Prenatal arsenic exposure interferes in postnatal immunocompetence despite an absence of ongoing arsenic exposure

Abstract

Arsenic (As) readily crosses the placenta and exposure of the fetus may cause adverse consequences later in life, including immunomodulation. In the current study, the question was asked how the immune repertoire might respond in postnatal life when there is no further As exposure. Here, pregnant mice (Balb/c [H-2^d]) were exposed to arsenic trioxide (As₂O₃) through their drinking water from time of conception until parturition. Their offspring, 4-week-old mice who had not been exposed again to As, were used for functional analyses of innate, humoral and cellular immunity. Compared to cells from non-As-exposed dam offspring, isolated peritoneal macro-phages (Mϕ) displayed no differences in T-cell stimulating ability. Levels of circulating IgG_2a but not IgG₁ were decreased in As-exposed dam offspring as compared to control offspring counterparts. Mixed-leukocyte reactions (MLR) indicated that CD4⁺ T-cells from the prenatal As-exposed mice were significantly less responsive to allogenic stimulation as evidenced by decreases in interferon (IFN)-γ and IL-2 production and in expression of CD44 and CD69 (but not CD25) activation markers. Interestingly, the Mϕ from the prenatal As-exposed mice were capable of stimulating normal allogenic T-cells, indicating that T-cells from these mice were refractory to allogenic signals. There was also a significant decrease in absolute numbers of splenic CD4⁺ and CD8⁺ T-cells due to prenatal As exposure (as compared to control). Lastly, the impaired immune function of the prenatal As-exposed mice was correlated with a very strong susceptibility to Escherichia coli infection. Taken together, the data from this study clearly show that in utero As exposure may continue to perpetuate a dampening effect on the immune repertoire of offspring, even into the early stages of postnatal life.

Keywords:

• Arsenic
• prenatal
• CD4+ T cell
• IL-2
• immunosuppression

Introduction

Arsenic (As) can induce a wide range of health effects and so is a major concern world-wide. As arsenic is reported to cross the placenta, the fetus is particularly vulnerable to toxic insult in utero (Concha et al. 1998; Davis et al. 2015). Immunotoxic effects of arsenic in early life have been widely studied in human and animal models. For instance, immunosuppression due to prenatal As exposure was identified (i.e. poor cellular immune function reported among cord blood cells) in human cohorts (Nadeau et al. 2014). Impaired function of T-cell subsets has also shown to be associated with prenatal As exposure. The presence of maternal cells in the cord blood limits the observation to be conclusive on immunosuppression in early childhood.

Substantial childhood morbidity that has been documented in As-polluted areas is probably due to compromised immune status. In utero exposure to As has been reported to reduce CD4/CD8 ratios in offspring (Soto-Pena et al. 2006). In preschool students in rural Bangladesh, As exposure appeared to cause reductions in cell-mediated immunity and in formation of T-helper (T_H) Type-1 cytokines (Raqib et al. 2017). Other studies reported alterations in fetal thymic function induced by oxidative stress and apoptosis (Ahmed et al. 2012). Such changes in immune function/thymic development have a potential to increase host susceptibility to infection, as well as to autoimmune diseases/chronic diseases in childhood and also later in life.

The prevalence of infectious diseases has been reported to be increased dramatically in As-impacted regions of the world (Ferrario et al. 2016; Gera et al. 2017). Chronic low-level As exposure has been reported to increase host risk of lung infections, influenza, diarrhea, etc. (Raqib et al. 2009; Rahman et al. 2011; Smith et al. 2012, 2013; Farzan et al. 2016), especially among children. While in many cases these outcomes might be associated with ongoing As exposure of the children, it should be recalled that As exposure can affect fetal development (Vahter 2008; Kippler et al. 2016; Attreed et al. 2017) and even cause long-term epigenetic re-programming of the fetus (Bailey and Fry 2014). One epidemiological study showed an association between in utero As exposure and development of adult diseases (Young et al. 2018). If one of the results of these changes was alterations in child immunocompetence, this might mean these hosts could be impacted even if they were no longer exposed to As in their environments.

