Modulatory Effects of Mercury (II) Chloride (HgCl2 ) on Chicken Macrophage and B-Lymphocyte Cell Lines with Viral-Like Challenges In Vitro.

Mercury (Hg) is a toxic trace metal ubiquitously distributed in the environment. Inorganic mercury (as HgCl2 ) can cause immunotoxicity in birds, but the mechanisms of action are still not fully resolved, especially with respect to responses to viral infections. To investigate the potential immunomodulatory effects of Hg2+ on specific cell types of the avian immune system, chicken macrophage (HD-11) and B-lymphocyte (DT40) cell lines were applied as in vitro models for the innate and adaptive immune systems, respectively. The cells were stimulated with synthetic double-stranded RNA, which can be recognized by toll-like receptor-3 to mimic a viral infection. The Hg2+ showed concentration-dependent cytotoxicity in both cell lines, with similar median effect concentrations at 30 µM. The cytotoxicity of Hg2+ was closely related to glutathione (GSH) depletion and reactive oxygen species induction, whereas the de novo synthesis of GSH acted as a primary protective strategy. Nitric oxide produced by activated macrophages was strongly inhibited by Hg2+ , and was also influenced by cellular GSH levels. Cell proliferation, gene expression of microRNA-155, and cellular IgM levels in B cells were decreased at noncytotoxic Hg2+ concentrations. The secretion of antiviral interferon-α was induced by Hg2+ in both cell lines. Overall, our results suggest that Hg2+ exposure can cause immunomodulatory effects in birds by disrupting immune cell proliferation and cytokine production, and might result in disorders of the avian immune system. Environ Toxicol Chem 2021;40:2813-2824.

INTRODUCTION

Mercury (Hg) is one of the trace metals that ubiquitously occurs in the environment, and human activities, such as mining and industry, have considerably elevated its environmental levels and global distribution (Dorea & Donangelo, 2006; Beckers & Rinklebe, 2017). Like other nonessential trace metals, Hg is regarded as a priority environmental contaminant due to its potential for bioaccumulation and biomagnification, and also its high toxicity (Mann et al., 2011). The major species of Hg include elemental mercury (Hg(0)), inorganic mercury (predominantly divalent mercury, Hg(II)), and organic mercury (mainly methylmercury, MeHg). Although MeHg is recognized as the most toxic form, Hg(II) is more commonly and abundantly found in the environment (Schlüter, 2000; Beckers & Rinklebe, 2017) and can also exert severe adverse effects on wild animals. Elevated Hg levels have been detected in wild birds, affecting multiple physiological parameters including immune functions (Thaxton & Parkhurst, 1973; Wolfe et al., 1998; Ackerman et al., 2016; Whitney & Cristol, 2018). For example, tree swallows in a Hg‐contaminated area showed compromised immune competence, with weaker skin swelling responses to phytohemagglutinin challenge (Hawley et al., 2009). Elevated levels of internal Hg were associated with decreased macrophage phagocytosis in black‐footed albatrosses (Finkelstein et al., 2007). The impact of Hg on the functioning of the immune system may result in higher risk of infections at individual levels and may cause prevalence of diseases (e.g., avian influenza) in the populations.

As just mentioned, for avian species, there are some in vivo and wildlife studies revealing the immune modulatory effects of Hg, but very limited knowledge on the cellular mechanisms and immune functional effects of Hg toxicity has been documented. More insights into the immunotoxicity of Hg were reported in studies using mammalian or human in vitro cell models (Loftenius et al., 1997; Silva et al., 2005; Das et al., 2008; Lehmann et al., 2011; Levin et al., 2020; Yang et al., 2020). Such in vitro studies can demonstrate the underlying mechanisms and can also reduce animal experiments. In vitro evidence has shown that the mechanisms of Hg‐induced toxicity in immune cells are related to the induction of oxidative stress, the inhibition of nitric oxide, and the disturbance of cytokine profiles (Kim et al., 2002; Kim & Sharma, 2004; Guzzi & La Porta, 2008). However, this is based on mammalian models, whereas little is known about the cellular mechanisms of Hg‐induced immunomodulatory effects in avian species.

To evaluate the effects of Hg exposure on avian immune functions, one should consider not only the baseline immunity but also the activated immunity, which is more relevant to vulnerability to diseases. Once infected by certain pathogens, avian immune cells recognize the pathogen‐associated molecular patterns with specific receptors and trigger multiple defence strategies (Mogensen, 2009). For instance, toll‐like receptor 3 (TLR3) can detect the double‐stranded RNA (dsRNA) produced by most viruses during their replication (Sen & Sarkar, 2005). After immune activation, macrophages can produce nitric oxide and undergo oxidative burst to kill the pathogens (MacMicking et al., 1997), and B‐lymphocytes can produce antibodies to neutralize the pathogens (Dörner & Radbruch, 2007). To mediate the immune responses and communicate with other cells, immune cells also produce cytokines with various functions, such as antiviral interferon‐α and β (IFN‐α and ‐β), interleukins ([ILs] e.g., pro‐inflammatory IL‐18) and chemokines (e.g., IL‐8; Kaiser & Stäheli, 2013). Apart from cytokines, microRNAs (miRNAs) also play an important role in regulating the immune responses; miRNAs such as miR‐155 and let‐7 are involved in the inflammatory response, cytokine production, and lymphocyte maturation (Gottwein et al., 2007; Lu & Liston, 2009; Staedel & Darfeuille, 2013; Kumar et al., 2015; Alivernini et al., 2018). An adequate immune response requires an appropriate population of functional immune cells and proper levels of immune mediators (namely, cytokines and miRNAs) for communication and cooperation among different immune cells.

