Dolutegravir potentiates platelet activation by a calcium-dependent, ionophore-like mechanism

Abstract Dolutegravir is a highly potent HIV integrase strand transfer inhibitor that is recommended for first-line anti-retroviral treatment in all major treatment guidelines. A recent study has shown that people taking this class of anti-retroviral treatment have a substantially higher risk of early-onset cardiovascular disease, a condition shown previously to be associated with increased platelet reactivity. To date, few studies have explored the effects of dolutegravir on platelet activation. Accordingly, the current study was undertaken with the primary objective of investigating the effects of dolutegravir on the reactivity of human platelets in vitro. Platelet-rich plasma, isolated platelets, or buffy coat cell suspensions prepared from the blood of healthy adults were treated with dolutegravir (2.5–10 µg/ml), followed by activation with adenosine 5’-diphosphate (ADP), thrombin, or a thromboxane A2 receptor agonist U46619. Expression of platelet CD62P (P-selectin), formation of heterotypic neutrophil:platelet aggregates, and calcium (Ca2+) fluxes were measured using flow cytometry and fluorescence spectrometry, respectively. Dolutegravir caused dose-related potentiation of ADP-, thrombin- and U46619-activated expression of CD62P by platelets, as well as a significant increases in formation of neutrophil:platelet aggregates. These effects were paralleled by a spontaneous, receptor-independent elevation in cytosolic Ca2+ that appears to underpin the mechanism by which the antiretroviral agent augments the responsiveness of these cells to ADP, thrombin and U46619. The most likely mechanism of dolutegravir-mediated increases in platelet cytosolic Ca2+ relates to a combination of lipophilicity and divalent/trivalent metal-binding and/or chelating properties of the anti-retroviral agent. These properties are likely to confer ionophore-type activities on dolutegravir that would promote movement of Ca2+ across the plasma membrane, delivering the cation to the cytosol where it would augment Ca2+-dependent intracellular signaling mechanisms. These effects of dolutegravir may lead to hyper-activation of platelets which, if operative in vivo, may contribute to an increased risk for cardiometabolic co-morbidities.


Introduction
Integrase strand transfer inhibitors (INSTI) are a class of antiretroviral agents that inhibit HIV-1 replication by blocking strand transfer reactions catalyzed by HIV-1 integrase (Smith et al. 2018). The INSTI have shown sustainable virological efficacy (Snedecor et al. 2019;Ambrosioni et al. 2020) with mild (depression, vertigo, and sleep disturbances) or no side-effects reported (Brehm et al. 2019). An improved metabolic profile in lipids (total cholesterol and low-density lipoprotein cholesterol ratio) has also been reported with combined anti-retroviral therapy (cART) regimens that contain INSTI compared to that observed with cART treatments that contain protease inhibitors (PIs) or efavirenz (Taramasso et al. 2018;Bagella et al. 2019). Dolutegravir (GSK-1349572) was the third INSTI approved by the U.S. FDA and regimens containing this drug have been recommended as the preferred first-and second-line anti-retroviral therapies (ART) for all populations (World Health Organization, 2019).
Despite the seemingly satisfactory safety profile of dolutegravir, Neesgaard et al. (2022) recently reported that initiation of anti-retroviral therapy with an INSTI in a large cohort of European and Australian, INSTI-naïve, participants living with HIV was associated with an early onset increased incidence of cardiovascular disease (CVD) -including myocardial infarction and strokes, and a need for invasive cardiovascular proceduresduring the first two years of exposure, which declined progressively thereafter. During the first six months of treatment with an INSTI, the adjusted incidence rate ratio for CVD events was 1.85 (CI: 1.44-2.39) compared with no exposure. However, indices of platelet activation and possible relationships with CVD events were not reported.
Platelet markers, such as mean platelet volume, circulating levels of microparticles, oxidation products, platelet-derived soluble CD62P (P-selectin), and CD40 ligand have been found to be elevated in people with cardiometabolic risk factors, such as obesity, dyslipidemia, or Type 2 diabetes mellitus (Barale and Russo 2020). Few studies to date have explored the potential effects of dolutegravir on platelet activation. One such study (Taylor et al. 2019) reported that dolutegravir inhibits ADPmediated platelet aggregation ex vivo following oral administration to healthy adult volunteers, as well as collagen-mediated activation of integrin aIIbb3 and up-regulated expression of CD62P and CD63 following exposure of isolated platelets from healthy volunteers to dolutegravir in vitro.
While this seemingly represents a potentially beneficial effect of dolutegravir, further exploration of the effects of this agent on human platelets is clearly needed. Accordingly, the current study aimed to investigate the effects of dolutegravir on platelet activation as measured by CD62P expression and formation of neutrophil:platelet (NP) aggregates following exposure of the cells to several different types of receptor agonists. From a mechanistic perspective, the effects of dolutegravir on alterations in platelet cytosolic Ca 2þ concentrations were also investigated.

