The causes of cancer are likely to involve a combination of genetic factors, exposures, and random chance. However, current evidence suggests between a third and a half of cancers may be preventable, and that exposures may play an important role in the etiology of cancer. In fact, it is possible that the most common type of cancer among children, childhood leukemia, may be a preventable disease. Current scientific knowledge suggests early life exposures, including exposures occurring in utero, play critical roles in the development of childhood leukemia. However, few risk factors have been established, and the underlying disease mechanisms remain elusive.

The overarching goal of this dissertation research was to address this gap in knowledge by performing an untargeted adductomics analysis of archived newborn dried blood spots (DBS) to identify risk factors for childhood leukemia resulting from in utero exposures. Metabolism of chemicals derived from the diet, exposures to xenobiotics, the microbiome, and lifestyle factors (e.g., smoking, alcohol consumption) produce electrophiles that react with nucleophilic sites in circulating proteins, notably Cys34 of human serum albumin (HSA). Since HSA has a residence time of 28 days, HSA-Cys34 adducts measured in archived newborn DBS reflect systemic exposures during the last month of gestation and allow us to explore potential risk factors resulting from in utero exposures.

Here, we have developed an untargeted adductomics method to detect HSA-Cys34 adducts in 4.7-mm punches from DBS (equivalent to 5 – 8 μL of whole blood). The workflow includes extraction of proteins from DBS, measurement of hemoglobin to normalize for blood volume, addition of methanol to enrich HSA by precipitation of hemoglobin and other interfering proteins, digestion with trypsin, and detection of HSA-Cys34 adducts via nanoflow liquid chromatography-high resolution mass spectrometry. As proof-of-concept, we tested the DBS-adductomics method with 49 archived DBS collected from newborns whose mothers either actively smoked or were nonsmokers during pregnancy. A novel normalization method (‘scone’) was used to remove unwanted technical variation arising from: HSA digestion, blood volume, DBS age, mass spectrometry analysis, and batch effects. A total of 26 HSA-Cys34 adducts were detected, including Cys34 oxidation products, mixed disulfides with low-molecular-weight thiols (e.g., cysteine, homocysteine, glutathione, cysteinylglycine, etc.), and other modifications. Using an ensemble of variable selection methods, including both linear and nonlinear models, we found that the Cys34 adduct of cyanide consistently discriminated between newborns of smoking and nonsmoking mothers with a fold change (smoking/nonsmoking) of 1.31 and a cross-validated area under the estimated receiver operating characteristic curve (cvAUC) of 0.79. Indeed, hydrogen cyanide is a component of cigarette smoke, and these results indicated that DBS-based adductomics is suitable for investigating in utero exposures to reactive electrophiles that may influence disease risks later in life.

We then applied the DBS-adductomics method to analyze 783 archived newborn DBS collected from childhood leukemia cases and matched controls participating in the California Childhood Leukemia Study. Childhood leukemia cases included the two main subtypes, acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), with additional molecular subtypes for ALL. After data preprocessing and normalization, a combination of linear and nonlinear models were used to identify adducts that discriminated between leukemia cases and controls. Of the 28 Cys34 adducts that were detected, none was predictive of ALL overall. However, several adducts showed increases in adduct abundances in subgroups of cases with T-cell ALL and B-cell ALL with t(12;21) translocations. For both subgroups, elevated levels of adducts of reactive carbonyl species in cases may suggest that oxidative stress and lipid peroxidation had influenced adduct production. Regarding AML, the Cys34 homocysteine adduct (with loss of H2O) was found to consistently discriminate between AML cases and controls with a fold change (case/control) of 0.66. Since homocysteine is an important intermediate in the folate-mediated one-carbon metabolism, this indicates alterations in epigenetic regulations and folate status may be involved in the etiology of AML. Moreover, because lower homocysteine levels were detected in newborn DBS collected years before AML cases were diagnosed, biological changes involved in the initiation of AML may be present at birth and that there may be avenues for preventing the disease. Since this was a hypothesis-generating study, these findings warrant replication in follow-up studies with larger sample sizes of the various subtypes. Future integrated analyses with other omics (e.g., genomics, metabolomics, epigenomics) will be key to obtaining a full picture of the disease mechanisms regarding childhood leukemia.