The aim of the present study was to investigate the modulatory effects and potential mechanisms of Hg2+ on the avian immune systems using chicken macrophage (HD‐11) and B‐lymphocyte (DT40) cell lines as in vitro models for the innate and adaptive immune systems, respectively, with and without a mimicked viral challenge with dsRNA. Both cell lines have been documented to express TLR3 and could be stimulated with dsRNA (Peroval et al., 2013; Quan et al., 2017). General toxicity (cytotoxicity, oxidative stress) and functional endpoints were quantified, including nitric oxide production by macrophages, cell proliferation, immune gene expression (receptors, cytokines, and miRNAs), and immune protein levels (cytokines and antibodies) as indicators for immunity after Hg2+ exposure and activation.

MATERIALS AND METHODS

Cell culture

The HD‐11 chicken macrophage cell line was a kind gift from J. van Baal (Department of Animal Sciences, Wageningen University, The Netherlands) and was cultured in RPMI 1640 medium (Gibco) with 10% fetal bovine serum (FBS; Sigma‐Aldrich) at 37  °C in a 5% CO2 humidified air incubator. Every 2–3 days, HD‐11 cells were subcultured by detaching the cells with 5 mM ethylenediaminetetraacetic acid (EDTA; Merck) and diluting the cell suspension with fresh culture medium. A washing step with phosphate‐buffered saline (PBS; Gibco) was performed before seeding the cells to remove the EDTA.

The chicken B‐lymphocyte cell line DT40 (ATCC® CRL2111™) was maintained in Dulbecco's modified Eagle medium (Gibco) containing 10% FBS, 10% tryptose phosphate broth (Sigma‐Aldrich), 5% chicken serum (Sigma‐Aldrich), and 0.05 mM 2‐mercaptoethanol (Gibco) at 37 °C in a 5% CO2 humidified air incubator. Cells were subcultured every 2–3 days by diluting the cell suspension with fresh culture medium.

Experimental setup

Effects of Hg2+ (as mercury (II) chloride [HgCl2]) exposure on general cytotoxicity and immune functions were investigated in HD‐11 chicken macrophages and DT40 chicken B‐lymphocytes. Both cell lines are able to be activated with dsRNA via the TLR3 pathway (He et al., 2007; Keestra & van Putten, 2008; Zou et al., 2017; Ma et al., 2019). Therefore, we used a synthetic analog of dsRNA, polyinosinic–polycytidylic acid sodium salt (poly I:C; Alexopoulou et al., 2001; Matsumoto & Tsukasa, 2008), to induce viral‐like challenges in all assays in the present study.

Metal ions including Hg2+ have a high affinity for thiol groups and can deplete the major intracellular antioxidant glutathione (GSH), resulting in oxidative stress and cytotoxicity (Hultberg et al., 2001; Nuran Ercal et al., 2005; Hossain et al., 2021). To reveal the possible underlying mechanism of Hg2+ toxicity related to GSH depletion, co‐exposure with the γ‐glutamylcysteine synthetase blocker L‐buthionine‐sulfoximine (BSO) was introduced to inhibit de novo GSH synthesis. The water‐soluble tetrazolium‐1 (WST‐1) assay was used to determine the median effect concentration (EC50) values after 24‐h exposure to Hg2+ in cells treated under different conditions, including immune activation with 25 µg/ml poly I:C (Sigma‐Aldrich) and inhibition of the de novo GSH synthesis by 200 µM BSO (Sigma‐Aldrich). These EC50 values served as the basis for defining the concentration ranges in follow‐up assays. Changes in intracellular GSH and reactive oxygen species (ROS) levels due to Hg2+ exposure were investigated to elucidate their potential role in the mechanism(s) of Hg2+‐induced toxicity.

Afterward, effects of Hg2+ exposure on immune functions were assessed. Production of nitric oxide by HD‐11 cells was quantified as a primary functional endpoint for macrophages. Cell proliferation was examined using both the WST‐1 assay for metabolic activity (cell activity), and the bromodeoxyuridine (BrdU) assay for DNA synthesis (cell division) with extended exposure time (48 h) in both cell lines. In addition, expression of immune function genes and miRNAs after Hg2+ exposure was profiled with quantitative reverse transcription polymerase chain reaction (RT‐qPCR) at the transcription level. Finally, enzyme‐linked immunosorbent assays (ELISAs) were employed to measure multiple immune functional protein levels, including the production of type I antiviral IFN‐α in the supernatant of both cell lines, and immunoglobulin M (IgM) levels in both supernatant (for secreted IgM) and cell lysate (for cellular IgM) of DT40 B‐lymphocytes. All the assays were performed in three independent experiments and carried out as described hereafter.

Exposure

For most of the assays, either HD‐11 macrophages or DT40 B‐cells were seeded at a density of 1 × 105 cells/well in a clear, flat‐bottomed 96‐well plate and incubated overnight before exposure. In proliferation assays, 1 × 104 cells/well were seeded, allowing a longer exposure time (48 h). For RT‐qPCR, 1 × 106 cells/well were seeded in 12‐well plates. A series of HgCl2 (Sigma‐Aldrich) stock solutions ranging from 0.2 to 2000 µM were prepared in sterile Milli‐Q water and diluted 20 times in culture medium to the final exposure concentration, ranging from 0.01 to 100 µM. Then 5 mg/ml poly I:C stock solution was prepared in PBS and stored at –20 °C. To activate the cells, poly I:C (Sigma‐Aldrich) stock solution was diluted 200 times in culture medium to a final concentration of 25 µg/ml in all assays. Also, to block de novo GSH synthesis via γ‐glutamylcysteine synthetase, 200 µM (final concentration) of BSO was added to the culture medium of the cells in the cytotoxicity, GSH, ROS, and nitric oxide assays. Exposure to HgCl2, poly I:C, and BSO started at the same time. Exposure time was 24 h for most assays, except for the proliferation assays, in which the exposure time was prolonged to 48 h.