Ethics statement
Permission to undertake this study, and to draw blood from healthy adult volunteers was granted by the Research Ethics Committee of the Faculty of Health Sciences (University of Pretoria). This study was performed in full compliance with the World Medical Association Declaration of Helsinki 2013 (Approval #116/2017 and 605/2018). Written informed consent was obtained from all blood donors, each of whom underwent a routine health check (including measurement of blood pressure) by an experienced, qualified nurse prior to the blood draw [16 females, 9 males, mean age (± SD) of 31 (±11) years].

Platelet-rich plasma and buffy coat preparations
To prepare platelet-rich plasma (PRP), blood (anti-coagulated with five units of preservative-free heparin/ml blood) was centrifuged at 250 Â g for 10 min at room temperature within 15 min of venepuncture as described previously (Nel et al. 2016). The essentially erythrocyte-and leukocyte-free upper layer of PRP was decanted and used in the experiments described below. Buffy coat suspensions prepared by sedimentation of heparinized venous blood were used in assays of NP aggregate formation (Nel et al. 2017).

Purified platelet suspensions
PRP was diluted four-fold with phosphate-buffered saline (PBS, 0.15 M, pH 6.8) containing ethylene glycol-bis(b-aminoethyl ether)-N-N-N'-N'-tetraacetic acid (EGTA, 0.6 mM final concentration) to minimize spontaneous platelet activation. Samples were then centrifuged at 250 Â g for 10 min at 20 C to deplete residual erythrocytes and leukocytes. The platelet-enriched supernatants were then centrifuged at 800 Â g for 20 min and the cell pellets re-suspended in PBS and counted using a Sysmex XN 20000 Automated Hematology Analyzer (Sysmex Corporation, Kobe, Japan). The median (IQR) platelet count of plateletenriched supernatants was 1205 (805 À 1505) Â 10 9 /L.

CD62P expression on platelets
In order to determine the expression of CD62P on platelets, the method of Anderson et al. (2020) was followed. In brief, to PRP (20 ll) in a final volume of 1 ml Hanks' balanced salt solution (HBSS, indicator-free, 1.25 mM calcium Ca 2þ , pH 7.4), either DMSO (control) or dolutegravir (at final concentrations of 5 and 10 mg/ml) were added. The tubes were then incubated for 10 min at 37 C after which the platelets were activated with ADP (100 mM) for 5 min. Flow cytometry was then performed to determine the magnitude of expression of CD62P according to the proportion (%) of CD42a þ platelets expressing the adhesion molecule. In an additional series of experiments, the effects of dolutegravir, at a fixed concentration of 10 lg/ml, on the expression of CD62P by platelets activated with thrombin (0.6 NIH units/ml) or U46619 (5 mM) were investigated.
Following incubation with the various test agents, the platelet suspensions were stained with 5 ml each of a murine anti-human platelet CD42a-phycoerythrin (PE)-labeled monoclonal antibody (Becton Dickenson, San Jose, CA) and an anti-human CD62Pfluorescein isothiocyanate (FITC)-labeled monoclonal antibody (Beckman Coulter, Miami, FL) to detect total and activated platelet populations, respectively. After 15 min of incubation in the dark, the samples were analyzed on a Gallios flow cytometer (Beckman Coulter) and the results expressed as the percentage of activated platelets, with 50,000 cells interrogated during each measurement. The gating and analytical strategies used in a typical experiment have been described elsewhere (Anderson et al. 2018).

