon J. Seidler (Oregon State
University) with significant contributions from Jane Rissler (University of
Maryland). Many Individuals have Influenced the form and composition of the
final draft. Special appreciation is extended to Charles Hendricks,
A1 Bourquin, Guenther Stotzky, Bob Brink, and Fred Betz for their critical
review and numerous editorial and technical suggestions. Appreciation is also
extended to members of the AAAS Workshop for their reviews, suggestions, and
encouragement. The coordination and development of the research plan is largely
due to the efforts of Harold Klbby.
46
-------
IX. LITERATURE CITED
1. Clay, D. 1983. Testimony before U.S. Congress House Subcommittee on
Science, Research, and Technology and House Subcommittee on Investigations
and Oversight. June 22, 1983. Washington, D.C.
2. OTS Research Needs: Background Material for the December ORD
Biotechnology Workshop, OTS, EPA, Nov. 1983.
3. Biotechnology Workshop Report. 1983. Report on ORD-sponsored
biotechnology workshop, December 14-15, 1983, Washington, D.C.
4. Lowrance, W. W. 1976. Of acceptable risk: science and the determination
of safety. Wm. Kaufmann, Inc., Los*Altos, CA, 180 pp.
5. National Research Council. 1983. Risk Assessment in the Federal
Government: Managing the Process. National Academy of Science. National
Academy Press, Washington, D.C.i 191 pp.
6. Powledge, T. M. 1983. Public says genetic engineers should proceed
cautiously. Biotechnology. October, pp. 645-646.
7. Powledge, T. M. 1983. Will restraints continue to loosen?
Biotechnology. June, pp. 322-328.
8. Sharpies, F. E. 1983. Spread of organisms with novel genotypes:
thoughts from an ecological perspective. Recorab. DNA Tech. Bull. _6:43-56.
9. Mllevski, E. 1983. Congressional hearing on the environmental
implications of genetic engineering. Recomb. DNA Tech. Bull. ^103-110.
10. Old, R. W., and S. B. Primrose. 1981. Principles of gene manipulation.
University of California Press, Berkeley, California.
XI. Clean-up microbes for the environment. Science News 119:246.
12. Kellogg, S. T., D. K. Chatterjee, and A. M. Chakrabarty. 1981. Plasmid-
assisted molecular breeding: New technique for enhanced biodegradation of
persistent toxic chemicals. Science 214:1133-1135.
47
-------
13. Pemberton, J. M., and R. H. Don. 1981. Bacterial plasmids of
agricultural and environmental importance. Agrlcul. Environ. _6_:23-32.
14. Stotzky, G., and V. N. Krasovsky. 1981. Ecological factors that affect
the survival, establishment, growth and genetic recombination of microbes
in natural habitats. In Levy. S. B., C. Cloves, and E. L. Koenig (ed.)
Molecular Biology, Pathogenicity, and Ecology of Bacterial Plasmids.
Plenum Press, New York, NT.
15. Schvien, U. and E. Schmidt. 1982. Improved degradation of
monochlorophenols by a constructed strain. Appl. Environ. Microbiol.
44:33-39.
16. Hardy, K. 1983. Bacterial plasmids. American Society for
Microbiology. Washington, D.C.
17. Boucher, C., B. Message, D. Debieu, and C. Zischek. 1981. Use of P-l
incompatability group plasmids to introduce transposans into Pseudomonas
solanacaearum. Fhytopathol. 71:639-642.
18. Ditta, G., S. Stanfleld, D. Corbln, and D. R. Helinski. 1980. Broad host
range DNA cloning system for Gram-negative bacteria: Construction of a
gene bank of Rhlzoblum melllotl. Proc. Natl. Acad. Scl. USA 77:7347-7351.
19. Barth, P. T., L. Tobln, and G. S. Sharpe. 1981. Development of broad
host-range plasmid vectors. _In_Levy, S. B., C. Clowes, and E. L. Koenig
(ed.) Molecular Biology, Pathogenicity, and Ecology of Bacterial
Plasmids. Plenum Press, New York, NY.
20. Kleckner, N., R. K. Chan, B. K. Tye, an D. Botstein. 1975. Mutagenesis
by insertion of drug resistance element carrying an Inverted repetition.
J. Mol. Biol. 97:561-575.
48
-------
21. Leemans, J., D. Inze R. Villarroel, G. Engler, J. P. Hemalsteens, M.
DeBlock and M. Van Montagu. 1981. Plasmid mobilization as a Cool for in
vivo genetic engineering. In Levy, S. B., C. Cloews, and E. L. Koenig
(ed.) Molecular Biology, Pathogenicity, and Ecology of Bacterial
Plasolds. Plenum Press, New York, NY.
22. Kirk, T. K. 1983. Biotechnology In utilization of wood.
Biotechnology. October, pp. 666-668.
23. Berlnger, J. E. 1979. The development of Rhlzoblum genetics. J. Gen.
Microbiol. 116:l-7.
24. Levlne, M. M., J. B. Kaper, H. Lockman, R. E. Black, and M. L. Clements.
1983. Recombinant DNA risk assessment studies In man: Efficacy of poorly
moblllzable plasmids in biologic containment. Recomb. DNA Tech. Bull.
j6:89-97.
25. Terrestrial Microcosms and Environmental Chemistry. 1977. _In Witt, J.
M., and J. W. Gillet (eds.). The Proceedings of Two Colloqula, June 13-
14, 1977, OSU, Corvalll8, OR. National Science Foundation, Washington,
D.C. 20550.
26. Terrestrial Microorganisms. 1977. ^In^ Gillet, J. W. and J. M. Witt,
(eds.). The Proceedings of the Workshop of Terrestrial Microcosms.
National Science Foundation, Washington, D.C. 20550.
27. Personal communication. Dr. Craig McFarlane. Corvallis Environmental
Research Laboratory, Toxic and Hazardous Materials Branch, U.S. EPA,
Corvallis, Oregon 97331.
28. Kleeberger, A., and W. Klingmuller. 1980. PIasmld-medlaced transfer of
nitrogen-fixing capability to bacteria from the rhizosphere of grasses.
Molec. Gen. Genet. 180:621-627.
49
-------
29. Krallkova, K., V. Krcmery, Sr., and V. Krcmery, Jr. 1983. Incidence of
gentamidn-reslstant bacteria and those with R. plasmids, transferable to
Z. coll K-12, in minlclpal waste waters. Recomb. DNA Tech. Bull. _6_:98-
100.
30. Schllf, W., and W. Klingmuller. 1983. Experiments with Escherichia coli
on the dispersal of plasmids in environmental samples. Recomb. DNA Tech.
Bull. 6_: 101-102.
31. Talbot, B. V., Jr., D. K. Yamanoto, M. V. Smith, and R. J. Seidler.
1980. Antibiotic resistance and its transfer among clinical and non-
clinical Klebsiella in botanical environments. Appl. Environ. Microbiol.
39:97-104.
32. May, S. N., and B. B. Bohlool. 1983. Competition among Rhizoblum
legumlnosarum strains for nodulation of lentils (Lens esculenta). Appl.
Environ. Microbiol. 45:960-965.
33. Kosslak, R. M., B. B. Bohlool, S. Dowdle, and M. J. Sadovsky. 1983.
Competition of Rhlzoblua melilotl strains In early stages of soybean
nodulation. Appl. Environ. Microbiol. 46:87Q-B73.
34. Tamamoto, D. K., and R. J. Seidler. 1981. Colonization of growing radish
plants by clinical and nonclinical Isolates of Klebsiella Inoculated onto
seeds. Curr. Microbiol. 289-293.
35. Kloepper, J. W., M. N. Schroth, and T. D. Miller. 1980. Effects of
rhlzosphere colonization by plant growth-promoting rhlzobacterla on potato
plant development and yield. Phytopathol. 70:1078-1082.
36. Schroth, M. N., and J. G. Hancock. 1982. Disease-suppresslve soil and
root-colonizing bacteria. Science 216:1376-1381.
50
-------
37. Johnston, A. W. B., G. Hombrecher, and N. J. Brevin. 1981. Rhlzoblum
plasmlds: their role in the nodulatlon of legumes. _In Levy, S. B., C.
Cloves, and E. L. Koenlg (ed.) Molecular Biology, Pathogenicity, and
Ecology of Bacterial Plasmlds. Plenum Press, New York, NT.
38. Johnston, S. W. B., J. L. Beynon, A. V. Buchanan-Wollaston, S. H.
Setchell, P. R. Hlrsch, and J. E. Berlnger. 1978. High frequency
transfer of nodulating ability between strains and species of Rhizoblum.
Nature 276:634-636.
39. Suslow, T. V., and H. N. Schroth. 1982. Rhlzobacterla of sugar beets:
Effects of seed application and root colonization on yield. Phytopathol.
72:199-206.
40. Suslow, T. V., and M. N. Schroth. 1982. Role of deleterious
Rhlzobacterla as minor pathogens in reducing crop growth. Phytopathol.
72:111-115.
41. Kloepper, J. V., and M. N. Schroth. 1981. Relationship of in vitro
antibiotics of plant growth-promoting rhlzobacterla to plant growth and
the displacement of root microflora. Phytopathol. 71 ;1020-1024.
42. Broadbent, P., K. F. Baker, and Y. Waterworth. 1971. Bacteria and
actinomycetes antagonistic to fungal root pathogens In Australian soils.
Aust. J. Biol. Scl. _24:925-944.
43. Terrestrial Microcosms. 1977. Proceedings of a workshop on terrestrial
microcosms. J. tf. Glllett and J. M. Witt (ed.). National Science
Foundation, Washington, D.C.
44. Torsvik, V. L., nad J. Goksoyr. 1977. Determination of bacterial DNA in
soil. Soil. Biol. Blochem. 10:7-12.
45. Torsvik, V. L. 1983. Isolation of bacterial DNA from soil. Soil. Biol.
Blochem. 12:15-21.
51
-------
46. Don, R. H., and J. M. Pemberton. 1981. Properties of six pesticide
degradation plasmids Isolated from Alcallgenes paradoxus and Alcallgenes
eutrophus. J. Bacterid. 145:681-686.
47. Rlssler, J. F. 1983. Research needs for blotic environmental effects of
genetically-engineered microorganisms. Draft document.
48. Federal Register. 1983. Guidelines for research involving recombinant
DNA molecules. 48 :24556-24581.
49. Gerhardt, Philllpp, Editor-in-Chief. 1981. Manual of methods for general
bacteriology. American Society for Microbiology, Washington, D.C.
50. Collins, C. H.< and Patricia M. Lyne. 1976. Microbiological Methods.
Buttervorth & Co. London, England.
51. Aaronson, Sheldon. 1970. Experimental Microbial Ecology. Academic
Press, Inc. New York, NT.
52. The Prokaryotes. 1981. A handbook on habitats, isolation, and
identification of bacteria. Volumes I and 11.^ Springer-Verlag, New York.
53. Meyers, J. A., Sanchez, D., Elwell, L. P., and Falkow, S. 1976. Simple
agarose gel electrophoresis method for the identification and
characterization of plasmid deoxyribonucleic acid. J. Bacterlol.
127:1529-1537.
54. Grunstein, M., and Hogness, D. S. 1975. Colony hybridization: a method
for the isolation of cloned DNAs that contain a specific gene. Proc.
Natl. Acad. Sci. USA 72:3961-3965.
55. Moseley, S. L., Huq, I., Allm, A. R. M. A., So., M., Samadpour-Motalebi,
M., and Falkow, S. 1980. Detection of enterotoxigenic Escherichia coll
by DNA colony hybridization. J. Infect. Dis. 142:892-898.
56. Masui, Y., T. Mlzuno, and M. Inouye. 1984. Novel high-level expression
cloning vehicles: 10^-fold amplification of Escherichia coll minor
protein. Biotech. 2:81-85.
52
-------
57. Kohn, G. K. 1980. Bioassay as a monitoring cool. Residue Rev. 76:99-
127.
58. Mulla, M. S., G. Majori, and A. A* Arata. 1979. Impact of biological and
chemical mosquito control agents on nontarget biota In aquatic
ecosystems. Residue Rev. 71;121-173.
59. Bourquln, A. V., P. H. Prltchard, R. Vanolinda, and J. Samela. 1979.
Fate of Olmllln In Estuarlne microcosms. Absts. Ann. Meeting. American
Society of Microbiology, Q76, page 232.
60. Prltchard, P. H ., A. W. Bourquln, H. L. Frederickson, and T. Mazianz.
1979. System design factors affecting environmental fate studies In
microcosms. In: Proc. Workshop microbial degradation of pollutants in
marine environments. US EPA, Gulf Breeze, FL. pp. 251-272.
61. Bourquln, A. tf. 1977. Effects of malathion on microorganisms of an
artificial salt-marsh environment. J. Env. Qual. _6:373-378.
62. Liang, L. N., J. L. Sinclair, L. M. Mallory, and M. Alexander. 1982.
Fate in model ecosystems of microbial species of potential use in genetic
engineering. Appl. Environ. Microbiol. 44:708-714.
63. U.S. E.P.A., 1983. Pesticides Assessment Guidelines, Subdivision M -
Biorational Pesticides. National Technical Information Service,
Springfield, VA. No. PB83-153965.
64. Betz, F., M. Levin, and M. Rogul. 1983. Safety aspects of genetically-
engineered microbial pesticides. Recomb. DNA Tech. Bull, bj 135-141.
65. Camann, D. E. 1980. A model for predicting dispersion of microorganisms
in wastewater aerosols, pp. 46-70. JLn Wastewater aerosols and disease.
U.S. EPA 600/9-80-028.
53
-------
B
-------
Biotechnology Health Risk Assessment Research Plan
by
Marvin Rogul
The Rogul Group
Washington, D.C.
John R. Fowle III
Office of Health Research, EPA
Washington, D.C.
David Kleffman
Office of Health Research, EPA
Washington, D.C.
Office of Health Research
U.S. Environmental Protection Agency
Washington, D.C.
October 1985
-------
TABLE OF CONTENTS
I. Introduction 1
II. Health Effects Work Group Panel Recommendations 3
A. Data Gathering and Information Management 3
B. Selection of Organisms for Validating Subpart M Test Approach....7
C. Protocol Development for Infectivity, Pathogenicity, and
Metabolic Characteristics of Recombinant Microorganisms 8
D. Bacterial Pathogenicity Categories 8
E. Establishment and Management of a Data Base of Characteristics
of the Potential Hazards of Genetically Modified Material 10
F. Selection and Assessment of Safe Hosts 10
G. Development of Molecular Probes 10
III. Discussion 12
A. Risk Assessment 12
B. Foundation Laid by the NIH Recombinant DNA Advisory Committee...14
1. E. coll studies which influenced the Development of the RAC
Tauidellnes 15
2. Experiments Simulating High Risk Conditions: Promoting and
Detecting Genetic Interchange 16
IV. References 18
-------
I. INTRODUCTION
Certain products of biotechnology fall within the regulatory purview of
the Office of Pesticides and Toxic Substances (OPTS) under the jurisdiction
of the Toxic Substances Control Act (TSCA) and the Federal Insecticide,
Fungicide and Rodenticide Act (FIFRA). In order to accomplish their mission
under these Acts, OPTS requested that the Office of Research and Development
(ORD) provide technical assistance in the development of models and data
bases for use to evaluate biotechnology products. This chapter presents
a research approach developed by the Health Effects Workgroup (HEWG) of
the Coolfont Workshop for using data and applying test methods to assess
the possibility of health risks associated with exposure to genetically
engineered microbes released into the environment. New test methods may
be developed as the need arises. This chapter presents a research plan,
not a research proposal. Specific proposals will be developed from this
plan. In all cases EPA will base its activities on the extensive body of
information concerning medical microbiology and will coordinate work with
other Agencies with health related missions, such as the National Institutes
of Health (NIH) and the Food and Drug Administration (FDA).
The Coolfont Health Effects Work Group (HEWG) recommendations are
described in this chapter with discussions of certain considerations about
risk assessment that were factored Into the development of the research
plan. The key role that the NIH Recombinant DNA Advisory Committee
(RAC) has played in assessing risks from certain genetically engineered
organisms and in providing guidance for future efforts Is recognized at
the conclusion of the chapter.
-------
-2-
This brief review points out the nature of the work EPA intends to
pursue with respect to the possible health effects of genetically
engineered organisms released into the environment.
-------
-3-
II. HEALTH EFFECTS WORK GROUP PANEL RECOMMENDATIONS
The major conclusion of the Cool font HEWG was that models and test
methods to predict the potential health effects arising from the acciden-
tal or deliberate release of biotechnology products are not scientifically
feasible at this time because of the variety of potential applications
of biotechnology and the types of organisms which could be used in
commerce. The panel emphasized that because of this, health assessments
of biotechnology products must be performed on a case-by-case basis.
Two major areas were identified where useful research could be con-
ducted: 1) the development of animal testing protocols to determine the
health effects of genetically altered organisms and 2) specific experiments
on agents of immediate concern to the Agency. The conclusions and general
recommendations provided by the HEWG are found in Table 1 and are discussed
below.
A. Data Gathering and Information Management
It was stressed that steps should be taken to insure access to a wide
range of pertinent, up to date information, including relevant computer
data bases, use of computers to catalog in-house experience, storage of
pertinent DNA sequence information, information about tissue banks and
monitoring data, etc.
As part of the emphasis to build upon past experience HEWG emphatically
recommended that tests be developed in accordance with the principles
contained in the Pesticide Assessment Guidelines of Subdivision M:
Biorational Pesticides which were developed by the EPA in 1982.* (See
Table 2 for an outline of the Subpart M Testing Approach.)
-------
-4-
Table 1. Conclusions and General Recommendations
1. Health research on the pathogenic potential of genetically engineered
microorganisms should be directed at answering questions for specific
organisms and specific situations. General situations do not presently
exist. An accumulation of data from many individual cases may eventually
be used to generate more general guidelines.
2. Since the prediction of potential health effects from genetically
engineered organisms is not completely attainable or scientifically
feasible in many cases, research aimed at developing generic and specific
models and test methods should be conducted in this area. For the present,
decisions on hazard and risk potential must be made on a case-by-case
basis.
3. Development of information systems may be useful for literature
searches and the cataloging of data on biotechnology, rather than as
an aid to predicting pathogenicity.
4. EPA must have ready access to existing literature on microbial
pathogenicity and biotechnology. An extensive library collection should
be developed.
5. The Agency has to develop in-house expertise and strong communications
with non-agency experts to review the health effects potential of new
microorganisms. Optimally, 1n-house experts should be centralized in
one location.
