Alterations of the NK cell pool in HIV/HCV co-infection

Background A relevant proportion of human immunodeficiency virus (HIV) infected patients is co-infected with the hepatitis C virus (HCV). HCV co-infection in HIV-positive patients is associated with faster progression of liver disease in comparison to HCV mono-infection. Natural killer (NK) cells critically modulate the natural course of HCV infection. Both HIV and HCV mono-infection are associated with alterations of the NK cell pool. However, little data is available concerning phenotype and function of NK cells in HIV/HCV co-infection. Methods A total of 34 HIV/HCV co-infected, 35 HIV and 39 HCV mono-infected patients and 43 healthy control persons were enrolled into this study. All HIV-positive patients were under effective antiretroviral therapy. NK cell phenotype, IFN-γ production and degranulation were studied by flow cytometry. Results NK cell frequency in HIV/HCV co-infection was significantly lower than in healthy individuals but did not differ from HIV and HCV mono-infection. HIV/HCV co-infection was associated with significantly decreased expression of the maturation/differentiation markers CD27/62L/127 on NK cells but increased expression of CD57 compared to healthy controls. Of note, expression also differed significantly from HCV mono-infection but was similar to HIV mono-infection, suggesting a pronounced impact of HIV on these alterations. Similar findings were made with regard to the NK cell receptors NKG2A/C and NKp30. More importantly, NK cells in co-infection displayed a highly impaired functional activity with significantly lower IFN-γ production and degranulation than in healthy donors as well as HIV and HCV mono-infection, suggesting a synergistic effect of both viruses. Conclusions Our data indicate that HIV/HCV co-infection is associated with significant alterations of the NK cell pool, which might be involved in the rapid progression of liver disease in co-infected patients and which mainly reflect alterations observed in HIV mono-infection.


Introduction
Due to similar transmission routes of infection a relevant proportion of human immunodeficiency virus (HIV)-positive patients is co-infected with the hepatitis C virus (HCV) [1]. HIV/ HCV co-infection is associated with a faster progression to liver fibrosis and cirrhosis, resulting in higher mortality compared to HCV mono-infected individuals [2][3][4][5][6][7][8]. Accordingly, liverassociated mortality has become a major cause of death in HIV-positive (HIV(+)) persons under combined anti-retroviral therapy (cART) [3].
Incomplete restoration of the immune system in HIV patients despite effectively blocked HIV replication is considered to importantly contribute to this phenomenon [9][10][11][12]. In this context, persistent dysregulation of the natural killer (NK) cell pool is of especial interest, as NK cells have been shown to effectively block HCV replication [13,14] and to display an antifibrotic activity [15].
Both HIV and HCV mono-infection are associated with significant perturbations of NK cells. For instance, a reduction of absolute NK cell numbers was found in HIV(+) as well as in HCV(+) patients [16,17] and both HIV and HCV mono-infection as well as HIV/HCV coinfection have been observed to be associated with the appearance of a highly dysfunctional subset of CD56 -CD16 + NK cells, characterized by a poor cytotoxic activity [18][19][20][21][22][23]. In addition, both viral infections are characterized by altered expression patterns of activating NK cell receptors (NKR) [11,24,25] although reports on NKR expression in chronic hepatitis C are controversial [24][25][26][27][28][29][30][31]. Besides these shared alterations in NK cell phenotype and functions observed in HIV and HCV mono-infection, there are also infection-specific differences regarding perturbations of the NK cell pool. As an example, expression of the inhibitory Ctype lectin receptor NKG2A has repeatedly been demonstrated to be increased in chronic hepatitis C [24,32], whereas HIV infection is associated with a decreased frequency of NKG2A [10,33]. Furthermore, Meier et al were able to detect significant differences in the production of IFN-γ between HIV and HCV [16]. Moreover, anti-viral [11,34,35] as well as anti-fibrotic [15] NK cell functions have been shown to be impaired in HIV patients even in the context of effective cART.
Although a large number of studies analyzed the impact of HIV and HCV mono-infection, respectively, on phenotype and functions of NK cells, little is known regarding alterations of this lymphocyte subset in HIV patients chronically co-infected with HCV.
Here, we show that HIV/HCV co-infection is associated with a significant dysregulation of the circulating NK cell pool, and present data suggesting that dysregulation mainly reflects alterations observed in HIV mono-infection.

