United States
                     Environmental Protection
                     Agency
Health Effects Research
Laboratory
Research Triangle Park, NC 2771 1
                     Research and Development
EPA-600/S1-83-005   Sept. 1983
&EPA          Project  Summary

                     Detection  of  Carcinogenicity
                     Based  on  Mutagenicity  in
                    Arabidopsis
                    George Redei and Gregona N. Acedo
                      Thirty-seven synthetic chemicals plus
                     two mycotoxins were tested for mu-
                     tagenicity in an Arabidopsis embryo
                     system. The results of this test, pro-
                     karyotic repair tests, bacterial mutation
                     assays, eukaryotic cell systems and in
                     vivo tests were compared to the car-
                     cinogenicity classifications of the
                     chemicals.
                      Thirty-two of the 37 chemicals tested
                     were correctly identified as either mu-
                     tagenic or nonmutagenic in the Arabi-
                     dopsis assay. To com pa re these results
                     with those  of the other  assays,  we
                     defined three criteria. "Sensitivity" in-
                     dicated the percentage  of tested car-
                     cinogens  that were  mutagenic  in a
                     system. Of 20 carcinogens tested in
                     the Arabidopsis assay,  1 9 were mu-
                     tagenic (sensitivity 95%). For the other
                     assays, sensitivities ranged from 16%
                     to  88%.  "Specificity" indicated the
                     percentage  of  noncarcinogens  that
                     were nonmutagenic in a system. Three
                     of twelve noncarcinogens were non-
                     mutagenic to Arabidopsis (specificity
                     25%). For the other assays, specificities
                     ranged from 20% to 100%.  The non-
                     carcinogens pyrene, 4-acetylamino-
                     fluorene, 1-naphthylamine, isopropyl-
                     N-(chlorophenyl) carbamate, azoxyben-
                     zene, and ascorbic acid were mutagenic
                     in  the Arabidopsis assay and  other
                     assays. "Accuracy" indicated the per-
                     centage of correctly identified chemi-
                     cals based on the predicted classifica-
                     tion. For the Arabidopsis assay, 22 of
                     32 chemicals were correctly identified
                     (accuracy 69%). Overall accuracies for
                     the other assays ranged from 57% to
                     62%.
                      The Arabidopsis assay was the most
                     sensitive and accurate short-term test
in our comparison. Therefore, Arabidopsis
can be used to supplement a battery of
short-term tests  for identifying car-
cinogens.

  This Project Summary was developed
by EPA's Health Effects Research Lab-
oratory, Research Triangle Park, NC. to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).

Introduction
  In an international program sponsored
by the British Medical Research  Council,
the U S. National Institute of Environmental
Health Sciences, the  U.S Environmental
Protection Agency (EPA), and the  National
Cancer Institute of Japan, workers from
63  laboratories evaluated the mutagenic
effects of 42 chemicals  using different
short-term biological assays, such as pro-
karyotic repair tests, bacterial mutation
assays, in vitro eukaryotic cell assays, and
in vivo tests with Drosophila and mice (the
results of these studies were published in
Evaluation of Short- Term Tests  for Car-
cinogens, F.D. deSerresandJ Ashby, eds.,
1981, Elsevier/North Holland, New York).
From the results of previous animal assays,
the chemicals were classified as proven
carcinogens, harmless analogues of car-
cinogens, and noncarcinogens This pro-
gram was intended to help researchers
select the most effective series of short-
term tests. Because no single  system
consistently identified all the carcinogens
and noncarcinogens, we  retested 37
chemicals using an Arabidopsis assay and
compared the results to those of the other
assays.

-------
  Arabidopsis thaliana (L) Heynh. seeds
were soaked  in water for 24 hr and then
treated for 1 5 hr with chemicals dissolved
in either medium with nutrients and minerals
or dimethylsulfoxide (DMSO). As a control,
some seeds were concurrently exposed
only to the solvents.  Approximately 250
seeds per treatment were planted on Promix
medium,  moistened,  placed  in covered
glass vessels, and incubated at 24° C
under a light intensity of 800 foot-candles.
After germination,  1 00 to 200 plants per
culture vessel were examined for mutations
(fruits with pale or white embryos) The
mutation frequency  for  exposed seeds
minus control frequency was recorded as
"mutation %."
  To  compare  the assays,  we defined
three criteria "Sensitivity" indicated the
percentage of tested carcinogens that were
mutagenic to a system. "Specificity" indi-
cated the percentage of  noncarcmogens
that were nonmutagenic to a system. "Ac-
curacy" indicated the percentage of correctly
identified chemicals based on the predicted
classification
  The Arabidopsis assay has many advan-
tages: (i)  it is a eukaryotic test, (n) forward
mutations are scored at more than 10,000
loci, which ensures a representative sample
of the genome, (in)  the  results  of the
embryo assays can be supplemented by
progeny tests, (iv) the assay does not rely
on any supplement for activation; the plant
seems to convert most promutagens into
ultimate  mutagens,  (v)  the  test is less
expensive than  most  higher  eukaryotic
assays,  and  (vi)  according  to literature
survey, 88% of the tested carcinogens are
mutagenic to Arabidopsis.