A few studies have indicated that As exposure is associated with changes in the immune system (Ferrario et al. 2016). In various models, direct As exposure caused suppressed immunoglobulin production (Selgrade 2007), decreased cytokine expression (Conde et al. 2007), defective antigen-driven T-cell proliferation, and reduced macrophage reactive oxygen species generation (Vega et al. 1999). It was also shown that As exposure in vitro can reduce IL-2 secretion by lymphocytes (Vega et al. 1999; Galicia et al. 2003). One mechanism proposed to explain this observation was that there is a delayed activation of T-cells due to As-induced accumulation of IL-2 in these cells (Galicia et al. 2003). There is also mounting evidences of an increased risk of infections due to prenatal As exposures (Ramsey et al. 2013). Thus, it is evident from the literature – from data derived from different experimental conditions and in different mammalian species – that all three arms of the host immune repertoire, i.e. innate, humoral, and cellular, can be affected by As exposure. While most of this data is based on de facto As exposures, less is known about what may happen as a result of in utero exposures to As.

In the investigation reported here, all three arms of immune system i.e. innate, humoral, and cellular, were evaluated to better understand how/if in utero As exposure influences the postnatal host immune repertoire. To address this question, As was given via drinking water to pregnant Balb/c mice and their resulting pups (who underwent no further As exposure) were examined.

Materials and methods

Reagents

Arsenic trioxide (As₂O₃) was purchased from MP Biomedicals (Irvine, CA). Concanavalin A (ConA), and most other general reagents were procured from Sigma (St. Louis, MO). Biotin-conjugated anti-mouse IgG_2a and anti-IgG₁ were also procured from Sigma. Cell culture reagents and fetal bovine serum (FBS) were purchased from Thermo Fisher Scientific (Waltham, MA) and Sigma. ELISA kits used for analyses of IL-2 and IFNγ were purchased from BD Biosciences (San Jose, CA). PerCP (peridinin-chlorophyll-protein)-conjugated-anti-CD4, APC (allophycocyanin)-anti-CD8, FITC (fluorescein isothiocyanate)-anti-CD25, APC-anti-CD44, FITC-anti-CD69, FITC-anti-CD11b, and PE (phycoerythrin)-anti-I-A^d were purchased from eBiosciences (San Diego, CA).

Animals

Male and female Balb/c (H-2^d) and C57BL/6 (H-2^b) mice (all 6-week-of-age) were procured from the ICMR-National Institute of Cholera and Enteric Diseases (Kolkata, India). All mice were housed in cages containing straw bedding held in pathogen-free facilities maintained at 24 °C with a 50% relative humidity and 12-h light:dark cycle. All mice had ad libitum access to standard rodent chow. After 2 weeks of acclimatization, the Balb/c (H-2^d) mice were bred by housing two females with a male. Once pregnancy was confirmed, the dams were given ad libitum access to drinking water containing 4 ppm As as described in He et al. (2007). The As-containing water was changed twice weekly. After birth, the mothers were then given ad libitum access to clean As-free water. For the experiments, when pups reached 4 week-of-age, groups were randomly collected, and processed for biomaterials. Age-matched juvenile mice whose mothers were never exposed to As were processed in parallel as controls. For each experiment, 5–6 juvenile mice were randomly chosen (without any sex bias) for evaluation (Aung et al. 2016).

All protocols were approved by the Institutional Animal Ethics Committee of ICMR-NICED (PRO/151/July 2018–June 2021). All experiments were done in accordance with the guidelines set by the committee for the purpose of control and supervision of experiments on animals (CPCSEA; Ministry of Environment and Forests, New Delhi, India).