General toxicity assays

Cytotoxicity assay

Cells were exposed to Hg2+ in four conditions, namely, with or without 200 µM BSO to block de novo GSH synthesis and with or without 25 µg/ml poly I:C activation. After 20 h of exposure, 10 µl (5% v/v) WST‐1 reagent was added to each well of HD‐11 or DT40 cells and incubated at 37 °C for 4 h before measuring absorbance at 440 nm and a reference wavelength at 620 nm with a SpectraMax M2 Microplate Reader (Molecular Devices). Cell viability was expressed as percentage of negative controls in the non‐BSO, nonactivated treatment group (–BSO –poly I:C group). Concentrations of Hg2+ used in the later assays were based on the results obtained for Hg2+ in cell viability assays.

GSH assay

ThiolTracker™ Violet Glutathione Detection Reagent (Life Technologies) was used to measure the intracellular GSH levels after Hg2+ exposure. Briefly, after 20 h of exposure, cells were washed twice with Dulbecco's phosphate‐buffered saline with calcium and magnesium (D‐PBS C/M; Gibco). Cells were then labeled with 20 µM ThiolTracker Violet dye working solution for 30 min at 37 °C. After they were washed for another two times with D‐PBS C/M, cells were resuspended, and fluorescence was measured at ex/em 404/526 nm with the SpectraMax M2 Microplate Reader. The GSH levels were shown as percentages relative to the negative controls in the –BSO –poly I:C group.

ROS assay

The ROS levels inside the cells were measured with 2′,7′‐dichlorofluorescin diacetate (DCFDA; Sigma‐Aldrich) reagent after 24 h of Hg2+ exposure. Once diffused into the cell and de‐esterified by cellular esterase, DCFDA is oxidized to highly fluorescent 2′,7′‐dichlorofluorescin (DFC) by ROS (Rosenkranz et al., 1992). The exposure conditions were identical to those of the cytotoxicity and GSH assays (also with or without poly I:C and BSO). Cells were first exposed for 20 h and then labeled with 25 µM DCFDA for another 4 h before fluorescence measurement at ex/em 485/535 nm. Results are presented as ratios relative to the negative controls in the –BSO –poly I:C group.

Immune functional assays

Nitric oxide (Griess) assay

Nitric oxide produced by HD‐11 macrophages under the same exposure conditions as the ones used in the cytotoxicity, GSH, and ROS assays was examined using the Griess assay (Sun et al., 2003). Briefly, after 24 h of exposure, 100 µl of exposed cell culture supernatant was transferred to a new 96‐well plate. Then 50 µl of 1% sulfanilamide (Sigma‐Aldrich) in 5% phosphoric acid (Merck) and 50 µl of 0.1% N‐(1‐naphthyl)ethylenediamine dihydrochloride (Sigma‐Aldrich) solution were added to the supernatant successively with a 5‐min interval. The plates were incubated for 5 min in the dark between the addition of the two solutions and before measuring absorbance at 540 nm with the SpectraMax M2 Microplate Reader. Nitric oxide production after treatment was calculated based on a calibration curve of NaNO2 ranging from 1.56 to 100 µM.

Proliferation assays

Two endpoints, namely, cell activity and cell division, were tested to evaluate cell proliferation levels after 48 h of Hg2+ exposure. The WST‐1 assay was performed as described previously in the Cytotoxicity assay section. The colorimetric BrdU assay (Roche) was applied according to the manufacturer's instructions. Briefly, during the last 4 h of exposure, newly synthesized DNA in the cells was labeled with BrdU. Floating DT40 cells were separated by centrifugation at 300 g for 10 min. Then both HD‐11 and DT40 cells were fixed and incubated with anti‐BrdU‐POD antibody, which bound to the BrdU and reacted with the substrate. The reaction was terminated by adding stop solution after incubation with the substrate for 30 min at room temperature. Cell division (DNA synthesis) was quantified by absorbance at 450 nm with a reference at 690 nm and expressed as a percentage relative to the negative control in the nonactivated group (–poly I:C group).

Gene expression

Expression of nine chicken genes (six for messenger [m]RNA and three for miRNA), including a housekeeping gene (glyceraldehydes‐3‐phosphate dehydrogenase [GAPDH]), the genes for type I interferons (IFN‐α, IFN‐β), interleukins (IL‐8, IL‐18) and toll‐like receptors recognizing viral pathogens (TLR7), reference miRNA (small nucleolar RNA, C/D box 68 [SNORD68]), and miRNAs involved in immune regulation (miR‐155 and let7) were tested in both cell lines by RT‐qPCR. All kits used were from Qiagen. Briefly, after 24 h of exposure, cells cultured in 12‐well plates were lysed with QIAzol Lysis Reagent, and RNA was extracted using an miRNeasy Mini Kit. The quality and quantity of extracted RNA were checked by Nanodrop (ND‐1000; Themoscientific). Then 300 ng total RNA was reverse‐translated to cDNA using the miScript II RT Kit with miScript HiFlex Buffer to get complementary (c)DNA for both mRNA and miRNA. Then RT‐qPCR was performed on a Rotor‐Gene® 6000 cycler with either the QuantiNova SYBR® Green PCR Kit for mRNA or the miScript SYBR® Green PCR Kit for miRNA. Commercially available primer assays were used for immune functional genes, including QuantiTect® primer assays (Gg_GAPDH_1_SG, Gg_IFNA3_1_SG, Gg_IFNB_1_SG, Gg_IL8L2_1_SG, Gg_IL18_1_SG, and Gg_TLR7_1_SG) for mRNA expression and miRCURY LNA miRNA PCR Assay for miRNA primers (dre‐miR‐155, hsa‐let‐7a‐5p, and SNORD68). The efficiency of qPCR amplification was checked for each primer using a 4‐fold serial dilution standard curve prior to sample measurement.