Neutrophil:platelet aggregate formation
Buffy coat suspensions (30 ll) in a final volume of 1 ml HBSS were incubated for 5 min at 37 C in the presence of either DMSO (control) or dolutegravir (at final concentration of 10 mg/ ml) after which the cells were activated with thrombin (0.6 NIH units/ml) for 5 min. The cell suspensions were treated with 5 ll of each of the following murine anti-human fluorochromelabeled monoclonal antibodies to detect neutrophils, platelets, and total leukocytes, respectively: anti-CD16-PE-cyanin 5 (Beckman Coulter), anti-CD42a-PE (Becton Dickenson, San Jose, CA), and anti-CD45-Krome Orange (Beckman Coulter), and then incubated for 15 min at room temperature in the dark. Cell suspensions were then analyzed using a Gallios flow cytometer. NP heterotypic aggregate formation was determined as CD16 þ / CD45 þ neutrophils co-expressing CD42a. Results were expressed as the relative mean fluorescence intensity (MFI) of CD42a as emitted by CD16/CD45 neutrophils as an index of the magnitude of the interaction of platelets with individual neutrophils using the gating strategy described previously (Nel et al. 2017).

Platelet cytosolic Ca 21 fluxes
Fluo-8 acetoxymethyl ester (Fluo-8-AM) was used as the fluorescent, Ca 2þ -sensitive indicator for these experiments (Grynkiewicz et al. 1985). In this series of experiments, purified platelets suspended in PBS were incubated at 37 C for 5 min followed by addition of Fluo-8-AM (2 mM) and a further incubation period of 25 min at 37 C. The platelets were then pelleted by centrifugation at 500 Â g for 15 min, and then re-suspended in PBS. For measurement of cytosolic Ca 2þ fluxes, 100-300 ml platelet suspension was added to HBSS (final volume of 3 ml containing a maximum of 1 Â 10 8 cells) and incubated for 10 min at 37 C. The Fluo-8-AM-loaded platelets were then transferred to disposable reaction cuvettes maintained at 37 C in a Hitachi 650 10S fluorescence spectrometer (Hitachi Ltd., Tokyo, Japan) with excitation and emission wavelengths set at 480 and 525 nm, respectively. Importantly, at these wavelengths, no evidence of dolutegravir-induced autofluorescence was detected. After 1-2 min, dolutegravir (2.5-10 mg/ml) or thrombin (1.0 NIH unit/ ml) as a positive control and a check for viability was added to the platelets and alterations in fluorescence intensity measured over a 5 min time course.
To evaluate possible involvement of mobilization of intracellular Ca 2þ by dolutegravir-treated platelets, the effects of selective inhibitors of phospholipase C (PLC), inositol triphosphate-sensitive Ca 2þ channels (IP3) and nonselective plasma membrane Ca 2þ channels (NSCC), on the alterations of cytosolic Ca 21 , were investigated. The cells were pretreated for 7 min at 37 C with the selective inhibitors of each of these mediators of Ca 2þ mobilization, i.e. U73122 (5 mM), 2-APB (50 mM), and SKF 96365 (50 mM), respectively. Following pre-incubation with U73122, 2-APB or SKF 96365, dolutegravir (10 mg/ml) was added to the cells and alterations in fluorescence intensity monitored over a 5 min time course.
In an additional series of experiments, platelets loaded with Fluo-8-AM were suspended in PBS with EGTA (300 mM), an extracellular Ca 2þ -chelating agent, to investigate dolutegravirmediated alterations in platelet cytoplasmic Ca 2þ concentrations in the absence of extracellular Ca 2þ . These results, from eight separate experiments using cells from seven different donors, are presented as the basal and peak Ca 2þ concentrations calculated as described previously (Grynkiewicz et al. 1985) and as representative traces.

Expression and statistical analysis of results
The results of each series of experiments are expressed as the median values and interquartile range (IQR) with the numbers of different donors (n) and replicate experiments clearly indicated. Statistical analyses were performed using Stata 17.0 SE (StataCorp, College Station, TX) software, with levels of statistical significance calculated using the Wilcoxon sign-rank test for comparisons of non-parametric data. A p-value < 0.05 was considered significant.

