1. IntroductionThe prenatal period is an important stage of human development that is susceptible to toxic environmental exposures, including toxic metals and metalloids [1]. During pregnancy, metal(loid)s, such as lead (Pb), mercury (Hg), arsenic (As), and to a limited extent cadmium (Cd), can cross the placental barrier resulting in fetal exposure [2,3,4]. Metal exposure can occur through diet and drinking water as well as from anthropogenic sources including cigarette smoke, fertilizers, industrial emissions, as well as occupational sources [5,6]. In Mexico City, the primary sources of Pb exposure include air pollution and diet, partially through the use of traditional Pb-glazed ceramics used to prepare and store food [7,8]. Exposure to metal(loid)s during the prenatal period is associated with adverse birth outcomes and poorer childhood health, including low birthweight, risk of adverse respiratory symptoms, and higher blood pressure [9,10,11]. Additionally, metal(loid)s are nephrotoxic elements that can affect kidney development altering glomerular or tubular function, which are vital for regulation of blood pressure, excretion of drugs, and maintenance of electrolytes, water, toxicants, and nutrient homeostasis, among other functions [12].Increasing severity of kidney damage, assessed by markers of tubular injury such as increased urinary neutrophil gelatinase-associated lipocalin (NGAL) [13], and increased urinary kidney injury molecule-1 (KIM-1) [14], are associated with a greater risk of chronic kidney disease (CKD) and end stage renal disease [15,16]. Additionally, markers that are freely filtered by the glomerulus such as serum cystatin C and serum creatinine can be used to assess kidney function [17]. Urinary kidney injury biomarkers, which include proteins secreted by the tubules such as beta-2-microglobulin (B2M), or that escape the glomerular filtration barrier such as albumin, provide sensitive indicators of kidney injury and dysfunction that can occur in the presence or absence of a rise in traditional clinical markers such as serum creatinine [18]. Assessment of urinary proteins such as NGAL, B2M, KIM-1, albumin, and their combinations, offer improved sensitivity compared to traditional indicators in diagnosing kidney injury and have been suggested to improve the prognosis of CKD [18,19]. Further, these protein biomarkers can be collected non-invasively in the urine, and may better detect nephrotoxic insults, including to metals such as As, Pb, and chromium [20,21,22].Several studies have examined the association between environmental exposures with pediatric kidney function and kidney disease [20,23,24,25,26]. However, there is a paucity of research on in utero exposure to metals and metalloids, a time of dramatic growth and development when exposure may have a severe and long-standing impact on kidney function. Exposure to metals, including when assessed as a mixture, has demonstrated moderate positive associations with traditional indicators of pediatric kidney function such as estimated glomerular filtration rate (eGFR) [27]. Similar findings in animal studies showed that metal mixtures in drinking water (including As, Cd, vanadium, and Pb), even at low environmental levels impaired kidney development in zebrafish embryos at early stages of pronephros development, a developmental window comparable to ~3–4 weeks’ gestation in the human fetus [12,28]. In the human fetus, metanephric kidney development (formation of the permanent functional kidney) is initiated at 4 to 5 weeks’ gestation to the beginning of the second trimester, with nephrogenesis and the progression of tubular functions occurring from 6 to 36 weeks’ gestation, with nephrons continuing to expand and mature beyond 36 weeks [29,30]. The impact of environmental factors during this period may lead to a reduced nephron number, increasing the risk of kidney disease in later life [31]; therefore, due to the timing of kidney and nephron development in utero, and evidence in prior work [32,33,34], we selected the second trimester to examine potential susceptibility to metal exposure. This study aimed to assess the associations between prenatal metal exposure during the second pregnancy trimester, a relevant window of kidney development, with novel urinary kidney injury biomarkers and eGFR assessed at preadolescence (ages 8–12 years). 4. Discussion

Overall, we found that prenatal urinary metal concentrations, both individually and as a mixture, were associated with altered urinary kidney injury biomarkers measured in healthy children. These associations were predominantly observed for second trimester urinary As and Cd and null associations were generally observed with blood metal concentrations.