6. EPA projects in biotechnology should go through rigorous peer review
similar to that performed at NIH. It is recommended that the Agency
convene a special subpanel of the Science Advisory Board and the Science
Advisory Panel to act as a review committee and that each laboratory
adopt an internal review mechanism and Institutional Biosafety Committee.
-------
-5-
7. Subdivision M guidelines for microbial pest control agents should
be reexamined and revised where necessary. The testing of viral agents
should be expanded and the guidelines should be validated with known
microorganisms (pathogenic and nonpathogenic).
8. To consider the potential pathogenicity of a new microorganism, it is
essential to first consider the nature of the source organisms and the
construct. If health effects testing is perceived to be necessary, it must
be performed in animal models which closely approximate human experience.
If data on human health effects are available, the data should be given
very high consideration when developing regulatory decisions.
9. It is recommended that the advice of a team of experts from certain
key disciplines be made available to manufacturers at early stages in
the development of products. The team of experts would work with manu-
facturers to assess possible risks and to suggest options and appropriate
tests prior to commercialization of a genetically engineered organism.
The team would include both EPA and outside experts.
10. It is recommended that funds be allocated to the Office of Research
and Development for the development of a continuing program to conceive,
develop, and evaluate the use of a variety of genetic markers for use
in tagging microorganisms (for identification purposes) and to advise
industry in their use. Genetic markers might include unusual resistance
patterns to antibiotics not used in human or animal medicine, unusual
fluorescence and metal ion resistance, or production of unique proteins.
Thus, tagged organisms could be identified or dismissed as etiological
agents of disease.
-------
-6-
Table 2. Outline of FIFRA Subpart M Microbial Pesticide Control Agent
Test Requirements
Tier I tests for Microbial Pest Control Agents:
1. Acute oral toxicity/infectivity in the rat (LD50)
2. Acute dermal toxicity/infectivity, rat or mouse (LD50)
3. Acute inhalation toxicity/infectivity in the mouse, rabbit
or guinea pig (LD50)
4. Intravenous intracerebral, and intraperitoneal toxicity/infectivity
using rabbits, new born or newly weaned mice and hamsters.
5. Primary dermal irritation on guinea pigs or rabbits
6. Primary eye irritation on rabbits
7. Hypersensitivity in hamsters or rabbits
8. Hypersensitivity incidents
9. Cellular immune responses in mice
10. Tissue culture tests for viruses
Tier II tests include:
The tests in this Tier are similar to Tier I but somewhat more
extensive in that they add a subchronic oral test, teratogenicity,
mutagenicity, and virulence enhancement tests. Dogs are added
as test animals and the number of animals used are increased.
Tier III tests:
1. Chronic oral test in the rat
2. Oncogenicity test in newly weaned mice and rats
3. Mutagenicity test in mammals
4. Teratogenicity test in 2 species from rat, mouse, hamster
or rabbit
-------
-7-
B. Selection of Organisms for Validating Subpart M Test Approach
The HEWG advised that a few key organisms should be studied using the
approach outlined by the Subpart M guidelines to determine its validity.
They suggested that the organisms developed to mitigate environmental
pollutants in EPA's Engineering Research Laboratories should be included
in this effort. The group suggested that Pseudomonas syringae "ice
minus" and Pseudomonas fluorescens containing the jJ. thuringiensis
toxin gene would be good candidates for study also.
C. Protocol Development for Infectivity, Pathogenicity, and Metabolic
Characteristics of Recombinant microorganisms
The HEWG recommended that EPA develop and validate more animal
pathogenicity test protocols and improve the testing component for viruses,
viral products, and, if appropriate, oncogenes in the Subdivision M
guidelines. The workgroup believed that a health testing scheme following
the tiered testing methods of the Subdivision M guidelines would provide
the Environmental Protection Agency with experience that might enable
the development of more specific and appropriate tests. Such tests
would employ non-human animal models or would test certain parameters of
importance to human physiology and could be used to demonstrate the
reasonable safeness of genetically engineered products and GEMS prior to
EPA approval for commercial production and application. One such test
could be the ability to grow at 37°C. This test would distinguish organisms
which could grow systemically at human body temperature. It would be
reasonable to assume that most organisms which could not grow at this
temperature would not be systemic pathogens. The production of toxic
-------
-8-
effects in human tissue culture and the production of exo- or endo-toxins
(such as toxin A or elastase in Pseudomonads, exotoxins in Corynebacteria)
or invasins (such as coagulase and laminin in Staphylococci) or oncogenic
material could also be added to Tier I testing protocols for the purpose
of excluding potentially harmful organisms from beiny released to the
environment with subsequent human exposure.
D. Bacterial Pathogenicity Categories
The Agency was advised to compile a list of the characteristics of
pathogens and nonpathogens that would have commercial application and to
advise industry on its use. The Agency was advised to develop bacterial
pathogenicity categories ranging from known pathogens to organisms incapable
of infecting humans, and if faced with the situation, to place special
emphasis on the microorganisms that might be constructed from oncogenic
viruses. It was recommended that the Agency support the evaluation and
validation of the Subdivision M pesticide guidelines. This should be
done for genetically engineered microbes, non-GEMS and a variety of
viruses and vectors. Expert advice should be sought from outside the
Agency to guide these efforts.
From the environmental health viewpoint, most risk assessment work
has been done with £. coli and is of limited value from the environmental
standpoint. There is little information on other organisms which are
being considered for commercial use in large scale contained facilities
or for release to the environment.
Except for allergies, and the rare infection due to accidential
exposure to the frank pathogens used in the production of toxins and
antigens for vaccines, there are no major problems in the traditional
biotechnology industries of baking, wine and beer making or other
-------
-9-
commercial activities using microorganisms. A number of companies
(Sybron Biochemical, Polybac and Flow Laboratories) have had at least a
decade of experience in degrading environmental waste by using micro-
organisms in bloaugmentation programs. No known health mishaps such as
infection have occurred.2 Although this does not guarantee that there
will be no human health hazards from environmental release of genetically
engineered organisms, it does show that our past experience with the
commercial use of microbes has not resulted in an unreasonable risk. On
this basis the HEWG opined that releases of genetically engineered microbes
are not likely to pose a health risk to humans unless they are pathogens,
related to pathogens, or contain a gene whose product is toxic to humans.
The large number of possible variations precludes a predictive model or
the ability to test each and every product. Thus, 1t was recommended
that each product be evaluated on a case-by-case basis and that advantage
be taken of the experience gained from each evaluation to Improve the
Agency's risk assessment capabilities. It is possible that relevant
health information will come from workers 1n the laboratory and at commer-
cial production facilities, where products of genetic engineering are
developed and made. Previous epidemiology studies of populations exposed
to microbial agents (e.g., sewage treatment plant workers, populations
exposed to wastewater aerosols) should be critically reviewed to determine
their relevancy to determining risks from exposure to GEMS. (Retrospective
epidemiology studies could be performed on workers who have been working
in traditional fermentation and other biotechnology firms. Prospective
studies could also be performed on these firms, the new genetic engineering
firms and biological pesticide applicators.)
-------
-10-
E. Establishment and Management of a Data Base of Characteristics
of the Potential Hazards of Genetically Modified Materials
Genetically engineered organisms should be evaluated using existing
data on parent, vector and host organisms and if needed, animal tests.
This should be used to support recommendation 1 above on infectivity,
pathogenicity and metabolic characteristics of recombinant microbial
organi sms.
F. Selection and Assessment of Safe Hosts
HEWG recommended that the EPA identify microorganisms that are consid-
ered to be relatively safe for use in developing new organisms and that these
"source" or host organisms be tested using the Subdivision M guidelines.
G. Development of Molecular Probes
The development of DNA/DNA, DNA/RNA, and RNA/RNA hybridization probes
as well as serological and other immunological tests are essential for
diagnosis and monitoring of interactions between genetically engineered
organisms and other species. These tests and their modifications, as
well as conventional monitoring techniques, are discussed in the monitoring
chapter of this report. The HEWG felt that the development and application
of such techniques are essential to any program evaluating potential
health effects from GEMS. One advantage of these approaches is that the
probes can specifically detect nucleotide sequences or gene products. A
new generation of avidin 3,4 complexing biotin probes holds special
promise for the future. The development of such approaches was thought
to have potential as an excellent diagnostic tool, and baculoviruses
were recommended as valuable research models for the development
-------
-11-
and application of such probes because of the use of baculovirus for
pesticide control purposes. Such work would have immediate benefit for
EPA decision-making.
-------
-12-
III. DISCUSSION
It was stressed by the HEWG that the use of animal surrogates is not
highly effective in detecting possible human opportunists or even pathogens.
That is why validation of these tests is extremely important.
A. Risk Assessment
To perform a risk assessment, data on (1) hazard identification and
characterization, (2) exposure and (3) dose/response relationships are
required. In the case of microorganisms it is assumed that after a
release to the environment risk is dependent on the survival of the
organism, and its growth and reproduction in a receptive environment;
the consequences can be either beneficial or deleterious. According to
Dr. Martin Alexander the construction of new genotypes and genetic exchanges
in the environment may lead to unexpected phenotypes and functions, and
because there is very little known about these processes, there is not a
sufficient data base from which to make adequate risk assessments.5
Prior to commercial production of GEMS, and their release to the
environment, some evaluation of potential health effects should be per-
formed based on existing literature or specific testing. The following
characteristics would be useful in assessing risks in experimental
animals and attempting to develop and validate extrapolation on approaches
to human risk. (The assumption in this line of reasoning is that generally
acceptable experimental animals models can be developed):
A. Nature and degree of pathogenicity/toxicity
1. Pathogenicity and virulence of viable organisms
a. route of infection
b. invasiveness in host
-------
-13-
c. replication in host
2. Toxicity due to viable or nonviable organisms or products
a. acute or chronic
b. mutagenicity
c. teratogenicity
d. oncogenicity
e. immune effects (toxicity and hypersensitivity)
B. Presence of other intracellular or extracellular agents (e.g.
viruses) in the product.
C. Degree of debility
1. auxotrophy
2. antibiotic sensitivity/dependence
3. pH sensitivity, light, temperature, (other physical factors)
D. Genetic stability
1. Under the conditions which it is to be used
a. environment (factory, outdoors)
b. host
E. Influence of the vector/insert an construct stability
F. State of the vector in the cell (integrated or non-integrated)
G. Mobility of vector
H. Expression of functions
I. Infectivity to other humans and animals
In genetically engineered microorganism there are two special concerns
which may relate directly or indirectly to human health. One concern
will be of genetic interchange between introduced organisms and those
which normally inhabit humans. The other consideration is the possible
interchange between DNA from GEMS and human DNA.
-------
-14-
Special consideration must be given to viruses, because they can
integrate into mammalian DNA and special probes should be made to
identify viral sequences if these viruses are to be used for commercial
purposes and released to the envirnment.
B. Foundation Laid by the NIH Recombinant DNA Advisory Committee
The members of the Recombinant DNA Advisory Committee (RAC) of the
National Institutes of Health (NIH) have had extensive experience
with risk assessment of genetically engineered organsisms. Although
most of the RAC experience has been medically oriented and limited to
work conducted with recombinant DNA (R-DNA) techniques to the exclusion
of all other forms of genetic engineering and manipulation, such as cell
fusion, transduction, transformation, conjugation and mutation, organiza-
tions such as the EPA can benefit from their approach and experience.
One of the major activities of the RAC was to formulate the NIH
Guidelines for Research Involving Recombinant DNA Molecules. The
guidelines have designated certain organisms as generally safe to
work with 1n the laboratory and usually exempt them from experimental
guidelines. The organisms are the attenuated Escherichia coli strain
K-12, an asporogenic Bacillus subtilis, and the yeast Saccharomyces
cerevisiae. These are very well characterized organisms and generally
believe to be safe with little chance of escaping from the laboratory;
incapable of surviving for any great length of time outside a laboratory
environment; or incapable of causing harm to humans or the environment.
In fact, NIH has funded a great deal of work to characterize the
potential human health risks of the medically important bacterium Escherichia
coli, the major bacterium for basic research in bacterial genetic engineering.
Their approach was discussed by the HEWG as a useful model for future efforts.
-------
-15-
1. E_. coli Studies which Influenced the Development of the RAC
Guidelines
Experiments were performed to determine whether certain strains of
debilitated coli strain K-12, could: (1) survive outside of the
laboratory, (2) survive in the mammalian gastrointestinal system, and
(3) transfer genetic information between strains in the mammalian gut.
The results indicated that E_. coli was a safe organism to use in that
although £. coli K-12 could survive in sewage, it could not persist in
the human alimentary system.6 In fact, when laboratory personnel who
worked with coli K-12 and R-plasmids (resistance factors; also called
R factors) were screened to determine whether their endogenous enteric
flora were contaminated with K-12 or R factors the results were negative.
There was no evidence of infection or R factor persistence in the gut.7
In human experiments greater than 10^ £. coli strain K-12 bacteria
were fed to volunteers. And although some of the organisms survived
passage through the digestive system, none of the organisms persisted in
the intestinal tract for more than a week or so, even under conditions
of extreme antibiotic pressure that would have been expected to aid
establishuent in the gut.8,9,10,11,12 it was also found that E. coli
K-12 was not capable of transfering poorly mobilized plasmid vectors
such as pBR325 in the presence of F-amp (fertility mobilizing factor) in
human volunteer studies. In fact, even a colonizing strain such as £. coli
HS could not transfer pBR325 unless antibiotic pressure somehow forced
and selected for the transferred vector in resident £. coli. However,
£, coli HS could participate in triparental conjugation in the human
intestine in the presence of tetracycline, and under these conditions
transfer the non-conjugative pJBK5 to the resident E. col 1.12
-------
-16-
From these findings it is evident that when microorganisms are
specifically debilitated and properly constructed, it is very difficult
to promote human infection and genetic interchange among them. The
exception appears to be when highly selective artificial environmental
conditions, such as antibiotic pressure are imposed upon the microorganisms.
Other organisms are also characterized by the RAC guidelines, but not
exempted from the guideline recommendations. These organisms are usually
characterized by their host characteristics and the genetic vectors they
contain. These host-vector (HV) systems are in turn assigned biosafety
levels comparable to their perceived level of risk. The greater the
danger of the host-vector, the higher the level of containment in the
biosafety system assigned.
2. Experiments Simulating High Risk Conditions: Promoting and
Detecting Genetic Interchange
Experiments sponsored by NIH have been designed to simulate high risk
conditions. In two of the most well known series of experiments, polyoma
tumor virus DNA was cloned into strains of £. coli K-12 with both plasmid
and phage vectors. In various fonms the cloned DNAs were introduced into
mice or cultured mouse cells. Evidence of polyoma virus infection was to
be assayed by measuring the immunological response to polyoma proteins.
Because of the potential dangers, these risk experiments were carried out
at the highest level of physical containment.
The results were consistent with the belief that these procedures
are extremely unlikely to produce evolutionarily fit, epidemic pathogens.
There was no immunological evidence for polyoma infection in mice when
the polyoma DNA was borne by a plasmid or by £. coli that were lysogenic
for 1ambda-polyoma prophage.
-------
-17-
On the other hand, in this same series of experiments, Immunological
evidence for biological activity was observed when dimers of the polyoma
DNA were carried by free-phage, and when the total DNA of the coli/
lambda-polyoma chimeras was introduced into the mouse.^ ^ However,
recent work has shown that prolonged feeding or inoculation of wild type
of laboratory strains of coli containing monomeric or dimeric forms
of polyoma virus to conventional, antibiotic-compromised and germ-free
mice did not demonstrate infection of the mice by polyoma virus.15 On
the basis of these findings 1t would be prudent to be especially mindful
of unexpected molecular interactions when dealing with any kind of mammalian
and retrovirus DNA. Preliminary data from EPA investigations in tissue
cultures warrant a close look at insect viral interactions with human
DNA (Kawanishi, C. personal communication).
-------
-18-
IV. REFERENCES
1. Betz FS, Beusch WR, Brlttin EB, Carsel R, Cohen SZ, Hoist RW, Keller A,
Mauer IN, Roessler W, Urban D, Vaughan A, and Woodrow W. Pesticide
Assessment Guidelines: Subdivision M, Biorational Pesticides. National
Technical Information Service. Springfield Va. 1982.
2. Genetic Control of Environmental Pollutants, ed. Omenn GS and
Hollaender A. pp. 331-349. Plenum Press, New York. 1984.
3. Lewin R. Gene Probes Become Ever Sharper. Science 221:1167, 1983.
4. Rlchman D et al. Summary of a Workshop on New and Useful Methods
in Rapid Viral Diagnosis. J Infect Ois. 150:941-951, 1984.
5. Alexander, M. Spread of Organisms with Novel Genotypes in Biotech-
nology and the Environment: Risk and Regulation, (eds) A. Telch, MA
Levin and JH Pace. AAAS, Washington, DC. 1985.
6. Sagik BP and Sorber CA. The Survival of Host-Vector Systems 1n Domestic
Sewage Treatment Plants. Rec DNA Tech Bull 2:55-61, 1979.
7. Petrocheilou V, and Richmond MH. Absence of plasmld of Escherichia
coli K-12 Infections among laboratory personnel engaged 1n R-plasmid
research. Gene 2:323-327, 1977.
8. Smith HW. Survival of orally administered E. coll K-12 in the alimentary
tract of man. Nature 255:500-502, 1975.
9. Anderson ES. Viability of and transfer of a plasmld from E. coli K-12
in the human Intestine. Nature 255:502-504, 1975.
10. Formal SB, Hornick RB. Invasive Escherichia coll. J Infect. Dis
137:641-644, 1978.
11. Levy SB, Marshall B, Rowse-Eagle D. Survival of Escherichia coll host-
vector systems in the mammalian intestine. Science 209:391-394, 1980.
12. Levine MM, Kaper JB, Lockman H, Black RE, Clements ML and Falkow S.
Recombinant DNA Risk Assessment Studies in Man: Efficacy of Poorly
Mobilizable Plasmids 1n Biologic Containment. Recombinant DNA Technical
Bulletin 6:89-97, 1983.
13. Israel, MA, Chan HW, Rowe WP and Martin MA. Molecular cloning of
polyoma virus DNA in Escherichia coli: plasmid vector system. Science
203:883-887, 1979.
14. Chan HW, Israel MA, Garon CF, Rowe WP, and Martin MA. Molecular
cloning of polyoma virus DNA 1n Escherichia coli: lambda phage vector
system. Science 203:887-892, 1979
15. Smith, C. Jr, E Milewskl and MA Martin. The effects of colonizing
mice with laboratory and wild type strains of Escherichia coli containing
tumor virus genomes. Recombinant DNA Technical Bulletin 8: 47-51, 1985.