Immune protein levels

Multiple immune functional proteins produced by the two cell lines, including IFN‐α by both cell lines as well as IgM (both released and cellular) by DT40 B‐lymphocytes, were quantified with commercially available sandwich ELISA kits. All the ELISA kits (chicken IFN‐α and chicken IgM ELISA kits) were obtained from ELISAGenie. Assays were conducted according to the instructions provided by manufacturer for each kit.

Supernatant and cells were separated after exposure by centrifugation at 1000 g for 20 min at 4 °C. The IFN‐α levels were measured in the supernatant samples from both HD‐11 and DT40 cells, whereas the IgM levels were only analyzed in the supernatant of DT40 cells. Cellular IgM content was also determined in cell lysate prepared through three repeated freeze–thaw cycles after removal of the supernatant. Quantities of each immune protein (in pg/ml for IFN‐α and ng/ml for IgM) were calculated by comparing absorbance at 450 nm with standard curves from each kit. Levels of IgM in cell lysate samples were expressed in ng/µg protein by normalizing with total protein content. Protein content was quantified with a Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific) according to the manufacturer's instruction.

Statistics

Statistical analyses and visualization of results were performed with GraphPad Prism 5. Concentration–response curves were fitted with nonlinear regression (variable slope, four parameters) to calculate EC50 values. One‐way analysis of variance with Dunnett's post hoc test was conducted to compare treatment groups with Hg2+‐free controls. Gene expression levels were normalized against the housekeeping gene GAPDH (for mRNA) or SNORD68 (for miRNA) and shown as log2 fold changes in comparison with the negative control of nonactivated cells by the –∆∆CT method (Schmittgen & Livak, 2008).

RESULTS

General toxicity

Cytotoxicity

Concentration–response curves for the effect of Hg2+ on both HD‐11 and DT40 cells were obtained and showed similar EC50 values of approximately 30 µM irrespective of the poly I:C activation when the cells were not treated with BSO (Figure 1). When the de novo synthesis of GSH was blocked by cotreatment with BSO, Hg2+ showed much higher cytotoxicity in both cell lines. In HD‐11 macrophages, the EC50 of Hg2+ for cells cotreated with BSO was 2.54 µM, and the EC50 for cells treated with both BSO and poly I:C was as low as 0.75 µM. A more potent effect of Hg2+ exposure was seen in BSO‐treated DT40 B‐lymphocytes, with EC50 values that were lower than 0.5 µM. An overview of all EC50 values is shown in Table 1. In HD‐11 macrophages, poly I:C increased the cell metabolic activity by approximately 50%, whereas no induction was seen in DT40 B‐lymphocytes (Figure 1). In addition, noncytotoxic levels of Hg2+ exposure resulted in a significant increase of cell metabolic activity in cells not exposed to BSO, especially in HD‐11 cells.

Figure 1

Cytotoxicity effects of mercury (Hg)2+ on nonactivated and activated (25 µg/ml polyinosinic–polycytidylic acid sodium salt [poly I:C]) HD‐11 (A) and DT40 (B) cells with or without glutathione synthesis inhibitor (200 µM L‐buthionine‐sulfoximine [BSO]) after 24‐h exposure. Results are expressed as mean ± standard error of the mean (n = 3), relative to negative control in nonactivated cells for cell viability. Significant differences between treatments and Hg2+‐free controls were checked with one‐way analysis of variance (*p < 0.05, **p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentrations (EC50s) for different treatment groups: red lines for the treatment groups with both BSO and poly I:C (+BSO +Poly I:C), orange lines for the treatment groups with BSO but without poly I:C (+BSO –Poly I:C), and black lines for the treatment groups without BSO because they showed similar EC50s.

Table 1

Overview of median effect concentration (95% confidence intervals) values of Hg2+ exposure for different endpoints

		–BSO –Poly I:C	–BSO +Poly I:C	+BSO –Poly I:C	+BSO +Poly I:C
Cytotoxicity	HD‐11	29 µM (24.08–34.11)	27 µM (17.02–41.77)	2.54 µM (1.05–6.15)	0.75 µM (0.29–1.96)
	DT40	30 µM (28.28–31.35)	29 µM (27.76–30.37)	<0.5 µM n.a.	<0.5 µM n.a.
GSH	HD‐11	16 µM (12.90–9.67)	15 µM (11.73–20.00)	2.28 µM (1.65–3.19)	0.82 µM (0.72 to 0.94)
	DT40	50~100 µM n.a.	50~100 µM n.a.	<0.5 µM n.a.	<0.5 µM n.a.
ROS	HD‐11	6.51 µM (5.16–8.21)	6.56 µM (4.78–9.02)	0.54 µM (0.43–0.68)	0.79 µM (0.48–1.29)
Nitric oxide	HD‐11	n.a.	3.85 µM (2.70–5.49)	n.a.	0.37 µM (0.17–0.78)
Proliferation: WST‐1	HD‐11	5.43 µM (3.74–7.87)	5.78 µM （3.94–8.47)	n.a.	n.a.
	DT40	5.60 µM (4.68–6.69)	4.28 µM (3.55–5.16)	n.a.	n.a.
Proliferation: BrdU	HD‐11	6.32 µM (5.04–7.93)	5.31 µM (5.02–5.62)	n.a.	n.a.
	DT40	2.82 µM (1.40–5.70)	2.67 µM (1.09–6.54)	n.a.	n.a.
IFN‐α	DT40	0.58 µM (0.10–3.24)	0.53 µM (0.35–0.79)	n.a.	n.a.

BSO = L‐buthionine‐sulfoximine; poly I:C = polyinosinic–polycytidylic acid sodium salt; GSH = glutathione; ROS = reactive oxygen species; WST‐1 = water‐soluble tetrazolium‐1; BrdU = bromodeoxyuridine; IFN‐α = interferon‐α; n.a. = not applicable.