Urinary As and Pb concentrations in this Mexico City population were similar to As (geometric mean: 15.21 μg/L), and higher, respectively, than urinary Pb (geometric mean: 0.92 μg/L) in a birth cohort in Greece, which reported associations between prenatal metal exposure and elevated blood pressure throughout childhood [47]. Second trimester metal concentrations in this population were higher than those reported in 1283 pregnant women enrolled in the National Health and Nutrition Examination Survey which reported Cd geometric mean blood levels of 0.27 mg/L and Pb blood levels of 0.62 μg/dL. One study of 909 healthy children aged 10–18 years in Sri Lanka reported urine concentrations of kidney injury biomarkers [48]. The concentrations reported in the Sri Lanka study (NGAL (2.86 ng/mL) and KIM-1 (0.11 ng/mL)) were lower than the median concentrations of NGAL (8.24 ng/mL) and KIM-1 (0.45 ng/mL) in our study [48]. There is limited information on reference levels of kidney injury biomarkers among healthy children, and often, adult reference data have been used to generalize pediatric reference intervals [49].Evidence supports that both occupational and environmental exposure to individual metals and metalloids can lead to an increased risk of CKD and tubular indicators of dysfunction [15,20,22,50]. Exposure to Cd causes dysfunction of the proximal tubule in the kidney, which may result in increased urinary excretion of low-molecular-weight proteins, including A1M, B2M, and RBP4; as such, prior cross-sectional studies have reported associations of low-level Cd exposure with tubular indicators of kidney disease [51,52,53,54]. Among 222 healthy male sugarcane cutters in Guatemala, a repeated cross-sectional study conducted over one year reported that low urine Cd concentrations (median range: 0.09–0.14 μg/L) were associated with higher urine NGAL excretion, with the observed associations at Cd levels below those previously associated with renal injury in prior studies [55]. A study of 490 Chinese women aged 35–54 years reported that increasing levels of urinary Cd were significantly associated with markers of tubular renal effects, as indicated by increased urinary N-acetyl-beta-d-glucosaminidase (NAG) and B2M [56]. Similarly, we observed that a doubling of urine Cd was associated with higher urinary B2M as well as comprising 51% of the metal mixture weights contributing to the association for B2M. Several studies have reported that urinary B2M concentrations, measured alone or in combination with other glomerular and tubular analytes, can be used to detect Cd-induced renal dysfunction at early stages [57]. We also observed that a doubling of urine Cd was associated with higher albumin, A1M, and RBP4. These findings herein also suggest that renal tubules may be affected by prenatal exposure to Cd, even at low exposure levels. Additionally, we report that a doubling of urine Cd and Pb was associated with higher TIMP1. TIMP1 regulates extracellular matrix production and inhibits collagen degradation enzymes (which includes matrix-metallo-proteinases), resulting in the development of tubulointerstitial fibrosis and worsening inflammation [58,59]. These findings could be impactful for advancing the development of biomarkers for diagnosing CKD progression.Our findings suggest an association between combined exposure to As, Cd, Hg, and Pb as measured in urine with kidney injury biomarkers that is in line with previous research, although to our knowledge no prior study has examined mixed prenatal metal exposures. In this study, urine Cd was weighted as the largest contributor to the metal mixture index in identified associations with six kidney injury biomarkers (albumin, B2M, RPB4, EGF, clusterin, and TIMP1). A mixture of As, Cd, Pb, and Hg levels measured in urine was associated cross-sectionally with higher eGFR and urine albumin levels in 12–19-year-old children in the United States [60]. In this prior study the association between the metals mixture and urine albumin was also driven by Cd (37%) in a similar proportion to what we have reported (50%) [60]. Another cross-sectional study among 1435 adults in the United States aged 40 years or older found that exposure to metal mixtures (e.g., cobalt, chromium, Cd, Hg, and Pb) in blood was associated with indicators of worse kidney function [61]. The findings included suggest that combined prenatal metal exposures may lead to subclinical glomerular or tubular damage assessed by urinary proteins in the absence of worse eGFR. While some metals such as