-------
c
-------
Environmental Engineering Research Support Proposal
by
John Burckle
Industrial Environmental Research Laboratory, EPA
Cincinnati, Ohio
Albert D. Venosa
Environmental Research Laboratory, EPA
Cincinnati, Ohio
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
Revised, 8 June 1984
-------
ORD/OTS BIOTECHNOLOGY WORKSHOP
Office of Environmental Engineering Technology
B1o-Eng1neer1ng Research Support Program
Industrial Envlronmental Research Laboratory-Cincinnati
I. LEGISLATION
The Toxic Substances Control Act (TSCA) mandates the regulation of chemical
substances, new or existing, which present an "unreasonable risk of
Injury to health or the environment". Section 5 of TSCA requires that
any person who Intends to manufacture a new substance other than for
Research and Development, must submit a notice containing certain Information
to EPA for review at least 90 days before manufacture. The Office of Toxic
Substances (OTS) has the authority to prevent manufacture, regulate use,
or permit unregulated use depending upon the findings of this review.
II. REGULATORY NEEDS
Because genetically engineered microorganisms (GEMS) have been determined
to constitute a "new chemical" by the Office of General Counsel of the Environ-
mental Protection Agency (EPA), the OTS required to conduct reviews of
Premanufacture Notices (PMN) for such substances. To conduct such PMN
reviews, the OTS needs appropriate technical Information and predictive
capabilities to assess the potential risks arising from the release of
and exposure to these substances, and to evaluate proposed alternatives
and the costs for preventing release and exposure.
In preparation for the ORD (Office of Research and Development) Biotechnology
Workshop I, (December 1983), OTS Issued extensive material for the purpose
of determining specific research needed 1n support of anticipated OTS
regulatory activities. Subsequently, the regulatory needs of OTS were
-------
-2-
stated 1n the draft document, "EPA's Proposed Approach for Providing
Technical Support In Biotechnology and Meeting the Research Needs of the
Program Offices". Evaluations of PMNs are to be the "primary thrust" of
OTS activity. Emphasis 1s to be placed on identifying existing data
bases and expanding capabilities for predictive risk assessment purposes.
In addition, the Workshop I specifically Identified the objective and scope
of the Workshop II (May 1984).
The focus of the workshop and hence this proposal, reflects OTS's
primary concern regarding deliberate and accidental release into the
environment and subsequent exposure of humans and the environment to such
substances. Genetically engineered microorganisms used as pesticides,
drugs, cosmetics, and food are specifically excluded from TSCA jurisdiction
and therefore are not addressed here. However, genetically engineered micro-
organisms "manufactured" for the production of substances so used may be
of concern and are addressed 1n the context of Industrial manufacturing
processes and plants.
Further, OTS has limited Its concern, for the purposes of this workshop,
to genetically modified microorganisms that are viruses, bacteria, fungi,
algae, or protozoa. It Is not within the scope to discuss the risks
associated with (1) accidental environmental release or worker exposure
to the vectors (e.g., plasmlds, ep1somes, viroids); or (2) exposure or
release of commercial products produced by genetically modified
organisms (e.g., enzymes used as catalysts). Chemical byproducts and
contaminants (e.g., toxins, companion viruses) which are not Intentionally
produced by the manufacturer as a commercial product but are associated
with the genetically modified organism are to be considered In workshop
-------
-3-
dlscussions because the risks due to by-products and contaminants may be
considered to be additional risks associated with the genetically modified
organism Itself.
In addressing these areas, the OTS has posed the following Information
needs In the previous workshop.
1. Are there other concerns regarding control technologies, Industrial
release, and worker exposure?
2. Are there concerns specific to any of the following modified
organisms: viruses, bacteria, fungi, algae, and protozoa?
3. For each of the above concerns, what information (In addition
to the following) will OTS need to evaluate its concerns?
a. identification of production processes, equipment, and
practices which may result 1n worker exposure or
environmental release;
b. Identification of control methods available to prevent
or reduce exposure/release for each media.
4. Are there predictive tools or test methods available for
obtaining the information needed to evaluate each concern?
5. If predictive tools or test methods are lacking, can ORD
develop predictive tools or test methods? If so, ORD should
describe the proposed research project 1n depth, including: objectives,
limitations, research methods, the end product of the project,
and how it will specifically address OTS needs; how long the
project will take to complete; and projected costs.
6. Are there certain manufacturing techniques or uses to which
OTS should pay particular attention?
-------
-4-
III. OVERALL PROGRAM APPROACH
The overall risk assessment model Is envisioned as composed of a series
of modules which, when linked together, will permit estimation of the risk
associated with the production and subsequent use 1n commerce of specific
genetically engineered microorganisms. These modules, likely to be
composed of a number of submodules, address the following aspects:
engineering - potential for accidental and deliberate release and
prevention and control of releases
environmental - potential for survival in the environment and adverse
effects
health - potential for adverse health effects
monitoring - techniques for detection and strategies for routine
monitoring to ensure that release 1s not occurring, and,
in the event of a release, that timely remedial actions
can be taken
While limitations exist 1n the technical Information data base required
for the preparation of the assessment modules, they are more related to lack
of knowledge of the potential behavior and characteristics of the specific
genetically engineered microorganisms to be evaluated. This aspect is
addressed by a recent congressional report (Environmental Implications of
Genetic Engineering - Staff Report, Committee on Science and Technology
U.S House of Representatives, 98th Congress, 1984) which examines the
issues of regulation. This report acknowledes the "newness" of this area
and states that..."Many of the standard approaches to the review of
"conventional" chemical substances would not be applicable". The problem
-------
-5-
1s that the effects of the substances cannot be adequately predicted 1n
advance because there Is no accumulation of knowledge to permit such
prediction as there Is with chemicals based upon structure activity
analyses. The report also points out that the OTS chemical staff conducting
PMN reviews would require ..."familiarity with containment practices for
pathological organisms and the associated worker protection techniques...".
Because the physical, chemical, and biological aspects of the environ-
ment affecting the viability, behavior, and interactions of genetically
engineered microorganisms 1n that environment (e.g., temperature, moisture,
accessibility and level of nutrients, oxygen, predators, pathogens, etc.)
and the specific properties resulting from the genetic modifications cannot
be specifically anticipated in advance, these must be determined through
the health and environmental effects assessment procedures. Once these
data are developed, they can be used 1n assessment protocols to estimate
risks.
IV. SUMMARY OF PROPOSED ENVIRONMENTAL ENGINEERING EFFORTS RELATED TO
REGULATORY NEEDS
A. Regulatory Needs
The OTS has expressed three major concerns in regard to
environmental engineering technology:
1. accidental or deliberate release of the genetically engineered
microorganism from the site of production (e.g., In effluents), during
transport (shipping), Intermediate storage and subsequent use;
2. availability and effectiveness of containment controls
or destruction techniques; and
3. worker exposure, particularly due to aerosols.
-------
-6-
B. Program Structure
The OTS has expressed the need for methodology for assessing the
accidental or deliberate release of genetically engineered microorganisms
which could result In subsequent worker exposure and environmental
contamination. The program proposed here addresses manufacturing
processes based on genetically engineered microorganisms and subsequent
use in other contained processes. Because genetically engineered
microorganisms have already been developed for applications requiring
deliberate release Into the environment, the proposed program also
addresses the development of procedures for assessing the engineering
aspects of introduction of genetically engineered microorganisms Into
the environment for a number of such applications 1n the form of
"scenarios" appropriate to the environmental conditions likely to be
encountered at representative sites. Applications to be considered for
evaluations Include: agricultural formulations; pollutant clean-up/
control (eg, spills, landfills, contaminated sediments, oil spills);
tertiary oil recover; 1n-situ mineral recovery (metals leaching, oil
shale), and other such operations not contained in chemical processing
equipment in the traditional sense.
The engineering assessment protocols for release and exposure can be
structured to account for several sets or combinations of various biological
properties, or subsets, and appropriate applications Involving deliberate
environmental release, which the Workgroup and Review Committees feel should
be addressed at this stage. Further, effort will be devoted to identifying
those specific data (chemical, physical, and biological) which will be
required as Inputs to the engineering risk assessment protocols so that the
OTS can specifically require the development and submission of such data
-------
-7-
as part of the PMN review procedures.
The proposed program is presented in detail 1n Section IV.D and is
comprised of two major subdivisions:
(1) Subsection IV.D.I: Methodology for assessing the potential
for accidental and deliberate release, exposure, containment and
decontamination of genetically modified organisms from biologically
based manufacturing processes, Including the production site,
shipment, intermediate storage and subsequent process use.
(2) Subsection IV.D.2: Methodology for assessing the potential for
accidental and deliberate release from the site of application, worker
exposure, containment and decontamination and for developing criteria
for safe use 1n the deliberate Introduction of genetically
engineered organisms Into a specific environmental area.
These items constitute the basic engineering elements of estimating
the potential for accidental or deliberate discharge, and, 1n the event of
process discharge, containment, worker protection, and decontamination.
Note that biological control, i.e., control based upon biological characteristics
Incorporated by design into the genetically engineered microorganism, is
considered as an activity integral to the development of the organism.
While a desirable attribute, biological control is therefore viewed as a
function more appropriate to the microorganism development process and 1s
not considered an environmental engineering pollution control function.
C. Proposed Approach
The engineering assessment protocol modules for release and exposure
for PMN review do not require the detailed data on the properties of the
specific genetically engineered microorganisms as required for the health
-------
-8-
and environmental assessment modules. It 1s recognized that a substantial
data base exists which deals with fermentation production technology,
physical containment methods, decontamination and destruction techniques,
and worker exposure and protection.
This data base includes past experience and practices with (1)
chemical engineering biotechnology based processes (particularly
fermentation, the most likely process to be used); (2) DOD, NASA, DHHS
(particularly RAC), and health care Industry technology for handling
potentially hazardous substances and wastes; (3) Industrial techniques developed
for fault analysis, redundant systems, component reliability testing and
failure analysis, and (4) equipment manufacturers. Further, the insurance
underwriting industry may be a valuable source of Information related to
the frequency of equipment failures resulting in losses and to loss
prevention technology. Based upon the examination of these sources,
coupled with a modest amount of pilot scale research to develop data on
potential release quantities and characteristics for specific operations,
we are confident that the methodology can be developed for: estimating
the nature and quantities of Industrial process release, potential worker
exposure, estimating quantities potentially released to the environment,
evaluating containment and worker protection technology, decontamination
technology, and monitoring strategies for environmental releases.
The project descriptions address the aspects of objectives, research
approach, and project output In such a way as to respond to the OTS needs
stated. Resource estimates are provided in the Program Summary provided for
-------
-9-
each major subdivision. It is estimated that 6 to 8 months will be
required to work out special arrangements with other agencies for cooperation
or participation, setting up Interagency Agreements where required,
preparing scopes-of-work and obtaining OTS reviews. The normal lead time
for acquiring contractor support varies from 6 to 8 months after the work
Is defined and authorized. However, this may be reduced through use of parallel
efforts such as announcement of Intent In the Commerce Business Dally,
etc, and using existing Level-of-Effort contracts where appropriate. In
general, the projects will take about 9-12 months of technical effort, 2
months for special arrangements for acquiring "sensitive" information,
and 3-4 months for preparation and review of the draft final reports.
Therefore, the projects will require about 18 to 24 months from Initiation
to completion of a report draft ready for publication.
-------
-10-
SECTION IV.D.
DEVELOPMENT OF ENGINEERING INFORMATION AND METHODOLOGY FOR RISK ASSESSMENT,
REDUCTION AND MANAGEMENT FOR GENETICALLY ENGINEERED MICROORGANISMS
IN BIOLOGICALLY BASED MANUFACTURING PROCESSES AND DELIBERATE ENVIRONMENTAL
RELEASE
PROGRAM AND RESOURCE SUMMARY
PROGRAM/PROJECT RESOURCES REQUIRED
IV.D
.1 Biologically Based Manufacturing
Processes
Positions
IFTE)
S&E
TW)
Xmural
UK)
Total
(*KJ
.1
Potential for Industrial Process
Release
0.8
48
625
673
.2
Potential for Worker Exposure
0.3
18
200
218
.3
Containment of Process Equipment
Releases
1.0
60
250
310
.4
Identify Monitoring Needs and Stragegies
0.3
18
175
193
.5
Decontamination Technology
1.0
60
350
410
.6
Worker Protective Equipment
0.6
36
200
236
(1) Personal Protective Equipment (0.3)
(2) Personnel Isolation & Decontamination (0.3)
(18)
(18)
(100)
(100)
Totals
4.0
240
1,800
2,040
IV.D
.2 Deliberate Environmental Release
.1
Site Profile Evaluation Procedure
1.0
60
400
460
.2
Site Containment Alternatives
0.5
30
250
280
.3
Monitoring Needs and Strategies
0.5
30
125
155
.4
Site Decontamination Alternatives
0.5
30
200
230
.5
Evaluation oof Applications (Innocula-
tlon) Technologies
0.5
30
200
230
Totals 3.7 224 1575 1799
-------
-11-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.l Accidental and Deliberate Release from Biologically Based
Manufacturing Processes
PROJECT: IV.D.1.1 Potential for Industrial Process Release
OBJECTIVE: Develop the procedures to predict the nature, source, and
amount of accidental and deliberate release of genetically
engineered microorganisms from Industrial manufacturing plants
manufacturing or utilizing such organisms.
APPROACH: The potential for industrial process release of genetically
engineered microorganisms will be evaluated through a multi-task
approach based primarily upon information available from the
literature and experts augmented by limited experimentation as
required to fill 1n gaps.
Task 1: Estimate the source and nature of emissions and
effluents for deliberate process releases which would occur in the
routine operation of a fermentation process. Another type
of deliberate process release occurs when equipment Is opened
for charging of feed materials or maintenance. These latter
Items are to be covered 1n Task 2 as they are intimately
related to the specific manufacturer's equipment design.
This task requires the analysis of the unit operations and
processes and associated types of process and pollution control equip-
ment which are likely to be used 1n the industrial biochemical
engineering processes; the feedstocks, Intermediates, by-products,
and impurities and their chemical and physical properties;
operating conditions; potential discharge points for deliberate
-------
-12-
releases; types of release (hydrosol or aerosol) and expected
composition and properties. This task will be developed on
the basis of available knowledge of the chemical and physical
properties and process operating conditions determined 1n lab
and process development work with genetically engineered
organisms as a first choice. Where such Information 1s
lacking, It will be developed on the basis of conventional
(I.e., non-genetically engineered organisms) biochemical process
Information.
This task has been partially completed. A draft report
has been produced (Industrial Process Profile for Environmental
Use: Industrial Applications of Recombinant DNA Technology -
Jacobs Englneerlng/Technichron, 1983) on the unit operations
and processes, lab operations, and identification of chemicals
used 1n lab RDNA experiments. This draft report 1s the
starting point of a more detailed and explicit Investigation
which will provide information relevant to identification of the
points and nature of deliberate releases of emissions and
effluents, quantities, and expected components, including those
relating to sampling for process monitoring and control and
product quality control.
Task 2: Estimate the potential for accidental or deliberate
release from processes and the characteristics of releases
for non-routine operations and equipment failures. For a
deliberate release, the specific equipment items must be
analyzed to estimate the potential release on the basis of
-------
-13-
specific equipment design. In regard to accidental releases,
for example, such as that occurring from a pump seal leak or
a catastrophic event such as a pressure vessel or safety vent
rupture, the most straight forward way to develop a "real" data
base is to enlist the cooperation of specific existing plants
(not necessarily using genetically engineered organisms) to
furnish access to maintenance records. As a secondary approach,
equipment failure analysis techniques might be applicable to
determining the point of failure. Alternative approaches
would be explored with the appropriate committees of profes-
sional and standards setting organizations.
Task 3: Estimates of the quantity of released material for
deliberate and accidental releases can be derived In a fairly straight-
forward manner through engineering analysis of the process
unit operations and the manufacturer's equipment designs for
both steady-state and start-up/shut-down conditions, Including
raw materials additions and maintenance.
RESOURCES: Task 1, to complete -
Intramural - 0.3 FTE, $ 18K S&E
Extramural -
Total
$150K R&D
0.3 FTE, $168K
Task 2
Intramural - 0.4 FTE, $ 24K S&E
Extramural
Total
$350K R&D
0.4 FTE, $374K
Task 3
Intramural
Extramural
Total
0.1 FTE, $ 6K S&E
$125K R&D
0.1, $131K
Grand Total 0.8 FTE. $673K
-------
-14-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGAM
IV.D.I Accidental and Deliberate Release from Biological Based
Manufacturing Processes
PROJECT: IV.D.1.2 Potential for Worker Exposure
OBJECTIVE: Develop the capability to estimate the nature of worker
exposure from the deliberate and accidental release of genetically
engineered microorganisms from Industrial manufacturing
plants.
APPROACH: The potential for worker exposure within the plant area
can be estimated by an analysis of the plant operational
procedures and related worker activities, eg., data logging,
equipment operation, materials charging, quality control
sampling, filter changing, and maintenance/repair activities.
The analysis will address the frequency, duration, and route
of potential exposure with estimates of the levels of maximum
and average exposure doses (assumes Project 1, Task 3 Is
completed). The analysis will Identify chemical, physical,
and other properties of the materials processed which will be
needed 1n the PMN review process to evaluate the appHcabilty
of specific alternatives.
This project will be coordinated with the National
Institute for Occupational Safety and Health (NIOSH). The
NIOSH (DHSS/USPHS/CDC) will be requested to join In
this research project because of: (1) their mission to develop
criteria for worker exposure as the basis for OSHA standards,
(2) their experience and expertise, and (3) their ability to
-------
-15-
acquire worker and Industry participation through the tripartite
committee approach. Models for the extent of environmental
contamination and general population exposure (transport,
fate, and receptor relationships) will be required to carry
this aspect of the risk analysis further. However, this type
of activity more appropriately 1s handled by OEPER, based on
the OEET plant emission/effluent estimates.
RESOURCES: Intramural - 0.3 FTE, $ 18K S&E
Extramural - $200K R&D
Total - 0.3 FTE, $218K
-------
-16-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.I Accidental and Deliberate Release from Biologically Based
Manufacturing Processes
PROJECT: IV.D.1.3 Technology for Containment (or Engineering Controls) of
Process Equipment Releases
OBJECTIVE: Evaluate the potential performance and costs of alternative
containment technologies for preventing/reducing the amount
of accidentally or deliberately released genetically engineered
microorganisms which would result In worker exposure or entry into
the environment.