Materials and methods
Patients RNA viral load as well as liver enzyme levels were retrieved from patients' files. A total of 108 patients, including 35 HIV mono-infected, 39 HCV mono-infected and 34 HIV/HCV coinfected individuals, all from the Cologne/Bonn area in Germany, were enrolled into this study. All HIV(+) patients were under effective cART, containing a combination of two nucleoside/nucleotide reverse transcriptase inhibitors (NRTI) and either a ritonavir-boosted protease inhibitor (PI) or a non-NRTI (NNRTI), with HIV RNA loads persistently below the level of detection. As a control group 43 healthy HIV(-)/HCV(-) individuals were studied. A detailed description of patient characteristics is given in Table 1 (Table 1). Informed written consent was obtained from all patients. The study had been approved by the local ethics committee of the University of Bonn (Bonn, Germany).

I) NK cell analysis: Frequency and phenotype
Cryopreserved PBMC were thawed, washed with phosphate buffered saline (PBS) and stained with anti-CD3 and anti-CD56 fluorochrome-labeled antibodies for detection of NK cells by fluorescence activated cell sorting (FACS). In addition, cells were stained with different combinations of the following fluorochrome-labeled antibodies: anti-CD27, anti-CD57, anti-CD62L, anti-CD69, anti-CD127, anti-NKG2A, anti-NKG2C, anti-NKG2D, anti-NKp30, anti-NKp46, anti-CD107a and anti-INF-γ (Table 2). Isotype controls were used to set up gates and to determine positive and negative populations, respectively (S1 Fig). Dead cells were excluded by use of Zombie Aqua Fixable Viability Kit (Biolegend, Fell, Germany). After incubation of the cells with 1 to 10 μl of antibody solution (depending on antibody) and washing with PBS, samples were analyzed on a FACSCanto II flow cytometer using CellQuest Pro (BD Biosciences) and  (Fig 1).
II) NK cell analysis: CD107a degranulation assay NK cell degranulation was assessed in response to HuH7 HCV replicon cells. To this end, cryopreserved PBMC were thawed and cultured at 37˚C in complete medium without any exogenously added cytokines. After 12 hours, PBMC were washed, resuspended and pre-stimulated with recombinant human interleukin (rhIL)-2 (25 U/mL; R&D Systems, USA). Then, cells were co-incubated with HuH7 HCV replicon cells at an effector/target (E:T) ratio of 1:1 in the presence of anti-CD107a to assess degranulation, as described before [37].

III) NK cell analysis: IFN-γ secretion
Cryopreserved PBMC were thawed, pre-stimulated with rhIL-2 (25 U/mL; R&D Systems, USA) followed by co-incubation with HuH7 HCV replicon cells at an E:T ratio of 1:1 at 37˚C for 5 hours. Brefeldin A (10 μg/mL; Sigma-Aldrich, St. Louis, MO, USA) was added after 1 hour of co-culture. Next, cells were harvested and washed, followed by intracellular staining with anti-IFN-γ and flow cytometry analysis.

Statistical analysis
Frequencies of cell sub-populations between the different cohorts were compared using Kruskal-Wallis tests. Test multiplicity was controlled by a false discovery rate (FDR) procedure accounting for dependency among statistical tests [38]. FDR-adjusted P values <0.05 were

Frequency of NK cells in HIV/HCV co-infection
First, we analyzed the frequency of NK cells among our study cohorts (Fig 2A). The frequency of circulating NK cells, assessed as percentage of CD3 -CD56 + cells among PBMC, was significantly lower in HIV/HCV co-infection compared to healthy controls but did not differ significantly from that observed in HIV and HCV mono-infected patients, respectively ( Fig 2B, left panel). HIV/HCV co-infection was associated with an increased frequency of CD56 bright NK cells, whereas its proportion of CD56 dim and CD56 bright cells was similar to both monoinfected groups, respectively ( Fig 2B, right panel).