GSH and ROS

Different patterns of intracellular GSH depletion were demonstrated in HD‐11 cells under different exposure conditions (Figure 2A). When the cell could still synthesize GSH (without BSO), a steep decrease in GSH levels was only found close to the EC50 of Hg2+ cytotoxicity, regardless of the activation. However, when the de novo synthesis of GSH was inhibited by BSO, GSH levels declined at much lower Hg2+ exposure concentrations, with an EC50 of 2.3 µM (with BSO but without poly I:C), which is approximately 5 times lower than the EC50 obtained without BSO (~15 µM). Cells cultured with both BSO and poly I:C were the most vulnerable, with an even lower EC50 of 0.82 µM Hg2+ and a significant decrease in their GSH level already at the lowest exposure concentration (0.5 µM). In HD‐11 cells, the EC50 values for GSH depletion were in the same range as cytotoxicity for the individual treatment groups (shown as the vertical dashed lines in Figure 2A). As for DT40 cells, poly I:C did not affect the GSH changes, whereas BSO played a major role (Figure 2B). After blocking of the synthesis of GSH, GSH levels inside the cells sharply dropped to background levels already at the lowest Hg2+ exposure concentration (0.5 µM). Nevertheless, GSH levels were not affected at noncytotoxic Hg concentrations in the absence of BSO in DT40 B‐cells.

Figure 2

Intracellular glutathione (GSH) levels after 24‐h exposure to mercury (Hg)2+ in nonactivated and activated (25 µg/ml polyinosinic–polycytidylic acid sodium salt [poly I:C]) HD‐11 (A) and DT40 (B) cells with or without GSH synthesis inhibitor (200 µM L‐buthionine‐sulfoximine [BSO]). Results are expressed as percentage (mean ± standard error of the mean, n = 3) in relation to the negative controls without GSH synthesis inhibitor and without activation (–BSO –poly I:C). Significant differences between treatments and Hg2+‐free controls of DT40 cells were checked with one‐way analysis of variance (*p < 0.05, **p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentrations (EC50s) of cytotoxicity after 24 h of exposure from WST‐1 assay for different treatment groups: red lines for the treatment groups with both BSO and poly I:C (+BSO +Poly I:C), orange lines for the treatment groups with BSO but without poly I:C (+BSO –Poly I:C), and black lines for the treatment groups without BSO because they showed similar EC50s.

Similar to the results for GSH, patterns of ROS induction by Hg2+ exposure were influenced by the inhibition of de novo synthesis of GSH by BSO. In HD‐11 cells, Hg2+ induced ROS up to approximately 6‐fold under all four exposure conditions (Figure 3A). The EC50 values in cells co‐exposed with BSO (without poly I:C 0.54 µM, with poly I:C 0.79 µM) were approximately 10 times lower than the EC50 values in cells that still could produce GSH (without poly I:C 6.51 µM, with poly I:C 6.56 µM). The poly I:C activation doubled the ROS levels in nonexposed HD‐11 cells but did not affect the EC50 values of Hg2+ exposure. In DT40 cells, no significant induction of ROS was observed due to poly I:C activation, whereas ROS induction caused by Hg2+ exposure mainly occurred at cytotoxic concentrations above the EC50 values derived for cytotoxicity (shown as the vertical dashed lines in Figure 3).

Figure 3

Intracellular reactive oxygen species levels after 24‐h exposure to mercury (Hg)2+ in HD‐11 (A) and DT40 (B) cells with or without glutathione (GSH) synthesis inhibitor (200 µM L‐buthionine‐sulfoximine [BSO]), and with or without immune stimulator (25 µg/mL polyinosinic–polycytidylic acid sodium salt [poly I:C]). Results are expressed as a relative ratio (mean ± standard error of the mean, n = 3) to the negative controls without GSH synthesis inhibitor and without activation (–BSO –Poly I:C, shown as the horizontal lines). Significant differences between treatments and Hg2+‐free controls were checked with one‐way analysis of variance (*p < 0.05, **p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentrations (EC50s) of cytotoxicity after 24‐h exposure from the WST‐1 assay for different treatment groups: red lines for the treatment groups with both BSO and poly I:C (+BSO +Poly I:C), orange lines for the treatment groups with BSO but without poly I:C (+BSO –Poly I:C), and black lines for the treatment groups without BSO because they showed similar EC50s.

Immune functional endpoints

Nitric oxide

Unlike the results in GSH and ROS, the patterns of nitric oxide production by HD‐11 macrophages after Hg2+ exposure largely depended on poly I:C and also on BSO co‐exposure (Figure 4). The poly I:C activation induced nitric oxide production from basal level to more than 10 µM, whereas nonactivated cells hardly generated any nitric oxide. Concentration‐dependent inhibition of nitric oxide production was caused by Hg2+ exposure in activated cells at noncytotoxic concentrations. Meanwhile, cells became more sensitive with BSO blocking the de novo synthesis of GSH, showing 1 order of magnitude lower EC50 (0.37 µM) than the one without BSO (EC50: 3.85 µM).

Figure 4

Nitric oxide production by activated (with 25 µg/ml polyinosinic–polycytidylic acid sodium salt [poly I:C]) and nonactivated HD‐11 cells with or without glutathione synthesis inhibitor (200 µM L‐buthionine‐sulfoximine [BSO]) were measured with the Griess assay after 24‐h Hg2+ exposure. Results are expressed as mean ± standard error of the mean (n = 3). Significant differences between treatments and Hg2+‐free controls were checked with one‐way analysis of variance (**p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentrations of cytotoxicity after 24‐h exposure from WST‐1 assay for different treatments: red line for the treatment group with both BSO and poly I:C (+BSO +Poly I:C), and green line for the treatment group without BSO but with poly I:C (–BSO +Poly I:C).