Results
Effects of dolutegravir on ADP-, thrombin-and U46619activated expression of CD62P by PRP The effects of dolutegravir (5 and 10 mg/ml) on expression of CD62P by ADP-activated platelets, relative to those of drug-free control cells, are shown in Figure 1. The CD62P expression levels of the drug-free, unstimulated control system (in the absence of ADP) were minimal, increasing significantly following addition of ADP. Pretreatment of PRP with dolutegravir (5 and 10 mg/ml) resulted in dose-related stimulation of ADP-activated up-regula-tion of CD62P expression, with significance levels of p < 0.0001 and p < 0.0001, respectively.
As shown in Figure 2, augmentation of CD62P expression by dolutegravir (10 mg/ml) was also observed following activation of platelets with either thrombin or U46619 (p ¼ 0.0001 and p ¼ 0.0002, respectively). The median percentages of stimulation of expression of CD62P by dolutegravir (10 mg/ml)-treated platelets activated with ADP, thrombin or U46619 were 75%, 195% and 69%, respectively. With respect to the effects of dolutegravir alone (in the absence of ADP, thrombin or U46619), a slight augmentation of expression of CD62P was observed at 10 mg/ml of this agent, the median (IQR) values for the untreated control system and system treated with dolutegravir being 0.23 (0.03-0.27) and 0.38 (0.12-0.49), respectively.

Effects of dolutegravir on neutrophil:platelet aggregation
The effects of dolutegravir (at 10 mg/ml) on the formation of NP heterotypic aggregates following addition of thrombin (0.6 NIH units/ml) to buffy coat preparations are shown in Figure 3. Relative to the drug-free, thrombin-activated control system, treatment with dolutegravir resulted in a significant potentiation of the formation of NP aggregates (p ¼ 0.0005), representing a mean percentage increase of 22%. Dolutegravir alone, in the absence of thrombin, caused only a slight increase in the formation of NP aggregates, the median (IQR) values for the untreated control system and the system treated with dolutegravir being 5.3 (4.6-5.9) and 5.9 (5.0-6.0), respectively.

Effects of dolutegravir on platelet cytosolic Ca 21 fluxes
As shown in Figure 4, addition of dolutegravir (at 2.5, 5. and 10 mg/ml) to plasma-free, isolated pure platelets pre-loaded with Fluo-8-AM resulted in an abrupt concentration-dependent increase in fluorescence intensity that quickly leveled off and remained elevated throughout the time course of each experiment, consistent with an immediate and sustained elevation in the cytosolic Ca 2þ concentration that was abolished when the cells were suspended in Ca 2þ -depleted medium. The median (IQR) increases in cytosolic Ca 2þ concentrations for dolutegravir-treated platelets (2.5, 5, and 10 mg/ml), were 109 (70-233 nM), 408 (179-444 nM), and 458 (356-562 nM), respectively.

