Differential Expression of Rac-1, CXCR4 and CCR5 on CD4 T-Cells at Different Stages of HIV-1 Disease Relate To Its Progression in Therapy Naive Individuals

Human Immunodeficiency virus, HIV-1 initiates infection via viral envelope (Env) glycoprotein gp120 interacting with cell surface CD4, followed by its association with a co-receptor that triggers the fusion of viral and host-cell membranes. Two chemokine receptors, CCR5 and CXCR4 are the predominantly known co-receptors for HIV-1 in vivo. All HIV-1 strains are classified phenotypically as R5 (Macrophage tropic), X4 (T-cell tropic), or R5X4 (dual tropic i.e. both M-tropic and T-tropic) depending on whether they preferentially utilize cysteinecysteine receptor 5 (CCR5) or cysteine-X-cysteine receptor 4 (CXCR4) or mixed [1]. This inter-conversion of tropism requires only a small number of changes in the Env V3 region. In early stage of HIV infection, only about 15% are dual/mixed tropic and is dominated by CCR5 tropism [2,3]. While in later stages of the disease, dual/mixed or pure CXCR4 tropism reaches up to 60% [4], and is highest in patients with CD4 ≤ 200 cells/μl [5,6].This may vary with the prevalence of HIV-1 subtype in a region. Why there is selective advantage of CCR5 tropic viruses in establishing HIV-1 infection, switching over to CXCR4 utilization in the advanced stage of disease, is not clearly understood yet. This tropism switch has been associated with rapid disease progression and poor clinical prognosis [7-9]. Besides preferential CXCR4-utilizing virus, there may be other factors also playing crucial role in faster disease progression and needs a closer look. It has earlier been indicated that Rac1,a small (~21 kDa) signaling G protein a member of the Rac subfamily of the Rho family of GTPases, is important to maintain the functional conformation of CXCR4 receptor [10], and could be one of the candidate molecules that could possibly play an important role in preferential co-receptor utilization by the virus.


Introduction
Human Immunodeficiency virus, HIV-1 initiates infection via viral envelope (Env) glycoprotein gp120 interacting with cell surface CD4, followed by its association with a co-receptor that triggers the fusion of viral and host-cell membranes. Two chemokine receptors, CCR5 and CXCR4 are the predominantly known co-receptors for HIV-1 in vivo. All HIV-1 strains are classified phenotypically as R5 (Macrophage tropic), X4 (T-cell tropic), or R5X4 (dual tropic i.e. both M-tropic and T-tropic) depending on whether they preferentially utilize cysteinecysteine receptor 5 (CCR5) or cysteine-X-cysteine receptor 4 (CXCR4) or mixed [1]. This inter-conversion of tropism requires only a small number of changes in the Env V3 region. In early stage of HIV infection, only about 15% are dual/mixed tropic and is dominated by CCR5 tropism [2,3]. While in later stages of the disease, dual/mixed or pure CXCR4 tropism reaches up to 60% [4], and is highest in patients with CD4 ≤ 200 cells/μl [5,6].This may vary with the prevalence of HIV-1 subtype in a region. Why there is selective advantage of CCR5 tropic viruses in establishing HIV-1 infection, switching over to CXCR4 utilization in the advanced stage of disease, is not clearly understood yet. This tropism switch has been associated with rapid disease progression and poor clinical prognosis [7][8][9]. Besides preferential CXCR4-utilizing virus, there may be other factors also playing crucial role in faster disease progression and needs a closer look. It has earlier been indicated that Rac1,a small (~21 kDa) signaling G protein a member of the Rac subfamily of the Rho family of GTPases, is important to maintain the functional conformation of CXCR4 receptor [10], and could be one of the candidate molecules that could possibly play an important role in preferential co-receptor utilization by the virus.
Normal cell functioning of host is altered by invading viral proteins to the benefit of the virus [11]. There have been constant efforts to understand these interactions between viral and cellular gene products which together determine the host's susceptibility to infection and disease progression in HIV-1 infection [12]. Working on few such molecules for their behavior in HIV-1 disease progression and M. tuberculosis co-infection, we have tried to analyze the role of HIV-1 co-receptors, CCR5 and CXCR4, variant 2 (shorter splice variant) of CXCR4, Rac1 (which is required to maintain CXCR4 in the responsive conformation) [10], NF-κB which has high affinity binding sight on long terminal repeat (LTR) enhancer region of HIV-1 and regulates HIV replication [13][14][15][16], and FoxP3 which enhances HIV-1 gene expression by modulating NF-κB occupancy at the LTR in human T-cells line [17].