Proliferation

As shown in Figure 5, prolonged exposure (48 h) to Hg2+ resulted in inhibition of cell proliferation at noncytotoxic concentrations (determined by the 24‐h WST‐1 assays, shown as vertical lines in Figure 5) in terms of both metabolic activity (WST‐1) and DNA synthesis (BrdU) in both cell lines (Figure 5). No significant differences were seen among EC50 values in both cells derived from both assays, ranging from 2.7 to 6.3 µM (Table 1). In addition, poly I:C treatment significantly stimulated metabolic activity as well as DNA synthesis in DT40 cells (Figure 5C and d, p < 0.05), whereas poly I:C significantly inhibited DNA synthesis in HD‐11 cells (Figure 5B, p < 0.05) without affecting metabolic activity (Figure 5A).

Figure 5

Cell proliferation of HD‐11 cells (A and B) and DT40 cells (C and D) with and without 25 µg/ml polyinosinic–polycytidylic acid sodium salt activation after 48‐h Hg2+ exposure. Results are expressed as relative absorbance to negative control in nonactivated groups (mean ± standard error of the mean, n = 3), and statistically compared with Hg2+‐free negative controls with one‐way analysis of variance (*p < 0.05, **p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentration of cytotoxicity after 24‐h exposure from the WST‐1 assay for HD‐11 cells (A and B) or DT40 cells (C and D). BrdU = bromodeoxyuridine; WST‐1 = water‐soluble tetrazolium‐1.

The effect of Hg2+ exposure on five immune functional genes and two miRNAs was analyzed. Only miR155 was found to be downregulated in activated DT40 B‐lymphocytes. (Figure 6A; detailed results shown in the Supporting Information, Figure S1).

Figure 6

Effects of mercury (Hg)2+ exposure with and without 25 µg/ml polyinosinic–polycytidylic acid sodium salt activation on expression of miR‐155 in DT40 B‐lymphocytes (A), IgM levels in DT40 cell lysate (B), and IFN‐α produced by HD‐11 (C) and DT40 cells (D). Results are expressed as mean ± standard error of the mean (n = 3), and significant differences between treatments and Hg2+‐free controls were checked with one‐way analysis of variance (*p < 0.05, **p < 0.01, ***p < 0.005). Vertical lines indicate the median effect concentration values of cytotoxicity after 24‐h exposure obtained from the WST‐1 assay. Horizontal lines show the detection limits of the IFN‐α enzyme‐linked immunosorbent assay (15.6 pg/ml). IFN‐α = interferon α.

Functional IFN‐α and IgM levels

For DT40 B lymphocytes, IgM levels were measured in both culture supernatant and cell lysate. Secreted IgM in culture supernatant did not respond to Hg exposure or poly I:C activation at concentrations below the EC50 for cytotoxicity (data not shown). The IgM detected in cell lysate consisted of the IgM in the cytoplasm and the IgM on the cell membrane functioning as B‐cell receptors. The poly I:C stimulated IgM levels in the cell lysate (Figure 6B). Cellular IgM levels in nonactivated DT40 cells significantly decreased after exposed to Hg2+, ranging from 0.01 to 1 µM, followed by a remarkable increase at 10 µM. Because the results were expressed per milligram protein content in the lysate samples, the increase in IgM levels at 10 µM could be due to the low protein content (Supporting Information, Figure S2).

The levels of antiviral IFN‐α secreted into the culture supernatant were measured with ELISA assays for both cell lines (Figure 6C and D). An increasing trend of IFN‐α with the increasing concentration of Hg2+ exposure was noted in both cell lines, especially in DT40 cells. Although significances were only found in high concentrations close to cytotoxic levels, the EC50 values for both nonactivated and activated DT40 cells were as low as 0.5 µM (Table 1). The poly I:C activation did not affect the IFN‐α production in either of the cell lines. Compared with DT40 cells, the secretion of IFN‐α by HD‐11 cells was less responsive to Hg2+ exposure, but the initial levels in the nonexposed cells were higher.

DISCUSSION

The objective of the present study was to investigate in vitro cytotoxic and immune functional effects of Hg2+ exposure on two chicken immune cell lines (HD‐11 macrophage and DT40 B‐lymphocytes as models for specific avian immune cell types), after a viral‐like challenge. Similar EC50 values for cell viability (~30 µM) were found for both cell lines after Hg2+ exposure for 24 h (Figure 1 and Table 1), which were in the same range of the EC50 values for Hg2+ from other in vitro studies using mammalian immune cells (Yamamoto et al., 1998; De Guise et al., 2000; Wataha et al., 2000; Kim & Sharma, 2004; Engin et al., 2017; David et al., 2020). The poly I:C induced cell metabolic activity only in HD‐11 macrophage cells but not in DT40 B‐cells (Figure 1), but after 48 h of exposure, only DT40 cells showed an increased metabolic activity after poly I:C challenge (Figure 6B and D). These cell‐ and time‐dependent responses might suggest that HD‐11 macrophages build a faster response to poly I:C challenge than DT40 B‐cells. The poly I:C challenge did not affect the sensitivity of either cell type to Hg2+ exposure, because similar EC50 values were obtained. Stimulatory effects on the activity of cells by low Hg2+ concentration exposure were seen, especially in HD‐11 cells (Figure 1A). Similar hormetic effects have been observed in Hg2+‐exposed mammalian cells in vitro, such as bovine leukocytes and a human mammary cell line (De Guise et al., 2000; Schmidt et al., 2004). This may point to cellular stress responses to detoxify the Hg2+ and could be important in case of chronic exposure scenarios (Damelin et al., 2000; Calabrese et al., 2007; David et al., 2020).

Compared with other trace metal ions, Hg2+ has a relatively high toxicity. For example, the toxicity of lead (Pb)2+ was evaluated in the same cell types with much higher EC50 values, namely, 480 μM for HD‐11 macrophages and 1700 μM for DT40 lymphocytes (Han et al., 2020). Therefore, although inorganic Hg2+ is generally regarded to be less toxic than MeHg, it could still be a trace metal ion of concern.