Discussion
The findings of the current study showed that dolutegravir, at concentrations of 5-10 mg/ml, caused statistically significant potentiation of several activities of human platelets in vitro. These included; (i) up-regulation of expression of the a-granulederived adhesion molecule CD62P measured in cells activated with ADP, thrombin, or U46619 (a thromboxane A 2 surrogate); (ii) formation of NP aggregates following exposure to thrombin; and, (iii) receptor-independent mobilization of intracellular Ca 2þ . The concentrations of dolutegravir at which augmentation of receptor-mediated platelet activation were observed are close to those reported to occur in plasma following oral administration of this anti-retroviral agent. These range from 3.34-6.16 mg/ml, depending on whether tablets or suspensions at a daily dose of 50 mg were administered (Cottrell et al. 2013).
Platelet CD62P (P-selectin) is widely recognized as being a mediator of platelet pro-inflammatory and pro-thrombotic activities that drive heterotypic aggregation and activation of neutrophils, monocytes, and endothelial cells (Th eorêt et al. 2011;Cerletti et al. 2012). CD62P via binding to its ligand, P-selectin glycoprotein ligand-1 (PSGL-1), plays a central role in the interactions between platelets, leukocytes, and endothelial cells (Yun et al. 2016). From a mechanistic perspective, dolutegravir appears to prime (sensitize) platelets to hyper-respond to receptor-mediated activation, irrespective of receptor type, including those activated by ADP (P2Y1/P2Y12), thrombin (PAR-1/PAR-4) and thromboxane A2 (TXA2). Mechanistically, elevations in cytosolic Ca 2þ concentrations represent a critical, early, and common event in platelet activation initiated by all of these receptor types (Varga-Szabo et al. 2009). In this context, dolutegravir-mediated spontaneous, receptor-independent increases in platelet cytosolic Ca 2þ concentrations appear to underpin the mechanism by which the anti-retroviral agent augments the responsiveness of these cells to ADP, thrombin, and U46619. Observations that dolutegravir at the concentrations used did not induce spontaneous, as opposed to receptor-mediated, activation of platelet CD62P expression may indicate that threshold levels of cytosolic Ca 2þ were not exceeded in this setting (Mazzucato et al. 2002).
Although prior literature on the effects of Ca 2þ mobilization in eukaryotic cells is extremely limited, two studies are, however, noteworthy. In the first of these, Al Mamun Bhuyan et al. (2016) reported that exposure of isolated human erythrocytes to dolutegravir, albeit for 48 h, at concentrations comparable to those used in the current study (4.77-19.08 mM equivalent to % 2-8 mg/ml), resulted in an influx of extracellular Ca 2þ that triggered oxidative stress, eryptosis, and hemolysis. The authors did not speculate on the mechanisms of dolutegravir-mediated Ca 2þ influx. More recently, Wang et al. (2020) investigated the effects of prolonged exposure (24 h) of a human non-small cell lung cancer cell line to a novel derivative of dolutegravir at concentrations of 5-20 mM on cell viability and alterations in cytosolic Ca 2þ concentrations. Those authors observed that exposure of the cancer cell line, but not a control non-cancerous cell line, to the dolutegravir derivative caused dose-related induction of cell death via a pro-apoptotic mechanism associated with both elevated concentrations of cytosolic Ca 2þ and cellular oxidative stress that was linked to inhibition of the Ca 2þ -sequestering endo-membrane Ca 2þ -ATPase (Al Mamun Bhuyan et al. 2016). These effects were not, however, observed following exposure of the cancer cell line to dolutegravir per se at a fixed concentration of 10 mM, possibly due to this being a threshold concentration for detection of effects on cytosolic Ca 2þ .
The findings of the current study differ from those reported by Taylor et al. (2019) who observed that collagen-mediated activation of platelets and expression of CD62P was inhibited in the presence of dolutegravir. These differences likely relate to the different types of activating stimuli used in both studies.
With respect to potential mechanisms of dolutegravir-mediated alterations in cytosolic Ca 2þ concentrations and cell viability, the current study differs in a number of aspects from those performed by Al Mamun Bhuylan et al. and Wang et al. These include differences in cell type, duration of exposure to dolutegravir, and apparent lack of cytotoxicity. The latter contention is based on our findings that rather than inhibiting platelet activation, which would have been the case in the event of cytotoxicity, the opposite was observed. Based on the results of our intracellular Ca 2þ -focused experiments, we believe that the most likely mechanism of dolutegravir-mediated increases in platelet cytosolic Ca 2þ relates to the combination of lipophilicity and divalent/trivalent metal-binding/chelating properties of this antiretroviral agent (Song et al. 2015;Grießinger et al. 2016;Thierry et al. 2017). These properties