Using data from the Northern California Childhood Leukemia Study (NCCLS) and the Fresno Exposure Study this dissertation shows that concentrations of chemical contaminants in residential-dust samples can be useful surrogates for indoor chemical exposures. This dissertation focuses on dust levels of four chemicals or classes which have been associated with childhood leukemia and/or developmental effects, namely polybrominated diphenyl ethers (PBDEs), polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAHs), and nicotine (a surrogate for tobacco smoke). Chapter 1 assesses the state of the science of residential-dust measurements, reviews global patterns in residential-dust levels of these chemicals, identifies known determinants of these chemicals' concentrations in residential dust, and estimates relative contributions of residential dust to the overall chemical intake of these chemicals in humans. Chapter 2 describes the analytical methods developed to measure PBDEs, PCBs, and PAHs in residential-dust samples. Chapters 3-6 compare residential-dust concentrations of these chemicals measured in California homes to levels reported in other studies from around the world. Chapters 3-5 also identify questionnaire-based predictors of residential-dust concentrations of nicotine, PAHs, and PCBs (chemicals for which sufficient data were available for statistical analyses). Chapter 7 investigates the variability of residential-dust levels of these chemicals within and between California households. A major finding of this work is the demonstration that current levels of nicotine, PAHs and PCBs represent indoor contamination from the distant past (e.g., over a period of years). This knowledge can be extremely useful to investigators who seek to perform retrospective assessment of exposures in studies of human health effects. The concluding Chapter 8 discusses the benefits and limitations of using residential-dust samples to estimate exposures to chemical contaminants, and presents ideas for future research in this field.

Chronic diseases such as cardiovascular disease, diabetes, and cancer are the leading causes of death among developed and developing countries, and account for approximately 75 percent of deaths worldwide. With the sequencing of the human genome and subsequent genomic studies, we now know genetic factors alone are responsible for a relatively small portion of these diseases. Specifically, cancer risk attributed to genetic factors is typically about eight percent. Thus, the vast majority of cancer risk likely lies within the realm of exposures (non-genetic factors) or a combination of genetic factors and exposures. The collection of exposures over an individual’s lifetime comprise the concept of the exposome, an epidemiological complement to the genome. The exposome is defined by measurement of both endogenous (inflammation, lipid peroxidation, microbiota) and exogenous (air pollutants, pesticides, drugs, diet, etc.) exposures within an individual.

Much exposure data is from non-individualized sources, such as air quality monitors or other spatial-temporal data, which have limited use in epidemiology. Individual exposure assessment consists largely of self-reported dietary and lifestyle data from interviews or questionnaires. In recent years, advances in analytical chemistry have permitted the simultaneous detection of thousands of molecules in biological fluids including urine, whole blood, plasma, and serum.

High resolution liquid chromatography mass spectrometry (LCMS) is a powerful technique to measure the accurate masses of molecules in biological fluids for high-throughput epidemiological studies. Chapter 1 of this dissertation details a method for the analysis of lipophilic molecules in plasma using specimens from 158 healthy volunteer subjects. The resulting data revealed levels of lipids and other molecules that differed between smoking and nonsmoking, white and black, and male and female subjects. A modified version of this LCMS method was used in the analysis of serum from subjects in a nested case-control study, described in Chapters 2 and 3.

Colorectal cancer (CRC) accounts for one fourth of all cancer deaths worldwide and less than about 15 percent of CRC risk is attributable to genetic factors alone. To investigate possible influences of exposures on CRC risk, serum from 190 subjects in the European Prospective Investigation of Cancer and Nutrition (EPIC) were extracted for lipophilic molecules and analyzed with high resolution LCMS. These prospective samples – collected up to 22 years prior to diagnosis - offered a unique opportunity to differentiate between CRC biomarkers related to disease causes and those that result from disease progression. Chapter 2 describes the testing of one class of lipids, ultra-long chain fatty acids (ULCFAs), that had been reported as a probable protective factor of CRC. Paired case-control differences were assessed with respect to the time period from when the serum was collected (study enrollment) to when the case was diagnosed. Since, case-control differences decreased with increasing time prior to case diagnosis, ULCFAs were likely depleted by cancer progression rather than by protective exposures.

Many of the features in LCMS profiling are unannotated (identity unknown) chemicals. Rather than relying on hypothesis-driven analyses of only known compounds, data-driven analyses of reliably detectable features can result in the generation of new hypotheses of possible disease-causing exposures. This untargeted methodology, used in Chapters 1 and 3, makes lipidomic and other exposure-related profiling a powerful tool in exposure assessment. In Chapter 3, the untargeted analysis of features in EPIC CRC serum samples revealed potentially relevant molecules associated with CRC causes and disease progression. As opposed to traditional p-value-centric analysis used in Chapters 1 and 2, a combination of regularized regression, random forest, and t-tests were employed in the feature selection for this untargeted analysis.

In Chapter 4, the lipophilic data from the healthy volunteer samples of Chapter 1 are studied once again. Using a method similar to the regularized regression technique described in Chapter 3, we determined which lipids were associated with levels of adductomic biomarkers (another methodology developed in our laboratory), which had also been measured in plasma from the same healthy volunteers. Analysis of the combined data from these two OMIC datasets found interesting correlations between particular lipids and adducts in these samples.

Aromatic hydrocarbons are ubiquitous pollutants. Many of the hydrocarbons such as benzene and benzo[a]pyrene are genotoxic and are known risk factors for cancers. The genotoxicity comes from the metabolic activation of the parent compound to a reactive electrophilic form, which binds to the nucleophilic DNA and albumin to form stable adducts. Because adduct represents the biologically effective dose and albumin has relatively long half-life (~ 1 month) and is abundant in blood (~45 mg/ml), albumin adduct becomes a promising candidate to assess exposure to aromatic hydrocarbons in cancer epidemiology. This dissertation can be roughly divided into two parts. In the first part, an antibody was used to quantitatively assess benzo[a]pyrene exposure. A sandwich enzyme-linked immunosorbent assay (ELISA), using a commercially available antibody, was developed to measure benzo[a]pyrene diol epoxide - albumin adduct (BPDE-HSA). The assay is much simpler and faster to use and is 10 times more sensitive than the traditional competitive ELISA to detect BPDE-HSA. The method was further revised to be directly compatible with plasma, hence eliminated all the tedious HSA isolation and purification steps. Finally, the revised assay was validated with two different sets of archived plasma samples. In the second part, another antibody was used in an enrichment step for the albumin adductomics. To characterize albumin adducts with liquid chromatography - mass spectrometry, they have to be digested and detected as peptide adducts. Cysteine (Cys34) is a known nucleophilic hotspot on the albumin that scavenges reactive aromatic electrophiles in blood (such as benzoquinone and naphthoquinones), therefore a semi-targeting antibody was designed to capture all peptides with electrophiles bound on the Cys34 (T3 adducts). After incorporating the antibody enrichment step into the adductomic workflow, results shown that the antibody enrichment has huge potential to increase the detection rate of low abundant T3 adducts. The major significance from this study is that the innovations are applicable to the exposure to other reactive chemicals. With a suitable capture antibody, the sandwich ELISA can be used to assess other exposures such as aflatoxin (in form of aflatoxin albumin adduct), while the semi-targeting antibody can be tailored to capture albumin peptide with a different locus of exposure significant, such as lysine.

Ambient air pollution and household air pollution are among the leading causes of mortality, responsible for 6 million premature deaths per year globally. The combustion of fossil fuels, including diesel and coal, constitutes a major source of air pollution. Polycyclic aromatic hydrocarbons (PAHs) are a group of compounds released during incomplete combustion of organic material which has been classified as carcinogenic to humans. Inhalational exposure to combustion emissions can be quantitated either externally such as by determining personal airborne PAH levels, or internally by determining the levels of a biomarker in the exposed population. Given the ease of sample collection, identifying suitable urine and blood biomarkers that accurately reflect exposure to air pollution is of significant interest. This dissertation examines the relations of the levels of biomarkers in urine and plasma with exposure determinants and with personal airborne measurements in humans exposed to emissions from diesel and solid fuel combustion. First, urinary PAH levels were measured using headspace solid-phase microextraction and gas chromatography-mass spectrometry in 28 nonsmokers in two studies before, immediately after and the day after short-term, controlled exposures to diesel exhaust (DE). Likelihood-ratio tests on linear mixed-effects models indicated that, after adjusting for other covariates, DE exposure did not significantly alter urinary PAH levels. Second, urinary PAH levels were reported in 163 Chinese female nonsmokers from Xuanwei and Fuyuan counties, where the highest lung cancer rates in the world are observed due to household air pollution from combustion of smoky coal. Mixed-effects models revealed that personal airborne PAH levels (over a wide range of exposure up to occupational levels experienced by coke oven workers), age and exposure characteristics such as fuel and stove types were significant predictors of urinary PAH levels. Third, using an untargeted nanoflow liquid chromatography and high-resolution mass spectrometry method, 50 adducts were detected at the nucleophilic Cys34 locus of human serum albumin (HSA) in 29 of the aforementioned Xuanwei and Fuyuan women and 10 local controls. Upon adduct annotation and quantitation, Wilcoxon rank sum tests showed that levels of S-glutathione (S-GSH) and S-γ-glutamylcysteine (S-γ-GluCys), adducts related to glutathione, an important antioxidant, were lower in exposed subjects compared to those in controls. After adjustment for age and personal measurements of airborne benzo(a)pyrene (a representative carcinogenic PAH) using multivariate regressions, the largest reductions in levels of S-GSH and S-γ-GluCys relative to controls were observed for smoky coal users, compared to users of smokeless coal and wood. These studies collectively demonstrate that urinary unmetabolized PAHs may be suitable biomarkers of wide-ranging exposures to fuel combustion emissions and that Cys34 adductomics is a promising tool for discovering biomarkers of longer-term exposures, owing to the one-month residence time of HSA.

This paper evaluates the decision by the Occupational Safety and Health Administration (OSHA) to base its Air Contaminants Standard on the threshold limit values (TLVs) of the American Conference of Governmental Industrial Hygienists. Contrary to the claim made by OSHA in promulgating the standard, the TLV list was not the sole available basis for a generic standard covering toxic air contaminants. The National Institute for Occupational Safety and Health (NIOSH) presented data indicating that the TLVs were insufficiently protective for 98 substances. NIOSH Recommended Exposure Limits (RELs) were available for 59 of these substances. The ratio of PEL to REL ranged up to 1,000, with a median of 2.5 and a mean of 71.4. OSHA excluded 42 substances from the standard altogether despite the availability of NIOSH RELs, solely because no TLV had been established.

Background

Methods

Results

Conclusions