Depletion of GSH and induction of ROS have been identified as major mechanisms of toxicity induced by trace metals with high affinity to the thiol group (Nuran Ercal et al., 2005). In the present study, we included BSO co‐exposure, blocking γ‐glutamylcysteine synthetase, to investigate the function of de novo synthesis of GSH in the Hg2+ exposure experiments. As shown in Figures 1, 2, 3, intracellular GSH levels dropped, ROS levels increased, and cell viability decreased in both cell lines after Hg2+ exposure without BSO, but when BSO inhibited the de novo GSH synthesis, the EC50 values were much lower. This indicates that depletion of GSH could be one of the major mechanisms of Hg2+‐induced toxicity in both cell lines and that the de novo GSH synthesis plays an import role in its detoxification. It is hypothesized that, at noncytotoxic levels, Hg2+ depleted endogenous GSH, which at the same time could be compensated for by the de novo production of GSH as a protective response. However, when the Hg2+ concentrations became too high, the capacity of the GSH de novo synthesis appeared to be not sufficient, resulting in a depletion of GSH, accumulation of ROS, and loss of cell viability.

Compared with HD‐11 cells, DT40 B‐cells showed higher sensitivity to Hg2+ exposure when BSO was co‐exposed in terms of cell viability and intercellular GSH and ROS levels (Figures 1, 2, 3). These differences were probably due to the nature of these two types of cells. Macrophages (HD‐11) are equipped with a robust and complex antioxidant buffer system, such as lipid mediators and metabolic reprograming, to survive from the high levels of ROS and nitric oxide produced by themselves after activation and consecutive oxidative burst (Brüne et al., 2013; Virág et al., 2019). Meanwhile, DT40 B‐cells apparently rely on de novo GSH synthesis more than HD‐11 macrophages, showing a sharp drop in GSH and cell viability already at the lowest Hg2+ exposure concentration (0.5 µM) after co‐exposure with BSO. However, without BSO, HD‐11 macrophages were more sensitive to Hg2+ exposure regarding GSH levels (Figure 2; with a EC50 at 15 µM) than DT40 B‐cells (EC50 values between 50 and 100 µM). Decreased GSH levels in DT40 cells were only seen above cytotoxic Hg2+ concentrations, suggesting that there might be other mechanisms inducing cell viability besides GSH depletion.

These cell‐specific effects demonstrated that under normal circumstances (without BSO) HD‐11 macrophages are probably more vulnerable to Hg2+‐induced GSH depletion and ROS induction resulting in decreased cell viability with lower EC50 values than DT40 B‐lymphocytes. Also, according to our results with BSO blocking the de novo GSH synthesis, macrophages are thought to be protected by a more complex antioxidative protective system against ROS, which likely includes other components than GSH, such as metabolic reprogramming bioactive and lipid mediators (Virág et al., 2019). These additional protective mechanisms are still effective after BSO exposure. However, the DT40 cell relies largely on GSH as an antioxidant.

In addition to cytotoxic effects, modulatory effects of Hg2+ on some functional endpoints were also detected in both cell lines at noncytotoxic concentrations. For instance, nitric oxide production was found to be strongly inhibited by Hg2+ in activated HD‐11 macrophages (Figure 4), as also reported in murine macrophages (Tian & Lawrence, 1996; Kim et al., 2002; Batista‐Duharte et al., 2018). Downregulation of inducible nitric oxide synthase in cells of the J774A.1 murine macrophage cell line by Hg2+ was demonstrated in both gene expression profiles as well as protein levels (Kim et al., 2002). However, increased nitric oxide levels were found in activated macrophage cell lines after noncytotoxic levels of cadmium (Cd)2+ and Pb2+ exposure (García‐Mendoza et al., 2019; Han et al., 2020), suggesting possible deviating mechanisms among divalent trace metal irons. Once the GSH synthesis was blocked by BSO, less nitric oxide was produced by activated HD‐11 cells exposed to Hg2+ between 0.5 and 5 µM (Figure 4). This could be related to the decreased cell viability (Figure 1A).

In addition, oxidized glutathione (in our case triggered by the treatment with BSO in activated cells) may induce S‐glutathionylation of the nitric oxide synthase enzyme (NOS) via exchanging thiol–disulphide, leading to uncoupling of biochemically active NOS dimers and a switch from the production of nitric oxide to other types of ROS (Chen et al., 2010; Dulce et al., 2011). This might be explained by the fact that the capacity of the protective cellular antioxidant determines how much nitric oxide macrophage cells can produce, to prevent cells from injuring themselves. Nitric oxide is involved in more immune functions than direct antimicrobial activities, such as regulation of cytokine production and lymphocytes differentiation (MacMicking et al., 1997; Bogdan, 2001; Tripathi et al., 2007). Hence, inhibited nitric oxide production after Hg2+ exposure in activated macrophages can cause an impaired antimicrobial response as well as disturbed immune signaling and might lead to higher risks of infections in vivo.

Proliferation of cells was supressed by Hg2+ exposure in both cell lines regardless of activation, affecting both DNA replication (Figure 5A and C) and metabolic activity (Figure 5B and D). The DT40 cells were slightly more sensitive to Hg2+ than HD‐11 macrophages, with approximately 2‐fold lower EC50 in BrdU assays (Table 1). Decreased immune cell proliferation due to Hg2+ exposure was also indicated in mice ex vivo, depending on age and organ (Silva et al., 2005). On the other hand, some studies have suggested stimulatory effects of low‐level Hg2+ exposure on murine lymphocyte proliferation (Pelletier et al., 1985; Jiang & Möller, 1996; Pollard & Landberg, 2001). Although hormesis was not found in the proliferation assays after 48 h of exposure (Figure 5), it was seen in the WST‐1 assays after 24 h of exposure (Figure 1). This might suggest that the effect on increased cell metabolic activities is a rather acute response. Interestingly, after poly I:C activation, HD‐11 cells tended to slightly decrease DNA synthesis, potentially to maintain cell homeostasis, whereas DT40 cells showed enhancement in both DNA synthesis and cell activity. In addition, similar EC50 values between unstimulated and stimulated cells suggest that poly I:C challenge did not affect the sensitivity of cells to Hg2+ exposure (Table 1). Our results indicate that Hg2+ exposure may cause a deficiency in functional immune cells and may disturb the immune cell composition in avian species in vivo.

Gene expression and immune protein

Gene expression and protein levels were measured as major immune parameters in the present study. The poly I:C was reported to be recognized by TLR3, triggering several pathways such as nuclear factor κB (NF‐κB) and interferon receptor factors (IRFs) pathways, resulting in generation of a spectrum of cytokines including proinflammatory cytokines (e.g., IL‐8 and IL‐18) via the NF‐κB pathway and antiviral type I IFNs (IFN‐α and ‐β) via IRFs pathways (Alexopoulou et al., 2001; Moynagh, 2005; Sen & Sarkar, 2005; Kawasaki & Kawai, 2014). Our results demonstrated a significant increase in IFN‐α levels due to Hg2+ exposure in both cell lines, especially DT40 cells (Figure 6C and D), similar to what was observed in Pb2+‐exposed HD‐11 and DT40 cells (Han et al., 2020). Although the expression of type I IFNs was documented to be lifted in HD‐11 cells after 2 h of incubation with poly I:C (Peroval et al., 2013), different IFN‐α levels after poly I:C activation were not found in the qPCR or the ELISA assays after 24 h of exposure in the present study. The prolonged stimulation time could have resulted in the degradation of the induced IFNs by poly I:C. In addition, because both HD‐11 and DT40 are immortal cell lines transformed with virus (MC29 virus for HD‐11 cells and avian leukosis virus for DT40 cells; Beug et al., 1979; Bezzubova & Buerstedde, 1994), they might have to some extent lost their capacity for type I IFN production after viral‐like challenge with poly I:C.

The HD‐11 cells were found not to produce more type I IFN after infection with the highly pathogenic avian influenza virus H5N1, whereas both ex vivo (in primary chicken splenocytes) and in vivo (in chicken) studies showed significantly increased type I IFN production after infection with the same virus (Liniger et al., 2012). This difference could be due to the presence of specialized type I IFN‐producing cells in primary splenocytes and in vivo, and might suggest a deficiency in type I IFN production in HD‐11 cells after viral challenge. Type I IFNs secreted by immune cells can pass the signal to the neighboring cells, which will develop an antiviral state to defend themselves against the infection. However, excessive type I IFNs may promote autoimmunity in vivo (Baccala et al., 2007). Therefore, the type I IFN induced by Hg2+ exposure might potentially result in a disorder of immunity.

Apart from cytokines, another group of crucial immune mediators is the miRNAs, which can regulate cellular activities posttranscriptionally by inhibiting translation and destabilizing mRNA (Bushati & Cohen, 2007). As a multifunctional miRNA, miR‐155 is involved in regulation of lymphocyte proliferation and differentiation, macrophage polarization, antibody production, and other functions (Alivernini et al., 2018). Our results indicated that miR‐155 was down‐regulated by 10 μM Hg2+ in activated DT40 B‐lymphocytes (Figure 6A). Decreased miR‐155 levels caused impaired B‐cell proliferation in mouse models (Rodriguez et al., 2007; Babar et al., 2012), which is in line with our results (Figure 5). In addition, cellular IgM content was found to be significantly decreased by noncytotoxic Hg2+ exposure in the cell lysate of nonactivated DT40 lymphocytes, but not in activated cells (Figure 6B). The IgM in the cell lysate referred mainly to the IgM on the cell membrane acting as B‐cell receptors (BCRs), which can recognize antigens and initiate humoral immunity (Friess et al., 2018). Therefore, Hg2+ affected B‐cell functions probably by suppressing miR‐155 expression after activation and decreasing membrane BCRs on the cell membrane without activation. This might result in a weakened pathogen defence and higher risk of infection in avian species.

CONCLUSIONS

In summary, the results of the present study revealed that Hg2+ had concentration‐dependent toxic effects on both chicken macrophage (HD‐11) and B‐lymphocyte (DT40) cells and affected their immune functions (Figure 7). De novo GSH production played a vital role in protecting the cells from Hg2+‐related GSH depletion and ROS induction, and meanwhile could influence the immune function of activated HD‐11 macrophages with respect to nitric oxide production. In addition, antiviral IFN‐α levels were elevated by Hg2+ in both cell types. In B‐lymphocytes, Hg2+ mainly downregulated the expression of miR155 after activation and decreased the cellular IgM levels, which probably act as BCRs, without activation. Based on the present study, it is likely that Hg2+ may disturb the homeostasis of the avian immune system, potentially resulting in an inadequate immune response after viral infections that might cause a higher risk of infection for individuals and even prevalence of diseases in the group. Although we only employed two in vitro cell models (macrophages and B‐lymphocytes) without considering the other parts of the immune system and interactions between immune cells, our results provide a starting point for future studies on molecular mechanisms, systematic immune responses, and effects of long‐term exposure to immune‐toxic trace metals in avian species after viral challenges.

Figure 7

Effects of mercury (Hg)2+ on the immune functions found in HD‐11 macrophage (A) and DT40 B‐lymphocyte (B). GSH = glutathione; ROS = reactive oxygen species; IFN‐α = interferon‐α; IL8 = interleukin 8; TLR3 = toll‐like receptor 3; poly I:C = polyinosinic acid‐polycytidylic acid; dsRNA = double‐stranded RNA; BSO = buthionine sulphoximine; miR‐155 = microRNA 155.

Supporting Information

The Supporting Information is available on the Wiley Online Library at https://doi.org/10.1002/etc.5169.

This article includes online‐only Supporting Information.

Supporting information.

Click here for additional data file.