Cardiovascular disease (CVD) is an increasing cause of morbidity and mortality in HIV+ individuals receiving combination antiretroviral therapy (cART) [1,2]. HIV infection is associated with an increased risk of cardiovascular conditions, including atherosclerosis [3,4] and coronary heart disease [5,6] and an approximately two-fold increased risk of myocardial infarction [7,8]. Importantly, increased CVD risk in HIV+ individuals is independent of traditional risk factors [5–10] and risk prediction algorithms based on these factors alone used in the general community setting underestimate atherosclerosis in HIV+ individuals . Although cART and specific antiretroviral agents may potentiate CVD [12–14], the increased CVD risk observed in HIV+ individuals who are treatment naive , undergoing treatment interruption , and in elite controllers  suggests the involvement of cART-independent factors. Taken together, these observations suggest that unique mechanisms exist in HIV+ individuals which act in addition to traditional factors to increase CVD risk, and these effects persist despite viral suppression.
In HIV+ individuals, CVD is associated with markers of inflammation [17–19] and is increasingly reported in association with markers of monocyte/macrophage activation. We have shown that monocyte and innate immune activation persist in HIV+ individuals despite viral suppression [20,21], and markers of monocyte/macrophage activation, including soluble (s)CD163 [22,23], sCD14  and cellular monocyte activation markers [25–27], are associated with atherosclerosis and CVD in HIV infection. Despite the significant links between chronic inflammation/monocyte activation and increased CVD risk in HIV infection, the mechanism remains unclear.
Proinflammatory cytokines such as tumour necrosis factor (TNF) potentiate atherosclerosis; they activate endothelial cells and monocytes, resulting in increased expression of adhesion molecules, release of chemoattractants by endothelial cells, lipid transcytosis and oxidation of low-density lipoprotein (LDL) . Activated monocytes are recruited to inflamed blood vessels and migrate into the vascular neointima, wherein they endocytose lipids and either exit via reverse transendothelial migration or develop into foam cells by phagocytosing highly inflammatory oxidized low-density lipoprotein (oxLDL; via scavenger receptors SR-A and CD36) . Foam cells have reduced migratory capacity  and accumulate in the neointima wherein they release proinflammatory molecules contributing to further recruitment of monocytes and progression of atherosclerosis . The ability of macrophages to exit the neointima and remove cholesterol via reverse transendothelial migration is critical to retarding the development of atherosclerotic plaques and promoting plaque regression .
We hypothesized that monocyte activation in the setting of chronic HIV infection  promotes atherosclerosis. We have previously developed a novel in-vitro model of the initiation of atherosclerotic plaque formation that couples transendothelial migration of primary human monocytes across an activated endothelium with foam cell formation [33–35]. Here, we adapted this model to investigate the atherogeneic potential of monocytes isolated from HIV+ individuals and determine whether inflammatory factors elevated in HIV infection influence early atherosclerotic events mediated by monocytes.
Recruitment and blood processing
Blood was obtained from HIV+ donors recruited from the Department of Infectious Diseases, The Alfred Hospital, Melbourne, Australia, and healthy control donors of a similar age following written, informed consent. Peripheral blood mononuclear cells (PBMCs) were isolated within 2 h of sample collection and were either used immediately (for migration assays) or stored in liquid nitrogen for later mRNA and cholesterol efflux analysis. Monocytes were further purified from PBMC via negative selection using magnetic beads (Miltenyi Biotec, Cologne, Germany) as per the manufacturer's protocol, which yields monocytes with a purity of 80–85% as determined by flow cytometry (not shown). This study received ethical approval from The Alfred Research and Ethics Committee and from the Royal Women's Hospital Ethics Committee, Melbourne, Australia.
Cell migration assay and analysis
Hydrated collagen gels were prepared in a 96-well format as described previously [33,36]. Gels were incubated at 37°C for 4–6 days until use. Primary human umbilical vein endothelial cells (HUVEC) were prepared as described  and used without further passage. 2 × 104 HUVEC were added to each collagen gel and incubated in Medium 199 (Life Technologies, Carlsbad, California, USA) containing 20% human serum for 3 days to allow confluent monolayers to form. Media were prepared using a single batch of pooled human serum (pHS) prepared from six HIV-seronegative blood donors (Australian Red Cross Blood Service, Sydney, Australia) or autologous serum (from the same donor as the monocytes) as indicated; all sera were heat inactivated at 56 °C for 30 min before use. Silver staining was performed on selected wells, in addition to routine phase-contrast microscopy, to verify HUVEC monolayer integrity (Supplementary Fig. 1A) . HUVEC were activated with 10 ng/ml TNF for 4 h  or left unactivated, then freshly isolated PBMCs (2 × 105/well) or purified monocytes (5 × 104/well) added for 1 h at 37 °C. Nonmigrated cells were removed by washing and cultures incubated for a further 48 h as described . For TNF blocking, 10 or 20 μg/ml anti-tumour necrosis factor receptor (TNFR) I and anti-TNFRII (R&D Systems, Minneapolis, Minnesota, USA), or respective isotype controls, were added immediately following monocyte migration. Forty-eight hours after monocyte migration, reverse-migrated cells were removed and collagen gels stained with Oil Red O as described . Gels were excised from wells, mounted on glass slides and foam cells counted by bright field microscopy (×40). Foam cells were defined as cells containing Oil Red O stained vesicles within the cytoplasm and determined as a proportion of total migrated cells within the counted area of the gel (Supplementary Fig. 1B). To investigate the phenotype of migrated cells within the collagen matrix, cells were extracted from the collagen by washing the gels without fixation and incubating in 1 mg/ml collagenase D for 20 min at 37°C, after which they were macerated manually and incubated further for 20 min at 37 °C. Resulting cell suspensions were filtered through 100 μm mesh prior to staining for flow cytometric analysis. Cells extracted from collagen gels were stained with 0.2 μg/ml BODIPY 493/503 (Life Technologies, Carlsbad, California, USA) in 150 mmol/l NaCl for 20 min at room temperature and anti-TNF-PE, CD36-FITC, CD14-APC and CD11b-PE (all from BD Biosciences, San Jose, California, USA) as described . Cells stained with anti-CD14 APC and BODIPY were sorted using a BD influx cell sorter. The specificity of phenotyping antibodies was verified via the use of isotype control antibodies on selected samples (Supplementary Figure 2).
