Dried blood spots for the identification of bio-accumulating organic compounds: current challenges and future perspectives

Abstract

The exposome is a concept that underlines the critical relationship between health and environmental exposures, including environmental toxicants. Currently, most environmental exposures that contribute to the exposome have not been characterized. Dried-blood spots (DBS) offer a cost-effective, reliable approach to characterize the blood exposome, which consists of diverse endogenous and exogenous chemicals, including persistent and bioaccumulating organic compounds. Current challenges involve prioritizing the identification by state-of-the-art mass spectrometry of likely up to tens of thousands of compounds present in blood; characterizing substances that represent a mixture of myriad constituent compounds; and detecting trace level contaminants, especially in quantity-limited matrices like DBS. This contribution reviews recent trends in DBS analysis of chemical pollutants and highlights the need for continued research in analytical chemistry to advance the field of exposomics.

1. Introduction

The exposome encompasses the totality of environmental exposures and over a lifetime^1,2. These exposures can be characterized using endogenous and exogenous chemical compounds that result from diet, lifestyle, radiation, stress, and environmental pollutants¹. Such factors likely play a greater role than genetics in the development of many chronic diseases, but the complex relationship between exposure and disease remains ambiguously defined at best. Unlike the genome, environmental exposures are dynamic and thus require frequent, cost-effective, and minimally invasive measurements to characterize variation over time.

In their pioneering study, Rappaport et al.³ explored the exposure-disease associations of 1,561 small molecules and metals derived from foods, drugs, endogenous processes and environmental pollutants in blood samples. The cumulative distribution of these compounds, see Figure 1, show that their blood concentrations span >11 orders of magnitude and that the concentrations of the pollutants are ~1000-fold lower than the remaining classes. This disparity has reinforced the view that the presence of drugs, food chemicals, natural toxins and other endogenous chemicals must be considered when assessing disease risks. The exogenous pollutants were primarily halogenated and exhibited common behaviour, such as environmental persistence, lipophilicity, toxicity and long-range transport potential^4–9. Since the early 2000s, signatories of the Stockholm Convention¹⁰ have imposed restrictions on the emission and production of over twenty-eight groups of these chemicals, coined persistent organic pollutants (POPs). The success of regulation is reflected by the steady decline in blood concentrations of some POPs^11,12. While continuing to monitor the effectiveness of regulating these compounds is important, one could argue that continued scrutiny of the impacts of dozens of regulated pollutants adds little to our understanding the etiology of disease, and that if we expect to reduce the burden of chronic disease, it is essential to develop methodologies that enable identification of unrecognized chemicals that potentially impact health¹³.

Figure 1.

The blood exposome³. Each curve represents the cumulative distribution of chemical species classified as pollutants (n=94), drugs (n=49), food chemicals (n=195) and endogenous chemicals (n=1223). Note that persistent organic pollutants, such as DDE (dichlorodiphenyldichloroethylene), PFOA (perfluorooctanoic acid), PCBs (polychlorinated biphenyls), PBDEs (polybrominated diphenyl ethers) and OCDD (octachlorodibenzo-p-dioxin) constitute a large fraction of the environmental pollutants. This figure was reproduced from Environmental Health Perspectives with permission from the authors.

Suspected POPs number in the thousands and few have been subject to environmental monitoring or epidemiological studies⁴. This realization was the outcome of seminal works by Howard and Muir⁶, Wania and Brown⁷, Strempel et al.⁸ and Scheringer et al.⁹, who employed predictive modelling to select potential POPs from among the c. 100,000 chemical substances produced in high volumes globally. Muir et al.⁴ recently performed a comprehensive evaluation of these data and assembled a list of 3421 chemicals that meet the same criteria as the Stockholm POPs, but have yet to be measured in the environment or human samples. The difference between the number of known and unknown environmental pollutants in the blood exposome may serve to reinforce the view that comprehensive exposure assessment to environmental pollutants must be considered when evaluating the risk of disease. Ideally, unknown environmental toxicants should be identified using an approach often referred to as “Nontargeted screening” (NTS)^1,5,13,14,15. Blood is a reservoir of endogenous and exogenous chemicals in the body^3,16. Thus, a small quantity of blood collected using dried blood spot sampling¹⁷ is an attractive matrix to sequence the exposome. However, Rappaport et al. posited that 90% of environmental pollutants with concentrations below ~0.1 μM in 50 μL of blood (Fig. 1, vertical dashed line)³ were not detectable using current analytical platforms designed for NTS, and thus a large gap in our knowledge of the exposome exists. Establishing the identities of environmental pollutants is the first critical step towards understanding their effects, which can depend on dose and may be revealed by changes in the concentrations of metabolites and biomolecules that reflect biochemical pathways^18,19. However, susceptibility to a chemical exposure can also vary over the course of a lifetime such that timing of the exposure can be more important than the dose, highlighting the need for frequent exposure measurements enabled by DBS screening²⁰.

This paper seeks to review DBS preparation and instrumental analysis; examine the critical barriers to further progress towards DBS screening of ultra-trace level contaminants; offer a forward-looking perspective as to how existing methodologies can be refined for the analysis of emerging pollutants; and highlight recent discoveries of persistent and bioaccumulating organic compounds that contribute to the blood exposome²¹.

2. Challenges

2.1. Blood spot sampling and storage

Dried blood spots^17,22–24 are obtained by pricking a finger or heel and collecting the blood onto filter paper²⁵. Difficulties can arise when collecting samples from neonates, manual laborers and at low temperatures due to vascular constriction²⁶. A circular punch of varying diameter (3–12 mm) is commonly taken from a DBS that has been collected on filter paper. (Note: a DBS with a diameter of 12mm corresponds to approximately 50 μL of blood). An inherent contamination risk exists from the skin surrounding the puncture site and during the drying time, which may vary between 15 minutes and many hours. Legacy POPs, such as organochlorine pesticides, polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) unsurprisingly have shown evidence of multiyear stability in DBSs during storage, although brominated substances may undergo photolytic Br loss if exposed to light²⁷. Blanks are essential to account for contamination during, transport and analysis²⁸ as indoor dust is a major source of exposure to persistent organics.

Other critical challenges of DBS collection include uneven spreading of the blood drop on the substrate; blood clotting during spotting; and non-uniform drying methods. Between patients, a different volume percentage of red blood cells, i.e. haematocrit, can affect the degree to which blood spreads: blood with high haematocrit will generally spread less than blood with lower haematocrit²⁴. Between samples from the same patient, these effects can be further amplified when multiple punches are collected from the same spot. Some studies have standardized punching procedures e.g. by punching spot centres and whole spot analysis has also been used to minimize haematocrit effects²⁹. Capiau et al.³⁰ have suggested that haematocrit bias may be corrected using non-contact diffuse reflectance spectroscopy.

Blood plasma, and not whole blood, is the preferred matrix for biomonitoring organic pollutants, but it is usually obtained by centrifugation, which complicates the logistics of sampling and requires larger quantities of blood. Hauser et al.³¹ have recently developed a dried plasma spot (DPS) device that provides a high yield of plasma in an exact volume, while also addressing issues related to haematocrit and spot-drying biases. The plasma is collected on a substrate that is enclosed, which may well prevent contamination from dust. Thus, the plasma spot approach is promising for future studies, but it needs further validation before adoption into biomonitoring studies. Moreover, the trove of archived DBS samples available for retrospective analysis will continue to drive method development aimed at addressing the limitations of sampling and analyzing DBS.

2.2. Sample analysis

Preparation of DBS samples²⁴ is typically performed by ultrasonic extraction²⁴. Handling and transferring small quantities of extract (20–50μL) can be laborious. Several novel methods introduced to address this issue have been applied to plasma extraction. Stir-bar sorptive extraction and similar methods that rely upon partitioning the analyte onto polymer substrates, like polydimethyl siloxane^32,33 and polyethersulfone¹³, have proven effective for the analysis of both polar and non-polar POPs. Non-polar POP concentrations in plasma are often normalized to lipid measurements, whereas DBS measurements are typically normalized to estimated whole blood volume. Estimating the volume of blood extracted from the blood spot is not trivial. The volume of blood present in the punch of a 12mm diameter blood spot is approximately 50 μL34, depending on the type and thickness of the filter paper^35,36. Sodium has been used as an indication of the plasma content of the spot and recently, Kadjo et al.³⁷ found that ring-disk electrode conductivity could be a useful nondestructive method of reducing errors associated with spotted blood volumes.

