Background
NK cells are effector cells of the innate immune system, capable of destroying virus-infected and tumour cells without prior sensitization [
1,
2]. The majority of human NK cells in peripheral blood are CD56
dimCD16
+ cells whereas CD56
brightCD16
+/− cells only constitute approximately 10% of the peripheral blood NK cell pool [
3]. In addition, CD56
bright NK cells have high surface density expression of type II membrane glycoprotein CD94, L-Selectin 62L, CD127 and lymph node homing receptor CCR7 but low expression of the low affinity IgG-Fc receptor III (CD16), killer cell immunoglobulin-like receptors (KIRs) and cytotoxic molecules such as perforin and granzyme B [
4]. Thus, NK cell subsets seem to perform distinct roles in the immune response. CD56
bright NK cells have more regulatory functions by means of cytokine production while CD56
dim NK cells are primarily cytolytic in function but produce significant amounts of cytokines when their activating receptors are engaged [
5]. This distinction is however not absolute [
6].
The chemokine receptor repertoire also differ among the NK cell subsets with the expression of CXCR3 at a much stronger density on CD56
bright subset than CD56
dim cells, while the latter express CXCR1 and CX3CR1 exclusively [
5,
7]. The differential repertoires of chemokine receptors and adhesion molecules endow NK cell subsets with divergent migratory properties: the CD56
bright subset preferentially homes to secondary lymphoid organs whereas the CD56
dim cells home to acute inflammatory sites [
5,
7].
Persistent infections by HCV and HIV present unique challenges to the immune system as they result in highly chronic viral persistence which lasts indefinitely without antiviral treatment [
8]. In HIV infection, several studies have shown functional impairment in NK cell cytokine secretion and cytotoxicity. Both acute and chronic untreated HIV infection is associated with alterations in NK cell subset distribution, with partial loss of CD56
+CD16
+ cells and expansion of CD56
−CD16
+ NK cells having impaired cytolytic function, increased activation markers and decreased cytokine production [
9,
10]. Similarly, HCV viral persistence has also been shown to be associated with defective NK cell responses in vivo both in the periphery and in the liver [
7,
11]. NK cell function was rescued after successful IFN-α therapy in chronically HCV infected patients [
12].
HIV infection is not only associated with changes in NK cell subpopulation but also with marked alterations in NK cell surface receptor expression and loss of function [
13]. In HIV viremic patients, there is an overall decrease in surface receptor density of NKp46 and NKp30 found on freshly isolated NK cells and dysfunction in NKp44
de novo expression upon stimulation in vitro resulting in an NCR dull phenotype [
14]. The proportion of NKp46 and NKp30 expressing CD56
dim NK cells and their cytolytic activity was shown to decrease with disease progression [
15].
Contrary to HIV infection, there is no consensus on NCR expression during HCV infection. There are reports that show increased proportions and density of NCRs including NKG2C, NKp44, NKp30 and NKp46 [
16,
17] while the earlier reports of decreased expression of NKp46 have not been subsequently confirmed [
7]. In addition, couple of recent studies have suggested that HCV infected cells may selectively down regulate NKp30 and impair NK cell function by this mechanism [
14,
18,
19].
Nearly all NK cells express NKG2D which is considered a potent activating receptor [
20] that has the ability to trigger cytotoxicity and at the same time capable of overriding signals provided by other inhibitory receptors. Similar to other NCRs, there is conflicting evidence with respect to NKG2D expression which has been reported from being up-regulated or down-regulated to being unchanged during chronic HCV infection [
16,
21].
A number of studies have revealed CD56
bright and CD56
dim NK cells as separate NK cell subsets rather than a homogenous population having unique roles in the innate immune response [
7]. By virtue of their ability to produce different cytokines, CD56
bright NK cells might play an important role in early immune responses as well as in shaping of the adaptive response [
5]. Not much is known about the impact of HIV and HCV co-infection on CD56
bright NK cells. In the current study, we therefore investigated the phenotype of CD56
bright NK cells in HIV-HCV co-infected subjects and compared these with HCV and HIV mono-infected patients as well as with healthy controls. We found that HIV-HCV co-infection is able to modulate the phenotype of CD56
bright NK cells in a complex way.
