Introduction to Exposomics
Elin M. Ulrich
U.S. Environmental Protection Agency
Center for Computational Toxicology and Exposure
The views expressed in this presentation are those of the author(s) and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.

-------
What is the Exposome?
-f First defined by Wild in 2005, the exposome includes chemical and non-chemical
stressors, from both internal and external sources across all life stages
Diet	Transportation
0
Products	Habits	Leisure and play
USD/	fl-Vhi
Occupation	General	Location
& i$*	AV S)	^ kd bhI^
CP Wild, A Scalbert, and Z Herceg, Environ. Mol. Mutagen. 54:480-499, 2013.

-------
Why Study the Exposome?
1) Understanding causes of disease
2) Ensuring chemical safety and
human/ecological health
"...70-90%	of disease
to differences	in environmen
GIVE A DOG A PHONE
Technology for our furry friends
EPIDEMIOLOGY
Environment and Disease Risks
Stephen M. Rappaport and Martyn T. Smith
Although the risks of developing
chronic diseases are attributed to
both genetic and environmental fac-
tors, 70 to 90% of disease risks are probably
due to differences in environments (1—3). Yet,
epidemiologists increasingly use genome-
wide association studies (GWAS) to investi-
gate diseases, while relying on questionnaires
to characterize "environmental exposures."
This is because GWAS represent the only
approach for exploring the totality of any risk
factor (genes, in this case) associated with dis-
ease prevalence. Moreover, the value of costly
genetic information is diminished when inac-
curate and imprecise environmental data lead
to biased inferences regarding gene-environ-
ment interactions (•#). A more comprehensive
and quantitative view of environmental expo-
School of Public Health, University of California, Berkeley,
CA 9472
-------
How to Study the Exposome
Bottom-Up
Top-Down
Measure Exposures
Within Relevant Media
Exposure surveillance
Present
Exposure forensics
Signature
Chemical prioritization
Relevant
Effect-directed analysis
Active
Biomarker discovery
Predictive
Measure Exposures
Within the Receptor
-f Need methods that can measure LOTS of chemicals
+ Something akin to in vitro toxicity assays but for exposure.
-------
What is Non-Targeted Analysis?
¦f Targeted Analysis
Standards, calibration curves
-f Suspect Screening Analysis (SSA)
Lists of compounds
4- Non-Targeted Analysis (NTA)
MS first principles
-------
Non-Targeted Analysis Workflow
Experimental Acquisition	Database & Library Matching Data Analysis & Computational Tools
Analytical Instruments
Chemical Databases
Computational Tools
High resolution accurate mass, mass spectrometry (QToF, Orbitrap)
CompTox Chemicals Dashboard, MassBank, PubChem
CPDat, media and retention time prediction, MetFrag, R/Python tools
-------
How does High Resolution MS work?
Atom
Natural
Abundance
Exact Mass

99.9885%
1.007825
2H
0.0115%
2.014102
12C
98.93%
12.000000
13C
1.07%
13.003355
14 N
99.632%
14.003074
15N
0.368%
15.000109
16Q
99.757%
15.994915
170
0.038%
16.999131
ISQ
0.205%
17.999159
19 F
100%
18.998403
32S
94.93%
31.972072
33$
0.76%
32.971459
34$
4.29%
33.967868
36$
0.02%
35.967079
35CI
75.78%
34.968853
37CI
24.22%
36.965903
Example: Fipronil
Molecular Formula: C^F^C^Fq^OS
Monoisotopic Mass: 435.938706
= (12.0000*12 Carbon) + (1.007825*4 Hydrogen) +
(34.968853*2 Chlorine) + (18.998403*6 Fluorine) +
(14.003074*4 Nitrogen) + (15.994915*1 Oxygen) +
(31.972072*1 Sulfur)
100
-------
What do we Learn?
Health
Chemical Safety
sftPt
4- Find chemicals of emerging concern
4- Sources contributing to exposure
Prioritization of chemicals
Assess toxicity/exposure overlaps
Predict exposures and risk
-------
Exposomics Experimental Design
Name
Example
Purpose
Tracers
Isotopically labeled standards: 13C3-Atrazine,
D3-Thiamethoxam, 13C4,15N2-Fipronil
Allows tracking of chromatographic
performance and mass accuracy
Replication
Triplicate injections of same sample vial
Removes risk of "one hit wonder"
Run order
randomization
8,	3,7,4,2, 1, 10,5,8,6, 9,2,5,4, 1,
9,	4, 7, 3, 8, 1,6, 10, 9, 6, 7, 5, 3, 2, 10
Minimizes/averages out batch or sample
order effects (e.g., carryover, temp &
instrument drift)
Pooled QC
sample
Combine 5 mg/pL from each of 10 samples (total
50 mg/pL) prior to extract to create pooled QC
Separate confirmation of presence with
different matrix, MS2 IDs
Blanks
Solvent, method, matrix, double blanks
Allows identification/subtraction/deletion of
interferences introduced in lab processes
Multiple lines of
evidence for ID
Retention time prediction/matching, Spectral
library/prediction matching, Data source ranking,
Functional/product uses, Media occurrence
Improves confidence in identification when
chemicals standards are unavailable
-------