APPROACH: These evaluations will be based upon engineering analyses of
biochemical engineering processes, storage and shipment
activities to determine alternative technologies for containment
of deliberate and accidental releases, Including ventilation
air, spills clean-up, and engineering controls for worker
protection. Estimates of the costs of containment methods
and analysis of the Impacts on process operablllty will be
prepared to aid those responsible for risk estimation and
risk reduction cost-benefit analysis 1n the Identification of
economically viable alternatives. A broad range of technologies
will be evaluated Including conventional chemical engineering
and pollution control techniques and also those techniques
employed in applications requiring extremely stringent controls
and safety measures, specifically the military CBW (chemical
and biological warfare) and other high-hazard biological activities.
Selection of equipment based on optimization of choice to
-------
-17-
mlnimlze the accidental and deliberate release and opportunities
for process modifications are additional alternatives which
will be addressed. For example, pumps driven through magnetic
Impellers would eliminate the problem of leakage through the
shaft seal or, alternatively, positive shaft seals or microfilters
can be employed. The analysis will be used to Identify
chemical, physical, and other properties of the material
processed as required 1n the PMN reviews to evaluate the
Integrity of materials of construction and effectiveness of
the containment systems. We would Invite NIOSH participation
in the area of engineering controls for worker protection.
The NIOSH engineering control group Is located in Cincinnati,
is currently working In this area, and has joined us 1n sponsoring
other studies (e.g. Industrial Process Profiles for Environmental
Use: RDNA; EPA Draft Report, 1982).
RESOURCES: Intramural
Extramural
Total
1.0 FTE, $ 60K S&E
S250K R&D
1.0 FTE, $31 OK
-------
-18-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.I. Accidental and Deliberate Release from Biologically Based
Manufacturing Processes
PROJECT: IV.D.1.4 Identify Monitoring Needs and Strategies
OBJECTIVE: Identify those plant, process, and worker aspects which represent
a known (deliberate discharge) or potential (accldential discharge)
Interface with the environment through which a pollutant of concern
might escape Into the general environment.
APPROACH: Operational analysis of the plant and worker activity patterns
as established 1n those tasks addressing (1) potential release
and worker exposure and (2) containment and decontamination
will be used to predict environmental Interfaces of concern.
A monitoring strategy for detection of the pollutants of
concern, or a suitable surrogate which can be used as an
Indicator of the potential for discharge, will be developed.
A strategy for employing simulants to test the system periodically
to ensure Integrity will be Included. This strategy can be
combined with the measurement methodologies developed by
OMSQA to serve as a guideline for PMN reviewers. The OMSQA and
other knowledgeable groups will be requested to provide for Peer
review of the strategies formulated.
RESOURCES: Intramural - 0.3 FTE, $ 18K S«E
Extramural - , $175K R&D
Total - 0.3 FTE, $193K
-------
-19-
IV.D.I
PROJECT:
OBJECTIVE:
APPROACH:
RESOURCES:
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
Accidental and Deliberate Release from Biologically Based
Manufacturing Processes
IV.D.1.5 Decontamination Technology
Evaluate alternative decontamination technologies which could
be used 1n the destruction of genetically engineered
microorganisms and hazardous/toxic chemical by-products from the
production/use process for probable effectiveness and environmental
acceptability.
Alternative technologies for decontamination (destruction of
genetically engineered microorganisms and chemical detoxification,
where required) of the substances recovered from various
containment alternatives after deliberate or accidental process
releases will be evaluated for effectiveness, estimated costs
for Installation and operation, secondary pollution and environ-
mental acceptability. Methods developed for use in the
conventional biochemical processing Industry, NASA, medical
disinfection, and DOD for biological decontamination will be
evaluated In conjunction with environmental pollution control
requirements, particularly hazardous waste disposal requirements
and predisposal treatment options. Some alternative technologies
Include: thermal treatment (e.g., Incineration, autoclaving),
chemical (acid/base), bacterial degradation (aerobic/anaerobic
destruction), various types of radiation, ozone, as well as
others.
Intramural
Extramural
Total
1.0 FTE, $60K, S&E
350K, R&D
1.0 FTE, $41OK
-------
-20-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IY.D.l Accidental and Deliberate Release from Biologically Based
Manufacturing Processes
PROJECT: IV.D.1.6 Worker Personal Protective Equipment
OBJECTIVE: Evaluate the probable effectiveness of available worker personal
protective equipment to provide adequate protection 1n the
event of an accidental or deliberate release of genetically
engineered microorganisms, Including accompanying chemical sub-
stances and contaminants, and blotlc or abiotic factors Involved
1n alternative decontamination procedures.
APPROACH: This project will combine Information from a number of sources
to evaluate the adequacy of worker personal protective equipment
and technology for resistance to chemical and biological agents
under expected contaminant release conditions. Protection against
chemicals, particularly new chemical substances subject to PMN
review are being examined under an ongoing project for OTS. Also, the
NIOSH, EPA, and Industry have conducted studies and developed
standards which will serve as a basis for chemical exposure
evaluation. The NASA, DOD, USPHS, and industry technology
for protection against pathogenic contamination will serve as
a valuable basis for biological exposure protection evaluation.
It will be necessary to evaluate the technology from the aspect
of protection from both chemical and biological agents under
the conditions expected to prevail in various Industrial exposure
scenarios. In addition, alternative management techniques for
containment and decontamination in the event of exposure will be
-------
-21-
aspects and limitations of the alternatives based on limitations
Imposed by chemical or biological agent properties. These will
serve to determine what chemical, physical, and other properties
of the genetically engineered microorganism and attendant substances
must be furnished for the PMN reviewer.
RESOURCES: Task 1 - Personal Protective Equipment
Intramural - 0.3 FTE, $18K, S&E
Extramural
Total
TOOK, R&D
- 0.3 FTE, $118K
Task 2 - Personnel Isolation & Decontamination
Intramural - 0.3 FTE, $ 18K S&E
Extramural
Total
$100K R&D
0.3 FTE, $118K
Grand Total - 0.6 FTE, $236K
plus NI0SH contribution
-------
-22-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.2 Deliberate Release Into the Environment
PROJECT: IV.D.2.1 Site Profile Evaluation Procedure
OBJECTIVE: Identify any characteristics of the site that would Influence
proper containment and decontamination (1f required) of the
genetically engineered microorganisms, Its genome, or the products
of Its metabolic processes.
APPROACH: In an approach similar to that employed 1n the preparation of an
Environmental Impact Statement, an evaluation methodology and
checklist of items to be considered In an evaluation of the
potential for escape of the genetically engineered microorganisms,
genome, or by-products from the field site will be developed.
The methodology will address the potential for off-site migration
1n the various media during application, biological activity period,
and decontamination activities through analysis of the site's envir-
onmental characteristics and Identification of the various routes of
potential migration (surface runoff, access to ground water, soil
porosity and permeability rate, wind velocities, wildlife vectors,
etc.) and specify specific migration routes of environmental concern
which would require mitigation. In the event that a breach in the
containment security system should occur which could permit an
accidental or deliberate release from the site, remedial
measures will be required. The Issues of rapid deployment
-------
-23-
of emergency containment and decontamination techniques (as
Identified under tasks IV.D.2.2 and .4) will be addressed.
RESOURCES: Intramural - 1.0 FTE, $ 60K S&E
Extramural - , $400K R&D
Total - 1.0 FTE, $460K
-------
-24-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.2 Deliberate Release into the Environment
PROJECT: IV.D.2.2 Site Containment Alternatives
OBJECTIVES: Determine suitable site containment technologies to prevent
uncontrolled migration of the genetically engineered microorganisms
from the treated site boundaries after application.
APPROACH: Based on the migration routes and environmental concerns which
are identified 1n project IV.D.2.1, alternative technologies and
operational procedures will be Identified and evaluated for
potential cost and effectiveness for preventing migration during
the genetically engineered microorganism activity and decontamination
stages. The susceptabil1ty of the various engineering materials
used for containment to environmental degradation, especially 1n
regard to the genetically engineered microorganisms, decontamination
agents, and native biota will be addressed also. Technology
developed in the DoD CBW field test programs, the Superfund
Remedial Action program, and the Emergency Response Program will
be primary Information sources.
RESOURCES:
Intramural - 0.5 FTE, $ 30K S4E
Extramural - , $250K R&D
Total - 0.5 FTE, $280K
-------
-25-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.0.2
Deliberate Release Into the Environment
PROJECT:
IV.D.2.3 Monitoring Needs and Strategies
OBJECTIVE: Develop requirements for monitoring to ensure site containment
Integrity, or, 1n the event of an uncontrolled release, trigger
contingency plans.
APPROACH: This project will Identify requirements for the development and
employment of monitoring procedures to ensure that the site 1s
secure 1n regard to pollutant migration via worker, environmental,
or wildlife transport. The requirements will be Identified
through the analysis of the outputs from Tasks IV.D.2.1, IV.D.2.2
and IV.D.4.1 and existing environmental monitoring technology. A
generic protocol will be developed that addresses Informational needs
to be satisfied, Including those related to environmental monitoring,
1n conjunction with the OMSQA. The OMSQA and other knowledgeable
groups will be requested to provide for peer reviews of the strate-
gies formulated.
RESOURCES:
Intramural - 0.5 FTE, $ 30K S&E
Extramural - , $125K R&D
Total - 0.5 FTE, $155K
-------
-26-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.2 Deliberate Release Into the Environment
PROJECT: IV.D.2.4 Alternative Site Decontamination Techniques and
Procedures
OBJECTIVE: Determine suitability of alternative techniques and applications
procedures for decontamination of the site.
APPROACH: Identify and evaluate suitable techniques for decontamination of
the field site upon completion of the genetically engineered micro-
organism activity stage. This task will draw upon the output
of Task IV.D.1.5, Decontamination Technology, and, 1n addition,
explore techniques which are specifically designed for field
applications (e.g., time release blocldes). Procedures for
application to specific scenarios based on various environmental
conditions of concern Identified In Task IV.D.2.1 will be evaluated
for effectiveness in decontamination, prevention of secondary
pollution, safety, and cost.
Prior to conducting a field application. It Is Important
to determine what abiotic and blotlc factors are effective in
controlling or Inactivating genetically engineered micro-
organisms that have been found to pose a potential medical
or ecological threat. Abiotic factors would include:
chemical disinfectants such as chlorine and other halogens,
chlorine dioxide, ozone, other oxidizing agents; physical
disinfectants such as ultraviolet light, electron beam
radiation, ultrasonics, etc.; field conditions, such as pH,
-------
-27-
temperature, moisture content, etc.; barriers. B1ot1c factors
Include effect of predators, parasites, antagonists and competitors.
It 1s also Important to assess the effects of nutrient availability
and concentration on the survival and perslstance of the genetically
engineered microorganisms In the proposed ecosystem.
The choice of the decontamination procedure depends on the type
of environment In which the field test will be conducted. For
example, for a closed environment, such as a wastewater treatment
plant that has been seeded with the genetically engineered
microorganisms, the chemical and physical disinfectants may be
the appropriate and likely choice. In an open environment,
however, such as a hazardous waste site, decontamination 1s more
difficult. Reliance on field conditions, barriers, blotlc
factors, nutrient availability, and time In such an open environment
may furnish an effective method of preventing, controlling and/or
the multiplication and persistence of the genetically engineered
microorganisms. The potential products, Including decomposition
products, resulting from decontamination require evaluation to
ensure that hazardous wastes either do not result or are properly
managed. Techniques which could be rapidly deployed In the
event of a release will be Identified. Also a 11st of Information
needs regarding specific properties of the gentically engineered
microorganism and the formulation 1n which it is applied will be
generated to determine physical, chemical, and biological property
data required In the PMN application for the reviewer (as necessary).
RESOURCES: Intramural - 0.5 FTE, $ 30K S&E
Extramural - , $200K R&D
Total - 0.5 FTE, $230K
-------
-28-
BIOENGINEERING RESEARCH SUPPORT PROJECT
for the
OTS BIOTECHNOLOGY PROGRAM
IV.D.2
Deliberate Release Into the Environment
PROJECT: IV.D.2.5 Application Technology for Genetically Engineered
Microorganisms
OBJECTIVE: Evaluate the use of alternative equipment and techniques to
Identify application procedures and conditions for Introduction
(or Innoculatlon) Into the environmental site that will be safe
for the operator and would prohibit escape of the genetically
engineered microorganisms from the site during application.
APPROACH: Identify various application techniques and equipment and then
evaluate for safety. Examples of approaches to be considered
are: (1) 1n spray application, the relation of droplet size and
wind velocity are Important to prevent wind drift; (2) the
suitability of encapsulated genetically engineered microorganisms
or genetically engineered microorganisms Incorporated Into an
Immobile substrate which could be deposited on the ground surface,
subsequently releasing the bloactlve substance upon exposure to
moisture.
RESOURCES:
Intramural - 0.5 FTE, $ 30K S&E
Extramural - , $200K R&D
Total - 0.5 FTE, $230K
-------
D
-------
Monitoring Techniques for Genetically Engineered Microorganisms
by
David Glaser*
Tim Keith
Peg Riley
Geoff Chambers
John Manning
Susan Hattingh
Ralph Evans
~Harvard University
Museum of Comparative Zoology
Cambridge, Massachusetts
May 30, 1985
-------
TABLE OF CONTENTS
I. Introduction 1
II. Sampling considerations 4
A. Introduction
B. Qualitative sampling
C. Desorptlon from sediments
D. Enrichment
E. Partitioning in the environment
F. Issues In sampling methods
III. Monitoring techniques 10
A. Conventional microbiological techniques
B. Immunological techniques
1. Standard methods of antibody production
2. Monoclonal methods of antibody production
C. Use of genetic markers
D. Molecular techniques
1. Restriction enzyme mapping
2. DNA probes
a. Restriction fragment hybridization
b. Colony hybridization
3. DNA/DNA hybridization
4. Genomic sequencing
IV. Microcosm tests for monitoring techniques 25
A. Microcosm construction
B. Microcosm methodology
C. Sample protocol
D. Containment
E. Points to consider
V. Quality assurance 31
A. Introduction
B. Testing for sensitivity and specificity
C. Testing for linkage between markers and rDNA
D. Summary
VI. Conclusions 34
A. Monitoring techniques
B. Scenarios for protocol development
C. Research needs
VII. Literature cited
-------
1
I. INTRODUCTION
The first field tests of genetically engineered microorganisms (GEM's) are
being scheduled for spring, 1985 (Bioscience News Update 1985)* Vltb the
approaching release of GEXs Into the environment, monitoring techniques will
have to be established to follow 1) the environmental fate of the GBi, 2) the
continued presence or absence of the recombinant DNA (rDNA) in the GB1, 3) the
continued functioning of the rDNA (that is, continued gene expression), and 4)
the transfer of the rDNA to new hosts. This information is required before the
appropriate regulatory committee can establish the potential risks resulting
from widespread release of the GB4.
It is the purpose of this document to discuss the design of monitoring
protocols and to suggest research needs for their development. Classical
microbiological and modern molecular techniques that will permit analysis of GEM
and rDNA fates in the environment are available. Standard microbiological
techniques will allow monitoring of GEM's and active rDNA, if there is a
suitable assay for the gene product, for example an enzyme assay. However,
these techniques will provide little information concerning the rDNA sequences
specifically. Molecular techniques to assay the DNA Itself will provide
information about the presence of the rDNA, changes in the rDNA sequence and Its
location in the genome, the functioning of the rDNA gene, and its transfer to
new hosts. Both molecular and microbiological techniques will be described in
the following sections with respect to their potential value, efficiency,
sensitivity, and relative cost.
A section on microcosm experiments has been added to this document to tie
together the various monitoring procedures. Typical host-vector systems can be
analysed in microcosm studies to test the accuracy of the proposed monitoring
protocols. These systems can then be used as standards which the industries can
-------
2
follow when designing microcosm and field tests for their specific host-vector
systems.
The specific monitoring techniques that will be most effective will depend,
to a large degree, on the specific organism being monitored. Therefore, it Is
useful to briefly discuss the types of recombinant organisms which have been
proposed for experimental release and commercialization. The different types of
recombinant organisms are:
1) organisms with normal genes deleted; an example of this Is Pseudomonas
avrlngae and Erwlnia herblcola. The latter, produced by mutagenesis, has
already been field tested by Mlcrolife Technics, Inc. These bacteria lack a
functional cell-surface protein that in wild-type bacteria catalyzes Ice-crystal
formation (Genewatch 1983).
2) organisms with new genes inserted using plasmid vectors, including:
a) plasmld-medlated gene transfer in bacteria; an example of this type
of GEM Is the strain of Pseudomonas cepacia that grows on 2,4,5-T as sole carbon
source. This was created by plasmld-asslsted molecular breeding, that Is, by
combining several strains of bacteria with different plasmlds coding for
degradatlve pathways for various chlorinated hydrocarbons and for resistance to
various antibiotics, along with several strains from waste dumping sites. These
were grown together in a continuous culture with 2,4,5-T and other plasmid
substrates. After eight to ten months, organisms capable of using 2,4,5-T as
sole carbon source were Isolated (Kellogg aL* 1981).
b) plasmld-medlated gene transfer to organisms other than bacteria; an
example of this type of GEM is the bacterial genus Agrobacterlum which can
transfer some of its plasmid DNA to infected dicot plant cells. There are two
species of Aarobacterlum of industrial interest, A.. tumefaclens which carries
the tumor inducing (Ti) plasmid and A.« rhlzocenes which carries the the root
-------
3
Inducing (Pi) plasmid. Portions of both Ti and R1 plasalds can become stably
integrated into dicot plant chromosomal DNA. These plasmlds have been modified
so their disease-causing genes have been activated thereby rendering them very
useful vectors. There have been several successful introductions and
expressions of bacterial genes in plants (Chilton ai.« 1983, Herrera-Estrella
fii. fli- 1983, Schell et al. 1983). In one case the cells have regenerated whole
plants in vhlch the foreign genes were still expressed (Fraley fit al« 1983).
Kemp (1983) reported the first transfer and expression of a plant gene from one
plant species to another; the seed storage protein from beans was successfully
introduced into both sunflower and tobacco cells.
3) transposition-mediated gene transfer; transposable elements are DNA
sequences that are able to replicate and Insert copies of themselves at new
locations in the genome. Experiments using transposable elements have been done
with corn cells (Peacock, 1983).
The following sections will detail currently available monitoring
approaches. It should be emphasized that there Is no one approach to monitoring
that will be suitable for all host-vector systems. Protocols will have to be
developed that are suitable for following the fate of each GEM produced.
-------
4
II. SAMPLING CONSIDERATIONS
A. Introduction
Developing an optimal sampling strategy is not trivial, because it depends
on the organism, the gene product of Interest and the medium from which it is to
be sampled. Thus, monitoring strategies must be tailored to each specific
case. However, there are considerations common to all monitoring problems.
