Roles and Mechanism of Natural Killer Cells in Clinical and Experimental Transplantation

Suraksha Agrawal; Piyush Tripathi; Sita Naik


Expert Rev Clin Immunol. 2008;4(1):79-91. 

In This Article

NK Cells in Transplantation

Allogenic stem cell transplantation (ASCT) is used widely for the treatment of leukemias and a variety of solid organ tumors.[41] Although the success of ASCT is crucially dependent on accurate matching of donor and recipient for the MHC loci because of the extensive polymorphism of these genes, there is an increasing compulsion to use allogeneic donors. The large number of allogeneic ASCTs being currently performed has provided opportunities to study the role of other genetic polymorphisms that are determining the success of the procedure.[42,43,44] The success of ASCT is limited by the occurrence of GVHD and, in the case of transplantations for treatment of malignancies, the relapse of the tumor itself. While regimes that prevent the occurrence or decrease the severity of GVHD also increase the rate of relapse; attempts to reduce the relapse rate results in increased rate of GVHD. ASCT across MHC-compatible pairs can still generate NK cell cytotoxicity due to KIR-ligand mismatch.[45] This is probably the explanation for the phenomenon of 'hybrid resistance' in which subsets of alloreactive NK cells cause rejection of homozygous parental grafts by the heterozygous F1 recipients.[46] The essential role of NK cells in the rejection of bone marrow grafts and transplantable lymphomas in mice,[47,48] as well as human MHC-mismatched allogeneic systems, has been well established.[49]

The role of NK cells in transplant rejection seems contradictory with observation that haploidentical ASCT with mismatch for KIR-ligand was associated with a significantly increased overall survival, better engraftment and reduced incidence of GVHD in patients with AML.[40] This is attributed to the NK cell-mediated clearance of residual leukemia cells (decreased relapse rate), host T cells (better engraftment) and host dendritic cells (reduced GVHD).[40] Thus, NK cells were actually beneficial and not detrimental to the graft. However, the experience has not been uniformly favorable and there are reports of decreased survival following KIR ligand-mismatched allografts.[50] These differences have been attributed to heterogeneity of treatment protocols and patient cohorts. Hence, the issue of using NK cells for adaptive immunotherapy after bone marrow transplantation (BMT) is still under debate.

In spite of the opposing role of NK cells in promoting rejection of grafts, successful ASCT protocols that deliberately use HLA mismatches to improve graft survival have been described.[51] In these situations, the NK cells are responsible for the graft-versus-leukemia effect, but do not participate and may even prevent the development of GVHD.[52] It has also been shown that very large doses of incompatible NK cell clones can be administered to murine transplant recipients without any GVHD.[53] This effect is mediated, at least in part, by the immunosuppressive cytokine TGF-β. BMT studies in mice also indicate that the beneficial effects of NK cells are optimal if they are administered soon after the transplant. Thereafter, NK cells and, more importantly, IL-2, which is used to activate them, are detrimental and can exacerbate the subsequent GVHD. Nevertheless, NK cells may add to target-cell damage through GVHD-initiated upregulation of surface molecules that activate NK cells through their NKG2D ligand.[54]

Conventional wisdom was that NK cells do not participate in rejection of solid organ transplants. This was based on the observation that depletion of NK cells in rat recipients of skin, heart or liver allografts did not alter the kinetics of rejection.[55] Conversely, these transplants into RAG1-/- or severe combined immunodeficient mice with intact NK cells survived indefinitely.[56,57] More directly, skin transplants from β2-/- donors lacking MHC class I expression were accepted without rejection.[58] However, more recently, NK cells have been implicated in the rejection of solid organ grafts.[59,60,61] For instance, NK cells infiltrate rat liver allografts and these cells are responsible for early chemokine production (MIP-1, ACCL1, CCL3, CXCL10 and CX3CL1), a key part of the acute rejection episode.[62,63] NK cells present in the grafts have also been implicated in amplifying the intragraft inflammation.[64] NK cell depletion has been shown to prolong the allograft survival in CD28-/- recipients, implicating them in the rejection that occurs despite costimulatory blockade. Thus, although NK cells may not be able to mediate rejection themselves, they participate by facilitating the action of alloreactive cells. This is assisted by the help NK cells provide for maturation of immature recipient cencritic cells[65] and also the IFN-γ produced by activated NK cells that may cause upregulation of MHC class II expression on graft endothelium.[66,67]

