Response to comment on: “Microplastic presence in dog and human testis and its potential association with sperm count and weights of testis and epididymis”

To the Editor, We have read the comments of Drs Uppu, Peijnenburg, and Hays on our paper “Microplastic presence in dog and human testis and its potential association with sperm count and weights of testis and epididymis.” We are grateful for their engagement on this topic, which clearly transcends the output of any single publication. As we continue to advance in the field of microplastics research, we recognize the complexity and technical challenges involved in detecting and quantifying microplastics, especially in biological tissues. To steal/modify a sentiment from the television show Ted Lasso, “[Bioanalytical assays] are never going to be perfect. The best we can do is to keep asking for help and accepting it when you can and if you keep on doing that, you’ll always be moving toward better.” There is no better example of this than the current world of microplastics research, where there is urgency to efficiently understand the nature of the problem, especially in human health where methods are lagging. But at the same time, micro- and nanoplastics are difficult to isolate and quantify, and the pursuit of precision is often at odds with the technical limitations of known approaches. Our intent in recent studies was to bring concepts of clinical biospecimen investigation into the realm of micro- and nanoplastics measurement. Hundreds of papers are published every year on controlled exposures to cells, animals, and plants, yet we have a very limited understanding of true body burdens in relation to exposure concentrations. We are not alone in seeking confident estimates of plastics in the human body, as numerous groups worldwide are striving for actionable, quantitative assays that will enable integration with health outcomes. Until this past year, assessments in human tissues were limited to visual spectroscopic methods (Fourier Transform infrared or Raman spectroscopy) that rarely permit visualization of particles less than 5 µm in diameter (including all nanoplastics). Such particle number/size data are challenging to link with metrics of pathology or risk of disease compared with mass concentration data. The authors of the Letter are correct that more information on the applied methods would be beneficial for the broader research community. However, the article was published as a Research Brief. Much of the details regarding tissue collections, digestion, and positive and negative controls were provided in the cited manuscript on plastics measurement in placentas (Garcia et al. 2024). Regarding the second point, the authors cite excellent papers that explore the validation of digestion methods, and they find that no perfect method exists currently. Some polymers (polypropylene, polystyrene) may be relatively easy to identify within a biological matrix, whereas others may be masked or augmented. Polyethylene (PE) is challenging as biological lipids may interfere with the PE pyrogram and create false positives. This was the main reason we chose the potassium hydroxide saponification method followed by ultracentrifugation, as this eliminated 99.0±0.9% of our original mass of sample. Although other methods with liquid–liquid extraction will increase partitioning lipids into the analyzed sample, our approach substantially removes biomatrix. Dawson et al. (2020) had notable success with this method. Similarly, Lykkemark et al. (2024) showed that polymers can be lost during the digestion process, and subsequent digestion/isolation steps can further reduce the final yield. Our approach aimed to significantly reduce the amount of matrix without risks of losing plastics through acid or organic solvent extractions, which could further impact the yield. Although we acknowledge that our approach is not perfect, it remains internally consistent and thus enables confident comparisons between different human and dog specimens. Notably, our digestion method rendered an average PE concentration of 121.0 µg/g in human testes samples and 69.9 µg/g in human placentas. A more detailed validation of digestion methods by Rauert et al. (2022) found that a chicken egg sample contained up to 240 µg/g PE and lamb meat contained 72 µg/g, whereas samples contaminated with lipids from an organic solvent liquid:liquid extraction method were far higher (>12,000 µg/g). Although we did not take the extensive validation conducted in that study, it is noteworthy that our human biospecimen results closely align with the values observed in human foods. Regarding visualization methods to confirm outcomes, we recognize that this is currently an unstated norm in the field, but our ongoing findings and other recent papers reveal a bias toward nanoplastic material (Fraissinet et al. 2024), which calls into question the value of characterizing particles larger than 5 µm. It has long been recognized in the air pollution field that nanoparticles have greater capacity to penetrate biological defense mechanisms. The bulk of our findings with TEM imaging of our isolated pellets reveal shard-like nanoparticles that are rarely longer than 200 nm (Campen et al. 2024). This finding is supported by a recent study of mussels, where a large mass concentration of plastics was observed in the 0 to 200 nm and 200 to 400 nm filtration bins (Fraissinet et al. 2024). Thus, although we agree that visual and spectroscopic confirmation adds to the weight of evidence, that area of research is far from perfect and likely not especially helpful for human health research. Similarly, in the inhalation toxicology field, although we have hypothesized that internalized particulates may drive health effects of air pollution or cigarette smoke, there are no viable approaches to count particles in biospecimens to link to health outcomes. Internal exposures are measured as single chemicals or metabolites, and visual detection has never been of clear value. Lastly, in the spirit of learning and growth, we would like to address one aspect of our approach that we will be changing to improve our outcomes, which is the idea that 12 polymers can be effectively measured in biological samples simultaneously. In reviewing our extensive data for testes, placenta, liver, kidney, and brain, we found that only 6 polymers provide sufficient confidence—which is correlated primarily with concentration—to merit inclusion in any health-based approach at this stage. These polymers—polyethylene, polypropylene, polyvinyl chloride, nylon, styrene-butadiene rubber, and polyethylene terephthalate—consistently represent 90% to 95% of the mass of plastics recovered from our samples and are typically achieve at least an 80% match for confidence in the FSearch and AMDIS pyrolysis GC/MS libraries. Although in our canine testes analysis we assessed linkages across fertility markers for all 12 polymers, future studies should more conservatively focus on polymers with high confidence or the sum of all high-confidence plastics. Based on the literature, it may not be possible to prepare samples with all polymers at once, as effectively removing all lipids may require secondary washes with organic solvents that remove soluble polymers. Furthermore, our knowledge of whether specific polymers may differentially drive health outcomes remain limited at this stage. It may well be that the physical presence of the plastics is the common toxicological issue, and the chemistry is of minimal concern—just as airborne particulate health effects are possibly influenced by their chemistry, but not so obviously as to drive a unique regulatory standard beyond the collective particulate matter (PM) 2.5. We thank the authors of the Letter and the Editor for providing an opportunity to clarify our methods and approaches in the article. The authors are correct that greater experimental details are required to move this research area forward. We remain committed to maintaining the highest scientific rigor in addressing the emerging environmental microplastics and health concerns. Microplastic concentrations were assessed in the Integrated Molecular Analysis Core of the UNM Center for Metals in Biology and Medicine (NIH P20 GM130422). Funding was also provided by the Academic Science Education and Research Training (K12 GM088021). This project was supported partially by pilot funding under NMINSPIRES-NIEHS (1P30ES032755). Conflicts of interest. None declared.