Expression of maturation/differentiation markers in HIV/HCV co-infection
Next, we studied the NK cell maturation/differentiation status in our cohorts ( Fig 3A). Compared to healthy controls we found HIV/HCV co-infection to be associated with a decreased frequency of CD56 bright NK cells expressing markers characteristic for less mature NK cells, such as CD27, CD62L and CD127 (Fig 3B), whereas the proportion of CD57 + NK cells was  Alterations of the NK cell pool in HIV/HCV co-infection

Expression of NK cell receptors in HIV/HCV co-infection
Then, we analyzed the expression of NK cell receptors (NKR) which have been shown to critically regulate NK cell functions (Figs 4A and 5A). As is shown in Fig 4B we found HIV monoas well as HIV/HCV co-infection to be associated with a significantly lower frequency of Alterations of the NK cell pool in HIV/HCV co-infection NKG2A-expressing NK cells than healthy controls whereas patients with chronic hepatitis C displayed a significantly higher frequency of NKG2A + NK cells compared to all other studied groups ( Fig 4B, upper left panel). Moreover, we found the frequency of CD56 dim NKG2C + NK cells to be significantly higher in HIV/HCV co-and HIV mono-infected patients than in healthy individuals ( Fig 4B, upper right panel). In addition, we observed the frequency of NKG2D + NK cells to be reduced in co-infected patients (Fig 4B, lower left panel), whereas NKG2D surface expression density (RFI) was not affected (Fig 4B, lower right panel). Similar observations were made in patients mono-infected with HCV. However, CD56 bright NK cells in HCV patients displayed an increased surface expression density of NKG2D. With respect to expression of the natural cytotoxicity receptors (NCR) NKp30 and NKp46 we found HIV(+) patients to display a significantly lower frequency of NKp30 + NK cells than HCV mono-infected patients and healthy controls, irrespective of HCV co-infection ( Fig 5B, left panel), whereas the frequency of NKp46 + NK cells was reduced in all patient cohorts (Fig 5B, right panel).

HIV and HCV infection affect NK cell functionality
Finally, we studied the NK cell activation status ( Fig 6A) and functional capacity ( Fig 7A). As is depicted in Fig 6B, the frequency of NK cells expressing the activation marker CD69 was not different in HIV/HCV co-infected patients from healthy controls, while HCV mono-infected patients showed the highest frequency in comparison to all other groups (Fig 6B). With respect to NK cell function, we observed degranulation of rhIL-2-stimulated NK cells following co- incubation with HuH7 HCV replicon cells to be significantly impaired in all virus-infected patient groups as compared to healthy controls (Fig 7B, upper right panel). Of note, this defect was most pronounced in HIV/HCV co-infected patients, which were found to have a significantly lower frequency of CD107a + NK cells than patients mono-infected with HIV and HCV, respectively. No such differences were seen when NK cell degranulation was studied in the absence of target cells (Fig 7B, upper left panel). Analyzing IFN-γ production of rhIL-2 stimulated NK cells we observed a significantly reduced percentage of IFN-γ + NK cells in HIV mono-infected patients compared to healthy and HCV mono-infected individuals (Fig 7B,  lower right panel). However, this functional defect was strongest in HIV infected patients coinfected with HCV. IFN-γ production of unstimulated NK cells was similar between the four study groups (Fig 7B, lower left panel).

Discussion
NK cells have been shown to importantly contribute to anti-HCV immune responses and to modulate HCV-associated liver fibrosis[10,12,15,29]. However, little is known regarding the effects of HIV/HCV co-infection on functions and phenotype of NK cells. Here, we show that HIV/HCV co-infection is associated with a significant dysregulation of the NK cell pool, including both phenotypic as well as functional alterations. NK cell development is considered to proceed sequentially from CD56 bright to CD56 dim NK cells [39][40][41]. This process is associated with a gradual decrease in CD27, CD62L and CD127, all highly expressed on circulating CD56 bright cells, together with a progressive increase in receptors characteristic of mature cells, such as CD57[18,42-50]. Of note, we found HIV/HCV co-infection to be associated with a shift towards a more mature NK cell phenotype as frequency of NK cells expressing CD27, CD62L and CD127 was significantly lower than in healthy controls, whereas the proportion of mature CD57 expressing NK cells was increased. Moreover, we found the frequency of NK cells positive for NKG2A, which has been shown to be highly expressed on less mature NK cells but to display a stepwise decrease during terminal . Thus, similar mechanisms may also play a role with respect to NK cells. NK cell maturation has been also observed in other viral diseases such as infection with the cytomegalovirus (CMV) and, thus, is not specific to HIV(+) individuals [47,58]. In HCV mono-infected patients, however, such a mechanism seems to be of minor if any relevance because the NK cell maturation status did not differ significantly from that found in healthy controls.
Several reports demonstrated significant alterations of the NK cell receptor expression in HIV and HCV mon-infection, respectively[11,21,24-26,28,30,59-62]. Of note, some of these perturbations have been found to differ significantly between both viral infections. The most prominent example relates to the expression of the inhibitory NKG2A receptor, which has been shown to be up-regulated in chronic HCV mono-infection[24,25,28,30,63] but reduced in HIV mono-infection [11,62,63]. These findings could be confirmed in the present study. More importantly, we found HIV/HCV co-infection to be associated with low expression of Alterations of the NK cell pool in HIV/HCV co-infection NKG2A similar to that seen in patients with HIV mono-infection. In addition, we found the expression of the activating receptors NKp30 and NKp46 to be down-regulated in both HIV mono-infected as well as in HIV/HCV co-infected patients. This is in line with earlier reports which also demonstrated a reduced surface expression of activating NK cell receptors in HIV (+) patients [59,60]. Activation of NK cells has been proposed to result in down-regulation of NKp30 and NKp46 [21,26]. Thus, low expression of NKp30 and NKp46 in HIV mono-infected and HIV/HCV co-infected patients might be due to HIV-induced NK cell activation. However, NK cell expression of the activation marker CD69 was not altered in HIV infected patients, irrespective of HCV co-infection, while elevated in HCV mono-infected patients, which rather argues against such a mechanism, since NKp30 expression in HCV monoinfected patients did not differ from that seen in healthy controls but was significantly higher than in HIV(+) individuals.
With respect to NK cell functions we observed HIV mono-infection to be associated with significantly impaired IFN-γ production, which is in line with previous reports[34, 64,65]. No such dysregulation of NK cell activity could be found in HCV mono-infection, whereas HIV/ HCV co-infected individuals displayed the lowest IFN-γ production. The degranulation rate of NK cells was reduced in all infected individuals while, similar to IFN-γ production, the most pronounced reduction could be found in co-infection. The exact mechanisms involved in impaired NK cell functions in HIV/HCV co-infection remain to be clarified. In HIV monoinfected patients, however, perturbed NK cell maturation as well as altered NK cell receptor expression have been suggested to cause dysfunctionality of NK cells [59,60,66]. Thus, it is conceivable that similar mechanisms might also play a role in co-infected patients.
Our study has several limitations. First, we only included HIV(+) patients under effective therapy and, thus, the effect of ongoing HIV replication on NK cells could not be assessed. Second, the number of cells that were available for our study was limited. Therefore, we could not analyze all markers in all patients. Moreover, a number of surface molecules important for NK cell functions and/ or differentiation, such as killer cell immunglobulin-like receptors (KIR), could not be tested. Third, we focused on CD56 + cells but did not examine the CD56counterpart. Accordingly, further studies are needed to characterize the full extent of alterations in the NK cell compartment in HIV/HCV co-infection.
Taken together, our data indicate that HIV/HCV co-infection is associated with significant alterations of the NK cell pool, which mainly reflect alterations observed in HIV monoinfection.