Histopathologic examination of offspring spleen and thymus

At 28 days-of-age, after their body weights were recorded, five mice in each group were selected, blood was drawn from their tail vein, and then they were euthanized by cervical dislocation. At necropsy, both macroscopic observations and weights of the spleen and thymus from each mouse were taken. Each organ was then fixed in 4% paraformaldehyde for 48 h at 4 °C. The fixed tissues were then dehydrated through graded alcohols, embedded in paraffin, and routine microtomy then carried out to generate 5-µM sections. The sections, in turn, were stained with hematoxylin and eosin for later microscopic examination.

Analysis of serum levels of IgG₁ and IgG_2a

Each collected tail vein blood sample was allowed to stand for 3 h at room temperature and then serum was isolated by centrifugation at 1800 rpm. Serum samples were then analyzed to determine total IgG₁ and IgG_2a titers. For this, an aliquot of each sample was diluted 1:10, 1:100, and 1:1000 with PBS (phosphate-buffered saline, pH 7.4) + 10% FBS solution. For each paired sets of diluted sample, an aliquot of the diluted serum was added to a 96-well plate whose wells had been pre-coated with biotin-conjugated mouse anti-mouse IgG₁ or mouse anti-mouse IgG_2a. The plates were then incubated for 1 h at room temperature before unbound serum components were removed by repeated washings with PBS (and centrifugation). Detection reagent (avidin-conjugated horseradish peroxidase) was then added to each well and the plates were incubated a further 1 h. The absorbance in each tube was then evaluated at 450 nm in an iMark™ Microplate Absorbance Reader (BioRad, Hercules, CA). Results were reported in terms of mean OD values. Each mouse sample was analyzed in triplicate.

Isolation of peritoneal exudate macrophages (PEMϕ)

For use in the MLR assays outlined below, sets of five 28-day-old Balb/c (As-exposed and unexposed) as well as naive C57BL/6 mice were each injected intraperitoneally (IP) once with 3 ml of a 4% (w/v) starch solution in water. After 48 hr, PEMϕ were collected by peritoneal lavage, pelleted, re-suspended in RPMI medium supplemented with 10% FBS, 100 U penicillin/ml, and 100 μg streptomycin/ml (the latter two from Gibco [Waltham, MA]), and then seeded into 24-well plates at 5 × 10⁴ cells/ml (0.5 ml/well) (Zhang et al. 2008). The cells were then cultured for 48 h at 37 °C in a humidified 5% CO₂ incubator to dampen/mitigate any residual effects of the starch. Non-adherent cells were removed thereafter by gentle washing of the wells with serum-free medium. The remaining adherent cells were then collected by gentle scraping with a plastic scraper for use in the MLR reactions (see below).

One-way mixed-leukocyte reactions (MLR)

Allo-antigens on antigen-presenting cells (APC) can be recognized by T-cells and studied in mixed-leukocyte reactions (MLR) using methods first defined by Lause et al. (1976). For the current study, Balb/C and C57BL/6 mice (MHC disparate [allogenic]) T-cells were utilized. In one sub-study, PEM from the 28-day old Balb/c hosts were cultured with allogenic normal CD4⁺ T-cells of C57BL/6 mice. Specifically, PEMϕ derived from the Balb/c mice (As-exposed and unexposed; see above) were treated in their wells with mitomycin c (10 μg/ml) for 6 h, and then washed thoroughly with complete RPMI medium (Sui et al. 2017). Thereafter, each well received aliquots of allogenic normal splenic CD4⁺ T-cells from C57BL/6 mice (at a ratio of 1:5; isolation of cells outlined below) and then were incubated for 72 h at 37 °C in a humidified CO₂ incubator. At indicated timepoints over this period, samples from each well were collected for later measures of IL-2 in order to assess CD4⁺ T-cell activation. In a parallel sub-study, naive C57BL/6 PEMϕ were cultured with allogenic CD4⁺ T-cells from the As-exposed or unexposed Balb/c mice. As above, the C57BL/6 PEMϕ were first treated with mitomycin c and then incubated with splenic CD4⁺ T-cells from the As-exposed or unexposed 28-day-old Balb/c mice. As a control, C57BL/6 PEMϕ were cultured with syngeneic T-cells in parallel. Culture samples were collected during the 72-hr period to permit measures of IL-2.