As, Cd, and Hg are directly toxic to podocytes in glomeruli [62], acute exposure to toxicants including Cd, Pb, and Hg can also occur via reabsorption in the apical membrane of the first zone of the proximal tubule, in addition to the loop of Henle, altering ion transport pathways with potential direct cellular toxicity [20,63,64]. Both acute and chronic toxicity can inhibit mitochondrial respiration and initiate apoptotic signaling cascades via the generation of reactive oxygen species [63,65]. Chronic exposure to toxic metals can lead to oxidative stress (via the depletion of glutathione or impaired metallothionein detoxification pathways) and inflammation that influence the progression of CKD or renal failure [65,66]. Yet, knowledge gaps remain in the developmental effects of metal(loid)s on renal function and maturation processes, as well as the cumulative effects in early adulthood.While traditional indicators of kidney function, including eGFR, serum creatinine, and albumin-to-creatinine ratio are used to diagnose acute kidney injury or CKD, newer biomarkers hold the potential of detecting renal damage at earlier stages. Urine biomarkers may be better able to predict renal function decline and CKD diagnosis than blood biomarkers, based on improved sensitivity and specificity [67], and urine collection is non-invasive, therefore more easily accessible in large population-based studies. The use of a single biomarker may not be ideal for predicting CKD progression because it may not fully characterize the complicated and compounded pathophysiological processes [68]. A multi-panel platform with biomarker specificity to nephron functional region (such as glomerular or specific tubular segments) may be more informative to determine critical sites of damage or treatment in renal insufficiency. Urinary proteomics or customized panels hold promise for biomarker discovery in this field [18].Our study had a few limitations. Prenatal metal concentrations and preadolescent kidney injury proteins were assessed at a single time point. We selected second trimester metal measurements to examine our hypothesis of exposure during a sensitive window of renal development; future studies may examine longitudinal kidney outcomes. Our assessment of urine proteins was limited to those on three pre-established panels of acute kidney injury. As with any observational study, we cannot rule out residual or unmeasured confounding due to unmeasured factors that could influence both prenatal metal exposure and protein concentrations in childhood. The timing of urine sample collection was not systematically recorded; however, the majority of urine samples were collected in the morning of each visit. We also did not speciate As or Hg metabolites which can vary proportionally by exposure sources, such as diet or geographic factors [69,70]. Toxicokinetic differences in metal and metalloid distribution, metabolism, and excretion also influence the selection of metal biomatrix and measured concentrations [71]. In this study we did not specifically account for the source of metal or the route of absorption which varies by study population. Lastly, as we did not identify a priori a single kidney biomarker as a primary outcome variable, we conducted multiple statistical testing on multiple biomarkers which may have increased the risk of false positive findings in our analyses. Our study also had many strengths. Metals were assessed in samples collected prenatally which enabled an assessment of the longitudinal relationship with kidney injury biomarker outcomes, thus limiting reverse causation bias that was of concern in previous cross-sectional studies [72]. The participants in PROGRESS are generally healthy with no history of clinical renal disease; thus, we did not anticipate directionality to be confounded by disease status. However, since our study population included relatively healthy women and children, it may not be generalizable to populations with CKD. This study also assessed metals in two biomatrices (urine and blood), which enabled a comparison of observed differences specific to each medium. Along with important covariates, we adjusted for hydration status. We further applied WQS, an established mixtures method, which allowed for an assessment of the joint effect of multiple metals on kidney injury biomarkers. By employing the WQS method, we accounted for collinearity among multiple predictors, as well as allowing for the detection of multi-metal contributors to the association with kidney injury biomarkers [73].