Certain scientific issues can be explored with test organisms to provide a basis
for conducting specific monitoring assessments most effectively; these are
discussed below.
B. Qualitative sampling
Two types of trade-offs exist in sampling strategy: (1) qualitative
quantitative, and (2) extensive j£&. intensive sampling. (1) The term
"qualitative" is used to describe sampling for presence/absence only.
"Quantitative" sampling refers to attempts to enumerate organisms, plasmlds or
DNA sequences. (2) The phrase "sampling extensively" refers to sampling a
large geographic area, while "sampling intensively" is used to refer to
searching only a small area, but attempting to cover as many different habitats
as possible, using closely-spaced samples.
Experience demonstrates that it is possible to enrich for most microbial
activities in most environments. In large measure this is attributed to the
fact that microbes disperse well and can last a long time under poor conditions.
Furthermore, under favorable conditions, few organisms can quickly give rise to
a large population. Beljerinck summarized this in the principle of microbial
ubiquity: "Everything is everywhere; the environment selects." (Atlas and
Bartha 1981, p. 5). Therefore, for the prediction of potential impact, it is of
-------
5
primary importance to find out where the GEM and rDNA sequence are transported
and to determine how long they can last after their purpose is fulfilled, that
is, to sample qualitatively and extensively. This is in contrast to the needs
of the producers and users of GEM's, to whom it is very important to know the
density and activity level of the G01 in the target field, that is to sample
quantitatively and intensively.
This suggests that research should concentrate on methods that are good at
very low densities to allow qualitative and extensive assays for the presence of
small populations.
C. Desprptlgn £csa sediments
To sample from sediments and soils one must deal with the problem that the
organisms are attached to surfaces. A sample of the community can be
transferred to fluid phase simply by placing a bit of soil in fluid, shaking
vigorously, and then spreading a sample of the fluid on an agar medium for plate
counts. However, organisms of interest may not dissassociate readily from
particles. Thus, it is necessary to maximize the ability to desorb cells from
surfaces. There has been much work on methods of desorption (see, for example,
Marshall 1976). The first research need is for compilation of existing methods.
Next, experimental research into whether desorption actually increases plate
counts is needed. If it is useful, it will then be necessary to tailor methods
to the types of organisms likely to be released.
D. Enrichment
An alternative to methods for plate counting low-density populations nay
turn out to be enrichment. Standard enrichment techniques aim to select for the
organism of interest. Por some GEM's, however, this requirement need not be so
-------
6
strict. For example, if an organism produces an unusual organic molecule,
enriching for the organism may increase the concentration of the product, and
this might occur independently of any Increases in other populations. Enrichment
methods can be quantified vlth Most Probable Number (MPN) methods (Russek and
Colwell 1983), although large samples are needed to achieve reasonable
precision.
Enrichment is sometimes more effective as a monitoring technique than plate
counting. For example, Ellbane fii. al. (1983) added a strain of bacteria that
could grow on 2,4,5-T as sole carbon source to a 2,iJ,5-T-contaminated soil and
watched the strain decline In numbers as the substance was catabolized. They
could not detect any organisms on plates after 6 weeks. At 12 weeks, 2,4,5-T
was again added to the soil, and by 15 weeks the 2,4,5-T-degradlng strain became
detectable again and increased dramatically in density. That is, the potential
for 2,4,5-T degradation and population growth remained, even though plate counts
were zero.
The problem of desorptlon of cells from particles in aquatic sediments and
soils is probably not as great using broth enrichment techniques, because the
sediment itself can be placed into the medium. Thus, two strategies present
themselves for sampling from solid phases: (1) dlssassociation of cells from
particles by physical or chemical means, followed by selective plating; and (2)
replicated enrichments, followed by replicated isolations from each enrichment
flask. Organisms Isolated in this manner can then be studied as necessary using
molecular techniques.
A great advantage of the second method is that enrichment is likely to pick
up OEM's at lower densities than Is possible with direct plate counts. This
suggests that the high densities present immediately following inoculation might
be sampled with plate oounts. If numbers then decline to levels undetectable
-------
7
with plate counts, broth enrichments can be used, quantified vltb MPN methods.
Molecular analyses can be performed on both types of samples.
E. Partitioning In the environment
One objective of low-density sampling Is to determine how added
microorganisms partition themselves In the environment. When an organism Is
Introduced to the environment, It tends to survive better in certain habitats or
mlcrohabltats than in others. For example, such habitats in a soil might
include leaf surface, rhlzosphere, the top centimeter of soil, anerobic
microsites, Invertebrate guts and external surfaces of Invertebrates.
Pseudomonaa fluorescens and £.. cepacia, strains of which have been engineered
for release Into the environment (see section VI), can act both as animal and
plant pathogens (Bergan 1981). Thus, sampling should Include potential local
animal hosts.
In non-target locations only small numbers of GBi's will probably exist.
This suggests that it may be worthwhile to sample initially very intensively and
frequently from the target field and a small adjacent area, Just following
release of the organism. For example, if it is found that Paeudomonas
fluorescens with a gene for a Bacillus thurlngenals endotoxin tends to survive
best in the top centimeter of grass rhlzosphere, then sampling effort in other
locales should ooncentrate (not exclusively) on the top oentimeter of grass
rhlzosphere, because this is most likely to give positive results at low
densities. The initial intensive sampling should seek to answer the following
questions: Does the organism tend to move towards or survive better in oertaln
types of habitats? What 1b the rate of Increase in numbers in adjacent areas?
The answers to these questions can be used to suggest where sampling effort
might be most effective in non-target fields with very low densities. Research
-------
B
night be directed towards studying the usefulness of this type of protocol in
test organisms.
P. Issues in sampling methods
Sampling methods have been studied for a long time, and many methods exist
and are veil-documented (for example see Atlas and Bartha 1981, Lynch and Poole
1979» Sieburth 1979). Effective sampling usually requires concentrating the
organisms, for example by filtration through an 0.2 micron filter. Conforms
are routinely sampled by filtering a standard quantity of fluid through a
filter, and then placing the filter directly onto M-Endo agar, which selects for
Gram negative species and turns metallic green when a rapid lactose fermenter
such as a coliform is present.
Vith respect to following the movement and survival of DNA sequences,
¦habitats" Include other organisms, potentially both animals and plants. Thus,
an important question is: How extensive should the search be for determining
the transfer of a certain DNA sequence? It is possible for genes to be
transfered from bacteria to plants, for example from AgrnhanteMum tumefaclens
to many species of plants (Atlas and Bartha 1981). In addition, it is possible
for plant pathogenic species to also be animal pathogens, for example
Pseudomonaa fluoresens and cepacia (Bergan 1981). A useful rule of thumb
would be to tailor each protocol to each individual case. These are Issues that
are not amenable to quantitative cost-benefit analyses. The cost of such
sampling must be balanced against the probability of such transfer occurring and
the possibility and likely magnitude of adverse effects.
It may be worthwhile to keep the following in mind: (1) An Important
problem in quantitative sampling is deciding on the size of the sample. A good
tool for designing an economical but sufficient strategy is the performance
-------
9
curve, which is a graph of density against sampling effort (for example numbers
of samples). This curve might rise and fall as sample size rises and then level
off at some larger sample size. To determine the best compromise between
expense and accuracy, one should perform a greenhouse test, taking a large
number of samples. As each sample is taken in sequence, the estimated density
is plotted against the number of samples taken so far. The optimum sample size
is the minimum number of samples required to reach the flat part of the curve.
(2) For some questions, relative numbers are Just as useful as absolute
densities and may be easier to measure with a desired level of precision. For
example, the rate of die-off of an added population oan be estimated with
relative densities.
-------
10
III. MONITORING TECHNIQUES
A. Conventional microbiological techniques
Conventional microbiology offers many methods for the Isolation of
microoorganisms from mixed culture samples. Many of these methods are for the
characterization of the total microbial population, and as such are of little
use in the recovery of a specific genetically engineered organism that has been
released to the environment. This section of the report vlll examine selective
isolation of organisms on solid media and in liquid culture.
Solid media techniques involve the immobilization of the desired liquid
enrichment media with agar. Mixed cultures are appropriately diluted in sterile
media, and then "streaked" onto the agar plate containing the Indicated media.
The individual bacteria grow into colonies on the plate, which allows the
analyst to pick the colony from the plate for further purification and
identification. This technique is used by aquatic and soil microbiologists to
identify species capable of growing on the chosen medium. Fermentation and
Industrial microbiologists use this technique to identify microorganisms capable
of growing on the medium or producing a valuable byproduct from it.
The species recovered are dictated by the medium chosen, that Is, its
components and their concentrations. Therefore, the nutritional requirements
and unique metabolic functions of the GEM should be well characterized.
To recover a GO! selectively on an agar medium, one can insert DNA
sequences that code for unusual metabolic traits, such as: (1) ability to grow
on an unusual compound; (2) resistance to a heavy metal, antibiotic or other
compound; and (3) production of a specific compound that can be assayed with
color changes in the medium. Iliese markers are considered in section IIIC.
As discussed in section II, liquid broth enrichment techniques are probably
-------
11
sore effective at recovering GEM's at low population densities and therefore are
likely to be of great importance. These can be quantified with most probable
number statistics (Russek and Colvell 1983), although large sample sizes are
needed for reasonable precision.
B. IgravnglPKlcal techniques
1. Standard Methods of Antibody Production
Fluorescent antibody techniques have been used for many years in medical
microbiology and pathology to examine tissue samples for infectious agents.
Recently, they have been applied in mixed culture environments to explore
ecological questions. The uses and limitations of Immunofluorescence techniques
have been discussed by Bohlool and Schmidt (1980).
To develop a fluorescent antibody a pure culture of organisms (the
antigens) are injected into a rabbit that then develops antibodies to the
organism. The animal is periodically bled to isolate antiserum to the injected
microorganism. After purification, the antibody is attached to a fluorescent
dye, commonly fluorescein lsothlocyanate (FITC). The labeled antiserum is then
added to a sample from the environment, which might be a small1 quantity of
soil, or a filter through which water has been passed. The antibody binds to
the antigenic microorganism. In theory, when the preparation is viewed under
epl fluorescence microscopy, only the antigenic organism with the attached
antibody can be seen.
Imunofluorescence techniques have been used to follow strains of Rhlzoblum
BMQiMf AfiPtgfeaCter. Bellerlnckla. Agosolrlllua. nltrlflers, sulfur- and
iron-oxidlzers, and various other types of organisms. Attempts have also been
made to look at fungi (Bohlool and Schmidt 1980).
However, there are several problems with the techniques: (1) Bohlool and
-------
Schmidt (1980) calculated that minimum countable densities in soils are about
6 7
10 to 10 organisms/g soil. This is far too Insensitive for studying low
population sizes of GEM's. The severity of the problem can be reduced by
increasing the efficiency by which cells are desorbed from the solid phase.
Therefore, research is needed on desorption techniques, as mentioned in section
11 for increasing the efficiency of plate counts.
In addition, the efficiency of recovery can be estimated in experiments in
sterile soils. This allows calibration against simple non-selective methods,
such as plate counts and epifluorescence microscopy with cells stained with dyes
such as acridlne orange (Bobble al. 1977) and 4,6-diamindino-2-phenyllndole
(DAPI; Porter and Feig 1980).
(2) The GEM cells may not be the only fluorescing particles in the
environment. There are several reasons for this: (2a) Inorganic particles
sometimes bind nonspeciflcally to the fluorescent antibody complex allowing soil
particles to interfere. (2b) Material in the environment can autofluoresce.
(2c) Other species may have surface proteins similar to those of the organism
of interest. There are also limitations in the method of production of
antibodies. The organisms injected into the rabbit may contain low levels of
contaminant proteins. The isolation procedure may not be perfect, resulting in
a mixture of different antibodies being added to the environmental sample.
Specificity can be Increased significantly by using monoclonally produced
antibodies, which are discussed in the next section. There is an Important need
for research into the specificity of fluorescent antibody techniques in mixed
oultures containing closely related species.
(3) On the other hand, GB4 cells may not fluoresce. (3a) Organic slimes
may prevent the immune reaction between the bacteria and the fluorescent
antibodies. For example, while attempting to determine the number of nitrifying
-------
13
bacteria in a fixed-bed wastewater treatment reactor, Szwerinakl fit. al.. (1985)
found large numbers of fluorescent cells In enrichment cultures tak$n from the
reactor. However, very few cells were seen in direct observations of
environmental samples. MPN counts of enrichment cultures gave such higher
counts than direct observations. This is likely to be a problem in any
environment with organic slimes, most importantly wastewater treatment
facilities. (3b) The antigen can become unstable under some growth conditions.
Bohlool and Schmidt (1980) site several examples suggesting that this may not be
a serious problem. However, stability should be tested for each antibody that
is used.
(ij) Finally, immunofluorescence may not tell us what we want to know.
(4a) Live and dead cells cannot be distinguished. This might be seen as an
advantage. Very low populations of live cells might be missed by these
techniques; having dead cells present increases the probability of finding small
populations, (lb) If the antigen is not a product of the rDNA, then this
technique will not be able to tell if the original antigenic oells have lost the
rDNA or if other strains have picked it up. Using two antibodies in the same
GEM, one to a normal cell protein and one to an rDNA product, can potentially
give information on rDNA transfer and loss.
In summary, fluorescent antibodies can potentially be extremely useful for
monitoring GBi's in the field. However, several problems exist in sensitivity
and specificity. As discussed in the next section, monoclonal antibodies can
potentially alleviate much of the specificity problem, as well as reducing
production oosts considerably.
-------
2. Monoclonal Methods of Antibody Production
Monoclonal antibodies (MAbs) can provide a very reliable and sensitive
method for identifying and monitoring the gene products of GEM's. MAbs are
prepared by injecting a mouse with a purified antigen, removing its spleen after
several weeks, isolating the antibody producing cells (the B lymphocytes), and
fusing them with mouse myeloma (tumor) oells. The products of this fusion are
grown in a selective medium which allows only the hybrid cells (hybrldomas) to
grow. These cells are able to grow permanently in cell culture due to the tumor
oells and produce a specific antibody due to the contribution of the B
lymphocytes. Hybrldoma clones that express the particular antibody of interest
can then be grown either in vitro or in the abdominal cavity fluid of mice for
the production of large quantities of monoclonal antibodies.
This method allows the production of large quantities of specific
antibodies against given antigens. Unlike conventional immunological methods,
antibodies produced by the monoclonal techniques are homogeneous and are
therefore more reliable (fewer false positives) in detecting a specific antigen.
However, as with conventional Immunological methods, disadvantages include the
inability to tell live from dead oells, autofluoresence of environmental
material and potential Interference from sources such as bacterial slimes in the
environment.
MAbs can therefore be extremely useful in detecting the products of genes
Introduced or altered by rDNA technoloy. The gene product of the G01 is
isolated, purified and used as the antigen to elicit an immune response and the
production of antibodies against it. Providing that the gene product is a
strong enough antigen to provoke an immune response, the use of MAbs will allow
a sensitive method for detecting its presenoe.
-------
15
Currently the Damon Biotech Corporation is developing a new technique for
the production of large scale amounts of MAbs by the technique of
microencapsulation. The method uses a porous carbohydrate capsule to
surround the hybridoma cells thereby retaining the antibodies produced while
allowing the circulation of nutrients and metabolic wastes. When the
encapsulated colonies are harvested after several days, the growth medium,
according to Damon, contains up to 40-509 by weight of MAbs. This new
production method should greatly reduce the cost of MAbs and make the use of
MAbs an economic way to detect the presence of a particular gene product.
The use of monoclonal antibodies should be considered a very sensitive,
reliable, and cost efficient means to Identify and monitor the gene products of
GDI's providing that antibodies to the gene product can be elicited through an
immune response.
C. The use of genetic markers
The use of marker genes may prove to be a very sensitive and inexpensive
way to follow GEW's upon release into the environment. There are several types
of marker genes available to genetic engineers:
(1) chromogenic markers; these are genes that produce a precursor to a
biochemical pigment. Using appropriate media one can produce pigmented
microorganisms. This confers a scoreable phenotype on the GEM (Masui al«
1984).
(2) resistance markers; these are genes that provide some type of
resistance to the microorganism, eg. antibiotic or heavy metal resistance. These
markers confer a selectable phenotype to the GEM; such genes may be favored in
some natural environments and their spread could have unexpected consequences.
-------
16
(3) rare sugar or other rare carbon sources; these are genes that confer
the ability to use unusual sugars or other carbon sources. One example would be
lactose utilization. Pew naturally occuring bacteria other than Escherichia coll
can ferment lactose, making lactose fermentation a Barker which would
differentiate a recombinant strain from most naturally occuring strains.
Markers used to label GQI's can be placed on the chromosome or on plasmid
DNA. One would want to link the marker gene and the rDNA gene to limit the
possibility that the marker is separated from the rDNA gene by recombination. A
marker situated on a plasmid will generally tend to be less stable for several
reasons; 1) conjugative plasmids may be transferred infectiously between the GEM
and other species In the environment, 2) if non-conJugative, the plasmid may be
mobilized by mobilizing plasmids that may be acquired by the host strain from
otber species, and 3) if the plasmid contains transposons, these elements may
lead to the transfer of genetic information between species. Some initial
studies have examined plasmid transfer In some commonly used host-vector systems
(Anderson,1975; Sagik and Sorber, 1979; Levy fiJLftl., 1980; Schiff and
Klingmuller, 1983; Levy, 1981).
Marker genes can be constructed with various degrees of specificity. As
mentioned previously, lactose fermentation may be an efficient marker gene to
distinguish between a OEM and naturally occurring bacteria (other than £.. coll).
Lactose fermenting microbes oan be identified very simply by using the coliform
oountlng technique. This Involves plating out samples from nature on M-Endo
agar. This agar contains lactose and dyes to inhibit gram positive bacteria.
The agar turns metallic green when the medium turns strongly acidic, as is the
case when £.. coll or any rapid lactose utilizer is plated on it.
When using any one metabolic characteristic for selective plating, one runs
the risk of recovering naturally oocurrlng organisms with the trait, as well as
-------
naturally occurring organisms to which the rDNA has been transfered from the
GQ1. To Increase specificity, more than one marker can be Inserted Into a GEM,
for example the ability to use an unusual compound and resistance to an unusual
antibiotic. In addition, If the rDNA is on a plasmid, then having one marker on
the plasmid and one on the chromosome allows one to trace separately the fates
of the plasmid and the GO!.
To make a marker gene more specific it could be placed under the control of
some promoter-operator sequence which would allow the gene to be Induced under
conditions unrelated to a need to ferment lactose. One example would be the trp
promoter-operator region, which would allow the Induction of the gene only when
tryptophan is present. A second example would be the phage lambda*s right
operator, which would allow induction only in response to UV radiation or high
temperature if a temperature sensitive repressor were used. Promoter-operator
units such as these can be engineered so as to allow Induction under known
laboratory conditions for monitoring purposes, but to prevent induction under
conditions leading to the selection for the marker in the field.