Analysis of splenic T-cell populations

For analysis of potential effects from maternal exposure to As on their T-cell populations, at 28-days-of-age, sets of As-exposed and unexposed mice were euthanized. At necropsy, each had their spleen aseptically removed and processed to generate single cell suspensions for use as indicated below (Ferrario et al. 2016). The numbers of nucleated cells were then counted using a hemocytometer. To estimate splenic levels of CD4⁺ and CD8⁺ T-cells, aliquots containing 10⁶ splenocytes/mouse were stained for 0.5 h on ice with a solution of PerCP-anti-CD4 and APC-anti-CD8 monoclonal antibody (1:500 dilutions) in PBS. The cells were then immediately analyzed in a FACS Aria II system using FACSDiva software (both Becton Dickinson, San Jose, CA). A minimum of 10,000 events per sample was acquired. Absolute numbers of splenic CD4⁺ and CD8⁺ T-cells were then determined as described in Helmby et al. (2000).

For their use in the MLR above and for analysis of activation marker expression, splenic CD4⁺ T-cells were isolated from among the starting sets of splenocytes (from Balb/c as well as from the naive C57BL/6 mice) using an Easy Sep CD4⁺ T-cell enrichment kit (Stem Cell Technologies, Vancouver, Canada), following manufacturer protocols. For the analyses of T-cell activation markers, parallel sets of purified CD4⁺ T-cells from the Balb/c mice were incubated with allogenic PEMϕ (5:1 ratio). After 24 hr of incubation, the T-cells were collected, and then stained and analyzed for CD44 and CD69 expression by flow cytometry (as above). In another experiment, the co-incubations were allowed to proceed for 72 h before the T-cells were collected, stained, and analyzed for CD25 expression by flow cytometry.

Host resistance to infection

To assess the impact of the prenatal As exposures on the intact host immune system, additional sets of control and prenatal As-exposed Balb/c mice (n = 3/group) were injected IP with a potent strain of Escherichia coli and then levels of the bacteria were assessed in several immunologically-active sites in the body (i.e. spleen, liver, peritoneal cavity). For the assay, E. coli strain E14 were grown overnight at 37 °C in Luria Bertani broth. After determining their concentration by optical density (derived from turbidity at 600 nm), the bacteria were washed by centrifugation and re-suspended in PBS for subsequent IP injection of 10⁸ E. coli (in 100 μl PBS). At 24 h post-infection, each mouse was euthanized by cervical dislocation, and their peritoneal cavity was rinsed with PBS (2 ml) that was in turn placed on ice. The liver and spleen were then aseptically removed and dissociated in PBS using a tissue homogenizer. Aliquots of each homogenate and of the peritoneal wash (20 µl wash/mouse) were then plated onto LB agar plates for estimation of bacterial load after overnight incubation at 37 °C. All data are reported in terms of absolute counting colony forming units (CFU) per g organ or per ml peritoneal wash.

Statistical analysis

All data (apart from contour plots) are reported as mean ± SE. All statistical analyses were performed using Prism-5 Software (GraphPad, San Diego, CA). All data were analyzed using a nonparametric t-test with 95% confidence intervals. A p < 0.05 was considered significant.

Results

Prenatal As exposure effects on body weight and thymic/splenic weight and histology

In this study, arsenic (As₂O₃) was given to the dams in drinking water during the length of their pregnancy (Figure 1(A)). The data showed that the As-exposed pups did not differ in body weight from control pups (Figure 1(B)), nor did weights of their spleen or thymus differ (Figures 1(C,D)). Histologic examination of each organ revealed no gross changes (Figure 1(E)).

Figure 1. Effect of prenatal As exposure on pup body weight, and splenic/thymic weights and histology. (A) Schematic diagram of experiment. (B) Body weights, (C, D) spleen and thymus weights, and (E) representative micrographs of the spleen and thymus of control and As-exposed offspring (H&E; 20× magnification). All samples were collected from mice at 28 days-of-age (n ≥ 7/group).

