Mercury leads to abnormal red blood cell adhesion to laminin mediated by membrane sulfatides

Abstract

Exposure to mercury is associated with numerous health problems, affecting different parts of the human body, including the nervous and cardiovascular systems in adults and children; however, the underlying mechanisms are yet to be fully elucidated. We investigated the role of membrane sulfatide on mercuric ion (Hg²⁺) mediated red blood cell (RBC) adhesion to a sub-endothelial matrix protein, laminin, using a microfluidic system that mimics microphysiological flow conditions. We exposed whole blood to mercury (HgCl₂), at a range of concentrations to mimic acute (high dose) and chronic (low dose) exposure, and examined RBC adhesion to immobilized laminin in microchannels at physiological flow conditions. Exposure of RBCs to both acute and chronic levels of Hg²⁺ resulted in elevated adhesive interactions between RBCs and laminin depending on the concentration of HgCl₂ and exposure duration. BCAM-Lu chimer significantly inhibited the adhesion of RBCs that had been treated with 50 μM of HgCl₂ solution for 1 hour at 37 °C, while it did not prevent the adhesion of 3 hour and 24 hour Hg²⁺-treated RBCs. Sulfatide significantly inhibited the adhesion of RBC that had been treated with 50 μM of HgCl₂ solution for 1 hour at 37 °C and 0.5 μM of HgCl₂ solution for 24 hours at room temperature (RT). We demonstrated that RBC BCAM-Lu and RBC sulfatides bind to immobilized laminin, following exposure of RBCs to mercuric ions. The results of this study may demonstrate clinical significance considering the potential associations between sulfatides, red blood cells, mercury exposure, and cardiovascular diseases.

1. INTRODUCTION

Mercury (Hg) is a heavy metal that is released into the earth’s atmosphere from natural sources, such as volcanic action, and human related sources, such as coal combustion, waste combustion, and smelting¹. Power plant flue gases are the largest source of Hg in the US². In the atmosphere, chemical reactions convert Hg to divalent form (Hg²⁺) that can bind to particulate matter. Both Hg²⁺ and particulate bound Hg ingress from the atmosphere into water sources, where a small portion of the Hg can be converted into methylmercury by microorganisms and chemical processes. In water bodies, Hg accumulates in fish. The most common exposure of humans to mercury is through the consumption of marine and freshwater fish^1,3,4. Historically, Hg has been used for medical purposes, such as in dental amalgam^5–7 and in vaccines and flu shots as a preservative between 1930–1999 in the USA⁸ as well as in industrial products such as skin-lightening creams, antiseptic facial products, mercury-containing laxatives, diuretics, teething powders, and in latex paint⁹. Mercury may also be ingested through high fructose corn syrup¹⁰.

Hg is biologically toxic in both divalent and methylated forms¹¹, at both acute and chronic exposure levels¹². After being ingested, Hg is absorbed into the bloodstream. Hg has been shown to affect protein stability¹³, mineral loss³, and to result in enzyme denaturation¹⁴, platelet activation¹⁵, and red blood cell (RBC) adhesion to endothelium¹². Earlier studies have reported on the effects of Hg²⁺ exposure on RBCs^12,16–20, where Hg²⁺ was shown to bind to outer surface of RBCs with a high affinity for sulfhydryl groups, lipid groups, and structural proteins^21–23. Moreover, Hg²⁺ exposure can cause neuronal disorders including autism spectrum disorder^24–27, immune disease²⁸, gout, liver and renal disorders²⁹, and cardiovascular disease (CVD)^30–32. The underlying mechanisms whereby Hg²⁺ contributes to CVDs have been linked with radical oxygen species (ROS) as well as attenuated activity of antioxidant enzymes^33–35 and LDL^36–38. Recent studies have also suggested the role of enhanced RBC adhesion in the development of cardiovascular related problems³⁹. Various cell surface and membrane alterations may take part in mediating this adhesion. For instance, an enhanced phosphatidylserine (PS) exposure on the outer leaflet of the RBC membrane has been associated with abnormal RBC adhesion to endothelial cells in vitro in the presence of Hg²⁺ toxicity¹². However, the role of other major components of the RBC membrane, such as BCAM-Lu or sulfatides, in this adhesion process has yet to be delineated. Furthermore, the majority of reports on the changes of RBC adhesivity and biomechanical properties due to high levels of mercury exposure have exclusively been performed using RBC suspensions prepared in various buffer solutions in the absence of plasma and other blood components, which may not accurately represent the physiological environment^12,17. Therefore, a better assessment of Hg²⁺ mediated RBC adhesion requires establishment of physiologically relevant conditions where direct and indirect effects of other blood cells and plasma proteins on the RBC adhesion dynamics may directly be taken into account.