Genetic engineers have on hand a variety of marker genes that can be made
as specific as desired. These marker systems oan then be very easily tested In
alcrocosm or greenhouse experiments and the efficiency and sensitivity of each
marker system analyzed directly. This technique may prove to be an inexpensive
first step in most monitoring protocols. It would then be necessary to examine
the GEH's isolated in the first step for oontinued presence of the rDNA gene.
d. Hclscul&r techniques
1. Restriction enzyme mapping
The use of restriction enzymes to generate a restriction map of host DNA
oan be a very useful technique for identifying and monitoring recombinant
-------
18
organisms, especially those with genes Introduced by plasmlds. The key to
restriction mapping lies In the ability of certain enzymes found In bacteria to
out double-stranded (ds) DNA at specific sites. Each restriction enzyme
recognizes a specific nucleotide sequence between 4 to 6 base pairs In length.
The enzyme cuts the DNA each place the specific sequence occurs. Different
enzymes recognize and cut different nucleotide sequences. A piece of DNA, cut
with a restriction enzyme, will produce a number of distinct fragments. These
fragments can be separated on the basis of Blze by gel electrophoresis. The
fragments can then be visualized by staining the gel with ethldlum bromide
(which binds to DNA) and photographing It under UV light (which causes the
ethldlum bromide to fluoresce). The size of the fragments can be determined by
running a control on the gel which consists of DNA fragments of known size. By
sequentially cutting the DNA with a series of restriction enzymes one can
construct a map of specific restriction sites (Meyers ££.&!.., 1976).
For use in monitoring, one would first have to construct a restriction map
for the nonrecomblnant and recombinant plasmlds. A comparison of the two maps
would show where the novel gene had been Inserted Into the plasmld DNA. Once the
restriction map of a plasmld with an inserted gene is known it should be fairly
easy to monitor this recombinant plasmld. One would first have to Isolate the
plasmld DNA from the host bacteria sampled In the wild and then use the
appropriate restriction enzymes to analyze It. The pattern of fragments produced
oan be compared to the original maps of both the normal and recombinant plasmids
to verify the continued presence of the rDNA Insert. Inserted and deleted DNA
would be recognized as changes in the size of particular DNA fragments.
The method is sensitive to the detection of Inserted and deleted genes and
is fairly rapid. The restriction enzymes themeselves are expensive but if the
manufacturer is required to do the initial analysis of the non-recomblnant and
-------
19
recombinant plasmlds, only a few enzymes should be necessary for subsequent
plasmld characterization.
2. DNA probes
Classical microbiological techniques will allow sampling of the environment
for novel organisms. It Is then Important to analyze the DNA of these organisms
to verify that the rDNA Is present, functioning, and In Its proper location. In
addition, tests may be required to examine the rDNA for transfer Into new hosts
(or new positions in the original host DNA). These analyses require specific
nucleotide sequence analysis. This oan be accomplished by using DNA probes. When
the rDNA is being characterized and inserted Into the host, a probe (DNA
sequence specific to the rDNA) can be constructed for use in future monitoring
events. A probe is essentially a complementary piece of DNA that will hybridize
(base pair) to the rDNA when both pieces of DNA are mixed together in
single-stranded form. The presence of this hybridization event can then be
visualized in two ways; through the use of radioactive or biotln labelling. The
former requires that the probe sequence Incorporate radioactive nucleotides
32
(usually P) into its sequence. The presence of the probe can then be followed
with the aide of x-ray film. The DNA of interest is attached in single-stranded
form to a filter, the probe DNA (in single-stranded form and with attached
radioactive nucleotides) can then be washed over the filter. A piece of film is
then laid over the filter, and wherever hybridization occurred a spot will show
up where the radioactive nucleotides exposed the film (Manlatls at al.., 1982).
This technique Is expensive due to the cost of radioactive nucleotides and
involves the handling of hazardous material. In addition, the probe is
short-lived (the radioactivity decays fairly rapidly).
Biotln labelling of probes is currently being developed as an inexpensive,
non-hazardous alternative to radioactive labelling. This technique involves the
-------
20
incorporation of biotin into the nucleotide sequence of the probe and takes
advantage of blotin-avidlne binding. First, avidin binds to the biotin. Next, an
en2yme binds to the avidin and oonverts a colorless soluble substrate into an
insoluble pigment. Spots of pigment on the filter correspond to places where
probe and rDNA sequences have hybridized (Levin, 1983; Langer ££.&!.•• 1981).
This technique will have the same specificity as the radlolabelllng.
The exact nature of the probe (length and sequence) will depend on the
host-vector (HV) system being employed. The length of the probe sequence will
have to be calibrated to ensure against both random hybridizations with the host
DNA and specific hybridizations with similar genes. This is a particular problem
when the rDNA is merely an altered gene normally present in the host. It will be
less of a problem if novel genes are being introduced. It may be appropriate to
require the rDNA manufacturer to analyze the probe specificity and document the
percentage of false positives/negatives.
The probe can be used for monitoring in two different experimental
approaches. These approaches are the focus of the remainder of this
section.
a. Restriction fragment hybridization
When characterizing the novel organism with fingerprinting (see previous
section), the DNA fragments from the agarose gels can be transferred to filters
and irreversibly bound. The DNA can be denatured (so that It is single-stranded)
and a probe can be washed over the filter. The probe will base-pair with the
rDNA and when visualized (aee previous section) the size of the exposed fragment
can be compared with the restriction map produced from the original novel
organism. If the rDNA is in the same position on the chromosomal or plasmld DNA
then the band that lights up will correspond in size to the restriction fragment
-------
21
produced when tbe gene vas originally Introduced Into the host. This technique
will provide evidence 1) whether the rDNA Is In tbe original position, 2)
whether the sequence has been altered, and 3) whether It has been deleted. The
technique may not be sensitive to minor sequence changes (point nutations) or
even gross changes (inversions). An additional set of tests must be performed to
ensure that the rDNA is still functional (see monoclonal antibodies section).
Restriction fragment hybridization is an expensive test since it involves not
only the use of restriction enzymes but also the creation and labelling of a
probe(s).
b. Colony hybridization
An alternative to fragment hybridization is the use of colony hybridization
(Grunstein and Hogness, 1975). This procedure involves the plating out of sample
cells (host cells) onto agar plates. The cells are lysed (cell walls broken) and
a filter Is placed over tbe plate so that the DNA from the host cells can be
bound directly from the plate surface. The filter can then be probed in the same
manner described above. Film is placed on tbe filter and spots on the film
oorresond to colonies on tbe original plate that contained DNA which hybridized
to the probe.
This test allows one to scan many cells from a sample to pick out those
with tbe rDNA incorporated. The cells producing spots on the film can then be
further characterized to be sure 1) the cells are the intended host of the rDNA,
2) the rDNA is still in the same position (fingerprinting), 3) the gene is
functioning.
This technique allows a rapid screening of many samples in a relatively
inexpensive fashion. It has recently been shown to be a highly sensitive
technique capable of detecting one colony in 10® oolonles of a nonhomologous DNA
-------
22
background (Sayler .afc. al.. 1985). Vitb appropriate amplifying media even lover
densities may be detected. It is likely to be the best technique to assay
directly for rDNA.
3. DNA-DNA hybridization
The technique of DNA-DNA hybrlzatlon entails the study of the temperature
dependent kinetics of the association of single stranded DNA molecules to form
double stranded structures (duplexes). The technique has been utilised mainly in
systematica studies and is capable of generating highly reliable data. However
It has limited resolving pover sinoe DNA sequences which are similar cannot be
discriminated as being different and sequences which are very divergent will not
reassociate.
It Is unlikely that the technique of DNA-DNA hybridization will be useful
for detecting recombinant DNA molecules In GBls released into the environment.
However, It may provide an estimate of the extent to which a particular sequence
has been altered following the release of some particular organism.
A detailed review of this technique has recently been published (Sibley and
Alquist, 1983). Using this method purified probe DNA from a reference source Is
sheared by sonlcation to a length of 500 base pairs. The probe is then
125
denatured by heating and subsequently labelled with I. Trace amounts of the
labelled probe are mixed with genomic DVA and the temperature at which It forms
duplex hybrids with whole genomic DNA from the organism under study is measured.
The difference in temperature between experiments with homologous and
heterologous combinations of probe and sample can be related to the degree of
sequence mismatch between the heterologous probe/sample pair. The best estimate
for this genetic distance is that a difference of 1 degree C corresponds to If
sequence divergence. The lower detection limit is around 0.5-1.01 divergence
-------
23
and tbe upper level of divergence that can be reliably estimated is around 30f.
The technique of DNA-DNA hybridization may be capable of identifying whether
genes inserted into GEM's are still present. Tbe main disadvantages inherent in
this technique are the need for substantial (>200 ug) samples of pure vhole
genomic DNA, the expense of the radioactive probe, and tbe limited resolving
power. The more direct method of hybridizing DNA probes to colonies or to
restriction digests is more sensitive and efficient.
4. Genomic sequencing
A new method of providing DNA sequence data from specific regions of
chromosomes direct from whole genomic DNA was recently reported (Church and
Gilbert, 1981). This technique offers a rapid and sensitive method for
screening released recombinant clones for DNA sequence stabilty.
The method is based on a combination of chemical sequencing of DNA (Maxam
and Gilbert, 1980) and Southern blotting (Southern, 1975). Vhole genomic DNA is
first obtained from a culture of the organism of interest. The DNA is cleaved
with a selected restriction enzyme which cuts the DNA within 100 bases of the
sequence to be investigated. The resultant mixture of DNA fragments is divided
into allquots which are then subjected to tbe separate chemical modification and
ohain cleavage reactions used in conventional chemical sequencing of DNA. The
products of the individual reactions are separated by electrophoresis and then
transfered to nylon membranes. The DNA can be permanently bonded la. situ by UV
irradiation and tbe bonded blots are stable.
The DNA sequence of interest Is cloned into M13 phage and single stranded
DNA Isolated. Radioactive probes (150 base pairs long) complementary to the
oloned sequences are then prepared and hybridized to the nylon screen blot. DNA
fragments binding the probe are revealed by autoradiography and the DNA sequence
-------
24
read from the photographs obtained.
Thus the presence or absence of an Introduced gene could be monitored by
determing whether any of the genomic DNA bound the radioactive probe prepared
from the originally introduced sequence. If present, the actual DNA sequence
could be compared with that of the probe to determine its fidelity.
This technique is very expensive and very new. It is potentially useful In
fine structure determinations, but less expensive, better characterized methods
exist, for example Sanger sequencing (Sanger si. al.. 1977)* Therefore, it is not
likely to be of use In the foreseeable future for routine monitoring of GEM'S in
the field.
-------
25
IV. MICROCOSM TESTS FOR MONITORING TECHNIQUES
A. HlcrPCQgB construction
A considerable amount of effort Is needed to apply the techniques described
in the previous section to environmental situations. In order to accomplish
this, microcosm experiments are necessary to evaluate the specific monitoring
protocols for test organisms. These must ultimately be tested in the field.
There has been much work on microcosm technology for many different habitat
types, and much of this has been summarized and referenced In a series of books
and symposia (Gillett and Witt 1979* Witt and Glllett 1979* Gelsy 1980,
Pritchard and Bourquin 1984, Kinne 1977, Draggan and Van Voris 1979). In
addition, there exist several large-scale In situ aquatic systems which can be
used as Intermediates between laboratory microcosms and the field (Grlce and
Reeve 1982). In Canada, the experimental lakes region provides opportunities
for perturbing and observing natural systems experimentally (Watson 1980).
It might be argued that in view of the shortage of funds for the problem of
assessing effects of genetically altered organisms, it would be best not to use
these funds for research into the design of new microcosm systems, because many
already exist. However, one important aspect of regulation is standardization,
and no standardized set of microcosms exists for any environment. In addition,
microcosms must form an Integral part of fate and effect studies. Thus, it
would be well-advised to conduct research into microcosm design as an integral
part of an experimental approach to monitoring techniques, fates and effects.
The problem of standardization of microcosm assay systems is a very
difficult one. The use of only one standardized system may be a desirable
approach, but it is not a scientifically credible one because of the limited
variability of any one experimental system compared with nature. This is a
problem because the techniques described above can perform differently under
-------
26
different natural conditions. Because of these considerations, several types of
construction should be used, for example the various designs described in Vitt
and Gillett (1979) and Gillett and Vitt (1979). In addition, geophysical
parameters such as soil Mater content and light and temperature regimes, should
be varied.
Eventually, a monitoring protocol may Include a standard series of
microcosms of varying construction, run under a set of standard conditions.
However, even this amount of standardization may be impossible, because the
environments into which GEM's are likely to be released vary greatly both
geographically and ecologically. Thus, appropriate microcosm design might have
to be decided for each case.
Research into microcosm design will undoubtedly be useful for other issues
in environmental protection as well, for example fates and effects of toxic
chemicals (Kimball and Levin 1985). Such efforts are likely to involve
long-term projects, because the design of microcosm systems for measuring
important ecological variables requires knowing which variables are important
and how to interpret changes in them. This requires much fundamental knowledge
of ecological processes which we do not now have. Thus, support for basic
studies of microcosm construction should be long-term and should Involve
multi-disciplinary teams so that physical, chemical and biological interactions
can be explored. Because of its relevance to other ecological problems, it may
be appropriate to support such research jointly with other programs in
environmental protection.
B. Microcosm methodology
There are two types of organisms to use in microcosm tests: (1) GEM's
that are likely to be released; and (2) organisms that already are added to
-------
field systems, whether they are GBI's or not. For example, a rhizobium with
various Barkers might be used to test monitoring methods. The advantage of these
organisms is that field trials can be used to test the realism of microcosm
tests.
The organisms used in microcosm tests can be marked in different vays.
Several strains of a test organism, differing only in specific markers
engineered into them, can be used to test the relative sensitivity of recovery
methods and stability of marking methods.
Different monitoring techniques should be tested together. Part of these
comparative studies might Include sterile and gnotoblotlc (known species
composition) systems. This will allow one to use simple methods such as
non-selective plating and general fluorescent stains to calibrate more specific
methods such as immunofluorescence.
C. 2affi£i£ Protocol
The usefulness of microcosm experiments lies in the ability to use
sensitive, expensive techniques to assay for GBt's and rDNA in order to evaluate
the various methods of recovery and to examine the fates of rDNA and GEM's in
model systems with greater precision than is possible In the field. A sample
protocol Is offered below as an example of an approach used to monitor a GEM.
1. a. Assay for live cells, if possible with selective plate counts.
b. Assay for rDNA activity with Immunofluorescence techniques.
2. If plate counts decline to unmeasurable levels, use enrichment techniques
to continue sampling for live cells.
-------
28
3. Simultaneously with 1. and 2., assay for the presence of rDNA vltb colony
hybridization methods. Comparison vlth the other techniques will help
answer the question: Is the rDNA gene still expressed?
4. If it becomes clear that rDNA sequences can be found, but that activity
cannot be measured, then the reasons must be found for the apparent lack of
gene expression. These Include:
a. The structural rDNA has been altered.
b. The control region for gene expression has been damaged.
c. The gene has moved to a site at which it is not expressed.
d. The rDNA has moved to different strains in which it is not expressed.
These reasons can be explored using the molecular techniques: Restriction
enzyme maps will tell if the gene has moved or if parts were deleted. The
use of labelled probes with maps will tell this more precisely. Finally,
genomic sequencing can be used to detect fine-scale changes.
One problem with microcosm assays for monitoring, fate and effects studies
is the containment of the GEM and rDNA. Preliminary monitoring and fate studies
should be done In the laboratory under oarefully controlled conditions,
involving the standard procedures for working with genetically-engineered
microorganisms. It is Important to test both monitoring techniques and
ecological effects at this stage, because the information gathered will be
needed to design greenhouse tests.
There are two aspects to oontalnment in greenhouse experiments: methods
-------
29
taken to protect against accidental release and a monitoring protocol to assay
for accidental release. The stringency of greenhouse controls should be
determined by a combination of Information on the basic biology of the GO! and
the function of Its rDNA, along with results from laboratory tests of potential
ecological effects. For example, preliminary toxicity tests can be performed In
the laboratory using single species and small microcosms. If results are all
negative, then larger-scale greenhouse tests can be performed with more complex
microcosms. Since any experimental test of ecological effects is insufficient
to prove the impossibility of adverse effects, laboratory results must be
combined with other biological knowledge. Two examples of GEM's designed for
release Into the environment are: (1) a strain of Pseudomonas cepacia designed
to degrade 2,4,5-T, a chlorinated hydrocarbon; and (2) a strain of Pseudomonas
fluorescens with a gene for the Bacillus thurlngensls endotoxin inserted,
designed to poison a species of pest caterpillar. As discussed In section VI,
general biological knowledge suggests that both the severity and likelihood of
potential adverse effects of the second are greater. This suggests that
controls in greenhouse tests should be tighter and that a monitoring protoool
should be well-tested in the laboratory before greenhouse tests of the £..
fluorescens are Initiated.
It is necessary to be able to monitor for the escape of GEM'a or rDNA from
greenhouses during fate and effects tests. The results of laboratory tests of
monitoring techniques should be used to design a greenhouse monitoring program.
This interim monitoring program may not be the same as that used in field tests
or in oommerclal applications. Thus, specific markers might be added to strains
for testing in the greenhouse; these markers might not be used In field tests If
better methods are developed during the preliminary tests. In addition,
specific traits might be added or subtracted from the GEM during the first
-------
30
greenhouse tests. These might Include debilitating the organism so as to
decrease the chances of population explosion in case of accidental release, for
example with sensitivity to temperatures seen in nature but not in the
greenhouse.
E. f
-------
31
V. QUALITY ASSURANCE
A. Introduction
Quality assurance is concerned with the degree of confidence one has in a
methodology. This Involves both the sensitivity (probability of false
negatives) and the specificity (probability of false positives) of the
technique. All of the monitoring techniques mentioned in section III may give
false negatives (the GQ4 is present but the technique does not detect it) and
false positives (the technique gives a positive result even though the GEM is
not present).
Prior to release into the environment, each monitoring technique must be
calibrated in order to determine Its ability to discriminate between the GEM and
other microorganisms In the environment. There are two types of calibration
experiments needed for each GEM, one to test for false positives and one for
false negatives.