Prenatal As exposure effects on circulating levels of IgG_2a and IgG₁

To assess potential shifts in future humoral immunity-related events, circulating levels of IgG_2a and IgG₁ in the As-exposed and unexposed Balb/c offspring were evaluated. No differences in IgG₁ status was noted between the groups (Figure 2(A)). However, IgG_2a levels were decreased significantly in the prenatal As-exposed Balb/c pups as compared to levels seen with control pups (Figure 2(B)). IgG_2a levels were ≈30% lower due to maternal exposure to As.

Figure 2. Effect of prenatal As exposure on circulating IgG levels. (A) IgG₁ and (B) IgG_2a in sera (dilution 10⁻³ presented) collected at 28 days-of-age. Data shown is representative of five mice/group (n = 5/group). Values shown are mean ± SE (OD). Value significantly different from control at *p < 0.05, **p < 0.01, ***p < 0.001.

Prenatal As exposure effects on splenic CD4⁺ and CD8⁺ T-cells levels

The frequencies of splenic CD4⁺ and CD8⁺ T-cells were evaluated in the mice. The data show that the absolute number of nucleated splenic cells remained unaltered as a result of the prenatal As exposures (Supplemental Table 1). In contrast, the absolute numbers of splenic CD4⁺ and CD8⁺ T-cells were decreased significantly in the pups whose dams underwent As exposure (Figure 3(A)). Specifically, there was a ≈ 20% reduction in CD4⁺ T-cells (Figure 3(B)) and a 5% reduction in CD8⁺ T-cells (Figure 3(C)) vs. levels in control pup spleens.

Figure 3. Effect of prenatal As exposure on splenic CD4⁺ and CD8⁺ T-cell frequencies. Splenocytes from individual mice were collected at 28 days-of-age, stained, and CD4⁺ and CD8⁺ populations evaluated using flow cytometry. (A) Representative dot-plots of CD4⁺ and CD8⁺ T-cells in control and prenatally As-exposed offspring. Percentage (B) CD4⁺ T-cells and (C) CD8⁺ T-cells (n = 5/group). Value significantly different from control at *p < 0.05, **p < 0.01.

Effects of maternal As exposures on offspring ex vivo mixed-lymphocyte reactions (MLR)

Mitomycin c-treated PEMϕ from As-exposed or unexposed Balb/c offspring were used to stimulate allogenic T-cells (from naive C57BL/6 mice) in a one-way MLR. After 48 h of co-culture, wells were sampled and IL-2 levels subsequently assessed. It was seen that there was no significant effect on IL-2 production from the T-cells regardless of PEM ϕ source (Figure 4).

Figure 4. One-way MLR. PEMϕ from exposed or normal Balb/c mice and allogenic C57BL/6 mice naive T-cells were used (MLR-I). PEMϕ either from As-exposed or normal Balb/c mice collected at 28 days-of-age were pooled, treated with mitomycin C (10 µg/ml) for 6 h, washed, and then combined with allogenic naive C57BL/6 splenic CD4⁺ T-cells at a ratio of 1:5 PEMϕ:T-cells. Cells were cultured for 48 h at 37 °C and supernatant was then harvested for IL-2 analysis. Values shown are mean (pg/ml)±SD. Value significantly different from control at **p < 0.01. All experiments were performed three times using pooled cells from each group (n = 5/group).

In the second MLR, splenic CD4⁺ T-cells from As-exposed or unexposed Balb/c offspring were incubated with PEMϕ from allogenic naive C57BL/6 mice for 48 hr and then levels of T-cell-derived IL-2 and IFNγ were evaluated. The results show allogenic stimulation caused significant IL-2 and IFNγ production compared to levels from syngeneic co-cultures. Allogenic stimulation of T-cells from prenatal As-exposed mice resulted in ≈50% less IL-2 production compared to that of T-cells from control pup counterparts (Figure 5(A)). Similarly, there was a 30% reduction in IFNγ production from As-exposed offspring T-cells compared to that of control pup T-cells (Figure 5(B)).