Live cell imaging
Purified monocytes were stained with PKH26 (Sigma-Aldrich, St Louis, Missouri, USA) according to manufacturer's instructions and added to HUVEC monolayers grown on collagen gels in 96-well optical plates (Greiner Bio-One, Frickenhausen, Germany). Cells were immediately incubated at 37°C in a humidified temperature-controlled chamber coupled to a DeltaVision deconvolution microscope (Applied Precision-GE Healthcare, Issaquah, Washington, USA) and images were captured in z-series on a charge-coupled device camera (AxioCamMRm Rev. 3, Carl Zeiss Microscopy, Jena, Germany) through a 20×, 0.45 NA lens. Detection and tracking of cell migration was performed using the tracking module of the Imaris software (Bitplane, Zurich, Switzerland).
Analysis of plasma lipids and soluble factors
Plasma TNF and CXCL10 levels were measured using commercial ELISA kits (human TNF-α ELISA, Cardinal Bio-research, New Farm, Australia, and Human CXCL10/IP-10 Quantikine ELISA KIT, R&D Systems, Minneapolis, Minnesota, USA). Cholesterol and oxidized LDL levels in plasma were measured using the HDL and LDL/VLDL Cholesterol Assay Kit and the Human Oxidized LDL ELISA Kit (CML-LDL) respectively (both from Cell Biolabs, San Diego, California, USA).
Cholesterol efflux assays
PBMCs (2.5–5 × 105) were incubated for 1 h at 37 °C in RPMI containing 10% pHS (RH-10), 1.5 μmol/l BODIPY-cholesterol (Avanti Polar Lipids, Alabaster, Alabama, USA), 6 μmol/l cholesterol and 300 μmol/l methyl-β-cyclodextrin (both from Sigma-Aldrich). Cells were washed twice with PBS (PBS without Mg2+ or Ca2+), resuspended in RH-10 and cholesterol was allowed to efflux from the cells for 30 min at 37 °C in either the absence or presence of 1 mmol/l of methyl-β-cyclodextrin. Cells were washed, fixed with 1% formaldehyde, immunostained as above and analyzed by flow cytometry.
Gene expression analysis of cholesterol efflux and metabolism intermediates
The expression of proteins involved in cholesterol metabolism and transport was assessed via quantitative real-time PCR (qPCR). Monocytes were isolated from frozen PBMC, rested for 4 h in RPMI containing 20% pHS at 37°C and total RNA was extracted. Potential DNA contamination was removed by DNase I treatment (37°C for 20 min, 75°C for 10 min, Roche, Basel, Switzerland). cDNA synthesis was performed using a combination (1 : 2) of oligo(dT) and random hexamer primers, respectively (Transcriptor First Strand cDNA synthesis kit, Roche) and qPCR analysis was performed using FastStart Universal SYBR Green Master Mix (Roche) for GAPDH, adenosine triphosphate (ATP)-binding cassette transporter A1 (ABCA1) and Acyl-coenzyme A: cholesterol acyltransferase (ACAT), and Brilliant SYBR II (Agilent Technologies, Santa Clara, California, USA) for ABCG1 and HMG-CoA reductase on a MX3005P qRT-PCR machine (Agilent Technologies). The primers used were as follows: GAPDH (Fwd 5′-CCATGGCACCGTCAAGGC-3′, Rev 5′-CCAGCATCGCCCCACTTG-3′), ABCA1 (Fwd 5′-GCACTGAGGAAGATGCTGAAA-3′, Rev 5′-AGTTCCTGGAAGGTCTTGTTCAC-3′), ABCG1 (Fwd 5′-CAGGAAGATTAGACACTGTGG-3′, Rev 5′-GAAAGGGGAATGGAGAGA-3′), ACAT (Fwd 5′-CAAGGCGCTCTCTCTTAGATGAAC-3, Rev 5′-GATAAAGAGAATGAGGAGGGCAATAA-3′) and HMG-CoA reductase (Fwd 5′-GGGACCAACCTACTACCTCAG-3′, Rev 5′-CGACCTGTTGTGAATCATGTGACTT-3′) at a final concentration of 280 nmol of each primer. PCR conditions were 95°C for 10 min then 40 cycles of 95°C for 15 s, 60°C for 1 min. Absolute quantification of gene expression was determined via comparison with a standard curve generated from serial dilution of a plasmid engineered to contain the amplicon of interest. Expression of all genes was standardized to GAPDH.
Data and statistical analysis
Nonparametric, two-tailed Mann–Whitney U test was used for unpaired data and Wilcoxon rank-sum test for paired data. Parametric unpaired t-test was used for data passing the D’Agostino–Pearson omnibus normality test. Statistical analysis was performed using GraphPad Prism software (GraphPad Software Inc., La Jolla, California, USA) where values of P < 0.05 were considered statistically significant.