Patients and controls
The study was conducted on different cohorts of individuals including 27 HIV-1 infected patients, 12 Mycobacterium tuberculosis (Mtb) infected patients (PTB), 8 HIV-1/Mtb co-infected patients (HIV-PTB), and 20 healthy controls (HC). HIV-1 infected patients confirmed by the blood sample being reactive by three rapid tests, as per National AIDS Control Organization (NACO) guidelines were enrolled from the Integrated Counselling and Testing Centre (ICTC) in the Department of Immunopathology, PGIMER, and Chandigarh, India. At the time of recruitment, the patients were interviewed by a counsellor to obtain informed consent and ascertain therapy naïve status. The status of disease was assessed by absolute CD4 cell count by flow cytometry using BD Tritest™ containing antibody conjugates CD3-FITC/ CD4-PE/CD45-PerCP with BD Trucount™ tubes (BD Biosciences, San Jose, USA). PTB and HIV-PTB patients confirmed positive for Mycobacterium tuberculosis by chest x-ray, sputum smear positivity and TB-PCR were recruited from DOTS centre, in our hospital.The study was approved by the Institutional Ethics Committee (IEC) of Postgraduate Institute of Medical Education and Research (PGIMER), Chandigarh, India and peripheral blood was obtained from enrolled subjects after an informed consent. The peripheral blood mononuclear cells (PBMCs) were isolated from heparinised blood by Ficoll-Hypaque density gradient centrifugation (HiMedia, Mumbai, India).

Treg Immunophenotyping
Freshly isolated PBMC were immunophenotyped for Treg number, FoxP3 expression per cell (MFI) and CCR5/CXCR4 expression using fluorophore-conjugated monoclonal antibodies (mAb): anti-CD4 PE/ FITC, anti-CD25 PE-Cy7, anti-CCR5 PE, and anti-CXCR4 PE (all from BD Bioscience, San Jose, CA) for cell surface markers in combination with intracellular Fork-head box protein 3 (FoxP3) mAb conjugated with Alexa488 (eBioscience, San Diego, CA). Samples were acquired into a flowcytometer (FACS Calibur, BD, USA) and analyzed using cell-Quest software (BD Bioscience). For further analysis patients were categorized on the bases of level of surface CD25 expression on CD4 + T-cells: CD4CD25 high (top 2% with highest CD25 expression), CD4CD25 intermediate (middle 15% with intermediate CD25 expression) and CD4CD25 low/negative (lower 83% with very low or no expression of CD25) cells ( Figure 1). Isotype controls were performed for each sample. Absolute counts for different populations were calculated from the total lymphocyte count in the whole blood.