B. Testing £fiL sensitivity and specificity
Microcosm experiments are necessary to test for sensitivity (false
negatives). These involve determining the smallest populations that can
reliably be detected with each technique as well as estimating precision and
accuracy. For example, the sensitivity of a DNA probe should be calibrated by
first determining whether it will hybridize to the rDNA and then determining the
alnlnum density of a specific GEM which will be detected in a background of
nonhomologous DNA.
Experimental oondltions for a soil microorganism should range from pure
culture studies in liquid medium, to pure culture studies In sterile soil, to
experiments oarrled out in soil cores taken directly from the field with no
-------
32
pretreatment. One Important aspect of these tests Is that It should be possible
to estimate population densities of the GEM accurately using another reliable
technique. For example, In a pure liquid culture, optical denBlty or
microscopic counts can be used as a calibrating technique.
Specificity (false positives) must be Btudled using field samples. For
example, testing Immunofluorescence techniques Involves observing stained and
unstained samples from the environment for autofluorescence of organic material
and specific or nonspecific binding of the antibody to other microorganisms (see
section IIIB1 on Immunological techniques).
Any attempt to use molecular probes for detection of GEM's first has to be
calibrated with known rDNA and DNA from non-recomblnant organisms. The probe
should be tested to determine the amount of cross-hybridization Interference due
to comnon DNA sequences. The specificity of DNA probes is critically related to
the stringency of the hybridization between the labelled probe and the tested
DNA. By changing the temperature of the hybridization and wash conditions one
can either increase or decrease the homology needed for heteroduplex formation.
The monitoring protocol must be engineered bo as to maximize sensitivity while
minimizing cross-hybridization. Finally, specificity of DNA probes must be
tested by assaying for homologous DNA in natural samples.
C. Testing for linkage between markers and rDNA
The ability of genetic markers to follow a GEM with a minimum of false
postlves and negatives critically depends on the degree of linkage between the
marker gene and the rDNA. If they are not closely linked on the bacterial or
plasmld chromosome, recombination will eventually destroy their association.
This results in both a failure to recognize the GB4 and misldentlfylng a
non-recoobinant organism as a GEM.
-------
33
Therefore, laboratory studies must be done to determine the degree of
linkage between the rDNA and the the marker. A crucial microcosm experiment
involves monitoring known densities of the GB4 with both the genetic marker and
another monitoring technique such as colony hybridization that allows the
investigator to detect the product of the rDNA or the rDNA itself. Such an
experiment provides information on the stability of the marker gene and the
rDNA.
D. ^""""fTY
In summary, all monitoring techniques require calibration tests to show
that they can accurately Identify the GEM in a background of non-recombinant
organisms and to determine the minimal densities of the GO! that can be
detected.
-------
34
VI. CONCLUSIONS
a. Monitoring techniques
The techniques for monitoring GQI's discussed in section III can be ranked
in a hierarchy in which cost increases with information received concerning the
rDNA. Cost and sensitivity to low population densities are not correlated.
Enrichment methods are likely to be more sensitive than plate counts at low
densities; both are inexpensive. All molecular methods require isolation on
plates, so sensitivity is a function of the proportion of GOl cells that grow
and the selectivity of the medium.
Selectivity can be Increased by adding markers for unusual growth
characteristics or product formation, but the cost here is not in monitoring,
but in modification of the G01. Adding components to the selective medium
entails minimal cost increase.
Thus, the techniques are (with numbers in Increasing order of information
received):
(1a) selective plating
(1b) selective enrichment
(2a) fluorescent antibodies
(2b) monoclonal antibodies
(3) colony hybridization
(4) restriction maps
(5) restriction maps with radioactlvely labeled DNA probes
(6) genomic sequencing
The suggested field monitoring protocol stresses qualitative, extensive
-------
35
sampling, concentrating on finding the few GQI's that sight be transported over
distances or those that Blight last a long time (see section II). Sampling might
involve a succession of techniques, starting with the least expensive and
proceeding to molecular techniques as simple methods begin to give negative
results. The latter will occur either as time passes or as distance from the
source Increases.
For example, the gene for the Bacillus thurlngenals endotoxin has been
added to a strain of Pseudomonas fluoreacens (see the second example below).
The Initial monitoring protocol might involve the use of a genomic marker, such
as the ability to produce a chromogenic substance on a certain agar medium. If
after awhile strains with the marker can no longer be found, then it might be
necessary to begin to assay Pseudomonas isolates with a DNA probe for the
endotoxin gene, using colony hybridization techniques. Other species of
Pseudomonas should be tested in case there Is gene transfer between them, and
other genera known to be able to receive plasmlds from Pseudomonas should also
be monitored for the rDNA sequence. Next, if colony hybridization produces only
negative results, it may be appropriate to map some Papudomonaa isolates with
restriction enzymes to check the sensitivity of colony hybridization.
B. Scenarios protocol development
Development of sampling protocols will require information on the
properties of the specific OEM's to be released, combined with general knowledge
of biological processes. The Important issues for protocol development can be
summarized in three questions:
(1) What are the possible adverse effects of release of the GEM?
A list of these will suggest:
(1a) In what geographical locations and habitats should monitoring
-------
36
efforts be concentrated?
(1b) What species of plants and animals should be monitored for
possible Impacts?
(2) How fast and by what mechanisms does the GDI and Its rDNA spread?
Answers to this will suggest:
(2a) Where should monitoring efforts be concentrated geographically?
(2b) In what parts of the environment (habitats) should should
efforts be concentrated?
(3) How might the GEM or its rDNA be modified, especially, how readily
does the rDNA separate from plasmld and genomic markers?
Answers to this will suggest! What combination of markers and direct molecular
assays for rDNA should be used to most effectively follow the GEM and the rDNA?
Three examples of protocol development are discussed below. Some of the
questions asked may already have been answered by the producers of the GEM's;
nevertheless they are presented as examples of the types of issues that are
important to regulatory agencies. The numbering and lettering follows the
scheme just mentioned.
The first example Is the "ice-minus* strain of Pseudomonas svrlngae studied
by Lindow and colleagues (Llndow 1983) and developed by Microlife Technics, Inc.
This strain Is intended for Introduction into crop fields to reduce frost
damage. When strains of Pseudomonas avrlngae isolated from crop leaf surfaces
are added to plant leaf surfaces in greenhouses, frost forms at relatively high
temperatures (Lindow 1983). A strain has been found that when sprayed on leaf
surfaces reduces frost damage (Lindow 1983). It probably acts by outcompeting
the lce-nucleatlon-actlve (INA) strains for space or nutrents on the leaf
surface (Cook and Baker 1983).
-------
37
Strains of Pseudomonas avrlngae are plant pathogens of a wide variety of
species, Including pear, cherry, wheat, lettuce, sugar beet, soybean, sunflower,
rye, rice, barley and lilac (Krieg 1981). They affect plants in at least two
ways; by toxin production (Daly and Deverall 1983) and by ice nucleation (Lindow
1983). Strains of £.. avrlngae are sometimes referred to as "pathovars"
(abbreviated "pv."; Krieg 1981), and sometimes as separate species (Agrlos
1978). The habitat of £. avrlngae is the leaf surface; continued association
with host plants is Important to survival (Schroth et al. 1981). It is widely
distributed; in a study of the distribution of £. avrlngae. strains were found
on 48 of 59 plant species examined (in 23 of 27 families) in North America
(Lindow fit. al.. 1977). In this study, all samples were devoid of obvious
symptoms of infection. However, the authors stated that the 800 isolates
obtained may have included pathogens of different plants.
(1) Possible adverse effects.
Based on the information given above, potential adverse effects include:
(1) Pathogenicity of the original strain used to produce the GEM towards
valuable species such as crops or endangered species; (li) accidental Induction
of or increase in pathogenicity towards valuable species during laboratory
culture and manipulation; (ill) accidental decrease in pathogenicity towards
undesirable species, that is, weeds or potential weeds.
These suggest that:
(1a) Monitoring for the GDI should concentrate on leaf surfaces, with
less attention paid to soil. This is because the potential adverse effects
occur on leaf surfaces and because survival is best on leaf surfaces.
Ideally, all local species should be monitored, because It is always
possible that a rare species is kept rare by pathogenic effects of £.
svrlngae. and with a less virulent strain the species may become a weed.
-------
38
However, it 1b not practical to test all species, so a decision oust be
made, based on feasibility of tbe test and likelihood of adverse effects.
The latter might be Judged from general knowledge of the biology of local
plant species.
(lb) All local plant species should be monitored for gross changes in
population density and visible indicators of disease. Monitoring should
concentrate on crops, endangered species and species Judged likely to be
weeds.
The above suggestions for monitoring can be made more effective by a series
of laboratory and greenhouse tests: (A) The growth and survival
characteristics of the G31 can be tested on local species of plants. If
survival is greater on certain species, these can be emphasized in the field.
(B) Experiments with the GDI and closely related pathogenic strains can be used
to discover which species of plants are most likely to show obvious signs of
disease. (C) Survival and growth In the soil can be tested using the specific
strain designated for release.
(2) Spread
The rate of movement of £. svrlngae in the soil is very amenable to
laboratory and greenhouse tests. Mechanisms can be studied by varying
environmental conditions, for example adding or subtracting earthworms,
arthropods or host plants. The effects of fertilization and of greenhouse
airflow patterns on migration can be studied. In general, these and other
factors Bhould be varied so bb to give the beBt possible opportunities for the
particular factor to have an effect. This hopefully gives an overestimate of
the importance of each factor, which allows one to develop worBt-oase scenarios.
(2a) Monitoring in the field should extend beyond the limits of
spread calculated in based on greenhouse experiments. Occasionally distant
-------
39
populations of disease-sensitive plants or species on vbicb £. avrlngae is
often found should be monitored for rare long-distance dispersal.
(2b) Monitoring should concentrate on the leaf surfaces found most
likely to give positive results in the greenhouse.
(3) CP1 or rDNA modification
The ice-minus property of the strains may be due either to deletions of the
nucleating proteins or to modifications of them. This difference Is crucial,
because modified proteins may recomblne or mutate back to nucleating forms,
whereas deletions are very unlikely to return to wild-type function. In
addition, if a natural protein Is only modified, then monoclonal antibodies for
the protein or molecular probes for the rDNA can be used.
If the rDNA is simply modified, then it may be modified further in the
field; this may cause molecular techniques to miss rDNA and may cause reversion
to ice nucleating activity. Modification of rDNA in the field can be tested in
the laboratory and greenhouse by Inoculating plants with the GEM and following
it through time with several different molecular assays. For example, colony
hybridization and restriction mapping with and without probes can be used to
measure rDNA density over time.
The molecular method giving the greatest apparent recovery Is not
necessarily the best one, because it could give false positives. This can be
tested in the laboratory by constructing several strains with similar but not
identical rDNA and observing how the probe results correlate with ice nucleating
activity.
Laboratory and greenhouse experiments oan be used to compare the power and
selectivity of different monitoring methods. Selective plating and enrichment
can be used to follow markers; fluorescent antibodies against surface proteins
can be used to follow the GDI; and molecular techniques can be used to follow
-------
40
the rDNA. The rates of die-off of the GEM measured by these different
techniques will Indicate the relative efficiencies of the methods.
In addition, these experiments can be used to test for separation of the
lce-nucleation Inactivity (INI) property from plasmid and chromosomal markers.
This includes both loss of markers and transfer of the INI property to other
strains of £.. svringae and other species of bacteria. For example, greenhouse
pots can be set up with host plants and several INI strains with different
markers. The pots are then sampled over time, and £. avrlnaae strains Isolated
and tested for the presence of the markers and INI. This will Indicate the rate
of die-off of the strains and whether the markers separate from the INI
property. The experiments can be repeated, adding INA strains with
distinguishable markers. Transfer of the INI property can thus be assayed.
These experiments vill Indicate how tightly linked are the INI property,
the rDNA and the markers. The results will suggest which combination of
molecular, immunological and standard microbiological methods follows the INI
property and the GEM most efficiently.
The second example concerns the strain of Paeuriomonas fluorescens that
contains the gene for the Bacillus thurlngensis endotoxin (Bioscience News
Update 1985). The aim of this introduction is to poison the black cutworm, a
parasite of corn roots, much as gypsy moth larvae are poisoned by £.
thurlngensis (Doane and McManus 1981). When £.. thurlngensis (hereafter B.t.)
sporulates, It produces a parasporal crystal, a protein crystal inside the cell
but outside the spore. The mixture of spores and crystals is sprayed onto
leaves and the caterpillars consume them In the process of consuming leaves.
The crystal dissolves in the alkaline gut of Lepldoptera, releasing toxic
proteins (Deacon 1983). These act to poison the Na,K-ATPase in the gut wall and
-------
41
eventually lead to the disintegration of the gut wall (Dr. Leigh English, pers.
comm., Harvard University). The spores Invade the body chamber, germinate and
grow. Feeding ceases soon after Ingestion of spores and crystals; death is
usually caused by the protein and sometimes by bacterial growth.
(1) Possible adverse effects
(1) If the added strain of £. fluorescens is transported to nearby areas,
it might have the same effect on Lepldoptera in those areas. If, for example, a
nearby grassland has an endangered species of butterfly, the transport of the
gene for the £.. thurlngensls endotoxin into the area might significantly affect
the survival of the species. (11) £. thurlngensls endotoxin kills Lepldopteran
larvae by poisoning their Na,K-ATPase in the alkaline environment of their gut.
Other orders of insects are usually not sensitive to this strain of
(Deacon 1983). However, it may be possible for the toxin to affect other
important bacterlovores, for example protozoa and nematodes. If the latter are
as important as they seem to be in nutrient recycling in the rhizosphere
(Elliott fijL ai.. 1979), then poisoning microbial food webs might have significant
effects on the entire soil ecosystem, including plants. (Ill) £. fluorescens
is known to be an animal pathogen (Bergan 1981).
(1a) Thus, development of an effective monitoring protocol should
involve answering such questions as are suggested by this scenario: Are
there other species nearby that might be adversely affected by transport of
the GEM? This includes effects of the B.t. toxin as well as pathogenic
effects of £. fluorescens itself on animals. If so, it may be necessary to
monitor for what will probably be low densities of the GB1 in the habitat
of the endangered species.
(1b) In addition, it may be necessary to do qualitative population
studies on the endangered species itself. Field monitoring might also
-------
42
Include estimation of population sizes of bacteriovorous species likely to
be affected. A relatively easy way to do this night Involve (1) counting
total densities of major groups, such as protozoa and nematodes, and (2)
counting certain species individually. The latter can be chosen on the
basis of ease of recognition or ecological Importance (for example plant
pathogenicity).
The value of such a sampling strategy must be explored In laboratory and
greenhouse experiments. These Include laboratory tests in which the GEM Is fed
to a variety of soil bacterlovores and if possible to other species of
Lepidoptera.
(2) Spread
(2a) As above, monitoring In the field should extend beyond the limits
of spread calculated based on greenhouse experiments.
(2b) £.. fluorescens is a soil microorganism, so monitoring should
concentrate in the soil. Greenhouse experiments should be done to
determine whether recovery Is better in the rhizosphere or In bulk soil,
and at what soil depth.
(3) GB1 or rDNA modification
Here, rDNA has obviously been added to the bacterium. Laboratory
experiments can be done with modified B.t. proteins to determine how
similar they must be to be picked up by the various molecular techniques.
As above, combinations of molecular, standard microbiological and
immunofluorescence techniques can be tested in the laboratory and
greenhouse to observe how tightly linked these traits are In the GEM, and
so to develop an effective, economical monitoring protocol.
The third example Is the 2,4,5-T-degradlng bacterium developed by Kellogg
-------
*3
fit Al» (1981) using plasmld-asslsted molecular breeding. This organism contains
a combination of genes, each of whlcb was already present In other genetic
backgrounds. The mixing of genomes was performed by combining bacterial
isolates, each with different combinations of plasmld-encoded pathways for
degradation of various chlorinated carbon compounds. This was done In a
chemostat with 2,4,5-T and other chlorinated compounds as energy sources.
After running for 8 to 10 months, bacteria that were able to use 2,1,5-T as sole
carbon source were Isolated.
(1) Possible adverse effects
This new combination of genes may degrade Important natural organic
compounds and thereby oause environmental damage, for example by poisoning a
plant species that contains a chlorinated organic compound, or by changing Its
rate of decomposition. In a recent review and listing of organic oompounds
found in plants, Robinson (1983) listed only three chlorinated molecules. In
addition, these degradative pathways are fairly specific for certain kinds of
molecules, so that the ability to use a very different chlorinated organic seems
unlikely. Thus, the possibility that this new degradative ability would extend
to a natural organic seems low.
The probability of such an effect occurring can be explored In greenhouse
tests. In such an experiment, a dense suspension of the GEM is combined with
other closely related strains In pots with plants known to contain a chlorinated
organic compound. This design has the effect of also testing for the creation
of new combinations of pathways by gene transfer in the pot. In the control
pots, equally dense suspensions of the GEM cured of the degradative plasmids are
Inoculated along with the same closely related strains. Differences between
treatment and control pots In plant health, growth rate and physiological
parameters are measured.
-------
The sane type of experiment can be used to determine if tbe GEM might
affect the rate of decomposition of plant material derived from species with
chlorinated compounds. These experiments can be done in soil and aqueous
microcosms.
Finally, £. cepacia, like £. fluorescens. is an animal pathogen (Bergan
1981). This can be tested In single-species laboratory toxicity tests.
Positive results would suggest that:
(la) Monitoring should concentrate near or on plant species that
contain chlorinated compounds. These might be found in references such as
Robinson (1983). Monitoring should Include dead plant material. If animal
pathogenicity is deemed potentially important in the strain due to be
released, then monitoring should also include possible animal hosts.
(1b) Local plant species with chlorinated compounds and potential
local animal hosts should be monitored for unusual changes In appearance or
population density.
(2) Spreatf
Rates of dispersal can be studied as described in the first tvo examples.
GEM or rDNA modification
Analysis of rDNA would be difficult because the components of the
degradatlve plasmlds are all or mostly all found in nature. Given this, along
with the fact that chlorinated compounds are very rare in nature, it might be
concluded that assaying for the rDNA in this organism might not be worthwhile.
However, if a certain portion of the rDNA Is not found naturally with the use of
DNA probes, then the appropriate probe can be used to follow this portion.
Tills suggests that molecular assays may not be very useful in this case.
For this organism, the easiest and most direct assay Is tbe ability to grow on
plates or in enrichment cultures with 2,4,5-T as sole carbon source (Kilbane £i.
-------
45
&1. 1983).
C. Research needs
The primary research need Is for mlcroooso tests of monitoring protocols.