Lipid rafts, which are membrane micro-domains, are composed of a lipid bilayer in which various membrane components, such as cholesterol and sphingolipids and contain specific membrane proteins including GPI-anchored proteins, flotillins, stomatin, aquaporin-1, and Gsa, are embedded. Lipid rafts have been implicated in cell signaling pathways in various cell types including RBCs⁴⁰. Lipid raft signaling is important in treatment of some diseases such as Alzheimer’s, Parkinson’s, prion, systemic lupus erythematosus, HIV, and CVD⁴¹.

The main objective of this study was to investigate the role of membrane sulfatide on mercuric ion (Hg²⁺) mediated red blood cell (RBC) adhesion to a sub-endothelial matrix protein, laminin, using a microfluidic system that mimics microphysiological flow conditions. For this purpose, we treated whole blood samples with both acute and chronic levels of Hg⁺² that spanned a concentration range of HgCl₂ varying between 0.1 μM and 50 μM, wherein low concentration samples (<1 μM) were incubated for 24 hours and high concentration samples (>5 μM) were incubated for 1 and 3 hours to simulate chronic and acute toxicity conditions, respectively. HgCl₂-treated whole blood samples were flowed through laminin-functionalized microfluidic channels under physiological flow conditions. Laminin is a sub-endothelium protein that supports adhesion through sulfatides⁴². Our results demonstrated that acute and chronic levels of Hg²⁺ ion increased the adhesion of RBCs to laminin. We further showed that elevated adhesion of HgCl₂-treated RBCs to laminin was mediated by BCAM-Lu and membrane sulfatides depending on the Hg²⁺ concentration and duration of the exposure of RBCs to Hg²⁺.

MATERIALS AND METHODS

2.1. Blood collection and preparation

Blood samples were obtained in Na-citrate (3.8%) containing vacutainers at the Hematopoietic Biorepository & Cellular Therapy Core at Case Western Reserve University under an Institutional Review Board (IRB) approved protocol. Whole blood samples were obtained from a total of eight healthy individuals. All experiments were carried out within 24 hours of venipuncture using pre-processing free whole blood samples. To prepare mercury-treated blood samples, a 500 μM mercuric ion (HgCl₂) stock solution was first prepared in phosphate buffered saline (PBS, pH 7.4). The stock solution was then added into whole blood to obtain final concentrations of 0.5, 1, 5 and 50 μM mercuric ions (Hg²⁺) in the samples. Next, the samples were incubated at 37 °C for 1 or 3 hours or at room temperature for 24 hours. Due to excessive hemolysis at 37°C with long term Hg exposure, 24 hour experiments were carried out in room temperature and at concentrations of 0.5 and 1.0 μM HgCl₂ (Fig. 6).

Figure 6. Effect of HgCl₂ on human RBCs hemolysis based on plasma color change and LDH activity.

(A) Plasma color turns red as hemolysis occurs. The tubes numbered 1 contain fresh blood samples (no incubation). The tubes numbered 2 contain control blood samples that were incubated at specific temperatures without the addition of HgCl₂. The samples in the tubes numbered 3 through 7 were incubated in RT or 37 °C in the presence of HgCl₂ at concentrations 0.1, 0.5, 1, 5, and 50 μM respectively. (B) LDH activity significantly increased in samples incubated at 37 °C for 24 hours.