T-cell proliferation assay
To examine the functional defects of T-cells in these cohorts, proliferative capability of T cells was assessed on stimulation with mitogen-PHA (phyto-heamagglutinin) by determining the extent of 3 H-thymindine incorporation. Briefly, 1 × 10 5 PBMCs/well were plated into 96-well plates in triplicates in complete RPMI 1640 medium (Sigma) supplemented with 10% heat-inactivated FCS, 20 IU/ml penicillin, 20 µg/ml streptomycin, 25 mM HEPES, and 2 mM L-glutamine for 96 hours at 37°C in 5% CO 2 , either in presence or absence of PHA (10 µg/ ml) as stimulant. Eighteen hours prior to termination of the culture, 1µ Ci of 3 H-thymidine (BARC, India) was added to each well. The cells were then harvested on glass fibre mats (Skatron, Belgium) and counted in liquid scintillation counter (Beckman, USA). All measurements were conducted in triplicates and the results were expressed as the stimulation index (SI), which represented the mean cpm (counts per minute) in the presence of mitogen divided by mean cpm in the absence of the mitogen.

Statistical analysis
Data are expressed as mean ± SE and ranges. Difference between groups was tested by one-way student's t-test. Correlation analysis was done using the non-parametric Spearman's Rank Correlation Coefficient. Linear Regression Analysis was carried out and the residuals were calculated. The level of significance was set at p<0.05.

Demographic and Immunological characteristics
The cross-sectional study included a total of 67 subjects: 27

High CD4+FoxP3 frequency correlates with poor T-cell stimulation in HIV-PTB co-infected and HIV advanced disease individuals
Within the HIV-1 infected group, there was significantly low frequency of Treg in individuals with CD4 count >500 cells/μl (mean ± SEM: 3.49 ± 0.31, p=0.0003) as compared to other groups and HC (Figure 2a) hence, indicating early selective depletion of Treg cells. This is followed by increase in the mean percent frequency of Treg as the disease progresses. Though statistically non-significant, we did see an increase in the Treg frequency in HIV-1 patients with Mtb coinfection when compared with only HIV-infected individuals with similar CD4 count (<250 cells/μl). Overall, there was a significant inverse relationship (p<0.0001) between percent Treg as well as FoxP3 expression with the disease progression in terms of CD4 count ( Figure  2b). Following PHA stimulation, mean proliferative index of T-cells from HIV-PTB co-infected individuals was significantly lower (mean ± SEM: 39.90 ± 10.00, p=0.0011) as compared to HC (mean ± SEM: 204.7 ± 30.76). Individuals with only PTB or HIV infected with CD4 count >500 cells/μl had higher proliferative index (mean ± SEM: 434.0 ± 138.8, p=0.0390 and 332.9 ± 97.74, p=0.1596 respectively). Overall, there was a significant negative correlation between T-cell proliferation and HIV-1 disease progression (Figures 2c and 2d). These results suggest that reduced T-cell proliferative potential in HIV-PTB co-infected as well as only HIV infected in advanced stage of disease is due to increased frequency of FoxP3+ Treg cells in these individuals.

HIV-1 subjects in early stage of disease had significantly higher number of CD4+CD25 high Treg cells expressing CCR5
CD4 T-cell sub-populations were examined for surface expression of CCR5, which were calculated as percentage of total analyzed CD4 T-cells in peripheral blood. All HIV-1 infected individuals, irrespective of Mtb co-infection, had significantly higher percentage of CD4 T-cells expressing CCR5 as compared to HC. This difference was not seen in individuals with PTB infection (Table 2). However, looking at the CD4+CD25 high Treg population within the CD4 T-cells, there was significantly higher percentage of cells with CCR5 expression in PTB individuals (mean ± SEM: 19.83 ± 2.52, p=0.0226) in comparison to HC (mean ± SEM: 12.72 ± 1.29) (Figure 3a). Similar results were seen in early stage of disease in only HIV-1 infected individuals having normal CD4 count (mean ± SEM: 32.17 ± 7.15, p=0.0021). Interestingly, there was a drop in the CCR5+ Treg cell numbers with decrease in CD4 count. There was significantly lower frequency of these cells in HIV patients patients (age range between 21-53 years and CD4+ T-cell count range: 32-854 cells/μl), 12 active PTB patients (age range between 21-65 years and CD4+ T-cell count range: 326-1368 cells/μl), 8 HIV-PTB coinfected patients (age range between 17-34 years and CD4+ T-cell count range: 30-273 cells/μl), and 20 HC (age range between 22-40 years and CD4+ T-cell count range: 444-1375 cells/μl). Table 1 summarizes their clinical characteristics.   with CD4 count <250 cells/μl (p=0.002) in comparison to those with CD4 count >500 cells/μl. In individuals with PTB co-infection, there was a significant increase in CCR5+ Treg number (mean ± SEM: 28.15 ± 9.41) when compared with only HIV-1 infected individuals with similar CD4 count (CD4 count <250 (p=0.0260)). There was however no significant change in CCR5 expression seen is CD4CD25 low/negative subpopulation among these groups (Figure 3a). This indicates that in the early stage of HIV infection there is higher number of Treg cells having very high CD25 expression co-expressing CCR5 also which may be the reason for their early depletion in these individuals.

