High-Confidence Metabolite Identification Maximizes Exposome Data to Shape Public Health Policy

5.0 Introduction

The last chapter touched on the importance of reliable metabolomics in exposome research. Here, we dive deeper into this topic by exploring the role of metabolite identification in the analysis and interpretation of exposome findings and how it impacts downstream decision-making. Biomonitoring plays a crucial role in understanding how exposure to chemicals, environmental factors, and other substances affect human health. The evidence generated from biomonitoring is often applied to public health policies and regulatory guidelines to avoid harmful exposures and poor health outcomes. Metabolite measurements are part of the evidence that regulators use to justify actions, and the interpretability of these metabolomics data rests on accurate identification of metabolites detected in biological samples. In this chapter we will discuss the levels of metabolite identity and criteria used to define them, as well as key instances where accurate metabolite identification led to policy changes that improved human safety by removing harmful chemicals from consumer products and the water supply.

5.1 Level 1 Metabolite Identification

Accurately identifying metabolites from untargeted profiling experiments is a complex and challenging process. The metabolome is comprised of numerous small molecules that encompass a wide range of chemical structures and ion features, and many of these metabolites exist as isomers, which have identical masses but different structures. Adding further complexity to metabolite identification is the reality that fragmentation spectra are highly specific highly specific to an LC-MS instrument and its parameters.

To promote best chemical analysis practices and maximize interpretability of metabolomics data, the Metabolomics Standards Initiative (MSI) has defined 5 levels of metabolite identification according to the available evidence that supports a biochemical’s identity (Figure 5.1).

5.1.1 Exposure to Bisphenol A

Bisphenol A (BPA) is an industrial chemical used to manufacture polycarbonate plastics such as baby bottles and protective coatings on the inside of food containers. Exposure to BPA is thought to occur mostly from ingesting food stored in BPA-coated containers. From the late 1990s to early 2000s several studies showed that exposure to BPA results in developmental and reproductive alterations in laboratory animals^3,4. This raised concern among environmental public health officials and prompted investigation into the concentrations of this compound in the U.S. population and potential impacts of its exposure on health.

In 2008, Calafat et. al. used isotope-dilution HPLC-MS/MS methods and Level 1 metabolite identification criteria to analyze urine samples from 2,517 participants in the 2003-2004 National Health and Nutrition Examination Survey⁵. They showed that BPA, as measured by free BPA and non-estrogenic metabolites of BPA, were detected in 92.6% of study participants. Follow-up epidemiological studies showed that adults with high BPA exposure were more likely to have coronary artery disease, diabetes, immune dysfunction, and liver enzyme abnormalities^6-8. Other studies reported association of high BPA levels with neurodivergent behavior in children¹⁰ and in utero exposure to BPA with low birth weight^10,11. These findings, together with toxicology and risk assessments, led to the 2011 decision by the European Union and China’s Ministry of Health to prohibit the manufacturing of baby bottles containing BPA^12,15. In 2012, this same policy change was implemented by the FDA for baby bottles and sippy cups sold in the United States^13,15. Since then, other Asian countries including Malaysia, Thailand, South Korea, and the Philippines have outlawed BPA in baby bottles, sippy cups, and infant formula¹⁵.

Given that BPA and its non-estrogenic metabolites share structural similarities, including two phenol rings linked by a central carbon bridge, Level 1 identification of those compounds was vital to generating data that could be compared across laboratories, institutions, and studies conducted years apart, to lead to an important policy change that will potentially mitigate harmful effects of BPA exposure on human health.

5.1.2 Exposure to Phthalates

Phthalates (i.e., plasticizers) are phthalic acid esters that impart plastics plastics with durability and flexibility. They are found in food packaging, pharmaceuticals, children’s toys, adhesives, and personal care products. Owing to their prevalence in consumer goods, steady levels of phthalates in the urine of children and adults have been detected since the 1970s¹⁶. Once ingested, phthalates are rapidly metabolized through a 2-step process (Figure 5.2). First, they are hydrolyzed into monoester metabolites by esterases and lipases. These monoesters can then be further metabolized through oxidation or conjugated with glucuronic acid, forming hydrophilic compounds that are readily excreted in the urine.

Recent studies have demonstrated a link between childhood exposure of certain phthalates to compromised health. For example, di-2-ethylhexyl phthalate (DEHP) and butylbenzyl phthalate (BBP) may increase the risk of asthma and eczema^17-20. Additionally, prospective studies from 4 different cohorts showed that gestational BBP, DEHP, di-butyl phthalate (DBP), and di-ethyl phthalate (DEP) exposures are associated with delays in the physical development of infants and toddlers^21,22 as well as parent-reported neurodivergent behavior^9,23.

In each of those studies, phthalates and their metabolic products were measured using validated targeted methods with isotope-dilution and reference materials to thereby achieve the criteria for Level 1 metabolite identification. High confidence in the biochemical species that were linked to concerning health outcomes, and the ability to compare data generated on different MS platforms in different laboratories, enabled a substantial amount of evidence on the harmful effects of those phthalates to be brought before regulatory authorities.

That evidence ultimately led to the European Union, the United States, and China limiting the concentrations of DEHP, DBP, and BBP to 0.1% by weight in children’s toys and childcare products^24-26. Overall, this example demonstrates how high-confidence metabolite identification is part of the bedrock of rigorous and impactful exposome research.

5.1.3 Exposure to Per- and Polyfluoroalkyl Substances (PFAS)

Per- and polyfluoroalkyl substances (PFAS) are a group of fluorinated chemicals widely used in industrial and commercial products including kitchenware, food packaging, clothing, and carpeting. They have been detected in the serum of several populations globally^27,28, and epidemiological studies have associated overexposure to PFAS with immunological health conditions³⁰, compromised fetal growth and development³⁰, and cancers^31,32 (Figure 5.3).

These clinical observations inspired significant efforts to understand how PFAS could drive the onset of those diseases. Many subsequent studies utilized validated LC-MS methods and biochemical standards for untargeted metabolic profiling. These studies showed that urea cycle/amino acid metabolism is affected by PFAS exposure, which has implications on insulin resistance and diabetes-related heart failure in older adults. Carnitines/acylcarnitines were also impacted by PFAS, showing impaired fatty acid oxidation and accumulation of lipids in tissue, which are major drivers of obesity and insulin resistance. Glycerophospholipid metabolism was another top-hit pathway associated with PFAS exposure, which has been previously linked to cardiovascular disease progression.

These studies, which identified the same PFAS-induced metabolic perturbations that could be linked to various diseases, along with biomonitoring and other mechanistic data, were influential in the recent wave of changes to national drinking water standards for certain PFAS. These new standards limit the amounts of the PFAS compounds PFOA, PFOS, PFNA, PFHxS, and HFOP-DA to 4 parts per trillion each. This example demonstrates the utility of untargeted metabolomics in characterizing cause-and-effect relationships between the exposome and health, and the value of accurate metabolite identification to biological interpretability and meaningful results.

As we mentioned earlier in this chapter, identifying exposome metabolites from untargeted profiling experiments is extremely challenging, and having robust analytical capabilities is essential to generating meaningful data. We note that Metabolon’s Global Discovery (i.e., untargeted) Platform has been validated against an independent targeted assay at Uppsala University⁴⁵, and demonstrated strong concordance in PFAS metabolite measurements, underscoring the platform’s analytical robustness. Such MSI Level 1 identifications provide the rigor and reliability necessary to support evidence-based policymaking for environmental exposures.

5.2 Conclusions and Metabolon’s Advantages

Level 1 metabolite identification is vital to generating rigorous and impactful findings that drive scientific discovery, and guide decision making for both public and individually based health concerns.

In untargeted metabolomics, a single MS1 peak can match multiple compounds in public databases because many metabolites share the same mass, and fragments can overlap between structurally related molecules. Most metabolomics providers report any potential database matches, which inflate their coverage numbers while leaving the investigator with a dataset containing ambiguous IDs, which undermines proper data interpretation. Furthermore, many providers do not distinguish which metabolites in a dataset were identified with Level 1 confidence from those identified with lower confidence.

Metabolon uses authentic biochemical standards, retention time confirmation, and biological plausibility filters to ensure each peak is assigned a single confident identification instead of multiple possible (i.e., ambiguous) annotations. Metabolon obtains evidence to support the identity of each metabolite, ranging from a biochemical standard (Level 1) to only molecular mass (Level 5), enabling accurate interpretation of the data and generation of robust hypotheses. Metabolon also has the largest library of Level 1 identified metabolites in the world, delivering industry-leading breadth and depth of metabolite coverage.

Metabolon’s library, coupled with our robust methods of annotation, gives our customers a decisive advantage over competitor metabolomics providers and arms them with the appropriate evidence to guide policy.