2.2. Microfluidic chip preparation and flow adhesion assay

The microfluidic chip fabrication process involved laser micro-machined (VersaLASER, Universal Laser Systems Inc., Scottsdale, AZ) poly(methyl methacrylate) (PMMA) top covers (McMaster-Carr, Elmhurst, IL) and double sided adhesive (DSA) (iTapestore, Scotch Plains, NJ) layers that were laser cut to desired channel geometry (4 mm width and 25 mm length). The DSA film was first attached on the PMMA cap that had the inlet and outlet holes (Fig. 1A). Next, the top PMMA cap was assembled on a functionalized microscope glass slide through the DSA layer (adhesion coating: APTES, 3-Aminopropyl Triethoxysilane, Electron Microscopy Sciences, Hatfield, PA). The microchannel height was determined by the thickness of the DSA film, which was approximately 50 μm (Fig. 1B). Design and fabrication of the microfluidic channels was described in detail in our previously published work^43–46. The assembled micro-devices were functionalized with laminin, a sub-endothelium associated protein, via covalent bonding to ensure consistent protein coverage of the microchannel surface at higher flow rates (Fig. 1C). Microchannels were initially loaded with N-g-Maleimidobutyryloxy succinimide ester (GMBS: 0.28% v/v), which is a cross-linking agent for protein immobilization, and incubated at room temperature for 20 minutes. Thereafter, the excess GMBS solution was removed by rinsing the microchannels with 100 % ethanol followed by a PBS wash step that was repeated twice. The surface-modified microchannels were perfused with a laminin solution (Sigma Aldrich, USA) at a concentration of 0.1 mg/mL and allowed to incubate for 1.5 hours at room temperature. Finally, the microchannels were blocked with 2% bovine serum albumin (BSA, Sigma Aldrich, USA) and incubated overnight in 4 °C to prevent non-specific binding.

Figure 1.

Assessment of mercury-induced RBC adhesion in a functionalized microfluidic platform under physiogical flow conditions. (A) The flow system consisted of three microchannels mimicking the size scale of post-capillary venules (height: 50 μm). (B) Hg-treated RBCs interact with the immobilized laminin on the microchannel surface. Some of the Hg-treated RBCs turn into echynocytes depending upon the HgCl₂ concentration and duration of the incubation. (C) Interaction between HgCl₂-treated red blood cell and surface immobilized laminin.

2.3. Adhesion inhibition experiments

For adhesion inhibition experiments, two different strategies were followed: inhibition of the specific molecule on the RBC membrane or inhibition of the specific adhesion site on the immobilized laminin. The former was carried out by adding heparin (50 U/mL) or lactadherin (10 μg/mL) directly into the samples–both control and HgCl₂-treated– and incubating at 37 °C for 30 minutes. In contrast, the specific adhesion sites on immobilized laminin were blocked by incubating the microchannels with bovine sulfatide (200 μg/mL), heparan sulfate (50 μg/mL) or BCAM-Lu (10 μg/mL) for 1 hour at room temperature based on published literature⁴⁷. The excess solutions were rinsed off by PBS prior to adhesion experiments that were performed as explained previously.

2.4. Scanning electron microscopy

Fresh human whole blood was collected in Na-citrate tubes. Control or treated samples were centrifuged at 500g for 5 minutes followed by removal of plasma and buffy coat. Pelleted RBCs were washed twice with PBS and suspended in 0.1 M phosphate buffer at a hematocrit of 40 %. Next, RBCs in the phosphate buffer (pH=7.4) were fixed with 2.5% glutaraldehyde (GA) for 1 hour in room temperature (primary fixation). The suspension was then centrifuged at 500g and 5 minutes to collect RBCs which were re-suspended and incubated in 0.1M cacodylate buffer (pH=7.4) containing 1% osmium tetroxide for 1 hour in room temperature (secondary fixation). Thereafter, the RBCs were rinsed twice with cacodylate buffer and dehydrated serially in 30%, 50%, 70%, 85%, 95% and two times in 100% ethanol while adjusting the hematocrit to 20%. Finally, the samples were dried with hexamethyldisilazane (HMDS) followed by mounting and coating with palladium and examining using a Helios NanoLab™ 650 (FEI, Field Emission SEM).

2.5. Lactate Dehydrogenase (LDH) activity measurements

The blood samples were treated with varying concentrations of HgCl₂ as explained in Section A, and LDH activity was measured immediately after incubation using clinical techniques in the Core Laboratory of University Hospitals Cleveland Medical Center.