Figure 5. Cytokine production in one-way MLR. CD4⁺ T-cells from normal and As-exposed Balb/c offspring were used as responder cells and naive C57BL/6 PEMϕ as stimulators (MLR-II). CD4⁺ T-cells were purified from spleens of the mice at 28 days-of-age and then pooled. PEMϕ were isolated from C57BL/6 mice, treated with mitomycin C (10 µg/ml) for 6 h, washed, and then combined with the Balb/c CD4⁺ T cells derived at a ratio of 1:5 PEMϕ:T-cells. A control assay was performed in parallel using CD4⁺ T-cells from each group in combination with syngeneic PEMϕ (Balb/c). After 48 h, culture supernatants were collected to permit analyses of released cytokines. Results are represented for both syngeneic and allogenic cultures. (A) IL-2. (B) IFNγ. Values shown are mean (pg/ml)±SD. Value significantly different from control at *p < 0.05, **p < 0.01. All experiments were performed three times using pooled cells from each group (n = 5/group).

Prenatal As exposure effects on splenic T-cell activation marker expression

In parallel with the MLR, expression of T-cell activation markers CD25, CD44, and CD69 were evaluated in the harvested splenic T-cells. Expression of each marker was reported in the context of both syngeneic and allogenic co-cultures. The data showed that when control T-cells were mixed with allogenic PEMϕ, while there were significant increases in CD44 and CD69 expression on the T-cells, this change was reduced on prenatal As-exposed offspring cells (Figure 6(A,B)). Accordingly, the frequency of double positive CD44⁺CD69⁺ cells was also decreased significantly (∼20%) as compared to in normal counterpart offspring T-cells (Figure 6(C)). In contrast, CD25 expression remained comparable between the two groups (Figure 6(D,E)).

Figure 6. T-cell activation in one-way MLR. CD4⁺ T-cells from normal or As-exposed offspring were used as responder cells and naive C57BL/6 PEMϕ served as stimulator cells (MLR-II). Protocols followed those outlined in Figures 3 and 4, using cells collected from mice at 28 days-of-age. Here, after 24 h of co-incubation, cells were analyzed for CD44 and CD69 expression; after 72 h, CD25 expression was assessed. (A) CD44⁺ and CD69⁺ expression in syngeneic and allogenic reactions (contour plot). (B) Representative histograms of stimulated CD44⁺ and CD69⁺ expression. Red = syngeneic reaction, green = allogenic stimulation. (C) Percentage CD44⁺CD69⁺ T-cells in syngeneic and allogenic co-culture. (D) CD25⁺ expression in syngeneic and allogenic reactions (contour plot). (E) Percentage CD25⁺ T-cells in syngeneic and allogenic reactions. Values shown are mean ± SE (n = 3). Value significantly different from control at *p < 0.05, **p < 0.01, ***p < 0.001. All experiments were performed three times using pooled cells from each group (n = 5/group).

Prenatal As exposure effects on host response to infection

To obtain a more “global” evaluation of host immunocompetence, host resistance to a pathogenic challenge was evaluated among pups in both groups. Mice were infected IP with 10⁸ viable E. coli (strain E14) and then organ parasite load was enumerated 24 h later. Bacterial load was expressed as cfu/g liver (Figure 7(A)), cfu/g spleen (Figure 7(B)), or as cfu/ml peritoneal wash (Figure 7(C)). In general, bacterial loads were significantly higher (∼20%) in the prenatal As-exposed pups (all sites) as compared to in the respective site in their normal control counterparts.

Figure 7. Effect of prenatal As exposure on susceptibility to septicemic E coli infection. At 28 days-of-age, mice in each group were infected intraperitoneally with 10⁸ septicemic E. coli strain E14 and then evaluated at 24 h post-infection. (A) Liver and (B) spleen were each isolated, homogenized, and plated to permit colony-forming unit (cfu) estimations. (C) Peritoneal cavity lavages were also plated. E. coli cfu in each organ or cavity lavage are reported as mean ± SE (n = 3 per group; cfu/g or cfu/ml). Value significantly different from control at *p < 0.05, **p < 0.01.