Discussion
HIV-HCV co-infected individuals progress more rapidly to fibrosis, cirrhosis, liver failure and hepatocellular carcinoma than patients infected with HCV alone [
7,
24]. Both HIV and HCV mono-infections have been shown to have impaired NK cell functions and skewed subset distributions [
7,
9]. However, the impact of HIV-HCV co-infection on phenotypic alterations of NK cell subsets is not well understood. Here, we investigated the phenotype of CD56
bright NK cells in HIV-HCV co-infected individuals and compared it with HIV or HCV mono infected patients as well as with healthy controls. In our study we found an expansion of CD56
bright NK cell subset in HIV-HCV co-infected cohort similar to reported for HCV mono-infection [
25,
26] although there is a study reporting no change in CD56
bright NK cell compartment in HCV mono-infection [
27]. In line with our findings, in a cohort of female patients chronically infected with HCV, the proportion of CD56
bright NK cells was increased as compared to HCV resolvers and healthy controls [
25]. The same increased percentage of CD56
bright NK cells was found among individuals with a positive tuberculin skin test compared with patients with overt tuberculosis and normal controls [
28]. The significance of increased CD56
bright cells in our study is currently unclear. Expansion of CD56
bright NK cells in HCV infection might have resulted from a decreased rate of differentiation towards CD56
dim NK cells in these studies. However, in our study the percentage of total NK cells and CD56
dim NK cells did not differ among all the groups (data not shown). The increased percentage of CD16 expression on CD56
bright NK cells in HIV-HCV co-infected patients indicated a phenotypic shift towards CD56
dim NK cells and their subsequent role in antibody dependent cellular cytotoxicity (ADCC). Increased expression of CD16 on CD56
bright NK cells is also reported in HIV-1 infected individuals which is partially restored after ART [
29].
NK cell function is primarily regulated by balance between activating and inhibitory receptors. NKG2D and NCRs are one of the main NK cell activating receptors that are required for NK cell effector functions. NKG2D and NCRs are differentially expressed by different subsets of NK cells [
20,
30]. We analysed the expression of NKG2D and NCRs (NKp30 and NKp46) on CD56
bright NK cell subset. NKG2D was upregulated in all the three infected populations as compared to healthy controls. The cytotoxic potential of all NK cell subsets increases when stimulated with cytokines like IL-2 or IL-12 in vitro [
7]. This might be one of the reasons for the increased expression of NCRs and NKG2D on CD56
bright NK cells in vivo as well in chronic viral infections.
In HIV viremic patients, there is an overall decrease of NKp46 and NKp30 on NK cells [
14] while in HCV infection couple of studies have shown reduced NKp30
+ NK cells. In our study, we did not observe a decreased NKp30 expression on CD56
bright NK cells in HCV mono-infection. Thus, reduced NKp46 and NKp30 on CD56
bright NK cells is a unique hallmark of HIV infection [
14,
18,
19]. However these results are not directly comparable or necessarily in conflict as our results are focussed on the different fractions of CD56
bright NK cell subset.
In our study NKp46 expression on HIV-HCV co-infected group differed significantly from both HIV and HCV mono-infected groups while NKp30 expression was increased on CD56bright NK cell subset in HIV-HCV co-infected cohort which differed significantly as compared to HIV mono-infected group. The finding of increased frequency of NKp30 and NKp46 expressing CD56bright NK cells in HIV-HCV co-infection as compared to HIV mono-infection also supports a shift of the populations either due to increased apoptosis of activated CD56bright NK cells or an increase of ‘HCV-like’ CD56bright NK cells in case of NKp30 expression while indicating a unique and an important role of NKp46 in HIV-HCV co-infection.
Immune activation is a hallmark of chronic viral infections like HIV [
31]. HIV infected patients showed an increased activation of CD56
bright NK cells as compared to HIV-HCV co-infected and HCV mono-infected groups as measured by the expression of CD69. This might be because of the ongoing residual viral replication even after one year of ART [
32]. The reduced expression of CD69 in HIV-HCV co-infected cohort suggests a decreased activation status of CD56
bright NK cells in HIV-HCV co-infection than HIV mono-infection. However, it is unlikely that HCV co-infection might ‘protect’ CD56
bright NK cells from increased immune activation. Alternatively, a loss of CD69 expressing CD56
bright NK cells could be either due to increased apoptosis or a relative increase of CD69 negative NK cells in the HIV-HCV co-infected cohort or both. We therefore analysed the expression of the apoptosis marker CD95.
Immune cells undergo apoptosis as a result of chronic activation or exhaustion during chronic viral infections. Fas mediated apoptosis (FMA) is one of the mechanisms by which cells undergo apoptosis [
33]. The expression of Fas receptor CD95 on CD56
bright NK subset was higher in HIV mono-infected patients. The same group also had the high expression of CD69 on the CD56
bright NK cells showing a chronic systemic immune activation status. The increased expression of CD69 as well as CD95 on CD56
bright NK cell subset in HIV treated patients may render activated NK cells more prone to undergo apoptosis. Enhanced susceptibility of NK cells to cell death by CD95-CD95L in HIV infected viremic individuals has already been reported by Kottilil
et al. [
34]. A number of reports show increased expression of CD95 on various lymphocyte subsets as a result of immune activation induced by HIV viremia thereby facilitating lymphocyte apoptosis [
35,
36]. The expression of CD69 on CD56
bright NK cells in HIV-HCV co-infected group was closer to HCV mono-infected group whereas the CD95 expression behaved more similar to HIV treated mono-infected group thus showing imprints of either infection.
CD127 (IL-7Rα receptor) is important for homeostatic proliferation, particularly in lymphopenic settings when remaining cells initiate a homeostatic response to repopulate the depleted compartment. The level of IL-7 and expression patterns of CD127 are believed to be skewed in HIV [
37,
38]. Our study also demonstrated a skewed expression of CD127 on CD56
bright NK cells in all the infected groups as compared to healthy controls, but the difference was not significant. Chronic viral infections may be associated with reduced CD127 in general.
Chemokine receptor CXCR3 and its ligands (particularly IP-10) are associated with a type 1 response and recruit lymphocytes to the liver [
39,
40]. Kimball
et al. [
41] have previously demonstrated an increased CXCR3 expression on CD8
+T cells from HIV-HCV co-infected patients as compared to healthy controls. A recent study showed HCV infection to be associated with significantly increased frequency of CXCR3
+ CD56
bright NK cells but these cells showed an impaired degranulation and IFN-γ secretion in response to hepatic stellate cells (HSCs) [
42]. In our study, we observed a decreased expression of the frequency of CXCR3
+ CD56
bright NK cells in HCV mono-infection as well as in HIV mono-infection, which was even more pronounced than in HCV mono-infection. In addition, our study showed an elevated expression of CXCR3 on CD56
bright NK cells in HIV-HCV co-infection as compared to HIV and HCV mono-infected groups which may reflect an effort to recruit more CD56
bright NK cells to liver in the wake of a co-infection.
One of the limitations of the current study is the lack of NK cell functional data in terms of NK cell cytotoxicity and cytokine secretion. Functional assays will reflect more clearly on the real impact of HIV or HCV mono-infection on HIV-HCV co-infection and may also help in deciphering the complex phenotype observed in HIV-HCV co-infection in our study.
In summary, the increased CD56
bright NK cells in HIV-HCV co-infection might be a compensatory mechanism to recruit more immunoregulatory CD56
bright NK cells. Even though NKp30 and NKp40 expression was increased in HIV-HCV co-infection, NKp40 expression was unique for HIV-HCV co-infection as it behaved neither like HIV nor HCV mono-infection. This observation may imply that among NCRs, NKp46 seems to play an important role in the cytotoxic potential of CD56
bright NK cells and in combating both HIV and HCV mono-infections. NKp30 expression was same for HCV mono-infection and HIV-HCV co-infection indicating that chronic viral hepatic infections do not have a dominant bearing on the expression of NKp30 on CD56
bright NK cells. While HIV and HCV mono-infections differed for NKp30 expression, they remained same for NKp46 expression on CD56
bright NK cells. Increased NKG2D expression in HIV and HCV mono-infections and HIV-HCV co-infection rather reflects a general feature of viral infections. Interestingly, CXCR3 expression increased in HIV-HCV co-infection vis-à-vis in HIV and HCV mono-infections - probably caused by an additional infection. It can also be speculated that increased CD69 and CD95 expression on CD56
bright NK cells in HIV mono-infection are a reflection of their exhaustive phenotype. A summary of various NK cell receptors expression is provided in Table
2.
Table 2
Summary of various NK cell receptors expression data
CD56bright NK cells | − | ↑ | − | ↑ | − | − |
CD16+CD56bright NK cells | − | ↑↑↑ | − | ↑↑↑ | ↑↑ | − |
NKG2D+CD56bright NK cells | ↑↑ | ↑↑↑ | ↑↑↑ | − | − | − |
NKp46+CD56bright NK cells | ↓ | − | − | ↑↑↑ | ↑ | − |
NKp30+CD56bright NK cells | − | − | − | ↑↑ | − | ↓↓ |
CD69+CD56bright NK cells | ↑↑↑ | ↑↑ | − | ↓↓↓ | − | ↑↑↑ |
CD95+CD56bright NK cells | ↑↑↑ | ↑↑↑ | − | − | ↑ | ↑ |
CD127+CD56bright NK cells | − | − | − | − | − | − |
CXCR3+CD56bright NK cells | ↑↑↑ | − | − | ↑↑ | − | − |
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SB and DMO conceived the study and SB, FA and DMO designed the experiments. SB and FA performed the experiments. HW, MC, JSzW, JvL provided the HCV infected samples. SB, FA, DMO analysed the data. SB wrote the manuscript. DMO, RES, SKS, HW, MC, JSzW, JvL and FA critically revised the manuscript. All authors read and commented on the manuscript.