2.6. Image processing and quantification

Phase contrast images of the microfluidic channel surfaces were obtained using an inverted Olympus microscope (model no: IX83) and commercial software (cellSense Dimension, Olympus Life Science Solutions, Center Valley, PA), which provided automated large field of view mosaic patches of designated areas. Microchannel images were processed using Adobe Photoshop software (San Jose, CA) for the quantification of adhered RBCs per unit area (32 mm²), which spanned one third (center) of the entire surface area.

2.7. Statistical analysis

One-sample Wilcoxon signed rank test was used to assess the statistical significance of differences between paired data sets. For non-paired data, we performed a one-way ANOVA test. The error bars throughout the figures indicate ± standard error of the mean.

2. RESULTS

3.1. Hg²⁺ induces RBC adhesion in a concentration dependent manner

We utilized a microfluidic platform to assess the effect of mercuric ion exposure on RBC adhesion to surface-immobilized laminin (Fig. 1A–C). All the adhesion experiments were carried out using whole human blood samples instead of buffered RBC suspensions in order to maintain the physiologically relevant environment during the experiments. Our results indicated that RBCs treated with HgCl₂ for 1 and 3 hours at 37 °C bind to laminin in a concentration dependent manner at high concentrations (5 and 50 μM) (Fig. 2A&B), but not at low concentrations (0.1 μM and 0.5 μM, data not shown). Treatment of whole blood samples with 0.1 μM and 0.5 μM of HgCl₂ for 24 hours resulted in enhanced RBC adhesion to laminin while incubation with 1 μM HgCl₂ for 24 hours did not yield significant increase in adherent RBC numbers (Fig. 2C, one-way ANOVA, p>0.05). In order to prevent excessive hemolysis during long term exposure, we incubated the low concentration samples at room temperature rather than at 37 °C.

Figure 2. Mercuric ion induced RBC adhesion to immobilized laminin.

(A) Adhesion tests were performed with whole blood samples that were incubated with HgCl₂ at concentrations of 0 (control), 5 μM, and 50 μM for 1 hour at 37 °C. Adhesion of RBCs to immobilized laminin was significantly mediated at increasing concentrations of HgCl₂. (B) Similarly, incubation of whole blood with HgCl₂ for 3 hours mediated concentration dependent RBC adhesion to laminin. (C) Whole blood samples were incubated at room temperature for 24 hours, rather than 37 °C, to eliminate severe hemolysis. RBC adhesion to laminin is enhanced at low-concentration of HgCl₂ while further increasing the concentration resulted in a significant drop in adherent RBC numbers. Error bars indicate the standard error of the mean (±SEM) from ten independent experiments. Each condition had its own control group in independently performed experiments. P values are calculated based on one-way ANOVA test.

3.2. Hg²⁺ induced RBC adhesion is modulated via BCAM-Lu and membrane sulfatide depending on incubation time

We carried out our inhibition experiments to identify the possible modulators of the Hg²⁺ driven RBC adhesion to LN (Fig. 3). For these experiments, the samples were first treated with 50 μM HgCl₂ for 1h (Fig. 3A) and 3h (Fig. 3B) or with 0.5 μM HgCl₂ for 24h (Fig. 3C). Then, the samples were incubated with lactadherin (10 μg/mL) or heparin (50 U/mL) to block posphadetylserine and sulfatide molecules on the surface of RBC respectively. To determine whether RBC membrane bound BCAM/Lu, heparan sulfate, or sulfatide would modulate Hg²⁺-mediated RBC adhesion, we first incubated the LN functionalized microchannels with these molecules at the concentrations of 10, 50, and 200 μg/mL respectively to saturate the relevant adhesion sites on LN and performed the adhesion experiments thereafter. Our findings indicated that BCAM-Lu significantly inhibited the adhesion of RBCs treated with 50 μM of HgCl₂ for 1 hour while the difference was borderline significant for the heparin treatment (Fig. 3A&B, p<0.05 & p=0.059, one-sample Wilcoxon signed rank test). Inhibition via both heparin and sulfatide significantly reduced the adhesion of 50 μM and 0.5 μM HgCl₂-treated RBCs to LN (Fig. 3B&C, p<0.05, one-sample Wilcoxon signed rank test). We also tested the effect of anionic polysaccharides, namely heparan sulfate, on Hg²⁺-mediated RBC adhesion as a negative control. In contrast to heparin and sulfatide, heparan sulfate had no statistically significant effect on Hg²⁺-mediated RBC adhesion to LN (Fig.3A–C, p>0.05, one-sample Wilcoxon signed rank test). Each molecule tested for the inhibition experiments had their own control group. Inhibition via both heparin and sulfatide significantly reduced the adhesion of 50 μM and 0.5 μM HgCl₂-treated RBCs to LN (Fig. 3B&C, p<0.05, one-sample Wilcoxon signed rank test). We also tested the effect of anionic polysaccharides, namely heparan sulfate, on Hg²⁺-mediated RBC adhesion as a negative control. In contrast to heparin and sulfatide, heparan sulfate had no statistically significant effect on Hg²⁺-mediated RBC adhesion to LN (Fig.3A–C, p>0.05, one-sample Wilcoxon signed rank test). Each molecule tested for the inhibition experiments had their own control group.