Discussion

There is mounting evidence describing the detrimental effects of arsenic (As) exposure on human health (Abdul et al. 2015). Studies in experimental animals have also yielded evidence of these effects (Santra et al. 2000). In humans, the durations of exposure and levels of As in the drinking water often dictate the breadth of ill effects induced. Many of these effects also seem to impact on children who may have been exposed in utero or as infants/neonates/young children. One study in children with varying levels of As exposure in Bangladesh found that total white blood cell (WBC) levels were most significantly reduced among those children who had the highest As exposures (Saha et al. 2013). Similarly, animal studies have shown that early-life As exposure gave rise to subsequent decreases in host resistance to influenza (Ramsey et al. 2013).

The World Health Organization (WHO) has indicated that a “safe” level of As in drinking water is 10 µg/L or 0.01 ppm for adults. In the experiments here, much higher levels of arsenic trioxide (As₂O₃) in drinking water were employed so as to mimic a scenario of highly exposed local populations in Bangladesh. Studies using animal models of As exposure have utilized different approaches to assess immunotoxicities induced by the metalloid. A prevailing approach to assess alterations in the immune repertoire caused by As exposure was to examine changes in disease susceptibility; Farzan et al. (2013) found that As exposures of young rats resulted in increased susceptibility to pathogen-based respiratory diseases. A study of preschool children in rural Bangladesh showed that repeated As exposures (in water) led to altered BCG vaccination efficacy, i.e. BCG-specific CD4⁺ T-cells that failed to respond to PPD, the antigen that had been seen earlier by the hosts (Ahmed et al. 2014). Based on these above findings, it is reasonable to conclude that As causes immunosuppression in exposed children.

A developing fetus is particularly vulnerable to toxic insult because of the rapid rate of in utero development. Similarly, gestation is a critical period for immune development (Dietert and Piepenbrink 2006). Accordingly, the current study attempted to address a specific point not evaluated systematically before, i.e. does prenatal exposure to As (in trivalent form) influence the immune repertoire of offspring at the innate, humoral, and/or cell-mediated immunity levels? Further, the study sought to assess whether immunomodulation developed and/or persisted during post-natal periods wherein there was no further As exposure.

One measure here of potential immunotoxic effects from As exposure during prenatal development were changes in immune system organ (thymus, spleen) weights and gross changes in their histological architecture. In the mice that no longer were in contact with As after birth, no significant changes were noted in their immune system organ weights or their histology. Studies have also noted that total numbers of splenic nucleated cells in male C57BL/6J mice were not impacted by As exposures (Xu et al. 2016). Those outcomes do not concur with findings in a study of preschool children from As-exposed areas of Bangladesh whose thymuses were altered by oxidative stress and apoptosis (Ahmed et al. 2012). Of course, that study was examining effects of ongoing As exposures and not one in which the exposure occurred well before the exposed hosts were analyzed (apart from also looking at spleen, not thymus and being in animal models).

Similarly, there were no structural/organizational changes in the splenic white pulp of the in utero As-exposed hosts (nor in their levels of nucleated cells), but there was a significant decrease in splenic CD4⁺T-cells and CD8⁺ T-cells. Interestingly, the reduction in CD4⁺T-cell levels was more pronounced than that of CD8⁺ cells. This change would be more in keeping with the findings of Kile et al. (2014) who noted that for each log₁₀ increase in drinking water As in Bangladesh, CD4⁺ T-cells in infant subjects were decreased by 9.2% (Kile et al. 2014). There is another report of As-exposure induced downward trends in IL-2 production among splenocytes stimulated with a mitogen in vitro (Conde et al. 2007); this finding implies either fewer numbers or functionality of CD4⁺ T-cells from As-exposed hosts. Still, the question remains as to why decreases in splenic CD4⁺ and CD8⁺ T-cells were not reflected in the white pulp architecture or in the absolute numbers of nucleated cells of the in utero As-exposed pups. Regarding the latter point, the present study did not quantify levels of other common cell types present in the spleen (i.e. B-cells, NK cells, macrophages, etc.). Whether or not these various immune cell subtypes were altered so as to yield no net negative change in overall splenic nucleated cell levels will be the subject of a follow-on investigation here. Regarding the white pulp, it is possible that gross architecture may not reflect any degree of change until a later time, i.e. if trends continue and T-cell levels fall off dramatically. Further studies (i.e. longer-term postpartum) are warranted.