In human transplantation, NK cells expressing cytolytic proteins, such as granzyme A and B, have been demonstrated in renal allograft during episodes of acute rejection.[32,68,69,70] A significant correlation was found between acute rejection episodes and donor-recipient KIR mismatches in human liver allograft recipients.[71] In liver transplants, it has been reported that monitoring of KIR2D(+)CD8(+) T cells, particularly KIR2DL1/S1(+)CD8(+) T cells at pretransplant, and both KIR2DL1/S1(+) and KIR2DL2/3/S2(+) T-cell subsets at early post-transplant period, could offer useful information for clinical follow-up of liver grafts.[72] In preoperative human renal transplant patients, a high correlation was noted between the presence of predicted NK cells alloreactivity and the ability of these NK cells to mount a donor-specific cytotoxic attack.[73] However, others have found no correlation between predicted NK cell alloreactivity and acute liver allograft rejection, despite the fact that recipients with predicted NK alloreactivity did have an expanded population of circulating NK cells 1 month post-transplant compared with recipients without predicted NK cell alloreactivity.[74] IFN-γ produced by NK cells infiltrates liver allografts immediately after transplantation and links the innate and adaptive immune responses.[63]

NK cells have been implicated in the process of chronic vascular rejection (CVA). The NK cell-activating combination of KIR2DS2 and its HLA-C ligand have been shown to be a risk factor for development of vascular problems in human rheumatoid arthritis[75] and similar mechanisms may be operative in the CVA of chronic solid organ rejection.[76,77] Interestingly, the proportion of CD56+ NK cells was increased significantly in heart transplant recipients with documented CVA as compared with those without CVA.[78]

Obviously, more studies in human acute and chronic rejection are warranted as immunosuppressive agents, such as ciclosporin[49,73,79,80] and FK506,[81] appear to be ineffective in suppressing NK cell responses.

Two phenotypically and functionally distinct subsets of NK cells have been described, CD16+ CD56dim and CD16- CD56bright. The CD56bright NK cells produce significantly greater levels of IFN-γ, TNF-β, granulocyte-macrophage colony-stimulating factor, IL-10 and IL-13 in response to monokine stimulation than do CD56(dim) NK cells.[82] CD56+bright NK-cells also express higher levels of several adhesion molecules, such as CD2, CD11c, CD44, CD56 and CD62L, compared with CD56+dim NK cells, making these cells more efficient in migration from blood to tissues and to form conjugates with target cells.[83] The CD56dim NK cells, by contrast, have more cytotoxic capacity then the CD56bright subpopulation and are believed to be a more advanced lineage proceeding toward death.[84]

The telomere length in CD56(dim)CD16(+) NK cells is significantly shorter than in CD56(bright)CD16(-) NK cells from the same donor and NK cells expressing activation markers, such as NKG2D and LFA-1, have significantly shorter telomeres.[85] CD56bright CD16- KIR- and CD56dim CD16+ KIR± NK cells are believed to represent sequential stages of differentiation and secondary lymphoid organs are believed to be sites where final maturation and self-tolerance acquisition occurs. NK cells collected from nonreactive lymph nodes display almost no KIR and CD16 expression, whereas NK cells derived from highly reactive states express significant amounts of KIR and CD16, suggesting that CD56(bright) NK cells acquire these molecules during activation and escape into the periphery as KIR(+)CD16(+) NK cells.[86]

Ruggeri et al. proposed that donor-derived NK cells lacking inhibitory KIRs for host MHC class I molecules mediate beneficial NK-versus-leukemia effects following an HLA haplomismatched BMT.[87] He investigated 20 donor-recipient pairs, in which the NK cell inhibitory KIRs of the haploidentical donor did not recognize one of the MHC class I ligands on the leukemic blasts of the patient. These patients had a lower incidence of leukemic relapse compared with those patients without KIR-MHC-class-I mismatch.[87] Donor NK cell clones were able to lyse target cells derived from patients or, in a few cases, lysed patient leukemic blasts directly. It was proposed that CD56+dim NK cells or Il-2-activated CD56+bright NK cells might have caused an enhanced NK-versus-leukemia effect. If all these mechanisms are well understood, selective block of these inhibitory components might allow for the therapeutic enhancement of NK cell effects.[83]