Results
  Thirty-two  of the 37 chemicals tested
with the Arabidopsis assay could be cor-
rectly identified as mutagens or nonmuta-
gens. Of these 32 chemicals, 28  were
mutagenic and four were nonmutagenic.
Nineteen of  20 carcinogens  were mu-
tagenic in the Arabidopsis assay (sensitivity
95%) The sensitivity ranges for the other
tests were  prokaryotic repair tests,  1 6%
to 73% , bacterial mutation assays,  27%
to 76%,  eukaryotic cell systems, 8% to
88%, and in  vivo tests, 1 7%  to 52%.
  Table 1 compares the sensitivities of all
assays combined,  bacterial mutation as-
says, and the Arabidopsis assay. Of the 2 5
carcinogens,  14 (Group I) were correctly
identified as mutagens by all assays com-
bined (sensitivity 60% or better) and bac-
terial mutation assays (sensitivity 70% or
better). The Arabidopsis  assay identified
10 of the  14 Group I  compounds as
mutagens (sensitivity 71 %) The identifica-
Table 1    A comparison of the sensitivities of all assays combined, the bacterial mutation tests,
          and the Arabidopsis assay for 25 carcinogens. Sensitivity is defined as the percentage
          of carcinogens that are mutagenic in a system. The symbols •, -, 7,  and NT indicate
          mutagenicity, nonmutagenicity, inconclusive results, and not tested, respectively
Carcinogens
Group I
fi-Propiolactone
4-Nitroqumoline-N-oxide
2-Acetylammofluorene
Benzo(a)pyrene
Epichlorohydnn
Methylazoxymethanolacetate
Methylene
bis(2- chloroaniline)
2-Naphtylamine
Cyclophosphamide
Hydrazine sulphate
Dimethylanthracene
Benzidme
Dimethylcarbamoyl chloride
Nitrosomorpholme
Group II
Auramme
Dimethylammobenzene
O- Toluidme
Hexamethylphosphoramide
Safrole
Urethane
Ethylenethiourea
Ethionme
Diethylstilbestrol
Chloroform
Ammotnazole

All assays

93
89
86
83
82
81
77

75
74
71
70
68
67
60

49
48
44
38
38
38
32
28
24
20
20
Sensitivity
Bacterial mutation

100
700
700
700
95
73
84

95
74
80
88
85
76
70

48
45
33
7;
20
26
20
14
14
11
70

Arabidopsis

•
NT
•
•
•
•
•

?
•
•
?
•
7
•

•
•
•
7
•
•
•
•
•
•
-
tion  of Group II carcinogens was more
difficult. The bacterial mutation and com-
bined assay sensitivities were below 49%
for these carcinogens. However, Arabidopsis
identified 9 of 1 1 Group II carcinogens as
mutagens (sensitivity 82%).
  Only three  of the 1 2 presumably  non-
carcinogenic chemicals were nonmutagenic
in the Arabidopsis assay (specificity 25%)
Specificity  ranges  for  the other  assays
were: prokaryotic  repair tests, 35% to
75%; bacterial  mutation assays, 59% to
82%, eukaryotic systems, 23% to 85%;
and in vivo tests, 20% to 1 00%. Although
classified as  noncarcmogens, pyrene, 1-
naphthylamme,  4-acetylammofluorene,
isopropyl-N-(chlorophenyl) carbamate,
azoxybenzene,  and  ascorbic acid were
mutagenic  in the Arabidopsis assay and
other assays.
  Table 2 compares the pooled accuracies
of four assay  systems with the accuracies
of \he Arabidopsis assay for 42 chemicals.
The Arabidopsis assay correctly identified
69% of the chemicals  and was the  most
accurate assay. Overall accuracies for the
other systems  were:  prokaryotic  repair
tests, 58%,  bacterial  mutation assays,
62%; eukaryotic tests, 60%; and in  vivo
tests, 57%.
Conclusion
  A higher percentage of carcinogens and
presumed noncarcmogens were mutagenic
to Arabidopsis than  to  the  prokaryotic
repair systems, bacterial mutation assays,
eukaryotic tests, and in  vivo tests. This
high sensitivity can be attributed to the
large number of target loci in the Arabidop-
sis genome.  Also, the Arabidopsis system
correctly identified carcinogens and non-
carcinogens more consistently than the
other assays. Therefore, the Arabidopsis
assay could be a  valuabe addition to a
battery of short-term tests for identifying
carcinogens
                                     2