are likely to confer ionophore-type properties on dolutegravir that would promote movement of Ca 2þ across the plasma membrane, delivering the cation to the cytosol where it is likely to augment Ca 2þ -dependent intracellular signaling mechanisms. This contention is supported by the observation that the abrupt dolutegravir-mediated increase in platelet cytosolic Ca 2þ was markedly attenuated following depletion of extracellular Ca 2þ . Furthermore, selective inhibitors of PLC, IP3 channels, and NSCC did not attenuate the dolutegravir-mediated increases in platelet cytosolic Ca 2þ concentrations which supports the contention that dolutegravir acts via an ionophore-like mechanism independent of receptor-mediated Ca 2þ signaling pathways. This observation is seemingly in contrast to the effects of ionomycin and A23187 on cell types (that did not include platelets) (Dedkova et al. 2000), indicating that dolutegravir at the concentrations used may act in a manner distinct from that of commonly-used ionophores.
Importantly, obesity and metabolic syndrome are associated with an accelerated atherothrombotic process, resulting from a combination of factors that includes platelet hyper-activity (Santilli et al. 2012). In the context of cART, associations with dyslipidemia (Tsai et al. 2017) and obesity (Bakal et al. 2018) have been described that were linked to an increase in platelet aggregation and reactivity, thus placing HIV-infected individuals at increased risk of CVD (Nkambule et al. 2017;Mesquita et al. 2018). As documented in recent reports, excess weight gain during therapy with dolutegravir (as well as some other INSTI) might therefore pose a serious risk for development of CVD and other cardiometabolic co-morbidities, such as insulin resistance and diabetes mellitus (Eckard and McComsey 2020;Gorwood et al. 2020;Taramasso et al. 2020;Hindley et al. 2021;Lahiri et al. 2021).
In addition, ongoing systemic inflammation observed in HIVinfected individuals, despite cART, has been identified as a risk factor for CVD that is associated with higher levels of D-dimer, as well as those of the inflammatory markers, interleukin (IL)-6 and C-reactive protein (CRP), as reported by the Strategies for the Management of Anti-retroviral Therapy (SMART) study (Borges et al. 2016).
Activated platelets also play a fundamental role in both inflammation and thrombosis that is associated with the release of granule-derived, pro-inflammatory mediators (Koupenova et al. 2018). Notably, HIV-infected individuals have impaired platelet function resulting from increased platelet activation, possibly as a result of microbial translocation and sustained exposure to microbial components in the circulation (Brenchley et al. 2006;Cassol et al. 2010). Although cART has been shown to are shown in the upper panel A as the magnitude of increase in cytosolic Ca 2þ following addition of dolutegravir (2.5, 5, and 10 mg/ml) to platelets suspended in Ca 2þ -replete medium and on the alterations of dolutegravir (10 mg/ml)-mediated Ca 2þ mobilization in the presence of selective inhibitors of PLC, IP3 and NSCC, namely U73122 (5 mM), 2-APB (50 mM) and SKF 96365 (50 mM), respectively. Representative traces from a single experiment are shown in the lower panel B. Dolutegravir (Dol, 10 mg/ml) as denoted by the arrow (#) was added to platelets suspended in Ca 2þ -replete (--) and -depleted (----) medium. An abrupt and sustained increase in cytosolic Ca 2þ was observed following addition of dolutegravir to platelets suspended in Ca 2þ -replete HBSS which was abolished in Ca 2þfree medium. effectively increase platelet counts following three months of treatment, platelet activation and dysfunction persist Mesquita et al. 2018;Madzime et al. 2021). Clearly, further activation of platelets by dolutegravir could have clinical consequences for cardiometabolic health, which should be explored in future studies.

Informed consent statement
Written informed consent was obtained from all blood donors who partook in the study.

Institutional review board statement
As noted, permission to undertake this study and to draw blood from healthy adult human volunteers was granted by the Research Ethics Committee of the Faculty of Health Sciences, University of Pretoria (Approval #116/2017 and 605/2018).

Author contributions
All authors contributed significantly to the conceptualization and planning of the study. MM, AJT, RA, GRT, HCS and JGN performed laboratory investigations; AJT, TMR, PWAM, RA and GRT contributed to analysis of the data and preparation of the figures, while all authors contributed to interpretation of data and compilation of the manuscript. All authors have approved this submitted version of the manuscript.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The study was supported from a grant awarded to HC Steel by the National Health Laboratory Research Trust of South Africa [Grant #004 94713]. C. Feldman was supported by the National Research Foundation of South Africa.

Data availability statement
Data can be made available upon reasonable request.