HIV infection increases foam cell formation by monocytes following transendothelial migration
We have previously shown that HIV infection of monocyte-derived macrophages in vitro impairs their ability to reverse migrate across a HUVEC-collagen gel model of early atherosclerotic events , suggesting these cells may have a greater propensity to differentiate into foam cells, and thus be retained in the collagen matrix. To investigate whether HIV infection in vivo is associated with increased propensity for monocytes to develop into foam cells, we utilized a refined version of this model to analyze the atherogenic properties of monocytes isolated by negative selection from fresh whole blood. Ageing is well documented to increase the risk of atherosclerosis, thus we investigated the effect of HIV infection in younger HIV+ individuals and uninfected controls of a similar age. Monocytes were isolated from seven healthy controls [all men, median age (range) 29.0 (25–56) years] and eight HIV+ men [median age 36.5 (26–47) years]. Of the HIV+ individuals, four were receiving cART and had low (80 copies/ml, n = 1) or undetectable (<40 copies/ml, n = 3) viral loads and the remaining four were not receiving cART and had a median viral load of 23 400 copies/ml (range: 300–254 800 copies/ml).
When added to unstimulated HUVEC in media containing 20% pHS, a significantly increased proportion of monocytes from HIV+ individuals differentiated into foam cells as compared with monocytes from controls (median foam cells 18.9 vs. 0%, respectively, P = 0.004, Fig. 1A; all samples tested using the same batch of pHS). To determine whether factors present in HIV+ serum may further influence foam cell formation, we analyzed foam cell formation using culture media containing autologous serum rather than pHS and found that although foam cell formation by control monocytes was not significantly altered in the presence of autologous serum, serum from HIV+ individuals significantly increased foam cell formation by autologous monocytes (P = 0.03, Fig. 1A).
Consistent with our previous findings, activation of HUVEC with TNF prior to monocyte migration significantly increased foam cell formation by monocytes from control donors (P = 0.016 for both pHS and autologous serum, Fig. 1A vs. B) and also HIV+ monocytes in the presence of pHS (P = 0.008). As seen with unactivated HUVEC, autologous serum from HIV+ but not controls, significantly increased foam cell formation when monocytes were added to TNF-activated HUVEC (P = 0.02 vs. 0.30, respectively, Fig. 1B). The increased rate of foam cell formation by control monocytes following endothelial cell activation meant that differences between control and HIV+ monocytes were only observed in the presence of autologous serum (P = 0.03, Fig. 1B). Similar levels of foam cell were formed by monocytes from young HIV+ individuals receiving cART with low/undetectable viral loads as compared with young untreated HIV+ individuals with viraemia (Supplementary Fig. 3A and B). We also analyzed foam cell formation in an expanded cohort of 17 HIV+ men aged 26–68 years consisting of 13 individuals receiving cART (12 virologically suppressed with plasma viral loads less than 40 copies/ml and one individual with 80 copies/ml) and the four viraemic individuals not receiving cART defined above. This analysis showed a nonsignificant trend towards increased foam cell formation with age; however, there were no differences between virologically suppressed and nonsuppressed HIV+ individuals (Supplementary Fig. 3C). Data from this expanded cohort of largely virologically suppressed individuals also confirmed the proatherogenic effect of autologous HIV+ serum (Supplementary Fig. 3D).
We then determined whether the proatherogenic effect of HIV+ serum was only evident in combination with autologous HIV+ monocytes or whether it was also capable of enhancing foam cell formation by monocytes from HIV-seronegative individuals. To investigate this, we analyzed foam cell formation in the model described above using monocytes from six HIV-uninfected controls in the presence of either pHS, or serum from five cART-treated, virologically suppressed HIV+ individuals. Foam cell formation by control monocytes was significantly increased by all five HIV+ sera used (Fig. 1C), confirming the presence of soluble factor(s) in HIV+ sera that promote foam cell formation. Significantly, the sera used in this experiment were all from patients with undetectable viral load, indicating the responsible factors remain elevated despite viral suppression. The above data indicate that HIV infection promotes the differentiation of monocytes into foam cells via a mechanism involving both soluble factors present in the serum and an intrinsic alteration to monocyte function.
Live cell imaging of monocyte migration in the transendothelial migration assay
To investigate the effect of HIV infection on monocyte movement postmigration, we labelled monocytes with the fluorescent membrane dye PKH26, and recorded individual cell tracks within the collagen layer using live cell imaging over 48 h (Fig. 2 and Fig2mov1 and Fig2mov2). Tracks obtained from monocytes from HIV+ individuals showed a greater number of slowly migrating cells (indicated by yellow foci in Fig. 2B vs. control donor in 2A), consistent with more monocytes developing into nonmotile foam cells. This behaviour was observed for monocytes from virologically suppressed HIV+ individuals receiving cART using either control serum or autologous serum (data not shown). We calculated the distribution of displacement length using tracks combined from at least three independent experiments and found that monocytes from HIV+ individuals migrated shorter distances than those from controls (evidenced by a shift to the left in the distance frequency histograms for HIV+ samples, Fig. 2C) and this effect was heightened in the presence of autologous HIV+ serum (Fig. 2C).
Although the initial migration speed of monocytes from HIV+ individuals was not different from control monocytes in the presence of pHS, the presence of autologous serum was associated with a significant increase in initial migration speed (Fig. 2D), suggesting factors in HIV+ sera accelerate monocyte migration at early time points after transendothelial migration. In contrast, tracks from HIV+ donors during the later stages of observation (i.e., 38–48 h) showed a lower migration speed as compared with control monocytes (Fig. 2E). Interestingly, the average migration speed of monocytes from HIV+ individuals was higher in the presence of autologous serum vs. pHS, although the speed was still significantly lower than that of control monocytes (Fig. 2E). Analysis of monocyte speed over the entire observation window (0–48 h) confirmed the overall average migration speed of monocytes from HIV+ individuals was lower than that of controls (data not shown). These data indicate that monocytes from HIV+ individuals have impaired movement post-transendothelial migration, consistent with their differentiation into foam cells, and that factors present in the serum of HIV+ individuals enhance this effect.