Significant decrease in the frequency of CXCR4+ cells in CD4+CD25 low/negative sub-population of HIV-PTB co-infected individuals
There was no significant difference seen in frequency of CXCR4+ cells within HIV-1 sub-groups with the disease progression. However, we found a decrease in the percentage of CXCR4+ cells in HIV-PTB co-infected individuals (mean ± SE: 86.43 ± 4.95) when compared with only PTB infected (mean ± SE: 94.99 ± 1.76, p=0.0232) and HC (mean ± SE: 95.28 ± 1.13, p=0.048) (Figure 4a)

Mtbco-infection
There was a trend of increase in the CXCR4 expression in HIV-1 patients with the disease progression ( Figure 5A, 5B, 5B). Further with the onset of AIDS the HIV-PTB co-infected individuals had higher CXCR4 expression (mean ± SEM: 126.0 ± 6.2, p=0.1201) in comparison to HIV-1 subjects with similar CD4 count (CD4 count <250 cells/μl) though statistically non-significant. Similar results were observed in CXCR4 variant-2 expression with significantly high expression in PTB and HIV-1 patients ( Figure 5B). Interestingly, there was significant decrease in the expression of variant-2 with PTB co-infection (mean ± SEM: 124.4.6 ± 7.2) when compared to only HIV-1 infected subjects with CD4 count <250 cells/μl (mean ± SEM: 103.6 ± 5.0, p<0.0175). This suggests that with the progression of HIV disease CXCR4 expression is increased both in only HIV infected as well as PTB co-infected individuals. Further, this increase in CXCR4expression positively correlated with NF-κB expression in HIV-1 subjects ( Figure 5E and 5f). This significant positive correlation of CXCR4 was extended with hemeoxygenase 1 (HO-1) but only with variant 2 ( Figure 5G and 5H).

Mycobacterium tuberculosis infection results in down regulation of Rac-1 expression in HIV-1 patients
Rac1 expression was significantly lower in PTB patients (mean ± SEM: 112.4 ± 4.7, p=0.0063), HIV-1 subjects with CD4 count >500 cells/μl (mean ± SEM: 109.7 ± 9.5, p=0.0283) and HIV-PTB co-infected individuals (mean ± SEM: 110.1 ± 2.3, p=0.0062) when compared to healthy controls (mean ± SEM: 131.7 ± 4.5). We saw a trend of initial decrease in Rac1 expression in early stage of HIV-1 infection followed by returning to normal expression as the disease progresses ( Figure  6a). Further, there was a significant decrease in the Rac1 expression in HIV individuals with onset of Mtb co-infection when compared with individuals infected with HIV-1 only with CD4 count between 250 and 500 cells/μl (p=0.0128) and <250 cells/μl (p=0.0325). This indicates that onset of Mtb infection down regulates the expression of Rac-1 gene making it unavailable to maintain the CXCR4 conformation suitable for HIV utilization. In addition, Rac1 significantly, positively correlates with immunosuppressive genes FoxP3 ( Figure 6F and 6G) and HO-1 ( Figure 6H and 6I) in HIV-1 individuals only (unpublished personal data).