-------
Table 2    A comparison of the pooled accuracies of four assay systems and the accuracy of the
           Arabidopsis assay for 42 chemicals Accuracy is  defined as the percentage of
           correctly identified chemicals based on the predicted classification Jhesymbols •, -,
           NT, and ? indicate carcinogen/city, noncarcmogenicity, not tested, and inconclusive
           results, respectively.

                                                    Accuracy (%)
Chemicals
Nitroquinolme-N-oxide
Methyl nit roquinol me
N-oxide
Benzidme
Tetramethylbenzidme
Dimethyl ammobenzene
D/methylammobenzene
sulphonic acid-Na
I'rPropiolactone
j-Butyrolactone
Benzofajpyrene
Pyrene
Ethionine
Methionine
Chloroform
Tnchloroethane
2-Acetylammofluorene
4-Acetylammofluorene
2-Naphthylamme
1-Naphthylamme
Nitrosomorpholine
Diphenylnitrosomme
Urethane
lsopropyl-N-(3-chlorophenyl)
carbamate
Methylazoxymethanol-
acetate
Azoxybenzene
Dinitrosopentamethylene
tetramme
Hydrazine sulphate
Hexamethylphosphoramide
Safrole
Diethy/stilbestrol
Cyclophosphamide
Epichlorohydrme
Auramme
Methylene bis(2-chloro-
aniline)
To/uidine
Ethylenethiourea
Aminotnazole
Dimethylcarbamoyl
chloride
Dimethylformamide
D/methy /anthracene
Anthracene
Sucrose
Ascorbate
Repair
tests
• 100

0
• 57
71
• 0

77
• 700
67
• 71
- 75
• 29
- 67
• 17
- 83
• 71
- 33
• 71
- 50
• 43
- 57
• 38

- 50

• 86
17

17
83
17
86
29
67
80
57

100
43
67
77

• 77
- 700
• 57
- 63
- 700
67
Bacterial
mutation
100

0
85
95
45

45
100
90
100
57
14
90
11
94
100
0
95
30
70
70
26

100

73
37

90
80
11
20
14
74
95
48

84
33
20
10

76
82
88
89
100
84
Eukaiyot/c
systems
91

5
68
88
64

65
92
71
77
53
35
76
31
67
73
50
75
36
56
65
41

38

83
46

56
21
48
39
26
73
20
50

54
64
38
46

64
76
70
92
75
58
In vivo
tests
20

60
40
88
40

100
33
100
83
100
40
100
0
100
100
100
29
78
50
80
100

100

100
67

75
0
88
20
33
100
20
0

66
0
0
0

40
100
0
100
100
100
Arabidopsis
assays
NT

NT
100
7
100

NT
100
7
100
0
100
100
100
0
100
0
7
0
100
0
100

0

100
0

7
100
7
100
100
100
700
700

700
700
700
0

7
0
7
100
100
0
Total
89

8
68
88
48

62
93
81
83
62
28
81
20
81
86
27
75
41
60
66
38

67

81
38

67
71
38
38
34
74
82
49

77
44
32
20

67
82
70
85
92
72
Accuracy of the systems
583
         62 1
                  595
573
                                      688
                    608

-------
       George Redei  and Gregoria N. Acedo are with the  University of Missouri,
         Columbia, MO 65211.
       Shahbeg S. Sandhu is the EPA Project Officer (see below).
       The complete report, entitled "Detection of Carcinogenicity Based on Mutagenicity
         in Arabidopsis," (Order No. PB 83-225 078,  Cost: $10.00, subject to change)
         will be available only from:
              National Technical Information Service
              5285 Port Royal Road
              Springfield,  VA 22161
              Telephone: 703-487-4650
       The EPA Project Officer can be contacted at:
              Health Effects Research Laboratory
              U S  Environmental Protection Agency
              Research Triangle Park, NC 27711
                                                   -,-US GOVERNMENT PRINTING OFFICE 1983-659-017/7186
United Stales
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use S300
                                U  5  tNVIK  KKUTtCriUN
                                      U*  5  H6KAKY
                                      o  D£AkbOK,v  SIRtEF
                                CHICAGO  IL  606U4

-------