Gas chromatography-high resolution mass spectrometry (GC-HRMS) is among the most sensitive and selective techniques used for the analysis of semi-volatile organic pollutants. For the last 25 years, many biomonitoring studies have revealed a steady decline in the blood concentrations of regularly monitored legacy pollutants³⁸, such as PCBs, organochlorine pesticides, and (mixed) halogenated dibenzo-p-dioxins^39,40 and dibenzofurans (PCDD/Fs)¹². GC-HRMS is an important technique for the analysis of brominated POPs, such as polybrominated diphenyl ethers (PBDEs)⁴¹ and other halogenated flame retardants that have since replaced the PBDEs. However, liquid chromatography (LC) is more appropriate for thermally labile POPs like hexabromocyclododecane (HBCDD) and polar toxicants like some perfluoroalkyl substances (PFAS)^36,42–44. Most studies employed targeted experiments for the analysis. While the advent of Fourier Transform and Time-of-Flight mass spectrometers (FTMS and TOFMS) promises to accelerate the identification of unrecognized contaminants, these techniques are inherently less sensitive^34,51.

The small blood volume present in DBS samples requires high instrument sensitivity. Batterman et al.²⁷ explored the suitability of DBS for measuring concentrations of various pollutants by comparing the ratio of concentrations measured in the National Health and Nutrition Examination Survey (NHANES) with estimated limits of quantification. Those PCBs, PBDEs and OCPs that were not suitable for analysis in DBS were characterized by a median NHANES blood concentration ratio of <1, either due to low blood concentrations or high background contamination of the filter materials. If the volume of blood required surpasses the volume that is typically associated with one punch, the entire spot or several punches from the spot may be taken. Generally, the quantity of PCBs and PBDEs detected in a 12-mm diameter spot is on the order of 10⁻⁵ μM, whereas PFOA and PFOS concentrations are higher by one order of magnitude and dioxins are at least 10-fold lower^3,20. Fetal exposure levels are much lower than those of adults⁴³. This should be considered when assessing whether a protocol offers adequate detection limits⁴⁵.

3. Future Perspectives

Published biomonitoring studies reflect the impact of pollutants that have been previously identified to be of concern. There is growing interest in investigating emerging contaminants which serve as replacements for regulated chemicals. These emerging contaminants include flame retardants such as chlorinated paraffins ^46–48, dechlorane plus^49,50, its analogues, and organophosphate esters (OPEs)⁵¹. Dechlorane plus has emerged as a global contaminant, having been identified in Canada, the United States, China and Europe (despite the absence of production sources in Europe)⁵⁰. Chlorinated paraffins are highly complex mixtures found in human blood, but continued monitoring of this class of compounds (and their mixed halogenated analogues) will require the development of sensitive and selective analytical methods. PFOS and PFOA levels in human samples are generally declining¹¹, but their replacements⁵² are ubiquitously found in commercial products, food packaging and firefighting foams. The emergence of GenX is a worrying example of the trend to replace known toxicants with structurally similar compounds. While it has not yet been found in human samples⁴⁸, its presence has been detected in various environmental matrices, which suggests that GenX is widespread and continued exposure may lead to a rise in blood concentrations over time. The effect of this exposure has not yet been established. Computer modelling of chemical inventories as well as recent experimental studies of wildlife strongly suggests that there are likely hundreds more pollutants that have not yet been identified in blood^4,13.