Figure 3. Modulation of HgCl₂-mediated RBC adhesion to immobilized laminin by heparin, sulfatide, BCAM-Lu, heparan sulfate, and lactadherine.

(A) BCAM-Lu significantly inhibited the adhesion of RBCs that had been treated with 50 μM of HgCl₂ for 1 hour at 37 °C. Effect of heparin on lowering the total number of adherent RBCs was borderline significant (p=0.059). (B) Blocking the laminin channels with heparin and sulfatide both resulted in decreased RBC adhesion after a 3 hour, 50 μM HgCl₂ treatment of the blood samples at 37 °C. (C) Similarly, RBCs from the samples treated with 0.5 μM HgCl₂ for 24 hours at room temperature had lower adhesion in heparin and sulfatide blocked channels. Each molecule tested had their own control group where HgCl₂ treated samples were analyzed in PBS-washed laminin channels. *Indicated statistically significant differences at p<0.05 based on a non-parametric one-sample Wilcoxon signed rank test. Error bars represent the standard error of the mean (SEM) from five to eight different experiments with fresh whole blood samples from individual healthy donors.

3.3. Hg²⁺ causes morphological changes in the RBC structure

SEM images of mercuric ion (Hg²⁺) treated RBCs at 5000 X magnification illustrate non-treated RBCs (Fig. 4A), treated RBCs with a final concentration of 50 μM HgCl₂ for 1 hour (Fig. 4B) and 3 hours (Fig. 4C) at 37 °C, and with a final concentration of 0.5 μM HgCl₂ for 24 hours at room temperature (Fig. 4D). As shown in Fig. 4B, C, and D, Hg²⁺ exposure changed the RBC morphology, resulting in loss of the characteristic biconcave shape and formation of echinocytes. However, we did not observe a meaningful contribution of these cells to total adhesion levels.

Figure 4. Scanning Electron Microscopy (SEM) images of normal and HgCl₂-treated RBCs at 5000X magnification.

(A) Non-treated RBCs (control) possess their normal biconcave morphology. (B) RBCs treated with 50 μM HgCl₂ for 1h in 37 °C, (C) 3h in 37 °C, and (D) 0.5 μM for 24h in room temperature show morphological differences in which some of the RBCs had turned into echynocytes at the end of the incubation period as indicated by the white arrows. Scale bars represent a length of 5 μm.

The number of adherent RBCs to immobilized laminin gradually increased at increasing Hg²⁺ concentrations (5 or 50 μM HgCl₂ for 3 hour incubation, Fig. 5A–C). Individual dots in Figures 5A through 5C illustrate the laminin-adherent RBCs from non-treated (control), 5 μM, and 50 μM HgCl₂-treated samples respectively. Each interrogation window corresponds to the unit area in which the quantification was performed. Microscopic images of individual RBCs are also shown with healthy-looking characteristic biconcave shape (Fig. 5A). Surface adherent RBCs from 5 μM, and 50 μM HgCl₂-treated blood samples are heterogeneous in terms of their morphology containing biconcave RBCs, early echinocytes, and echinocytes (Fig. 5B&C).