Having seen that splenic T-cell frequency was impacted by maternal As exposures, and that decreases in levels of CD4⁺ cells were more pronounced than those of CD8⁺ cells in the As-exposed offspring, studies on CD4⁺ T-cell function were undertaken. Specifically, one-way Mixed Leukocyte Reactions (MLR) were done using PEMϕ from prenatal As-exposed pups as stimulator cells to drive responses among responder allogenic normal T-cells (C57BL/6). It was observed in the MLR reaction that PEMϕ (regardless of the source being As-exposed or normal hosts) were capable of driving allogenic T-cells to produce IL-2. This indicated that prenatal As exposure did not induce any generalized defect in PEMϕ function in the context of stimulating T-cells. This observation clearly indicated that there was, at least in this one functional regard, no long-term defect induced in the innate function of PEMϕ due to the in utero As exposure.

In contrast, when PEMϕ from naive allogenic mice (C57BL/6) were used to stimulate T-cells from As-exposed Balb/c pups (responders), it was seen that these cells failed to produce IL-2 and IFNγ and there were reductions in inducible activation markers CD44 and CD69. Oddly, CD25 expression was not affected by the host in utero As exposures. An explanation for this could be that all surface molecules/markers were not equally affected by dam As exposure, i.e. each can differ in how they are assembled in lymphocyte membrane bi-layers. Although we have not yet studied the potential impact of in utero As exposure on membrane fluidity, it is known that As-induced “stress” can cause changes in cell membrane fluidity (Ghosh et al. 2018).

Because it was clear that the in utero As exposure led to some longer-term effects on T-cells/compartment composition, the present study also looked for any potential impacts on B-cell-related outcomes in the offspring. For this, non-stimulated host circulating levels of IgG_2a and IgG₁ were evaluated in the offspring as these isotypes have been accepted as markers for potential T_H1 and T_H2 type responses, respectively (Mountford et al. 1994). Despite no impact on IgG₁ levels, there were decreases in IgG_2a levels, suggesting to us that in utero As exposure potentially impacted on T_H1 responses well after daily encounters with As had discontinued. This suggestion gained support from the findings of the studies here of allo-stimulation of T-cells from the As-exposed mice, i.e. they failed to produce IFNγ important for T_H1 cell expansion.

Conclusions

Development of T_H1 responses is essential for cell-mediated immunity and, ultimately, host resistance against pathogenic infections. Various animal studies have shown that early-life As exposure decreases host resistance to influenza (Ramsey et al. 2013). In the present study, an defect in T-cell repertoire development due to As exposure manifest as increased bacterial burdens in several organs/sites in As-exposed mice, even when the exposure had only occurred in utero. This clearly indicated to us that As induced in utero a generalized defect in immune function that then persisted – even in the absence of any further As exposures.

Acknowledgments

The authors are grateful to Prof. Syamal Roy (NIPER, Kolkata) and Dr. Shanta Dutta (ICMR-NICED, Kolkata) for support, helpful discussion, and suggestions for the manuscript. A grateful acknowledgement is also made to Dr. Sulagna Basu for providing the E. coli sE14 strain. MC is a recipient of a Council of Scientific and Industrial Research (CSIR) fellowship.

Disclosure statement

The authors declare no conflicts of interest. The authors alone are responsible for the content of this manuscript.

Additional information

Funding

This work was funded by the ICMR-National Institute of Cholera and Enteric Diseases.