This should Involve testing specific sets of techniques with representative
organisms, preferably GBI's due to be released. With the exception of genomic
sequencing, all of the techniques mentioned In section III are well-established
laboratory procedures. Thus, research support is needed for their application,
not their development. Research should stress comparative studies of the
techniques and studies of their application to natural systems. Long-term
research on microcosm design for assessing fate and environmental impacts is
necessary either to bring us closer to the goal of a standardized set of
microcosms with standardized measurements, or to ascertain that such
standardization is not possible given natural variability. Repllcablllty
between laboratories is an Important Issue for EPA to consider in supporting
research; it may be worthwhile to support similar projects in more than one
laboratory.
Microcosm tests should Include those mentioned in section V on quality
assurance. These Involve testing microcosm populations of GEM's for sensitivity
of the techniques. In addition, laboratory and environmental samples must be
tested for false positive results.
Two important aspects of the application of these techniques to natural
situations are: (1) Partitioning of GQ4's into different parts of the
environment. This can be explored experimentally with GEM's which are to be
released or with very similar strains. The purpose is to see if GEM's survive
or grow better in different habitats, for example different depths of the soil,
water column or aquatic sediments. This information might be useful in
-------
46
designing sampling strategies for low-density populations.
(2) Desorptlon of cells from surfaces for subsequent plating and molecular
analysis. There Is already much literature, so the first step should be the
compilation of existing techniques. The second step should then be to explore
the different methods experimentally.
The research needs described here can be divided up into short- and
long-term efforts. Short-term efforts Involve testing existing microcosm
systems, monitoring techniques and desorptlon methods. The purpose of this is
to establish an interim set of protocols. Much Information can be gathered in a
matter of one or two years, using GEM's already prepared for release.
Three organisms that may be useful for short-term research are: (1) the
lce-mlnus Pseudomonas svrlngae with a gene deletion; (2) the £. fluorescens
with a single gene for the £.. thurlngensls endotoxin added; and (3) the 2,4,5-T
degrading strain of £. cepacia with several added genes. These three are quite
different in their ecological effects and In the nature of the genetic
manipulation, so they form a good starting point. Consideration should also be
given to other types of organisms planned for release.
Long-term projects Involve microcosm design for simultaneous assessment of
monitoring techniques, fate and effects, new techniques for desorptlon, new
marking methods, and further studies on how GEM's partition themselves in
different parts of the environment. Advances in molecular techniques for
examining changes in genomic structure should be followed and applied as
appropriate to monitoring protocols.
-------
*»7
VI. LITERATURE CITED
1. Agrios, G.N. 1978. Plant Pathology, second edition. New York, Academic
Press.
2. Anderson, S.S. 1975. Plasmid transfer in £.• coll- J. Infec. Diseases 175:
686-687.
3. Atlas, R.M. and R. Bartha. 1981. Microbial Ecology: Fundamentals and
Applications. Reading, Mass., Addlson-Wesley Publishing Company.
4. Bergan, T. 1981. Human- and animal-pathogenic members of the genus
Pseudomonas. In.: The Procaryotes, Volume 1, edited by M.P. Starr al..
Chapter 59. New Tork, Sprlnger-Verlag.
5. Bioscience News Update. Gene-splicing field trials schedule. Bioscience 35:
1U7.
6. Bohlool, B.B. and E.L. Schmidt. 1980. The Immunofluorescence approach in
microbial ecology. Adv. Microb. Ecol. 4:203-21(2.
7. Chilton, M.D., A. DeFramond, R. Fraley, ££.&!.• 1983. Ti and Ri plasmlds as
rectors for genetic engineering of higher plants. Abstract, 15th Miami
Vlnter Symposium, Miami, January 17-21, 1983.
8. Church, G.M. and W. Gilbert. 1981. Genomic sequencing. Proc. Nat. Acad.
Sci. USA. 81: 1991-1995.
9. Cook, R.J. and K.F. Baker. 1983. The Nature and Practice of Biological
Control of Plant Pathogens. St. Paul, Minnesota, The American
Phytopathologlcal Society.
10. Daly, J.M. and B.J. Deverall, eds. 1983. Toxins and Plant Pathogenesis.
New Tork, Academic Press.
11. Deacon, J.V. 1983. Microbial Control of Plant Pests and Diseases.
-------
Washington, D.C., American Society for Microbiology.
12. Doane, C.C. and M.L. McManus, eds. 1981. The Gypsy Moth: Research Toward
Integrated Pest Management. Washington, D.C., U.S. Department of
Agriculture.
13. Draggan, S. and P. Van Torls. 1979. The role of microcosms in ecological
research (An Introduction to this special issue on microcosms). Int. J.
Env. Studies 13: 83-85.
14. Elliott, E.T., C.V. Cole, D.C. Coleman, R.V. Anderson, H.W. Hunt and J.F.
McClellan. 1979. Amoebal growth in soil microcosms: A model system of C, N,
and P trophic dynamics. Int. J. Env. Studies 13: 169-174.
15. Fraley, R.T., R. Borsch and S.G. Rogers. 1983. University of California at
Los Angeles - Keystone Meeting, Keystone, Colorado, March 1983.
16. Geisy, J.P., ed. 1980. Microcosms in Ecological Research. Washington, D.C.,
U.S. Department of Energy.
17. Genewatch. 1983. Ice-enhancing bacteria. Genewatch 1(3-4): 11.
18. Gillett, J.W. and J.M. Witt, eds. 1977. Terrestrial Microcosms.
Proceedings of a workshop on terrestrial mlcrooosms, 1977*
19. Grlce, G.D. and M.R. Reeve, eds. 1982. Marine Mesooosms. New York,
Springer-Verlag.
20. Grunsteln, M. and D.S. Hogness. 1975. Colony hybridization: A method for
the isolation of cloned DNAs that contain a specific gene. Proc. Nat. Acad.
Sci. 72: 3961-3965.
21. Herrera-Estrella, L., A. Depicker, M. Van Montagu, et al. 1983. Expression
of ohlaeric genes transfered into plant cells using a Ti-plasmid-derived
vector. Nature 303: 209-213.
22. Hobble, J.E., R.J. Daley and S. Jasper. 1977* Use of Nuclepore filters for
eountlng bacteria by fluorescence microsoopy. Appl. Envlr. Microb. 33:
-------
1225-1228.
23* Kellogg, S.T., D.K. Chatterjee and A.M. Chakrabarty. 1981. Plasmid-assisted
molecular breeding: Mew technique for enhanced biodegradation of persistent
toxic chemicals. Science 214: 1133-1135.
24. Kemp, J. 1983. UCLA - Keystone Meeting, Keystone, Colorado, March, 1983.
25. Kilbane J.J., D.K. Chatterjee, and A.M. Chakrabarty. 1983* Detoxification
of 2,4,5-trichlorophenoxyacetlc acid from contaminated soil by Pseudomonaa
cepacia. Appl. Envir. Mlcrob. 45: 1697-1700.
26. Kimball, K.D. and S.A. Levin. 1985. Limitations of laboratory bioassays:
the need for ecosystem-level testing. Bioscience 35: 165-171*
27. Kinne, 0. 1977. International Helgoland Symposium "Ecosystem Research":
Opening Address. Helg. wlssen. Meeresunt. 30: 1-7*
28. Krieg, N.B., ed. 1981. Bergey*s Manual of Systematic Bacteriology, Volume
1. Baltimore, Maryland, Williams and Wilkins.
29. Langer, P.R., A.A. Waldrop and D.C. Ward. 1981. Enzymatic synthesis of
biotin-labeled polynucleotides: Hovel nucleic acid affinity probes. Proc.
Nat. Acad. Scl. 78: 6633-6637.
30. Levy, S.B. 1984. Survival of plasmlds in £. coll. In Arber, W., K
Illmensee, V.J. Peacock and P. Starling (eds.) Oenetlc Manipulation.
Impact on Man and Society. Cambridge, Cambridge University Press.
31. Levy, S.B., B. Marshall, A. Orderenk, and D. Rouse-Eagle. 1980. Survival of
Escherichia coll host-vector systems in the mammalian intestine. Science
209: 391-391.
32. Levin, R. 1983. Genetic probes become ever sharper. Science 221: 1167*
33• Llndow, S.E. 1983. Methods of preventing frost injury caused by epiphytic
lce-nucleatlon-actlve bacteria. Plant Disease 67: 327-333*
34. Llndow, S.E., D.C. Amy and C.D. Upper. 1977* Distribution of epiphytic
-------
ice-nucleation-active strains of Pseudomonas svrlngae. Proc. Amer.
Phytopath. Soc. 4:107.
35. Lynch, J.M. and N.J. Poole, eds. 1979. Microbial Ecology: A Conceptual
Approach. Oxford, Blackwell Scientific Publications.
36. Maniatis, T., E.F. Fritsch and J. Sanbrook. 1982. Molecular Cloning: a
Laboratory Manual. Cold Spring, New Tork, Cold Spring Harbor Laboratory.
37* Marshall, K.C. 1976. Interfaces in Microbial Ecology. Cambridge, Mass.,
Harvard University Press.
38. Masul, Y., T. Mlzuno and M. Inouye. 1984. Novel high-level expression
cloning vehicles: 104-fold amplification of Escherichia coll minor protein.
Biotech. 2: 81-85.
39. Maxam, A.M. and V. Gilbert. 1980. Sequencing end-labeled DNA with
base-specific chemical cleavages. Methods Enzymol. 65: 499*560.
40. Meyers, J.A., D. Sanchez, L.P. Elwell and S. Falkow. 1976. Simple agarose
gel electrophoresis method for the identification and characterization of
plasmld deoxyribonucleic acid. J. Bact. 127: 1529-1537*
41. Peacock, V.J. 1983. Gene transfer in agricultural plants. Abstract, 15th
Miami Winter Symposiuma, Miami, January 17-21, 1983*
42. Porter, E.G. and T.S. Felg. 1980. The use of DAPI for identifying and
counting aquatic microflora. Llmnol. Oceanog. 25: 943-948.
43. Pritchard, P.H. and A.V. Bourquin. 1984. The use of mlcrcosms for
evaluation of interactions between pollutants and microorganisms. Adv.
Microb. Ecol. 7: 133-216.
44. Robinson, T. 1983. The Organic Constituents of Higher Plants. North
Amherst, Massachusetts, Cordus Press.
45. Russek, E. and R.R. Colwell. 1983. Computation of most probable numbers.
Appl. Env. Microb. 45: 1646-1650.
-------
46. Sagik, B.P. and C.A. Sorber. 1979. The survival of H-V systems In domestic
sewage treatment plants. Recomb. DNA Tech. Bull. 2: 55-61.
47. Sanger, F. &£.&!.. 1977. Proc. Nat. Acad. Scl. USA. 74: 5463.
48. Sayler, G.S., M.S. Shields, E.T. Tedford, A. Breen, S.V. Hooper, K.M.
Sirotkin, and J.W. Davis. 1985. Application of DNA-DNA colony hybridization
to the detection of catabolic genotypes in environmental samples. Appl.
Envir. Microb. 49: 1295-1303.
49. Schell, J., M. Van Montagu, J.P. Hernalsteens, et al. 1983. Ti plasmlds as
experimental gene vectors for plants. Abstract, 15th Miami Winter
Symposium, Miami, January 17-21, 1983*
50. Schiff, V. and V. Klingmuller. 1983* Experiments with Escherichia coll on
the dispersal of plasmlds in environmental samples. Recomb. DNA Tech.
Bull. 6: 101-102.
51. Schroth, M.N., D.C. Hildebrand and M.P. Starr. 1981. Phytopathogenic
members of the genus Pseudomonas. In; The Procaryotes, Volume 1, edited by
M.P. Starr sL. Chapter 60. New York, Springer-Verlag.
52. Sibley, C.G. and J.E. Alquist. 1983. Phylogeny and classification of birds
based on the data of DNA-DNA hybridization. Current Ornithology 1: 245-292.
53* Sleburth, J.M. 1979. Sea Microbes. New Tork, Oxford University Press.
54. Southern, E.M. 1975. Detection of specific sequences among DNA fragments
separated by gel electrophoresis. J. Mol. Biol. 98: 503-517.
55. Szwerinskl, H., S. Galser and D. Bardtke. 1985. Immunofluorescence for the
quantitative determination of nitrifying bacteria: Interference of the test
In biofllm reactors. Appl. Microbiol. Biotechnol. 21: 125-128.
56. Vatson, J. 1980. Foreword to the special issue on the Experimental Lakes
Area. Can. J. Fish. Aq. Scl. 37: 311-312.
57. Witt, J.M. and J.V. Gillett, eds. 1979. Terrestrial Microcosms and
-------
52
Environmental Chemistry. Proceedings of two colloqula, held June 13-14,
1977, at Oregon State University, Corvallls, Oregon.
-------
£
-------
AAAS/EPA Biotechnology Workshop
29 April - 1 May, 1984
LIST OF PARTICIPANTS
Workshop Chairman
Gilbert S. Omenn, Dean
School of Public Health and Community Medicine
University of Washington
Seattle, WA
Health Effects Review Group
D. Michael Gill (Chairman)
Department of Molecular Biology and Microbiology
Tufts University Medical School
Boston, MA
Susan Gottesmann
Laboratory "of Molecular Biology
National Cancer Institute
Bethesda, MD
Dennis J. Kopecko
Department of Bacterial Immunology
Walter Reed Army Institute of Research
Washington, DC
Richard Novick
Department of Plasmid Biology
Public Health Research Institute
New York, NY
Mark Townsend (Rapporteur)
Office of Toxic Substances, EPA
Washington, DC
David Kleffman (Author)
Office of Research and Development, EPA
Washington, DC
Clinton Kawanishi
Health Effects Research Laboratory, EPA
Research Triangle Park, NC
Michael Waters
Health Effects Research Laboratory, EPA
Research Triangle Park, NC
Daphne Kamely
Office of Research and Development, EPA
Washington, DC
-------
Mark Segal
Office of Toxic Substances, EPA
Washington, DC
Stephen L. Johnson
Office of Pesticide Programs, EPA
Washington, DC
Tina Levine
Office of Toxic Substances, EPA
Washington, DC
William Schneider
Office of Pesticide Programs, EPA
Washington, DC
Irving Mauer
Office of Pesticide Programs, EPA
Washington, DC
Environmental Effects Review Group
R. Darryl Banks
New York State Department of
Environmental Conservation
Albany, NY
Philip Regal
Department of Ecology and Behavioral Biology
University of Minnesota
Minneapolis, MN
Frances Sharpies (Chairman)
Office of Institutional Planning
Oak Ridge National Laboratory
Oak Ridge, TN
Guenther Stotzky
Department of Biology
New York University
New York, NY
Frederick Betz (Rapporteur)
Office of Pesticide Programs, EPA
Washington, DC
Charles Hendricks (Author)
Office of Research and Development, EPA
Washington, DC
A1 Bourquin (Author)
Environmental Research Laboratory, EPA
Sabine Island
Gulf Breeze, FL
-------
Harold Kibby
Environmental Research Laboratory, EPA
Corvallis, OR
J.W. Falco
Office of Research and Development, EPA
Washington, DC
Robert Brink
Office of Toxic
Washington, DC
Arthur Stern
Office of Toxic
Washington, DC
Jane Rissler
Office of Toxic
Washington, DC
Herb Manning
Office of Pesticide Programs, EPA
Washington*. DC
Allen Vaughan
Office of Pesticide Programs, EPA
Washington, DC
Carl Grable
Office of Toxic Substances, EPA
Washington, DC
Ramon Seidler (Author)
Professor of Microbiology
Oregon State University
Corvallis, Oregon
Monitoring and Quality Assurance Review Group
Victor J. Cabelli (Chairman)
Department of Microbiology
University of Rhode Island
Kingston, RI
Karl M. Johnson
U.S. Army Medical Research Institute of
Infectious Diseases
Fort Detrick, MD
David Pramer
Waksman Institute of Microbiology
Rutgers University
Piscataway, NJ
Substances, EPA
Substances, EPA
Substances, EPA
-------
Gary S. Saylor
Graduate Program in Ecology
Department of Microbiology
University of Tennessee
Knoxville, TN
William G. Wells (Rapporteur)
Office of Public Sector Programs, AAAS
Charles Plost
Office of Research and Development, EPA
Washington, DC
Christon J. Hurst (Author)
Environmental Monitoring Systems Laboratory, EPA
Cincinnati, OH
John A. Santolucito
Environmental Monitoring Systems Laboratory, EPA
Las Vegas, NV
Nancy Chiu
Office of T-oxic Substances, EPA
Washington, DC
Michael Callahan
Office of Toxic Substances, EPA
Washington, DC
Andrew Jovanovich
Office of Pesticide Programs, EPA
Washington, DC
Amy S. Rispin
Office of Pesticide Programs, EPA
Washington, DC
Michael Dellarco
Office of Research and Development, EPA
Washington, DC
Containment and Control Technologies Review Group
Aileen Compton
Smith, Kline and French Laboratories, Inc.
Philadelphia, PA
Charles Cooney (Chairman)
Department of Chemical Engineering
MIT
Cambridge, MA
-------
Hester Kobayashi
Standard Oil of Ohio Research Center
Cleveland, OH
Seth Pauker
Biogen Research Corporation
Cambridge, MA
Barry D. Gold (Rapporteur)
Office of Public Sector Programs, AAAS
N. Dean Smith
Industrial Environmental Research Laboratory, EPA
Research Triangle Park, NC
John Burckle (Author)
Industrial Environmental Research Laboratory, EPA
Cincinnati, OH
Albert D. Venosa (Author)
Environmental Research Laboratory, EPA
Cincinnati, OH
Bala Krislirian
Office of Research and Development, EPA
Washington, DC
Elizabeth Ward
Office of Toxic Substances, EPA
Washington, DC
Larry Longanecker
Office of Toxic Substances, EPA
Washington, DC
Organizers and Staff
John R. Fowle, III
Office of Research and Development, EPA
Washington, DC
Anne Hollander
Office of Toxic Substances_, EPA
Washington, DC.
Morris A. Levin
House Committee on Science & Technology
Washington, DC
Albert H. Teich, Project Director
Office of Public Sector Programs, AAAS
Jill H. Pace, Project Associate
Office of Public Sector Programs, AAAS
Mary I. Haddock, Administrative Secretary
Office of Public Sector Programs, AAAS
-------
Special Guests*
Carl Gerber
Office of "Research and Development, CPA
Washington, DC
William Waugh
Office of Toxic Substances, EPA
Washington, DC
Carl Mazza
Office of Toxic Substances, EPA
Washington, DC
Donald Clay
Office of Toxic Substances, EPA
Washington, DC
Bernard Goldstein
Office of Research and Development, EPA
Washington, DC
Scott Baker
Office of Research and Development, EPA
Washington, DC
^Opening Session
------- |