Figure 5. A typical overview and Close up view of RBCs adhered to laminin-immobilized channels.

(A) Non-treated RBCs with negligible adhesion to laminin, (B) 3h treated with final 5 uM mercuric ion, (C) 3h treated with final 50 uM mercuric ion. The number of adhered RBCs were increased by increasing mercuric ion concentration. RBCs with characteristic biconcave shape along with RBCs lacking biconcave morphology and turn to echinocytes are shown in small boxes which is high-resolution phase-contrast images of adhered RBCs inside the microchannels. The coloured dots represent an individual adhered RBC. Large scale bar reperesents a length of 1 mm, small scale bar represents a length of 5 μm.

3.4. Increased LDH activity indicates significant hemolysis following 24h treatment of blood with HgCl₂

Finally, we quantitatively evaluated the effect of HgCl₂ on the hemolysis of human RBCs by analyzing the plasma color change and through a direct measurement of LDH activity. Plasma color turns red as hemolysis occurs. The photographs in Figure 6A demonstrate plasma colors of blood samples treated in varying conditions. The tubes labeled 1 contain fresh blood samples as control. The tubes labeled 2 contain control samples incubated in the absence of HgCl₂. The tubes numbered 3 through 7 were treated with HgCl₂ at concentrations of 0.1, 0.5, 1, 5, and 50 μM respectively (Fig. 6A). We did not observe significant color changes following incubation except in samples incubated at 37 °C for 24 hours – both in control and treated samples. Level of LDH activity in these samples also significantly increased, suggesting elevated hemolysis as shown in Figure 6B. Notably, the difference in LDH activity between control and treated samples was negligible (Fig. 6B).

3. DISCUSSION

Adhesion of healthy RBCs to the endothelium or relevant matrix proteins is negligible. However, in certain pathological conditions, RBC adhesion significantly increases, which is linked to the disease type that can be of hematologic⁴³, inflammatory⁴⁸, and/or endothelial⁴⁹ origin. This elevated adhesive phenotype may trigger additional complications, leading to a deteriorated disease state. Recently, different forms of mercury toxicity have been increasingly recognized as a potential cardiovascular risk factor³⁵. The toxic effects of mercury on the cardiovascular system have been attributed to its ability to enhance the production of free radicals and ROS as well as increased LDL oxidation and translocation of PS towards the outer cell membrane^17,31,50. More recently, aberrant RBC adhesion to endothelial cells induced by mercury exposure has been demonstrated, which may also be regarded as a potential mechanism in the initiation and propagation of mercury-related CVD¹².

Sulfatides play a critical role in CVD⁵¹. For example, serum sulfatide levels have been suggested to predict the incidence of CVD in patients with end-stage renal disease⁵². The role of membrane sulfatides in the nervous system, immunity, platelet adhesion and aggregation, bacterial and viral infections is also well recognized⁵³. However, the role of sulfatides in cellular adhesion and its relation to CVD is not yet clearly understood. It has been shown that sulfatides are located on the outer surface of the RBC membrane and they have binding affinity to three proteins, vWF (von Willebrand factor), TSP (thrombospondin) and laminin^54,55. In this study, we describe a novel mechanism in which Hg²⁺-mediated adhesion of RBCs to an extracellular matrix protein, laminin, takes place via the sulfatides and may contribute to the development of the mercury toxicity related CVD.

Laminin is an adhesive molecule found within the sub-endothelial matrix and does not interact with blood under normal circumstances. In contrast, exposure of endothelial cells to toxic levels of mercury may induce damage to the endothelial lining and thus allowing direct contact between laminin molecules and blood cells^33,37. Therefore, it is plausible to postulate that mercury induced adhesion of RBCs to laminin in vivo may deteriorate the endothelial health in the context of CVD by initiating further signaling events due to locally impaired flow dynamics. Additionally, laminin is an ideal candidate in this study because an adhesive mechanism has already been established between sulfatides and laminin.