Plasma tumour necrosis factor and CXCL10 levels are increased in HIV+ individuals
HIV infection is associated with dyslipidaemia because of the combined effects of HIV and certain antiretroviral drugs which may influence the atherogenic properties of monocytes and sera from HIV+ individuals. To determine whether altered cholesterol levels in autologous serum were responsible for the increase in foam cell formation by monocytes from HIV+ individuals in this study, we measured plasma levels of high-density lipoprotein, LDL and oxLDL by ELISA. No significant differences were observed in plasma levels of any of these forms of cholesterol between HIV+ and control individuals (Supplementary Fig. 4), suggesting that the heightened foam cell formation by HIV+ monocytes observed here is unlikely to be due solely to altered levels of these cholesterol species in the blood.
Inflammatory factors activate monocytes and can also drive foam cell formation independent of dyslipidaemia ; therefore, we measured plasma levels of the proinflammatory cytokine TNF and the monocyte activation biomarker CXCL10 in our cohort. Both CXCL10 and TNF levels were significantly elevated in HIV+ as compared with controls (Fig. 3A and B, P = 0.01 and 0.03, respectively). Taken together, these data suggest that plasma markers of monocyte activation and inflammation, but not plasma cholesterol levels, are significantly increased in these HIV+ individuals who also show increased foam cell formation.
Anti-tumour necrosis factor receptor blocks foam cell formation independent of human umbilical vein endothelial cell activation
We have shown here and previously  that TNF activates endothelial cells to increase foam cell formation, and plasma TNF levels have been found to be elevated in HIV patients irrespective of cART . We therefore considered whether TNF promotes foam cell differentiation independently of its effects on HUVEC. Monocytes from control donors were allowed to migrate through TNF-activated HUVEC, then blocking antibodies to TNFR I and II were added to media containing either control pHS or serum from virologically suppressed HIV+ donors and cells incubated for 48 h. Monocytes from each control donor were incubated with serum from three different HIV+ donors, with serum from a total of six HIV+ donors utilized for these experiments. Anti-TNFR antibodies reduced foam cell formation in a dose-dependent manner compared with cells treated with isotype control antibodies (Fig. 3C), and this inhibition was observed when either control serum (pHS) or HIV+ serum was used. These data suggest that endogenous TNF in serum, and/or TNF produced by activated monocytes during foam cell differentiation, enhances foam cell formation independently of its effects on endothelial activation, and that increased levels of TNF in individuals with HIV likely contributes to the proatherogenic effect of HIV+ serum.
Foam cells downregulate markers associated with immunity and endothelial attachment
To characterize the foam cells generated in our model following monocyte transendothelial migration, we labelled migrated cells extracted from the collagen with the fluorescent lipophilic dye BODIPY to enable detection of foam cells and macrophages via flow cytometry. To confirm the ability of BODIPY staining to identify foam cells, we analyzed labelled cells using imaging flow cytometry and found that cells with high BODIPY staining exhibited a morphology consistent with foam cells (Supplementary Fig. 5). By gating BODIPYhigh cells, the percentage of foam cells was determined based on fluorescence intensity (rather than visually which may be subjective). This gave similar values to those determined by Oil Red O staining and light microscopy analysis assayed in parallel (Supplementary Fig. 5). Taken together, these data confirm that both BODIPY and Oil Red O staining were able to identify similar populations of foam cells and macrophages within our model.
Having validated the use of fluorescent BODIPY staining for identification of foam cells, we labelled cells extracted from the collagen matrix with both anti-CD14 and BODIPY and analyzed them by standard flow cytometry. This revealed two distinct populations of CD14+ cells; large, BODIPYhighCD14low cells with high granularity, consistent with foam cells, and a population of smaller BODIPYlowCD14high cells with relatively low granularity, consistent with monocytes/immature macrophages (Fig. 4A). The characterization of these two novel populations was confirmed by microscopic analysis of cells sorted by fluorescence-activated cell sorting (FACS) and was consistent in experiments using monocytes from 14 independent donors (Fig. 4B and C).
We further analyzed the phenotype of macrophages and foam cells by measuring expression of the β2 integrin subunit CD11b and the oxLDL receptor CD36. Foam cells had lower expression of both receptors compared with macrophages, consistent with a role in lipid storage as opposed to migration and active lipid accumulation (Fig. 4D and E). Moreover, foam cells had significantly higher intracellular levels of TNF compared with macrophages (Fig. 4F), indicating a proinflammatory phenotype. These observations show that foam cell formation is associated with increased cell size, reduced surface expression of CD14, CD11b and CD36 but an increase in the production of the proinflammatory cytokine TNF.
Monocytes from HIV+ individuals showed impaired cholesterol efflux and reduced expression of cholesterol transporters
The above data indicate that monocytes from HIV+ individuals have an increased potential to form foam cells, independent of the atherogenic effects of serum, suggesting that HIV is associated with intrinsic changes to monocytes that may alter lipid transport. We therefore assessed the ability of monocytes from virologically suppressed cART-treated individuals to efflux cholesterol using an ex-vivo assay. PBMC from HIV+ individuals and controls of a similar age [n = 8 each, all men, median age (range) 43.5 (33–51) and 36.5 (28–52) respectively] were loaded with BODIPY-labelled cholesterol and the subsequent cholesterol efflux from monocytes determined via flow cytometry. This analysis revealed that monocytes from HIV+ individuals were significantly impaired in their ability to efflux cholesterol from the cell, and this impairment was evident both in the absence and presence of the exogenous cholesterol scavenger methyl-β-cyclodextrin (Fig. 5A and B, respectively).