Discussion
The results in the present study elucidate some of the cellular factors manipulated by HIV-1 at different stages of disease progression and in the event of M.tuberculosis co-infection. It has earlier been revealed that CCR5 and CXCR4, the major co-receptors critical for viral entry into the host cells, are differentially expressed on CD4+ T-cells, and their relative expression is closely associated with disease pathogenesis [18][19][20][21][22][23][24]. We studied their expression on different subsets of T-cells including T-regulatory cells and evaluated their role in changing the frequency of these cells with disease progression. We also evaluated the change in Rac1 expression, critical for functional CXCR4 receptor [10], for it's possible role in M5 to T4 HIV-1 viral shift based on availability of co-receptor.
We found a significantly low frequency of FoxP3+CD4 cells in early stage of disease followed by a gradual increase with the progression of disease. These observations prompted us to look at the expression of CCR5 and CXCR4 on Treg cell sub population for their possible role in influencing the change in Treg cell frequency with HIV disease progression. Frequency of CD4 T cells expressing CCR5 was found to be up-regulated in HIV-1 infected individuals when compared to the healthy controls, irrespective of the stage of the disease, which is in concordance with the previous findings [25][26][27]. There was higher percentage of CD25 high Treg cells expressing CCR5 in comparison to CD25 low/negative effector cells (p<0.0001) and which could make them preferential targets for R5 viral attack in the initial stages of disease ( Figure 3a). On the other hand, in advanced stage of the disease with CD4 count <250 cells/μl, the number of CCR5 positive Treg were significantly lower when compared to HIV-1 individuals in early stage of disease (p=0.0022). With previous reports in non-subtype C infections, the depletion of CCR5 expressing memory cells in acute infection may drive viral evolution towards CXCR4 [28,29], and recent reports showing about 30% of CXCR4-utilising viruses in untreated and treated HIV-1 subtype C infected adults from South Africa, Malawi, and Zimbabwe, suggesting a shift in viral properties [30][31][32]. This change in CCR5 expression might be responsible for increase in the frequency of Treg cells in the advanced stages of the disease.
On the contrary we observed a lower frequency of CD4 T-cells expressing chemokine receptor CXCR4, in the later stages of HIV disease when compared to healthy controls, although statistically nonsignificant (Figure 4a). This is in concordance with previously reported observation in HIV subjects [25][26][27]. It is interesting to note that there is a reverse scenario when we considered the quantitative expression of CXCR4 per cell (MFI) on CD4 cells in these subjects (Figure 4b). There was an increase in the CXCR4 expression in advanced stages of HIV-1 when compared to asymptomatic early stage HIV-1 subjects. This could explain decrease in the frequency of CXCR4 positive CD4 T-cells in later stages of the disease due to X4 tropic viral shift and killing of these cells. Again, contrary to CCR5, there was significant decrease in percent CD4 T-cells expressing CXCR4 (p=0.0033) even quantitatively (MFI, p=0.0004) in CD25 high Treg when compared with CD25 low/negative cells in healthy controls. This could be a possible explanation for the decrease in percent of CXCR4 expressingCD4 T-cells in CD25 low/negative subpopulation of cells within HIV-1 groups negatively correlating with disease progression but with no effect on Treg cell number with low expression of CXCR4.
However the situation was totally different if we look at the data from patients infected with only Mtb or co-infected with HIV. It is well known that Mtb co-infection in HIV may create a microenvironment enhancing the productive infection of T-cells by HIV [33][34][35][36]. We observed that HIV-1 individuals with active pulmonary tuberculosis displaying increased CCR5 (p= 0.0364, Figure 3b) and not CXCR4 (p=0.1914, Figure 4b) expression per cell (MFI) on CD4 T-cells when compared with HIV-1 subjects with similar CD4 count (<250 cells/ μl) suggesting preferential utilization of CCR5 rather than CXCR4 by virus in co-infected individuals. This falls in line with reports