Previous studies on Hg²⁺-mediated RBC adhesion were based on endothelial-RBC interactions or performed in the absence of in plasma and other blood cells (WBCs and platelets) using buffered RBC suspension¹². Although we have observed similar findings regarding Hg²⁺-mediated adhesion of washed and isolated RBCs in our system (Fig. S1), one of the main goals of this study was to assess this phenomenon in the presence of white blood cells and plasma, which would more accurately represent the physiological conditions. Therefore, we used blood samples collected in Na-citrate containing vacutainers for the Hg²⁺ experiments. It should be noted that Na-citrate has a binding affinity for the Hg²⁺ ions (Hg binding constant for citrate is 10.9d, for Ethylenediaminetetraacetic acid (EDTA) 21.5d and for cystein 14.4d⁵⁶) and may have reduced the available ion amount for RBCs during incubation. Further, metal binding to plasma proteins, such as serum albumin, may also affect the dynamics of interactions between Hg²⁺ and RBCs when using whole blood samples. Nevertheless, we have observed significantly increased RBC adhesion even at the least HgCl₂ concentration (0.1 μM) used in this study.

Acute symptoms of mercuric ion toxicity were simulated by addition of high doses of HgCl₂ into whole blood, while chronic conditions were generated via adding a lower concentration of HgCl₂ and extending the incubation period. Reported ranges of blood mercury levels in vivo are 16,000 μg/L (58 μM) to 11 μg/L (0.055 μM) in blood⁵⁷. Therefore, the range of blood HgCl₂ concentrations in this study has been determined to be between 0.1 μM to 50 μM, which correspond to chronic and acute toxic conditions respectively. Hg and membrane interactions were reported in physiologically relevant experimental conditions (i.e., 100 mM NaCl, pH 7.4), since salt concentration and pH govern chemial speciation of Hg as HgCl2, HgCl3− and HgCl4−^58,59. In this study, we did not control or quantify Hg speciation, as the main goal was to interrogate the impact of mercuric ions on RBC properties and adhesion.

Mercuric ion induced RBC adhesion to laminin was concentration dependent for all incubation times. We observed RBCs with characteristic biconcave shape that transformed to echinocytes upon mercury exposure, consistent with the previous findings in literature¹². A potential reason for this transformation is the preferential interaction of Hg2+ with phosphatidylcholine which is located in the outer hemilayer of the RBC membrane. This interaction alters the morphology of the RBC leading to the generation of echinocytes¹⁸. Formation of echinocytes was particularly dominant in samples incubated with 1 μM HgCl₂ for 24 hours as shown in Figure 4. The transformation of RBCs into echinocytes may have impaired the adhesion characteristics of those cells to LN due to a mechanism that may be out of the scope of this study (Fig. 2C).

In order to investigate membrane sulfatide mediated adhesion in the presence of whole blood proteins, we used intact laminin, as sulfatide binding requires an intact three-dimensional laminin structure⁶⁰. However, LG domain of laminin interacts with other cell surface molecules such as heparin, BCAM-Lu, and heparan sulfate. laminin does not bind to neutral glycolipids or gangliosides; however, it binds to certain glycolipids especially with high affinity for sulfatide⁶⁰. To identify which cell surface molecules would alter the adhesion driven by mercuric ions, we performed inhibition experiments. Our results revealed that BCAM-Lu and RBC membrane sulfatide have an important role in the initial adhesion of Hg²⁺ exposed RBCs to laminin (Fig. 3). Potentially, RBC spectrin-BCAM/Lu interaction was disrupted indirectly by Hg^{2+ 61,62} and caused BCAM-Lu mediated adhesion to immobilized laminin⁶³ (Fig. 7). Intermediate high concentration (50 μM HgCl_2, 3 hours) or prolonged lower concentration (0.5 μM HgCl_2, 24 hours) of mercuric ion exposure increased adhesion that was inhibited by heparin or sulfatide blockade. Laminin has heparin binding domains, and sulfatide binds to laminin via the same domain⁶⁴. Furthermore, heparin blocking indicated the charge-related interactions between laminin and sulfatide⁴² since laminin has positive functional groups and sulfatide has a negative functional group at physiological pH levels. Inhibition of PS via lactadherin had no effect in RBC adhesion levels (Fig. 3A–B), suggesting that Hg²⁺ mediates RBC adhesion to laminin in a fashion independent of increased PS translocation.