To investigate the mechanistic basis for this impairment, we analyzed the expression of key genes involved in cholesterol synthesis and transport. RNA was extracted from monocytes purified from virologically suppressed HIV+ individuals on cART and age-matched controls and gene expression analyzed via qPCR. This analysis revealed significantly reduced expression of the key cholesterol efflux transporter ABCA1 in monocytes from HIV+ individuals (Fig. 5C, P = 0.02), whereas expression of ABCG1 showed a similar although not statistically significant downregulation (Fig. 5D). Expression of HMG-CoA reductase, the rate-determining enzyme for cholesterol biosynthesis, was also significantly reduced in monocytes from individuals with HIV (Fig. 5E, P = 0.02), whereas monocyte expression of the cholesterol esterase ACAT was not altered by HIV infection (Fig. 5F). These data indicate that HIV infection is associated with a decreased expression of key cholesterol efflux genes in monocytes and an impaired ability to efflux cholesterol from the cell, which likely contributes to the increased potential of monocytes from HIV+ individuals to form foam cells.
Using a model of atherosclerosis that couples primary human monocyte transendothelial migration and foam cell differentiation, we show that monocytes from HIV+ individuals have a higher rate of foam cell formation, which is because of both intrinsic changes to monocyte function and a proatherogenic effect of soluble factors present in HIV+ serum. These findings may provide some of the first mechanistic evidence regarding the pathogenesis of increased atherosclerosis and CVD in the setting of HIV infection.
In this study, we adapted our in-vitro model which couples monocyte transendothelial migration and foam cell formation to determine the functional properties of human monocytes purified from an HIV+ patient cohort. The strengths of this model include the following: monocytes migrate through an activated primary human endothelial monolayer in response to chemokines secreted in situ and into a fibrous collagen matrix that mimics the subendothelial neointima; it measures foam cell formation by migrated monocytes, rather than in vitro differentiated macrophages; and exogenous modified lipoprotein is not added, therefore, foam cells form in response to endogenous lipids in the human-serum containing media and/or generated in response to endothelial activation. This model is therefore a physiologically relevant system for studying the early stages of foam cell formation in vitro by migrating human monocytes.
Although HIV infection and certain antiretroviral drugs including protease inhibitors can cause dyslipidaemia , plasma levels of LDL, high-density lipoprotein and oxLDL in the HIV+ individuals who contributed to this study were not different from controls, suggesting that heightened foam cell formation observed here cannot be explained by lipidaemic HIV+ serum alone. Endothelial activation and/or monocyte transendothelial migration may affect levels of oxLDL within the gels, however, which could influence foam cell formation. Significantly, serum from HIV+ individuals had elevated levels of TNF and foam cell formation was blocked by anti-TNFR antibodies added after monocyte migration, supporting a proatherogenic role for TNF in these individuals via mediating monocyte activation and increased foam cell formation. Our data were qualitatively supported by live cell imaging showing monocytes from HIV+ individuals moved more slowly within the collagen matrix of the model, particularly at later times, consistent with differentiation into foam cells. These novel data regarding the proatherogenic properties of monocytes from HIV+ individuals may provide a mechanistic explanation for increased atherosclerosis in this population.
In addition to the proatherogenic effects of HIV+ serum, our data suggest HIV infection induces changes to monocytes that predispose them to foam cell formation. Indeed, we found an altered expression of critical cholesterol synthesis and transport proteins in monocytes from HIV+ individuals. The reduced gene expression of ABCA1 may translate to an impaired ability to remove cholesterol from the cell, consistent with our observation of impaired cholesterol efflux in monocytes from individuals with HIV. Feeney et al. also examined the expression of cholesterol metabolism genes in monocytes and found reduced levels of ABCA1 mRNA in viraemic but not virologically suppressed HIV+ participants, whereas expression of the lipid uptake receptors CD36 and LDL-receptor was reduced in both HIV+ groups. The different result regarding ABCA1 expression may be because of differences in the methodology of monocyte isolation (positive selection in the Feeney et al. study as compared with negative selection used in our study), differences in the cohorts used (50% women and 64% non-White vs. 100% White men in our study) or the smaller sample size in our study. Thus, these findings need to be confirmed in a larger cohort adequately powered to examine the effect of demographic factors including sex and race. The reduced expression of the cholesterol synthesis enzyme HMG-CoA reductase in monocytes shown here is consistent with the previously reported study and may represent a cellular response to increased intracellular cholesterol levels.
The foam cell data presented here were generated from a combination of virologically suppressed and nonsuppressed individuals. Although analysis of foam cell formation in an expanded HIV+ cohort aged 26–68 years did not reveal any significant differences associated with detectable levels of viraemia, this requires confirmation in a larger sample set. The fact that serum from virologically suppressed individuals enhanced foam cell formation suggests the persistence of proatherogenic factors in serum in individuals receiving effective cART. Furthermore, our findings of impaired cholesterol efflux from monocytes were also made in virologically suppressed individuals, consistent with our findings that increased monocyte and innate immune activation associated with HIV infection are maintained following viral suppression by cART [20,21]. It remains possible that antiretroviral agents may contribute to this phenotype in cART-treated individuals; however, this requires investigation. Given only a very small proportion of monocytes are found to be infected with HIV in vivo (<0.1%), the proatherogenic monocyte phenotype observed here is likely secondary to HIV-associated immune activation and inflammation and not a result of direct infection, although it remains possible that nonproductive/latent HIV infection may also alter monocyte phenotype. Our data suggest monocytes from virologically suppressed individuals receiving cART exhibit a proatherogenic activity and that this may be a pathogenic factor underlying increased atherosclerotic risk in HIV+ individuals.