which demonstrated that R5 viral variants were preferentially recovered from patients with active TB [37]. Furthermore, increased frequency of CD4 cells expressing CCR5 and up-regulated expression limited to CD25 high and CD25 intermediate populations only in co-infected individuals with no significant change in CD25 low/negative subpopulation indicated the influence of MTB on these two subpopulations only making them more vulnerable to HIV-1 infection. Whereas, a significant decrease in frequency of CXCR4+ T-cells in co-infected individuals could mainly   be attributed to a significant decrease in CD25 low/negative CD4 T-cell subpopulation. On the other hand, significantly increased CXCR4 expression on CD25 low/negative cells during Mtb infection makes it the main population affected in co-infected individuals as evidenced by their low frequency in these individuals. Thus, the data suggests that M.tuberculosis co-infection in HIV-1 individuals seems to be selectively affecting the differential expression of these chemokine receptors on different subpopulations of cells thereby leading to their selective depletion during different stages of disease.
Over the course of infection, the co-receptor usage of HIV changes from CCR5 to CXCR4 in 50% of the infected individuals [1,7,[38][39][40][41]. It is generally believed that relatively late appearance of X4 virus variants is driven by evolution of virus population during the course of disease [42]. This assumption is in conflict with viral property of rapid turnover of about 10 10 to 10 12 virions every two days combined with the high mutation rate of HIV [42][43][44][45][46][47][48][49][50]. This should result in emergence of X4 variants fairly early during infection, especially because, in some cases, only two mutations are thought to be necessary for co-receptor switch. Hence factors that influence this evolution of the virus remain largely unknown. It has been demonstrated recently that Rac1 is specifically involved in regulating the conformation of CXCR4, thereby controlling signaling efficiency of the receptor. Furthermore, the conformation adopted by CXCR4 after Rac1 inhibition blocks HIV-1 infection, most probably by interfering with virus binding and subsequent entry into the host cells [10]. It was interesting to note in our study that Rac1 expression was significantly lower in patients in early stage of disease when compared to healthy controls ( Figure 6A). With the disease progression we observed an increase in its expression in individuals infected with only HIV-1. These results indicate and support the other previous findings that initially there is R5 virus predominance as there is low Rac1 expression and X4 viruses even present are unable to use CXCR4 co-receptor. Rac1 has no effect on CCR5 conformation or its utilization by the virus [10]. Though, in advanced stage of HIV-1 infection we found an up-regulatedRac1 expression which would make CXCR4 conformationally favorable to HIV-1 attachment and possibly is one of the reasons for R5 to X4 viral shift with HIV disease progression. We found a significant positive correlation of Rac1 with NF-κB ( Figure  6D) and CXCR4 ( Figure 6B) expression in healthy controls. Though this correlation of Rac1 and CXCR4 was not obvious in HIV-1 patients ( Figure 6C), NF-κB retained significant positive correlation in HIV-1 subjects indicating this pathway not being affected by the disease condition ( Figure 6E). However, with M. tuberculosis co-infection, the Rac1 was significantly down-regulated making available CXCR4 receptors non-utilizable by the virus. This observation is in line with the finding that R5 viral variants were preferentially recovered from HIV infected patients with active M. tuberculosis [51].
Overall, we have recognized novel inter-relationships of genes and their behavior in HIV-1 therapy naïve individuals, highlighting how the virus manipulates the host machinery during Mtb co-infection to its advantage. Though, significant correlation of NF-κB with Rac1 in both healthy and HIV-1 subjects was observed, a highly significant correlation between FoxP3, HO-1 and Rac-1 (unpublished personal data) was seen in HIV-1 subjects with CXCR4 variant 2 showing positive correlation with HO-1, these findings need further attention to look at their clinical implication and disease outcome.