Figure 7. A model for Hg²⁺ induced red blood cell membrane changes and adhesion to laminin.

(0) Intact red blood cell membrane structure including lipid and non-lipid raft is shown. (1) In this model, RBC BCAM-Lu and spectrin interaction may be disrupted indirectly by Hg²⁺. (2) Mercury exposure may activate PLA2 (phospholipase A2), Ca²⁺ independent PLA2 (iPLA2) and lipoprotein-associated PLA2s (lp-PLA2), also known as platelet activating factor acetylhydrolase (PAF-AH), leading to release of PAF that activates sphingomyelinase, which may result in the formation of ceramide by hydrolysis of sphingomyelin interacting with cholesterol and glycosphingolipid⁴⁷. (3) This mechanism may result in ceramide rich platforms in the lipid rafts and ceramides have very poor affinity for cholesterol dislike sphingomyelin^{48, 49}. Cholesterol holds raft together. Poor affinity of ceramides for cholesterol and changes in membrane composition and distribution of various lipids in the lipid rafts may cause direct or indirect sulfatide accessibility.

Sulfatides, distributed on the outer leaflet of the plasma membrane, may interact with adjacent functional molecules that are located in the lipid rafts, which are detergent insoluble membrane fractions and enriched in cholesterol and sphingolipids and contain specific membrane proteins^65,66. By changing membrane structure or composition, movement¹³ or degradation⁶⁷ of gangliosides may increase exposure of sulfatides⁶⁸. Furthermore, mercuric ion exposure is known to activate PLA2 (phospholipase A2), Ca²⁺ independent PLA2 (iPLA2) and lipoprotein-associated PLA2s (lp-PLA2, also known as platelet activating factor acetylhydrolase (PAF-AH))⁶⁹. Mercury also induces the release of PAF, which activates sphingomyelinase, leading to the hydrolysis of sphingomyelin, forming ceramide which is in contact with cholesterol and glycosphingolipid⁷⁰. This mechanism yields ceramide rich platforms in the lipid rafts, which may deplete cholesterol^71,72 and may cause membrane modification, resulting in sulfatide accessibility (Fig. 7).

The current literature regarding the role of mercuric ions in mediating RBC adhesion to endothelial cells or endothelium-associated proteins are mainly explained by PS-related mechanisms. RBC membrane sulfatides bind to laminin with high specificity and affinity and have been accounted for enhanced RBC adhesion in sickle cell disease. Here, we show that mercuric ions lead to increased RBC adhesion to LN via membrane sulfatides, and notably, this adhesion was not mediated by PS as incubating samples with lactadherin (a PS-binding protein) did not reduce RBC adhesion.

In future studies, the in vitro platform presented here can be expanded to assess the role of mercury exposure in contributing to abnormal adhesion of other blood cell types, including platelets, as it has been shown that mercuric ion activates platelets¹⁵ and sulfatides are thought to initiate platelet aggregation via a pathway involving P-selectin⁷³. Mercuric ion exposure may trigger sulfatide mediated platelet adhesion on laminin, as well as perturbations in red cell structure and function. Implications of platelet activation in potentially enhanced RBC-platelet interactions are subject to a future study.

In conclusion, we have observed that treatment of whole blood with varying concentrations of HgCl₂ up to 24 hours caused abnormal RBC adhesion to a sub-endothelium protein, laminin, in a microfluidic adhesion assay. We found that the effect of HgCl₂ on RBC adhesion was mediated by BCAM-Lu within the first hour of exposure while RBC membrane sulfatides did not play a role in this process. Adhesion of HgCl₂-treated RBCs longer than 1 hour took place via a novel mechanism whereby mercuric ions facilitated the accessibility of membrane sulfatide groups that were recognized by laminin. Notably, BCAM-Lu did not significantly contribute to the adhesion of RBCs to laminin when the samples were treated with Hg²⁺ longer than 1 hour. Overall, these results increase our understanding of the toxic effects of mercury on the adhesive behavior of RBCs to a biologically relevant substrate, laminin, which may contribute to the development of certain diseases such as thrombosis and CVDs.