Endothelial activation plays an integral role in atherosclerosis and monocyte transendothelial migration via the expression of adhesion molecules , and soluble endothelial activation markers are elevated in HIV+ individuals and independently associated with carotid intima-media thickness, a surrogate measure of atherosclerosis . Zietz et al. examined the aortic endothelium of patients largely with advanced HIV disease and found structural abnormalities which were accompanied by increased expression of adhesion molecules and mononuclear cell infiltration. Thus, the proatherogenic effects of HIV+ monocytes observed here may well be potentiated by heightened endothelial cell activation in vivo. Proinflammatory cytokines are well documented to be associated with increased cardiovascular risk factors and outcomes in both HIV+ and seronegative individuals ; however, viral factors in the blood may also contribute to increased atherosclerosis in HIV infection. HIV infection and HIV-derived single stranded RNAs increase foam cell formation by monocyte-derived macrophages in vitro[45,46]. The HIV accessory protein Nef promotes foam cell formation in these cells by inhibiting cholesterol efflux , and high concentrations of Nef persist in plasma of HIV+ individuals receiving cART . Taken together, it is therefore likely that changes to both monocyte activation and gene expression together with inflammatory cytokines and virus-specific factors in serum contribute to the proatherogenic potential of monocytes observed here. Heightened foam cell formation by monocytes/macrophages in vitro is consistent with increased prevalence of atherosclerotic plaques in HIV+ as compared with uninfected individuals [4,22,48], and it would be of interest to determine whether this is associated with altered plaque morphology in HIV+ individuals in vivo.
We devised a technique to differentiate between monocyte-derived macrophages and monocyte-derived foam cells produced in our model based on CD14 and BODIPY staining and determined that foam cells have an inflammatory phenotype, as defined by higher TNF and downregulated CD11b expression similar to that described by others . Higher intracellular TNF levels compared with migrated macrophages suggests foam cells may have a positive feedback role in activating the endothelium and promoting oxidation of LDL during the development of atherosclerotic plaques, as well as acting on nearby foam cells to enhance cellular activation. We also found foam cells downregulated expression of CD14, CD11b and CD36, the latter two being critical for monocyte attachment and lipid uptake, respectively. Downregulation of CD36 expression was unexpected as CD36 contributes to foam cell formation . Although reduced CD36 expression in response to uptake of modified LDL particles has been reported by others , the majority of studies suggest oxLDL particles increase CD36 expression [51,52] and CD36 is highly expressed in foam cells obtained from mature human plaques . The kinetics of CD36 expression during the process of foam cell formation is poorly understood, and the decreased CD36 expression we observed may be the result of receptor internalization due to endocytosis of lipid, or may be a transient response to lipid uptake immediately after migration. Indeed, our model analyses foam cells formed within 48 h of migration, whereas foam cells extracted from human plaques likely developed over longer periods. This, combined with findings from parallel studies indicating that foam cells extracted from our model are smaller and contain less cholesterol ester than mature foam cells extracted from atherosclerotic plaques (unpublished findings), suggests they likely represent immature foam cells or foam cell precursors. Studying the behaviour of these cells is therefore highly relevant for strategies aimed at preventing or reversing the early stages of foam cell formation.
Our study reports novel data indicating that monocytes from HIV+ individuals, a population characterized by chronic inflammation and immune activation, have a greater propensity to differentiate into foam cells and that elevated levels of TNF in plasma of these individuals further enhances this effect. This may explain the increased risk of atherosclerosis in these individuals after adjusting for viral and traditional risk factors and warrants further clinical investigation of the relationship between monocyte activation, atherogenic behaviour and the development of atherosclerosis in HIV+ individuals receiving cART.
The authors wish to thank The Alfred Hospital Infectious Disease Unit staff for patient recruitment, Dr Clare Westhorpe for technical expertise and training, Maelenn Gouillou for assistance with statistical analysis, Dr Bill Kalionis and research nurses at the Royal Women's Hospital, Melbourne, and Nguyen Thanh Thao Le for help with cord collection and HUVEC preparation, respectively, the AMREP Flow Cytometry Core and Burnet Cell Imaging Facilities, Dr Annie Yan and Ben Hibbs from the Melbourne Materials Institute, Melbourne University for assistance with imaging flow cytometry and all participants who generously donated blood for this project. The authors gratefully acknowledge the contribution to this work of the Victorian Operational Infrastructure Support Program received by the Burnet Institute.
Author contributions: A.M., A.C.H., T.A.A., C.F.P., J.Z. and M.D.Y.S. generated experimental data for this study in the laboratory of S.M.C. and A.J.; C.F.P., S.M.C., W.A.M. and A.J. had input into study design, whereas A.M., C.F.P., A.C.H. and A.J. performed data analysis and manuscript preparation.
Source of funding: The work was funded via NHMRC project grants to A.J. A.M. is supported by the Postdoctoral Programme of the German Academic Exchange Service (DAAD) and an Occupational Trainee Scholarship of the Burnet Institute. S.M.C. is a recipient of a National Health and Medical Research Council of Australia (NHMRC) Principal Research Fellowship. W.A.M. is supported by grants R01 HL046849 and R37 HL064774 from the NIH.
Conflicts of interest
There are no conflicts of interest.