In order to make further progress, the obstacle noted by Rappaport et al.³ will need to be surmounted, viz. that 90% of environmental pollutants with concentrations below ~0.1 μM in 50 μL of blood are not identifiable using modern FTMS¹³ and TOFMS⁴⁵ instruments. Comprehensive two-dimensional gas chromatography (GCxGC) and cryogenic zone compression (CZC) ¹² are elegant approaches to improve separation and detection of contaminants by GC-MS. For this approach, the column effluent is trapped using a device called a modulator, and subsequently re-injected into a secondary column. During this process, the eluting peak becomes narrow, due to cryogenic focusing, and the signal is enhanced as shown in Figure 2a. The contour plot in Figure 2b is the result of multidimensional separation, which occurs when one employs a secondary column with a stationary phase that is orthogonal to that of the primary column. A similar effect may alternatively be accomplished using flow modulation⁵³. CZC was originally conceived as a targeted approach, but Brasseur et al.⁴⁹ coupled it with full-scan TOFMS. This increased sensitivity towards DP in human samples and demonstrated, for the first time, the potential to collect signal enhanced full-scan mass spectra with femtogram level concentrations. Recent work⁵⁴ by the group of Schoenmakers suggest that compound detectability may be improved for LC-amenable compounds using an analogous LCxLC approach called active modulation. Though promising, the approach has not yet been applied to DBS or POPs.

Figure 2.

(a) Modulated gas chromatogram obtained from 2378-tetrachlorodibenzo-p-dioxin, resulting in 10-fold signal-to-noise (S/N) enhancement. Similar S/N enhancement can be achieved using either thermal or flow modulation; (b) Comprehensive two-dimensional gas chromatography (GCxGC) affords unparalleled separation of organic pollutants from in plasma.

Another challenge is related to recognizing which of the myriad signals detected from a blood spot corresponds to a suspected pollutant. It is impractical to comprehensively monitor all endogenous, exogenous and anthropogenic chemicals. One strategy is to employ suspect screening databases: Exposome-Explorer (v. 2.0) was a recently developed as an online database combining all available information (including the nature of the compound, their concentrations, study populations and analytical techniques) on exposure biomarkers and pollutants²¹. An alternative approach is to create analytical metrics that distinguish between those chemicals that persist and bioaccumulate from those that do not. Figure 3 displays 22,043 high-volume chemicals used in North America (in blue) along with the subset of these chemicals that are suspected of exhibiting POPs-like behaviour (in red). Since most known POPs are halogenated, it is unsurprising that the majority of suspected POPs (in red) cluster together according their respective mass-to-charge (m/z) and isotope ratios. This simple example shows that the environmental significance of a compound detected by mass spectrometry can be assessed based on MS measurements without beforehand knowledge of its structure or effect. In other words, an unknown compound detected in the region of Figure 3 that is densely populated by red dots may be assumed to have a high potential to persist and bioaccumulate. Indeed, the environmental and toxic effects of a chemical are typically prescribed by its structure, which in turn is reflected by the appearance of its mass spectrum⁵⁵. Exploring these relationships may well reveal opportunities for analytical and physical chemists to collaborate with and complement a diverse range of researchers in environmental health, with the goal of sequencing the range of exposures that contribute to the exposome.

Figure 3.

(a) The compositional space of chemicals of commerce⁵. Suspected POPs generally cluster together in regions bound by similar mass spectrometry metrics, e.g. mass and isotopologue ratios (A+2 and A-2) relative to the most abundant isotopologue; (b) Even polyfluorinated substance can be recognized by isotope ratio measurements without a priori knowledge of their structure. This figure was reproduced from Environment International with permission.

In summary, environmental exposures that comprise the chemical exposome remain largely unknown. DBS promises to be a cost-effective, reliable approach to characterize unrecognized or unknown pollutants. But, meaningful progress towards this goal will require a transition from targeted to non-targeted methodologies that are currently under development. To our knowledge, no study to date has attempted non-targeted screening of DBS and very few pioneering NTS studies have been performed with larger blood or plasma sample quantities^13,45. Online separation and pre-concentration techniques, such as CZC and GCxGC, may well be critical to addressing the challenge of identifying femtogram level pollutants in DBS.