1. Introduction
Despite the introduction of novel diagnostic and therapeutic modalities for the management of cytomegalovirus (CMV), it remains a significant cause of serious posttransplant complications, with high rates of mortality and graft loss [
1]. Donor (D) and recipient (R) CMV serostatus is an important determinant of the risk of CMV infection posttransplantation. Although serostatus is useful, it does not allow us to precisely predict who is at risk for infection. Despite the implementation of prevention strategies that are based on pretransplant D/R serostatus, a high incidence of late-onset CMV has occurred in all studies that have evaluated universal prophylaxis. Apart from pretransplant serology, other risk factors for CMV infection include certain types of transplant (e.g., lung), shorter courses of prophylaxis, disease severity, higher levels of immunosuppression, transplant failure, allograft rejection, and the presence of immune reconstitution [
2,
3,
4,
5].
The immune response to CMV infection is complex. The role of both adaptive and innate responses has not been well established in transplant recipients. Natural killer (NK) cells are effector cells that constitute a crucial part of the innate immune system and provide the first-line of defense against and control of CMV infection [
6]. The inhibition and activation of the NK cell response is modulated by signals that include the interplay between polymorphic killer-cell immunoglobulin-like receptors (KIRs) and their ligands, human leukocyte antigen (HLA) class I molecules. There are two types of KIR molecules (activating and inhibitory) that are expressed on NK cells and minor subpopulations of T cells. KIRs that have long cytoplasmic tails (killer-cell immunoglobulin-like receptor-2DL1 (
KIR2DL1),
KIR2DL2,
KIR2DL3,
KIR2DL5,
KIR3DL1,
KIR3DL2, and
KIR3DL3) suppress effector functions. KIRs that have short intracytoplasmic tails (
KIR2DS1,
KIR2DS2,
KIR2DS3,
KIR2DS4,
KIR2DS5, and
KIR3DS1) activate lymphocytes through an associated DAP12 molecule [
7,
8,
9].
KIR2DL4 encodes a receptor that performs both inhibitory and activating functions. Inhibitory KIRs recognize HLA class I molecules on the surface of target cells. Activating KIRs have lower affinity for HLA, and their natural ligands are less well documented. There are two categories of KIR haplotypes. Group A haplotypes carry mostly inhibitory KIRs; only
KIR2DS4 and
KIR2DL4 are activating ones. B-type haplotypes are more variable and contain more than one activating KIR gene other than
KIR2DS4 [
10]. KIR and HLA genes have been identified on two separate chromosomes (chromosomes 19 and 6, respectively) and, thus, are inherited independently. Therefore, an individual may lack the corresponding HLA ligands for KIRs. Depending on KIR and KIR ligand genotypes, people may differ substantially in their NK response. The lack of ligands for inhibitory KIRs and the presence of activating KIRs in the recipient have been associated with a protective effect on the rate of CMV infection after kidney transplantation [
11,
12].
The aim of the present study was to analyze the association between post-kidney transplant CMV infection and the recipient’s KIR genotype and evaluate other possible risk factors for the occurrence of CMV infection in this patient population.
3. Discussion
The present study analyzed the association between KIR genotype diversity in kidney transplant recipients and the occurrence of CMV infection. The key finding of this study was the association between the lack of activating
KIR2DS2 and posttransplant CMV infection. This phenomenon was also described in patients after hematopoietic cell transplantation (HCT). In the study by Zaia et al., the presence of
KIR2DS2 and
KIR2DS4 in the donor genotype was associated with lower CMV infection in HCT recipients [
14]. This finding was supported by the fact that upregulation of the KIR expression of 2DS2 and 2DS4 protected against CMV reactivation in the early post-HCT time.
KIR2DS2 and
KIR2DS4 expression was elevated in individuals after HCT compared with donor expression prior to transplant and in recipients with CMV DNAemia compared with non-viremic recipients [
15]. In another study of a D+/R− organ transplant recipient cohort, the presence of
KIR2DS2 in combination with
KIR2DL3 was associated with protection against CMV viremia. However, this protective effect was only present if neither the donor nor the recipient expressed any HLA-C2 molecules [
16]. In immunocompetent individuals,
KIR2DS2 was found to be negatively associated with the risk of infection, in which it was more frequent in controls than in those with CMV infection (OR = 0.14) [
17]. Additionally, the binding of
KIR2DS2 to different peptides that derive from diverse pathogens and to different HLA class I molecules is particularly interesting given its protective role in acute viral encephalitis and chronic hepatitis C virus infection [
18,
19,
20]. Interactions between viral epitopes, KIRs that are expressed on NK cells, and HLA ligands can lead to either the activation or inhibition of cytotoxic NK cell activity. KIR and HLA genotypes segregate independently of each other. Therefore, a proportion of KIRs likely has no HLA ligand. Conversely, individuals may possess HLA ligands for which they have no KIR. Group 1 HLA-C (HLA-C1) allotypes are characterized by an asparagine residue at position 80. They are ligands for the inhibitory receptors
KIR2DL2 and
KIR2DL3, which segregate as alleles of a single locus.
KIR2DL2 binds HLA-C1 with greater affinity than
KIR2DL3 [
21]. Our results showed that subjects who had
KIR2DL3 and
KIR2DL2–HLA-C1 were more likely to develop CMV infection. These findings are consistent with previous reports of KIR and HLA genotyping, indicating that the absence of HLA ligands for inhibitory KIRs could be associated with a significant reduction of CMV infection [
11]. The cumulative evidence shows that
KIR2DL2–HLA-C1 and
KIR2DL3 may be molecular biomarkers of virus-induced diseases [
22,
23,
24].
To date, it has been difficult to demonstrate the binding of
KIR2DS2 to HLA-C1. The sequences of
KIR2DS2 are very similar to
KIR2DL2. Despite shared sequence homology, one key difference between
KIR2DS2 and
KIR2DL2 is the presence of a tyrosine (compared with a phenylalanine) residue at position 45, and this is thought to substantially affect the binding of
KIR2DS2 to HLA-C [
25]. Steward et al. examined the binding of activating KIR to cells that were infected with human herpes viruses, including CMV, inducing no detectable
KIR2DS2 ligand at the cell surface [
26]. A very low affinity of the
KIR2DS2–HLA-C interaction that does not have the same degree of peptide selectivity as other KIR–HLA interactions was the reason that we did not analyze
KIR2DS2–HLA-C. We are aware that the lack of information concerning KIR2DS2 ligands may be a limitation of the present study.
Other authors did not find that KIR and HLA genotypes are related to protection against CMV infection in kidney transplant recipients [
27]. One possible explanation for this finding is that the previous study analyzed a patient cohort with a high risk of CMV infection (i.e., exclusively CMV-negative recipients of a CMV-positive donor kidney).
KIR2DL2 and
KIR2DS2 belong to the KIR B haplotype. The majority (74.8%) of our recipients were CMV-seropositive before transplantation. Evidence indicates that KIR B haplotypes modify the risk of CMV viremia solely in patients who have already been exposed to CMV before organ transplantation, suggesting that “primed” or “memory-like” NK cells are cellular correlates for this protective effect [
28].
The incidence of CMV infection or reactivation was proposed to be increased by the heritance of multiple activating KIR genes [
12,
29]. We observed a trend toward a lower incidence of CMV infection in recipients who carried the KIR B/X genotype, but no significant difference was found between both the KIR A/A and KIR B/X genotypes (
p = 0.065). The degree of protection against CMV was also suggested to increase with the number of activating KIR genes [
12]. We analyzed the potential additive protective effect of the number of activating KIRs relative to CMV occurrence, but we did not confirm such an association.
In the present study, a protective effect against CMV infection in recipients who had the
KIR2DS2 genotype was found in 73.6% of the individuals, but 26.4% who had this protective genotype still had CMV infection. This may be attributable to the fact that the KIR genotype profile does not necessarily correlate with concurrent expression. The expression of
KIR2DS2 was previously shown to vary among genotype-positive healthy individuals, in which some were expressers and some were non-expressers [
15]. We did not investigate the expression of KIRs, which may be considered a limitation of our study.
Similar to previous reports, the present study confirmed that donor and recipient (D/R) CMV serostatus and allograft dysfunction were key predictors of the risk of CMV infection after kidney transplantation [
2,
3,
5,
30]. Renal insufficiency is associated with immune dysfunction, including lymphocytopenia, which contributes to the high prevalence of infections [
31]. The present study confirmed our previous findings that total lymphocyte count before posttransplant day 90 is a risk factor for posttransplant CMV infection [
32]. This phenomenon has also been described in liver transplant patients [
33,
34].
Interestingly, in the present study, an earlier time of antiviral prophylaxis initiation was not significant in the univariate analysis, but it became a significant predictor of posttransplant CMV infection in the multivariate analysis. Antiviral prophylaxis that is initiated promptly after transplantation could impair the development of an adequate CMV-specific response of cytotoxic and helper T lymphocytes, which seems to be crucial for protection against CMV infection. This finding is consistent with another study that found that a 14-day delay in the initiation of long-term prophylaxis in (D+/R−) organ transplant recipients could prevent the development of late CMV infection [
35].
5. Conclusions
In summary, our results suggest that genotyping recipients for the KIR2DS2, KIR2DL2, and KIR2DL3 genes may help predict the occurrence of posttransplant CMV infection. Other factors, such as graft function, time of antiviral prophylaxis initiation, lymphocyte blood count, and pretransplant serostatus, were independent predictive factors for the occurrence of CMV. Currently, the risk stratification for posttransplant CMV infection is primarily based on the assessment of D/R CMV serostatus prior to transplantation. In our opinion, this strategy could be improved by implementing other biomarkers that are predictive of CMV infection after kidney transplantation, which would allow for more personalized management. The assessment of KIR2DS2 prior to transplantation could potentially be used in combination with D/R serostatus to more accurately predict the risk of CMV infection, including the precise identification of transplant individuals who require a longer duration of antiviral prophylaxis therapy in individuals who do not have KIR2DS2. In conclusion, our findings confirm that KIR/HLA genotypes play a significant role in anti-CMV immunity and suggest the contribution of not only environmental but also genetic factors in the incidence of CMV infection after kidney transplantation.