1. Marin B, Thiebaut R, Bucher HC, Rondeau V, Costagliola D, Dorrucci M, et al. Non-AIDS-defining deaths and immunodeficiency in the era of combination antiretroviral therapy
2. Rodger AJ, Lodwick R, Schechter M, Deeks S, Amin J, Gilson R, et al. Mortality in well controlled HIV in the continuous antiretroviral therapy arms of the SMART and ESPRIT trials compared with the general population
3. Hsue PY, Hunt PW, Schnell A, Kalapus SC, Hoh R, Ganz P, et al. Role of viral replication, antiretroviral therapy, and immunodeficiency in HIV-associated atherosclerosis
4. Lo J, Abbara S, Shturman L, Soni A, Wei J, Rocha-Filho JA, et al. Increased prevalence of subclinical coronary atherosclerosis detected by coronary computed tomography angiography in HIV-infected men
5. Currier JS, Taylor A, Boyd F, Dezii CM, Kawabata H, Burtcel B, et al. Coronary heart disease in HIV-infected individuals
. J Acquir Immune Defic Syndr
6. Obel N, Thomsen HF, Kronborg G, Larsen CS, Hildebrandt PR, Sorensen HT, et al. Ischemic heart disease in HIV-infected and HIV-uninfected individuals: a population-based cohort study
. Clin Infect Dis
7. Triant VA, Lee H, Hadigan C, Grinspoon SK. Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human immunodeficiency virus disease
. J Clin Endocrinol Metab
8. Freiberg MS, Chang CC, Kuller LH, Skanderson M, Lowy E, Kraemer KL, et al. HIV infection and the risk of acute myocardial infarction
. JAMA Intern Med
9. Hsue PY, Lo JC, Franklin A, Bolger AF, Martin JN, Deeks SG, et al. Progression of atherosclerosis as assessed by carotid intima-media thickness in patients with HIV infection
10. Lorenz MW, Stephan C, Harmjanz A, Staszewski S, Buehler A, Bickel M, et al. Both long-term HIV infection and highly active antiretroviral therapy are independent risk factors for early carotid atherosclerosis
11. Parra S, Coll B, Aragones G, Marsillach J, Beltran R, Rull A, et al. Nonconcordance between subclinical atherosclerosis and the calculated Framingham risk score in HIV-infected patients: relationships with serum markers of oxidation and inflammation
. HIV Med
12. Friis-Moller N, Sabin CA, Weber R, d’Arminio Monforte A, El-Sadr WM, Reiss P, et al. Combination antiretroviral therapy and the risk of myocardial infarction
. N Engl J Med
13. Islam FM, Wu J, Jansson J, Wilson DP. Relative risk of cardiovascular disease among people living with HIV: a systematic review and meta-analysis
. HIV Medicine
14. Currier JS, Lundgren JD, Carr A, Klein D, Sabin CA, Sax PE, et al. Epidemiological evidence for cardiovascular disease in HIV-infected patients and relationship to highly active antiretroviral therapy
15. El-Sadr WM, Lundgren J, Neaton JD, Gordin F, Abrams D, Arduino RC, et al. CD4+ count-guided interruption of antiretroviral treatment
. N Engl J Med
16. Pereyra F, Lo J, Triant VA, Wei J, Buzon MJ, Fitch KV, et al. Increased coronary atherosclerosis and immune activation in HIV-1 elite controllers
17. Hsue PY, Scherzer R, Hunt PW, Schnell A, Bolger AF, Kalapus SC, et al. Carotid intima-media thickness progression in HIV-infected adults occurs preferentially at the carotid bifurcation and is predicted by inflammation
. J Am Heart Assoc
2012; 1:jah3-e000422doi: 10.1161.
18. Duprez DA, Neuhaus J, Kuller LH, Tracy R, Belloso W, De Wit S, et al. Inflammation, coagulation and cardiovascular disease in HIV-infected individuals
. PLoS One
19. De Luca A, de Gaetano Donati K, Colafigli M, Cozzi-Lepri A, De Curtis A, Gori A, et al. The association of high-sensitivity c-reactive protein and other biomarkers with cardiovascular disease in patients treated for HIV: a nested case-control study
. BMC Infect Dis
20. Hearps AC, Maisa A, Cheng WJ, Angelovich TA, Lichtfuss GF, Palmer CS, et al. HIV infection induces age-related changes to monocytes and innate immune activation in young men that persist despite combination antiretroviral therapy
21. Martin GE, Gouillou M, Hearps AC, Angelovich TA, Cheng AC, Lynch F, et al. Age-associated changes in monocyte and innate immune activation markers occur more rapidly in HIV infected women
. PLoS One
22. Burdo TH, Lo J, Abbara S, Wei J, Delelys ME, Preffer F, et al. Soluble CD163, a novel marker of activated macrophages, is elevated and associated with noncalcified coronary plaque in HIV-infected patients
. J Infect Dis
23. Subramanian S, Tawakol A, Burdo TH, Abbara S, Wei J, Vijayakumar J, et al. Arterial inflammation in patients with HIV
24. Kelesidis T, Kendall MA, Yang OO, Hodis HN, Currier JS. Biomarkers of microbial translocation and macrophage activation: association with progression of subclinical atherosclerosis in HIV-1 infection
. J Infect Dis
25. Westhorpe CL, Maisa A, Spelman T, Hoy JF, Dewar EM, Karapanagiotidis S, et al. Associations between blood monocyte markers and carotid atherosclerosis in HIV-positive patients
. Immunol Cell Biol
26. Baker JV, Hullsiek KH, Singh A, Wilson E, Henry K, Lichtenstein K, et al. Immunologic predictors of coronary artery calcium progression in a contemporary HIV cohort
27. Funderburg NT, Zidar DA, Shive C, Lioi A, Mudd J, Musselwhite LW, et al. Shared monocyte subset phenotypes in HIV-1 infection and in uninfected subjects with acute coronary syndrome
28. Lusis AJ. Atherosclerosis
29. Febbraio M, Silverstein RL. CD36: implications in cardiovascular disease
. Int J Biochem Cell Biol
30. Park YM, Febbraio M, Silverstein RL. CD36 modulates migration of mouse and human macrophages in response to oxidized LDL and may contribute to macrophage trapping in the arterial intima
. J Clin Invest
31. Galkina E, Ley K. Immune and inflammatory mechanisms of atherosclerosis (*)
. Annu Rev Immunol
32. Llodra J, Angeli V, Liu J, Trogan E, Fisher EA, Randolph GJ. Emigration of monocyte-derived cells from atherosclerotic lesions characterizes regressive, but not progressive, plaques
. Proc Natl Acad Sci U S A
33. Muller WA, Weigl SA. Monocyte-selective transendothelial migration: dissection of the binding and transmigration phases by an in vitro assay
. J Exp Med
34. Randolph GJ, Furie MB. A soluble gradient of endogenous monocyte chemoattractant protein-1 promotes the transendothelial migration of monocytes in vitro
. J Immunol
35. Westhorpe CLV, Dufour EM, Maisa A, Jaworowski A, Crowe SM, Muller WA. Endothelial cell activation promotes foam cell formation by monocytes following transendothelial migration in an in vitro model
. Exp Mol Pathol
36. Westhorpe CL, Zhou J, Webster NL, Kalionis B, Lewin SR, Jaworowski A, et al. Effects of HIV-1 infection in vitro on transendothelial migration by monocytes and monocyte-derived macrophages
. J Leukoc Biol
37. Angelovich TA, Hearps AC, Jaworowski A. Inflammation-induced foam cell formation in chronic inflammatory disease
. Immunol Cell Biol
2015; doi: 10.1038/icb.2015.26.
38. Cassol E, Malfeld S, Mahasha P, van der Merwe S, Cassol S, Seebregts C, et al. Persistent microbial translocation and immune activation in HIV-1-infected South Africans receiving combination antiretroviral therapy
. J Infect Dis
39. Souza SJ, Luzia LA, Santos SS, Rondo PH. Lipid profile of HIV-infected patients in relation to antiretroviral therapy: a review
. Rev Assoc Med Bras
40. Feeney ER, McAuley N, O’Halloran JA, Rock C, Low J, Satchell CS, et al. The expression of cholesterol metabolism genes in monocytes from HIV-infected subjects suggests intracellular cholesterol accumulation
. J Infect Dis
41. Maslin CL, Kedzierska K, Webster NL, Muller WA, Crowe SM. Transendothelial migration of monocytes: the underlying molecular mechanisms and consequences of HIV-1 infection
. Curr HIV Res
42. Ross AC, Rizk N, O’Riordan MA, Dogra V, El-Bejjani D, Storer N, et al. Relationship between inflammatory markers, endothelial activation markers, and carotid intima-media thickness in HIV-infected patients receiving antiretroviral therapy
. Clin Infect Dis
43. Zietz C, Hotz B, Sturzl M, Rauch E, Penning R, Lohrs U. Aortic endothelium in HIV-1 infection: chronic injury, activation, and increased leukocyte adherence
. Am J Pathol
44. Hearps AC, Martin GE, Rajasuriar R, Crowe SM. Inflammatory co-morbidities in HIV+ individuals: learning lessons from healthy ageing
. Curr HIV/AIDS Rep
45. Mujawar Z, Rose H, Morrow MP, Pushkarsky T, Dubrovsky L, Mukhamedova N, et al. Human immunodeficiency virus impairs reverse cholesterol transport from macrophages
. PLoS Biol
46. Bernard MA, Han X, Inderbitzin S, Agbim I, Zhao H, Koziel H, et al. HIV-derived ssRNA binds to TLR8 to induce inflammation-driven macrophage foam cell formation
. PLoS One
47. Raymond AD, Campbell-Sims TC, Khan M, Lang M, Huang MB, Bond VC, et al. HIV Type 1 Nef is released from infected cells in CD45(+) microvesicles and is present in the plasma of HIV-infected individuals
. AIDS Res Hum Retroviruses
48. Fitch KV, Srinivasa S, Abbara S, Burdo TH, Williams KC, Eneh P, et al. Noncalcified coronary atherosclerotic plaque and immune activation in HIV-infected women
. J Infect Dis
49. Maiguel D, Faridi MH, Wei C, Kuwano Y, Balla KM, Hernandez D, et al. Small molecule-mediated activation of the integrin CD11b/CD18 reduces inflammatory disease
. Sci Signal
50. Kapinsky M, Torzewski M, Buchler C, Duong CQ, Rothe G, Schmitz G. Enzymatically degraded LDL preferentially binds to CD14(high) CD16(+) monocytes and induces foam cell formation mediated only in part by the class B scavenger-receptor CD36
. Arterioscler Thromb Vasc Biol
51. Feng J, Han J, Pearce SF, Silverstein RL, Gotto AM Jr, Hajjar DP, et al. Induction of CD36 expression by oxidized LDL and IL-4 by a common signaling pathway dependent on protein kinase C and PPAR-gamma
. J Lipid Res
52. Hirano K, Yamashita S, Nakagawa Y, Ohya T, Matsuura F, Tsukamoto K, et al. Expression of human scavenger receptor class B type I in cultured human monocyte-derived macrophages and atherosclerotic lesions
. Circ Res
53. Nakata A, Nakagawa Y, Nishida M, Nozaki S, Miyagawa J, Nakagawa T, et al. CD36, a novel receptor for oxidized low-density lipoproteins, is highly expressed on lipid-laden macrophages in human atherosclerotic aorta
. Arterioscler Thromb Vasc Biol