Natural revenge over cytomegalovirus

In this issue of Blood, Foley et al show that, in allogeneic hematopoietic transplant recipients, donor-derived natural killer (NK) cells respond to cytomegalovirus (CMV) by acquiring features that are reminiscent of the specificity and memory of adaptive (T cell) immune responses.¹ 

Unfortunately, after allogeneic hematopoietic transplantation CMV disease is still, even in the era of the pre-emptive ganciclovir therapy, a major cause of death. Susceptibility to CMV reactivation is linked to immune incompetence as induced by conditioning regimens, immune suppressive measures to prevent/control graft-versus-host disease (GVHD), and the time needed for immune system maturation.

CMV has evolved escape mechanisms to evade both adaptive (CD8⁺ T cells) and innate (NK cells) immune responses. Virus-encoded genes down-regulate class I human leukocyte antigen (HLA) expression, which impairs antigen presentation and diminishes T-cell recognition but renders infected cells susceptible to NK-cell lysis (reviewed in Foley et al¹ ).

NK cells are well recognized for their ability to provide a first line of defense against viral pathogens. Their function is regulated by a balance between activating and inhibiting receptors. Clonally distributed inhibitory killer cell immunoglobulin-like receptors (KIRs) recognize allotypic determinants shared by certain groups of HLA class I alleles. On interaction with self-HLA molecules, NK cells, which express inhibitory KIRs for self, are “licensed/educated” to become fully functional, exert “missing self-recognition,” and therefore, kill target cells that bear down-regulated HLA class I molecules, such as CMV-infected cells.²  To escape NK-cell control, CMV encodes viral glycoproteins that mimic class I HLA (eg, UL-18) and interfere with expression of ligands for the activating NK receptors NKG2D and DNAM-1 (reviewed in Foley et al¹ ).

Foley and colleagues show a specific NK-cell subset expands and persists over time in response to CMV reactivation, suggesting a long-lasting specific memory had developed. As NK cells are the earliest immune cells to recover after transplantation, this observation is significant because it suggests they may contribute to controlling viral reactivation early after transplantation. Specifically, the authors show that NK cells expressing the NKG2C activating receptor expand, display a mature CD56^dim/CD57⁺/NKG2A⁻ phenotype, produce IFN-γ, persist over time, and express KIR for self (donor)–HLA class I. NKG2C⁺ NK cells that expanded after CMV reactivation responded more robustly (ie, they produced more IFN-γ) when they expressed KIR that was specific for self-HLA class I molecules. The authors concluded that NKG2C⁺ NK cells were “licensed” through self-HLA recognition and proposed CMV-favored NK-cell education. The authors suggest NKG2C was probably directly involved in the response to human CMV.

Thus, their observations may be biologically very relevant as they are providing a model of virally induced NK-cell education in the setting of allogeneic hematopoietic transplantation. These data are in line with observations of a relationship between CMV and NKG2C expression. Normal immunocompetent CMV-seropositive individuals have an increased proportion of NK cells expressing NKG2C; this percentage remains high for years after CMV infection.³  Moreover, after solid organ transplantation, CMV reactivation is associated with expansion of NKG2C⁺ NK cells.⁴  Further investigation is needed to demonstrate whether the NK-cell NKG2C activating receptor specifically recognizes CMV and, if so, the exact nature of the putative CMV ligand. Most importantly, it remains to be seen whether, and to what extent, NK cells contribute to attenuate CMV reactivation in allogeneic hematopoietic transplant recipients.

Intriguingly, experimental evidence in mice showed that transfer of NK cells recognizing the CMV m157 protein from CMV-exposed mice protected naive mice from CMV challenge.⁵  Thus, NK cells are now known to display specificity and memory, that is, properties that are conventionally attributed only to T cells. Very interestingly, in murine models and in clinical trials, NK cells have been clearly demonstrated not to cause GVHD, even when specifically alloreactive across major MHC barriers.^6,7  Thus, unlike T cells that carry the risk of causing lethal GVHD, NK cells are safe as a form of cellular immunotherapy in the allogeneic hematopoietic transplant setting. Furthermore, alloreactive NK cells are strong mediators of graft-versus-leukemia effects in T cell–depleted haploidentical hematopoietic transplantation for acute myeloid leukemia.^6,7  In this regard, it is worth noting that after unmanipulated HLA-matched sibling or unrelated hematopoietic transplantation, CMV reactivation was associated with a lower risk of leukemia relapse.⁸  The effect might have been because of CMV-induced NK-cell maturation and consequent elimination of residual leukemic cells.

Thus, an exciting potential clinical application of the data presented by Foley et al might be the use of donor NK-cell immunotherapies after transplantation to help control not only leukemia relapse, but also CMV reactivation.