In a study of 52 patients transplanted with HLA-identical bone marrow (by Rhoades et al.),[88] there was a profound decrease in absolute number of CD3- T cells and an increase in CD3- CD16+, CD56+ (a subset that coexpresses CD8+ 'dim') NK cells in the peripheral blood lymphocytes. Concurrent with the emergence of acute GVHD (aGVHD), CD4+ T and CD20+ B lymphocytes failed to recover within 90 days in the patients with grades II-IV aGVHD. Ex vivo partial T-cell depletion in itself did not significantly impair T-cell recovery compared with that in non-T-depleted recipients, unless aGVHD occurred. Although leukocyte cellular infiltration in the skin was generally sparse, CD16+ NK lymphocytes were significantly increased in grades II-IV aGVHD. By contrast, there was no increase in CD3+, CD4+ or CD8+ lymphocytes in these lesions compared with skin biopsies obtained from BMT patients without aGVHD or from normal skin. Taken together, these findings suggest that NK cells may be important in the pathogenesis of human aGVHD.

Some recent studies have investigated the effect of immunosuppressive drugs on NK cell subsets in the outcome of allotransplantation. Wang et al. studied the effect of ciclosporin A (CsA), which is commonly used to prevent GVHD on NK cell activity.[89] They cocultured NK cells with IL-2 and IL-15 in the presence or absence of CSA for 1 week and found that CSA-treated cultures showed fewer CD56(+)CD16(+)KIR(+) NK cells and a reciprocal increase in CD56(+)CD16(-) KIR(-) cells. These changes were mainly due to a reduced proliferation of the CD56(dim) NK cell subpopulation and a relative resistance of CD56(bright) NK cells to CSA.[90] Their results indicated that CSA influences NK cell function and phenotype, which may have important implications for graft-versus-leukemia effects.

Natural killer T (NKT) cells are a special subset of T cells that are characterized by coexpression of the NK cell receptor-CD161 and an invariant TCR-α chain (V-α 14J α28). They are generated in the thymus, represent less than 0.1% of peripheral blood lymphocytes[91] and are most abundant in the liver, spleen and bone marrow.[92] Activation of NKT lymphocytes lead to rapid elaboration of either IFN-γ or IL-4,[93] thus these cells have the capability to mediate both proinflammatory and anti-inflammatory immune responses.

NKT cells have been implicated in the control of several autoimmune diseases in mice and humans, including Type 1 diabetes, experimental autoimmune encephalomyelitis, systemic lupus erythematosus and systemic sclerosis.[94] The importance of NKT cells in transplantation was highlighted by the finding that NKT cells acted as 'natural suppressor cells' and prevented GVHD after bone marrow transplantation.[95] NKT cells are also required for the induction of cardiac-transplant tolerance by costimulatory blockade of CD28/B7 and LFA-1/ICAM-1 and tolerance induction by tolerizing antibodies.[96,97] Anti-CD4 monoclonal antibodies allowed rat islet xenografts to be transplanted into C57BL/6 mice without any immunosuppressive drugs.[96] This beneficial effect appeared to be associated with reappearance of CD41 NKT cells 14 days post-transplantation. Anti-CD4 monoclonal antibody treatment did not induce tolerance toward transplanted allograft in val 4-deficient mice; however, adoptive transfer of val 14 NKT cells into these animals restored acceptance of rat islet cell xenografts.[97] Thus NKT cells may have an important role for induction of systemic tolerance in various conditions.

NKT cells and neutrophils have also been shown to play an important role in the innate immune response in case of renal ischemia reperfusion injury (IRI), which is a CD4+ T-cell-dependent event. When mice were administered with antibodies to block CD1d or deplete NKT cells, or in mice deficient of NKT cells, the immune response was attenuated markedly.[98] These effects were associated with a significant decrease in renal infiltration and activation of NKT cells, and a decrease in IFN-γ-producing neutrophils. The results support the essential role of NKT cells and neutrophils in the innate immune response of renal IRI by mediating neutrophil infiltration and production of IFN-γ.

Long-term survival of corneal allograft is also NKT cell dependent, probably through the induction of regulatory T cells. The results obtained with NKT cells may be due to the heterogeneous nature of these cells. Indeed, CD4+, CD8+ and CD3+ CD4- CD8- subsets of NKT cells have been identified recently. Further understanding of these subpopulations of NKT cells will be required in order to manipulate these cells as regulatory cells in various pathological immune responses.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: