Pregnancy Exposome A New Frontier In Personalized Medicine
By Deborah Borfitz
December 13, 2022 | The importance of the pregnancy exposome on health was explored in depth by Barbara Cohn, Ph.D., director and senior research scientist at the Public Health Institute (Oakland, California), at the Mayo Clinic’s recent Individualizing Medicine Conference. She has been an integral part of a groundbreaking, multigenerational study examining the ripple effects of chemical exposures on pregnant women and their offspring and the potential for metabolomics to drive individualized prevention and treatment.
A leading expert in her field, Cohn’s research program focuses on how environmental chemicals affect reproductive health and exposure to certain chemicals in the womb influences a child’s risk of disease many years later. The beauty of the study approach is that three generations—the mother during the critical period when her body is adapting to pregnancy (F0s, now about 80 years old), the fetus whose organ systems are developing (F1s, now hitting 60), and the germline within the fetus that will create the grandchildren (F2s in their 20s and 30s)—are exposed to the same exposome but at different stages of development.
Her talk focused on the biology-dependent responses to those exposures, and the rationale and data sources for multigenerational studies. Cohn also presented outcomes from several targeted and untargeted exposome studies of breast cancer involving dichlorodiphenyltrichloroethane (DDT), an insecticide widely used in agriculture in the U.S. before being banned in 1972.
The “extra wrinkle,” as she points out, is that the choice of populations included in an exposome study matters a great deal. “Simply collecting hundreds of thousands of samples is not wise unless you stratify carefully.”
Pregnancy exposome studies being done today are possible because of an earlier study launched by Kaiser Permanente of Northern California that enrolled 98% of all pregnancies between 1959 and 1967, Cohn reports. The perinatal serum from those mothers is still available, together with 60 years of follow-up data across the three generations. The fourth generation (F3s) are just about to be born.
Cohn began by discussing work published in 2007 investigating DDT-associated breast cancer in mothers diagnosed before age 50. Researchers found a “dramatic decline in the relationship between DDT and breast cancer as the age of first exposure approaches the age of puberty,” she reports. “This is very important because... if your population contains the wrong birth cohorts you would not be able to see this association, and this is one of the reasons why we need to consider windows of susceptibility in environmental research.”
A further complication is that the exposure isn’t all that matters—so does the outcome window, which Cohn believes is a classic characteristic of endocrine-disrupting chemicals. The results changed when the outcome period was extended to age 54. For individuals first exposed to DDT in utero or between ages 3 to 13, the association between DDT exposure and breast cancer was strong (odds ratio of 5). People whose first exposure was post-puberty (ages 14 to 25) began to move into a period of risk while the group exposed in infancy through puberty were at risk in both periods.
The daughters were observed long enough to look at their risk of breast cancer in relation to the perinatal DDT exposure of the mothers, says Cohn. Here, researchers found that the o,p isomer, a contaminant of the commercial DDT mixture, was associated with an almost fourfold increased risk of breast cancer among mothers in the highest quartile of exposure.
In the most recent study in the granddaughters, the research team found that the grandmothers’ exposure to o.p’-DDT was correlated with early menarche and a doubling of their breast cancer risk if they were in the top tertile of exposure and this association wasn’t modified by the grandmother’s body mass index, Cohn says. These granddaughters were also at an almost threefold risk of obesity when the grandmother had a higher blood level of o.p’-DDT, but only when the grandmother was of normal weight.
“So, here we see the opportunity for interaction between the endogenous status—perhaps the metabolome, perhaps the way in which this chemical is handled across the placenta—that is partly a function of the body habitus of the grandmother,” she says. Again, the effects were seen only because the study samples were stratified.
In addition to early menarche and more obesity in the granddaughters (F2s), the multigenerational effects of DDT include higher breast density in the daughters (F1s) in the perimenopausal and premenopausal period, says Cohn. Collaboratively with the lab of Dean Jones, Ph.D. (Emory University), the Public Health Institute followed up on these findings with an untagged metabolome-wide association study (MWAS) to assess connections between metabolites and p,p’-DDT, o,p’-DDT, and p,p’-DDE levels (Reproductive Toxicology, DOI: 10.1016/j.reprotox.2019.05.059), which suggested a potential role for serine and fatty acid metabolism on the causal disease pathway.
“DDT is not just one exposure but a mixture in a commercial jar and it likes to become the metabolite DDE, which is highly persistent,” Cohn explains. Based on the variable pathways associated with the three chemical compounds in the study, it is probably not a good idea to combine them or have one stand in for the other to determine the “DDT effects” on health.
Findings of the MWAS “supports the epidemiology where most studies show absolutely no DDE associations with risk,” Cohn notes. Researchers thereafter found that p,p'-DDE was associated with increased levels of non-essential amino acids important in cell proliferation, especially in cancer. It also may be involved in fatty acid metabolism associated with mitochondrial dysfunction, toxic responses, and disease mechanisms.
While researchers at the Public Health Institute were looking at metabolic associations of p.p’-DDT on the F0 cohort, elsewhere scientists were conducting an experiment where mice were given the compound at a dose analogous to the human exposure levels—and succeeded in recapitulating its effects on the uterine cycle and metabolism, Cohn continues.
Researchers elsewhere have investigated global associations between the metabolome and various environmental exposures. It has consequently been posited that a small number of metabolic phenotypes may account for responses to a large class of environmental chemicals. “This is really exciting because it seems to suggest that we might someday be able to reduce this big mess to... [biomarkers of exposures based on] what people were exposed to in the past,” says Cohn.
It also brings science one step closer to understanding disease mechanisms that might be interrupted.
As it turns out, certain metabolic communities appear to correlate with chemical exposures, including per- and polyfluoroalkyl substances (PFAS) and DDT compounds, as well as each other, Cohn says. The compounds also share many of the same pathways.
Moreover, “the metabolic phenotypes within these pathways are shared between the pathways,” she continues, meaning a handful of them may account for the response for a large class of environmental chemicals. “The network model revealed that most metabolite communities are not specific to an association with a particular exposure,” and this is a promising method of analysis she credits to work done by Shuzhao Li, Ph.D. (Emory University and The Jackson Laboratory for Genomic Medicine). Li is now exploring whether the identified metabolic communities are also associated with breast cancer.
Meanwhile, in a preliminary study of the first 50 cases of breast cancer in the F1 daughters in relation to what was in the perinatal serum of the F0 mothers, Cohn and her team have found “significant associations” with higher levels of metabolites and certain environmental exposures. “This begins to show the power of these methods, [but] we have a long way to go,” she says.
Early, unpublished study results presented at the annual conference of the International Society of Environmental Epidemiology (Athens, Greece) in September suggests that in addition to significant metabolic disruption, exposure to methylmercury and/or organophosphate insecticide during pregnancy is associated with subsequent breast cancer risk. The work was led by researchers at Emory University and was a first for untargeted exposomics in pregnancy and maternal breast cancer, notes Cohn.
The background here is again the archived pregnancy samples collected decades ago by the Child Health and Development Studies in California, she points out. The search for the link between the pregnancy exposome and maternal breast cancer has primarily been classical epidemiological studies that have found evidence for third-trimester estrogens as well as DDTs and polychlorinated biphenyls (PCBs). “We have also found evidence for markers of placental function, including size, shape, rate of weight gain, and rate of blood pressure change [associated with] breast cancer risk in the mothers.”
Computation tools such as xMSanalyzer and liquid chromatography–mass spectrometry have also been employed to identify metabolic features associated with breast cancer in both second and third trimester pregnancies and environmental chemicals, she says. These are revealing novel environmental risk factors never measured before as well as potential pathways to breast cancer via the pregnancy estrogens that have been previously calculated.
Among the findings are that benomyl, a fungicide introduced in 1968 by DuPont, and polybrominated diphenyl ethers, a brominated compound used in the plastics industry in the 1950s and 60s (and as an additive flame retardant today) are significantly higher in breast cancer cases in both the second and third trimester serum, says Cohn. “We are still trying to sort all this out... looking for a predictor.”
Cohn and her colleagues have already published a paper looking at the three main forms of pregnancy estrogens—estrone (E1), estradiol (E2), and estriol (E3)—to see whether any environmental chemicals could be found in a matching quantity in the blood of pregnant women as it relates to breast cancer. Benomyl was positively correlated with these estrogen levels, as was an old chemical of concern already thought to be an estrogen disruptor.
The findings are interesting given that estrogen levels in pregnancy are highly regulated by the placenta and synthesized from androgens, she says. E3 production also is partially regulated by the fetal adrenal gland. “This raises the possibility that these chemicals are interfering with classic pregnancy levels of estrogen and potential exposures in both the fetal and maternal compartment.”
A MWAS of the endogenous metabolites for benomyl in the third trimester revealed numerous correlations with compounds that include fungicides with a history of interfering with the cell-division cycle and benomyl whose use is already banned due to concerns about its health effects, Cohn says. Moreover, benomyl was positively associated with pathways linked to inflammation in the third trimester, including increased prostaglandin, long-chain fatty acids, and disturbed lipid metabolism that may have a link to cancer.
“We are seeing the potential for these environmental chemicals and inflammatory metabolites in pregnancy that could be risk factors for breast cancer,” says Cohn. “They could operate through disordered placental development, altered estrogen synthesis of metabolism in pregnancy, and perhaps they are also entering the fetal compartment and disturbing normal development, including androgen control by the adrenal gland of the fetus, precursor for estriol, which is also correlated with [fetal] brain development.”
Old data remain relevant today, concludes Cohn, highlighting the ongoing obesity epidemic. “We are going to try to use this to understand mechanisms and discover biomarkers of risk and intervention targets using high-resolution metabolomics.” It’s a critical step in individualized prevention if not treatment and will require investment in multiple, long-term studies.