Donor-specific antibodies against HLA-C, HLA-DP and HLA-DQ and their implications in kidney transplantation

Abstract

HLA-C, HLA-DP and HLA-DQ are thought to be benign due to low expression and few initial negative studies. Historically, most allocation programs used HLA-A, HLA-B and HLA-DR antigens for matching. With the advent and use of single-bead antigen assays, more was learned about donor-specific antibodies (DSAs) against these antigens. Interest in these antigens and antibodies grew when cases of acute antibody-mediated rejection (AMR), mixed rejections, chronic AMR, and reduced graft survival were reported with DSAs against these antigens. Although the deleterious effects of these DSAs are more pronounced in retransplants, harmful effects have also been observed in first-time recipients. DSAs against each of these antigens can trigger rejection alone. Their combination with DSAs against HLA-A, HLA-B and HLA-DR can cause more damage. It has been shown that strategies that reduce mismatches for these antigen lead to fewer rejections and better graft survival. There is a need for greater consensus on the universal typing of these antigens prior to transplantation for better patient and graft outcomes. This review focuses on the interaction of these antigens with lymphocytes and killer immunoglobulin receptors, arguments for not typing them, detailed analyses of the literature about their harmful effects, potential strategies moving forward, and recommendations for the future.

Key Words: Kidney transplantation; HLA-C; HLA-DP; HLA-DQ; Donor specific antibodies; Review

Core Tip: HLA-C, HLA-DP and HLA-DQ are capable of initiating alloimmune responses and donor-specific antibodies (DSAs), which can cause antibody-mediated rejection (AMR), mixed rejection, chronic AMR and reduced graft survival. Better matching for these antigens could reduce production of DSAs, leading to better graft and patient outcomes. This review explores the reasons for not typing these antigens in the past, examines evidence for their harmful effects, and proposes a way forward that will help clinicians to avoid or minimize the harmful effects of these DSAs.

INTRODUCTION

Kidney transplantation is one of the viable options for patients with end-stage renal disease. It provides a better quality of life and improves survival. HLA compatibility between donor and recipient is of utmost importance, and HLA plays an important role in the selection of patients to undergo transplantation[1]. The HLA genetic system consists of three regions containing class I, II and III regions located on chromosome 6[1]. The class I area has genes that encode HLA-A, HLA-B and HLA-C. The class II region has genes that for HLA-DR, HLA-DQ and HLA-DP. Class III region lies between classes I and II and encodes molecules that are not targets for allorecognition[1]. All nucleated cells express HLA class I molecules (A–C) which are recognized by CD8⁺ T lymphocytes. In contrast, HLA class II molecules (DR, DQ and DP) are only expressed on the surface of antigen-presenting cells such as B lymphocytes, dendritic cells, macrophages, monocytes, Langerhans cells, endothelial cells, and thymic epithelial cells, and are recognized by CD4⁺ T lymphocytes[2,3]. There are nine clinically relevant HLA alleles (three HLA class I and six HLA class II) that an individual inherits from each parent[4]. All three HLA class I genes (HLA-A, HLA-B and HLA-C) and six HLA class II genes (DRB1, DRB3/B4/B5, DQA1, DQB1,DPA1 and DPB1) are encoded on chromosome 6[4,5].

Historically, most allocation protocol use HLA-A, HLA-B and HLA-DR loci for matching. HLA-C, HLA-DQ and HLA-DP were not routinely typed universally. As a result, donor-specific antibodies (DSAs) against these loci, even if present, are often missed. Various studies have reported anti-HLA-C antibodies are present in 42%–56%[5,6], and that of HLA-DP-specific antibodies in 42% of patients with DSAs[7]. Similarly, in post-transplant follow-up, among 16% of recipients who developed DSAs against HLA class II, 52% were against HLA-DQ[8]. DSAs against HLA-C, HLA-DQ and HLA-DP have been reported to cause rejection[9-11]. This review focuses on HLA-C, HLA-DQ and HLA-P loci, DSAs against these molecules, their impact on graft outcomes, and the way forward.

HLA-C AND ITS POTENTIAL IMPLICATION FOR RENAL ALLOGRAFT OUTCOME

The HLA genetic system is located on chromosome 6, and the class I region contains genes that encode HLA-C molecules[1]. In 1970, extensive serological testing led to discovery of HLA-C[12]. These initial experimental serological tests revealed that, of the two, HL-A loci are not firmly attached to each other, and reside on different molecules. Incubation of antisera with lymphocytes led to a third segregant series initially named as AJ series and further probing led to identification of HLA-C[12]. HLA-C has many serologically defined specificities including Cw1, Cw2, Cw3, Cw4, Cw5, Cw6, Cw6, Cw7, Cw8, Cw9 (w3) and Cw10 (w3)[13]. HLA-C can affect graft outcome by three mechanisms.

Interaction with CD8⁺ T lymphocytes

Like other HLA class I molecules HLA-C is expressed by all nucleated cells and is recognized by CD8⁺ T lymphocytes[14]. Antigen-presenting cells, such as dendritic cells, present antigens in combination with class I molecules, leading to the activation of CD8⁺ cytotoxic cells. Upon antigenic stimulation, they proliferate into effector cells, with some differentiating into memory cells for future encounters. CD⁺ T lymphocytes play a crucial role in combating viral infections, killing cancer cells, and may also be involved in alloreactive responses during allograft rejection. Cell markers such as CD27, CD28, CD57, or T-box expressed in T cells, can identify specific memory subsets[15]. CD8⁺ cells kill their targets by releasing preformed cytotoxic proteins. These cytotoxic proteins include granzyme, which induces apoptosis and perforin, which kills by creating holes in its target cells[16]. The presence of CD8⁺ cells has been reported in renal and other solid organ transplant rejections[17-20]. In a study on kidney transplant recipients, the phenotype and molecular signatures of CD8⁺ T cell subsets in biopsy-proven T-cell-mediated rejection was investigated. Through use of flow cytometry and microarray it was found that C-C chemokine receptor (CCR)7⁺CD8⁺ T cells significantly decreased, but CD28^nullCD57⁺CD8⁺ T cells and CCR7CD45RA⁺CD8⁺ T cells showed an increase in patients with T-cell-mediated rejection[17]. In a novel study on kidney allograft biopsies, gene transcript levels were determined for granzyme B, perforin, and T-cell intracellular antigen 1. It was found that granzyme and perforin transcript levels were higher in patients with acute T cell rejection[19]. The roles of granzyme and perforin in renal allograft rejection have also been reported in several other studies[16,17,21]. Using monoclonal antibodies for human granzyme A or B in immunohistochemical staining techniques was applied to renal biopsies of acutely rejected kidneys and it detected high numbers of granzyme A and B positive lymphocytes infiltrating tubular epithelium, and vascular and glomerular structures[21]. Allorecognition of HLA-C mismatches by CD8⁺ cell in hemopoietic stem cell transplantation has been recognized. Investigating this interaction in kidney transplantation would be beneficial.

Interaction with killer immunoglobulin like receptor

Human killer cell immunoglobulin-like receptors (CD158) are a family of transmembrane glycoproteins. Killer immunoglobulin receptor (KIR) is expressed on natural killer (NK) cells and a subset of T cells[22]. KIR consists of polymorphic activating and inhibitory receptors that regulate function of human NK cells. HLA-C like other MHC class I molecules binds KIR. The inhibitory receptors bind to class I molecules, then KIR inhibits NK cells and thus prevents damage to normal cells[23]. The downregulation of class I molecules leads to activating receptors that activate NK cells that then play a potential role against tumors and viral infections.

Recently, interest has grown in the role of NK cells and their interactions with dendritic cells and B and T lymphocytes in adaptive immunity, leading to progressive renal allograft injury[24]. HLA-C is a ligand for KIR, which regulates NK cells. The HLA-C2 molecule is more inhibitory to NK cell function than HLA-C1 is. Hanvesakul et al[24] genotyped 760 kidney transplant recipients to examine the impact of HLA-C1 and HLA-C2 interaction with KIR[24]. In vitro cocultures of dendritic cells and NK cells were also conducted. Interleukin (IL)-15 activated donor-derived NK cells, which caused differential HLA-C genotype-dependent dendritic cell maturation. Contrary to HLA-C2, dendritic cells with HLA-C1 were associated with proinflammatory cytokine synthesis [tumor necrosis factor (TNF)-α, IL-12p40/p70], more potent maturation, and upregulation of CCR7 expression. The study concluded that HLA-C1 and KIR interaction is a major determinant of graft failure[24]. Coculture of NK cells and dendritic cells leads to the production of multiple cytokines, including interferon- (IFN)-γ, TNF-α, IL-12, IL-15 and IL-18. These responses are modulated by the interaction between HLA-C and KIR, leading to the activation of T cells via T helper (Th)1 responses[25]. Ischemia–reperfusion injury in the immediate post-renal transplant period results in the release of proinflammatory cytokines, including IL-15[26,27], which leads to monocytes and dendritic cells infiltration. Donor-derived NK cells interact with recipient dendritic cells, upregulating Toll-like receptor 4 expression on dendritic cells, which enhances antigen presentation and alloimmune responses[26]. In another study, a higher risk of chronic rejection after kidney transplantation was observed in patients transplanted from donors homozygous for HLA-C1. These patients had a significantly higher risk of chronic rejection than patients transplanted from donors homozygous or heterozygous for HLA-C2 or with epitopes belonging to the HLA-Bw4 ligand group[28]. The authors observed higher chronic rejection in recipients and donors who completely lacked the two functional ligand combinations of rKIR2DL1/dHLA-C2 and rKIR3DL1/dHLA-Bw4[28].

DSAs against HLA-C

HLA-C is not routinely typed, and the frequency of anti-HLA-C antibodies is lower than that of anti-HLA-A or HLA-B. The frequency of anti-HLA-C DSAs varies between 42% and 65% in sensitized patients[4,6]. Duquesnoy and Marrari[29] analyzed 45 HLA-C mismatched kidney rejection cases that underwent graft nephrectomy and found that DSAs against HLA-C were present in 60% of these patients. Work by Duquesnoy and Marrari[29] suggests that HLA-C mismatches can lead to humoral sensitization and that HLA-C eplets can induce DSAs. The authors showed the importance of HLA matchmaker in understanding DSA reactivity patterns and selecting appropriate patients and selecting appropriate candidates for transplant[29].

HLA-C DSA have been reported to cause antibody-mediated rejection (AMR) in various case reports[30-33]. DSAs against HLA-C can cause positive flow crossmatches and severe AMR, leading to graft loss[31]. Bachelet et al[31] reported a case of AMR resulting in graft nephrectomy, identifying DSAs against HLA-Cw6 and Cw17. Analysis of renal tissue showed focal necrosis of the cortical parenchyma, vascular lesions, and positive peritubular C4d staining. A fragment of the removed graft was analyzed with a single bead antigen assay, which revealed reactivity against donor antigens Cw6, Cw17 and the cross-reacting Cw antigens. This reflects that DSAs against HLA-C can activate complement and cause severe AMR[31]. The association between DSAs against HLA-C and AMR have been reported in several retrospective studies[9,34]. In a case–control study of 22 patients with DSAs against HLA-Cw compared with 88 sensitized control patients, AMR was observed in 27.3% of patients with DSAs against HLA-C, compared to 9.1% in the immunized control group. Additionally, the patients with DSAs against HLA-C had a higher mean fluorescence intensity of 4966 in the rejection group as compared to 981 without rejection[9]. Another important study found that DSAs against HLA-C carry a similar risk for AMR compared to DSAs against other class I HLA-A and HLA-B[34]. These findings support the growing consensus that DSAs against HLA-C can trigger AMR.

HLA-DP AND POTENTIAL IMPLICATION FOR RENAL ALLOGRAFT OUTCOME

HLA-DP was initially referred to as secondary B-cell antigen and was discovered through a secondary stimulation assay known as the primed lymphocyte test[35,36]. The HLA-DP region consists of two loci, HLA-DPA1 and HLA-DPB1, which encode the HLA-DP protein. HLA-DP1 region contains 491 alleles and encodes 233 proteins, while the HLA-DP has 2221 and encodes for 1325 proteins[13].

Interaction with CD4⁺ T lymphocytes

Like other MHC class II molecules, HLA-DP presents processed antigens by antigen-presenting cells to CD4⁺ T lymphocytes. This step is critical for initiating immune responses against specific antigens[37]. CD4⁺ T lymphocytes have various subsets including Th1, Th2, Th17, Th22, regulatory T (Treg) cells, and T follicular helper cells[38]. One of the major functions of CD4⁺ T cells is to help B cells proliferate and differentiate into antibody-producing plasma cells. This occurs through the interaction between antigen-binding B cells with Th cells, leading to the expression of the B-cell stimulatory molecule CD40 ligand (CD40L) on the Th cell surface. This interaction causes Th cells to secrete B-cell stimulatory cytokines such as IL-4, IL-5 and IL-6. These cytokines lead to the proliferation and differentiation of B cells into antibody-secreting plasma cells[39]. Th1 cells also activate macrophages through two signals. The first signal involves the secretion of IFN-γ, which binds to IFN-γ receptors on the macrophage surface. The second signal involves the display of the costimulatory protein CD40L, which binds to CD40 on the macrophage, activating it and leading to release of lysosomal enzymes and free radicals that kill intracellular organisms or any other foreign particles[40]. Moreover, CD4⁺ T cells also play a crucial role in the maturation of CD8⁺ T cells[41].

CD4⁺ T cells play an important role in rejection, either directly or indirectly, by activating CD8⁺ T cells or macrophages. CD4⁺ T cells have been implicated in AMR[42] and mixed cellular rejection[43] in experimental studies. The role of helper functions of memory CD4⁺ T cells was evaluated in a mouse model of renal transplantation. Recipients with donor-reactive memory CD4⁺ T cells rapidly lost graft function. High serum titers of DSAs were observed in sensitized mice. Biopsy of the implanted kidney in mice showed minimal T-cell infiltration, intense CD4⁺ deposition in the grafts of sensitized recipients fulfilling criteria of AMR[42]. Neutralization of IFN- and depletion of both CD4⁺ and CD8⁺ cells failed to prevent rejection. However, depletion of B cells successfully reduced DSA levels and improved graft survival, signifying that memory CD4⁺ T cells exert their deleterious effects via DSAs rather than through T cells[42]. The role of CD4⁺ T cells was further evaluated in another well-planned mouse model. A mouse model of mixed AMR though revealed CD8⁺ T cell in majority as compared to CD4⁺ T cells. Depletion of CD4⁺ T cells prevented graft loss contrary to the CD8 cell depletion where no beneficial effect was observed. Further evaluation by enzyme-linked immune absorbent spot identified that CD4⁺ T effectors responded to donor alloantigen by both the direct and indirect pathways of allorecognition. Laser capture microdissection and immunostaining studies confirmed that CD4⁺ T cells infiltrating the graft synthesized molecules capable of harming the graft. Bioluminescent imaging revealed that CD4⁺ T cells in areas of graft having immune reactivity[43].

Similarly, the roles of CD8⁺ T cells[44] and macrophages[45-47] in renal allograft rejection have been well described. However, the potential role of HLA-DP in activation of CD4⁺ T cells and subsequent renal allograft rejection is less well studied and could represent a valuable area of future research.

DSAs against HLA-DP

HLA-DP protein can stimulate CD4⁺ T cells, which through helper signals, can in turn stimulate B cells to differentiate into antibody-producing plasma cells. Initial studies suggested that DSA against HLA-DP may not be harmful[48,49]. However, primed lymphocytes due to infection, ischemia–reperfusion injury, or retransplantation can produce antibodies that may trigger AMR. Over time, multiple reports have suggested an association between anti-DP DSAs with allograft rejection and graft loss, both in primary kidney transplant recipients and in retransplantation[10,50]. In a retrospective analysis, 48 patients with preformed DSAs against HLA-DP and HLA-Cw were compared to those with no DSAs and those with preformed DSAs against-A, B, DR and DQ. The HLA Cw/DP DSA group exhibited a more frequent positive flow crossmatch than the no-DSA group. Flow crossmatch positivity was similar to the A/B/DR/DQ DSA group. Biopsy-proven acute rejection-free survival was worse in the Cw/DP and A/B/DR/DQ DSA groups than in the no-DSA group[10]. De novo DSA against HLA-DP can also arise in first-time transplants and lead to late-onset AMR, even in the absence of other DSAs. This suggests that these antibodies are capable of activating complement, resulting in AMR upon first exposure[50]. Therefore, the previous assumption that DSA against HLA-DP is benign has shifted over time.

HLA-DQ AND POTENTIAL IMPLICATIONS FOR RENAL ALLOGRAFT OUTCOME

Earlier serological and cellular methods for HLA typing failed to differentiate between HLA-DR and HLA-DQ regions. The introduction of molecular methods for HLA typing led to the discovery of the HLA-DQ antigen in the MHC class II region[51]. HLA-DQ is composed of two subunits that combine to form an alpha and beta heterodimer. The DQ subunit is encoded by the HLA-DQA1 gene while the DQ β subunit is encoded by HLA-DQB1[52]. HLA-DQA1 has 508 alleles and encodes 244 proteins, while HLADQ-B has 2330 alleles and encodes 1455 proteins[53].

Interaction with CD4⁺ T lymphocytes

Like all other class II region loci (HLA-DP and HLA-DR), HLA-DQ locus molecules present processed antigens via to CD4⁺ T cells. As described earlier, the activation of various CD4⁺ T cell subtypes leads to transformation of B lymphocytes into antibody-producing plasma cells[39], activation of macrophages[40], and the proliferation of CD8⁺ T cells[41]. CD4⁺ T cells, CD8⁺ T cells, along with macrophages, play important roles in mediating renal allograft rejection[42-45,47] as explained earlier. As a result, the HLA-DQ locus, like other HLA class II loci, may play a significant role in rejection.

DSAs against HLA-DQ

HLA-DQ protein can also stimulate CD4⁺ T cells, which, through helper signals, can stimulate B cells to differentiate into antibody-producing plasma cells. It was initially thought that HLA-DQ mismatches between donor and recipient, in the presence of HLA-DR compatibility, did not affect graft outcome[54]. However, early studies on the response of cultured human kidney capillary endothelium to immunological stimuli suggested that expression of HLA-DQ may be upregulated and could be a potential target for HLA-DQ antibodies[55]. Definitive evidence of the pathological role of DSAs against the HLA-DQ locus was established by two studies in 2012, which showed that de novo DQ DSAs are associated with a significant risk of AMR, transplant glomerulopathy, and poor graft outcomes[56,57]. Willicombe et al[56] were the first to provide important information about DSA against HLA-DP. They found a significant discrepancy between matching at the HLA-DR and HLA-DQ loci, enhanced immunogenicity with both HLA-DQ and HLA-DP mismatches, and that anti-DQ DSAs are the most frequent de novo antibodies associated with AMR, transplant glomerulopathy, and renal allograft loss[56]. Subsequent studies reinforced these findings, linking chronic-active AMR[11] and inferior graft function in patient with DSA against HLA-DQ[58].

ARGUMENT FOR NOT TYPING HLA-C, HLA-DQ AND HLA-P

Underexpression of HLA-C

Historically, the serological polymorphism for HLA-C was poorly defined compared to HLA-A and HLA-B. HLA-C is expressed on cell surfaces at a lower level (10%) compared to HLA-A or HLA-B[1-3]. Although HLA-C is synthesized intracellularly in similar amounts to of HLA-A and HLA-B, it is expressed less on the cell surface. Investigations have shown a variety of reasons for this. One explanation for the low expression is the increased turnover of heavy chain mRNA[59]. Another reason for low expression is the absence of nuclear factor B binding sites, which leads to weaker induction of HLA-C transcription by inflammatory cytokines such as IFN- and TNF-. Inefficient association of the heavy chain with β₂-microglobulin also contributes to low expression of HLA-C on the cell surface. Additionally, a significant portion of HLA-C molecules is retained in the endoplasmic reticulum, where they are unable to be transported due to stable interaction with the transporter associated with antigen processing[60]. As a result of this lower expression, it was initially assumed that the immunogenicity of HLA-C would be reduced, leading to a lower prevalence and strength of DSAs against HLA-C, which could not provoke rejection. However, the notion that HLA-C is less expressed should be taken with caution as DSA against HLA-C can bind to donor cells and can activate complement and can lead to graft dysfunction.

Underexpression of HLA-DP

Like HLA-C, routine typing for HAL-DP is not universally performed, as it was also thought to be less expressed. HLA class II antigens are differentially expressed in developing lymphoid cells, with HLA-DR and HLA-DP are expressed earlier than HLA-DQ, which is expressed later[61]. HLA-DR is expressed more abundantly than HLA-DP and HLA-DQ[61]. The expression of class II antigens in the kidneys has also been studied. It was shown that HLA-DR is constitutively expressed on the endothelium of glomerular and peritubular capillaries, while HLA-DP and HLA-DQ are not. However, mRNA transcripts for HLA-DP were detected in renal vasculature[62], prompting further investigation. Subsequent research demonstrated that all three class II molecules can be expressed on glomerular endothelial cells freshly isolated from normal human kidney biopsies[63]. An experimental study showed that IFN- exposure stimulates the expression of HLA-DR more than HLA-DP[64]. However, despite the stimulation of HLA-DR and HLA-DP expression, IFN- exposure failed to support T cell responses indicating that additional factors are needed for full activation[64]. The expression of HLA-DP and HLA-DQ in response to inflammatory cytokines is less intense and delayed as compared to HLA-DR. IFN- and TNF led to more upregulation of mRNA with subsequent HLA-DR expression in vascular endothelial cells, as compared to HLA-DP and HLA-DQ expression where lower responses were observed[65]. The upregulation and expression of HLA-DR in glomerular endothelial cells is rapid and occurs in 18–24 h as compared to 4–7 d in case of HLA-DP and HLA-DQ[66]. This lower expression of HLA-DP may provide false reassurance to transplant clinicians. In reality, DSAs against HLA-DP have a significant deleterious effect on renal allografts.

Underexpression of HLA-DQ and its masking by linkage disequilibrium

HLA-DQ was not initially typed due to its low expression. However, over time, evidence has evolved and more facts have resurfaced about HLA-DQ. Among class II antigens, HLA-DQ is the last to be expressed in developing lymphoid cells[61], and like HLA-DP, it is the least expressed antigen when compared to HLA-DR[61]. The expression of HLA-DQ in response to inflammatory cytokines and inflammatory conditions is less intense and delayed compared to HLA-DR[65,66]. Another challenge is that HLA-DQ alleles can be expressed at different levels[67], and there is significant variability in the cell surface expression of different HLA-DQ molecules[68]. These facts make HLA-DQ expression more complex to understand. It is important to note that like other class II antigens HLA-DQ can express itself in glomerular endothelium[62]. Gene expression and microarray evaluation of preimplantation biopsies have shown that high HLA-DQB1 RNA expression is associated with poorer renal allograft outcomes[67]. Sensitizing events such as retransplantation, infection and ischemia–reperfusion may lead to brisk and intense expression, which can trigger the production of DSAs that may provoke rejection.

Another important factor is the masking of HLA-DQ expression by linkage disequilibrium. Linkage disequilibrium refers to the nonrandom association of alleles at different loci[68]. It was previously assumed that if HLA-DR compatibility was maintained serologically, HLA-DQ would not affect kidney transplant outcomes due to the strong linkage disequilibrium between these loci[69-72]. However, with the advent of molecular typing, more was discovered about the HLA-DQ locus and DSAs against these loci. Molecular techniques have shown that HLA-DQ is discordant in 15%–26% of HLA-DR-matched patients[73,74]. Willicombe et al[56] analyzed the effect of linkage disequilibrium between HLA-DR and HLA-DQ in kidney transplant recipients. Out of 505 patients analyzed, 108 patients (21.4%) were matched at both HLA-DR and HLA-DQ. On the other hand, 284 patients (56.2%) were mismatched at both HLA-DR and HLA-DQ loci. Thirty-eight (7.5%) of the patients were matched at HLA-DR but mismatched at HLA-DQ, and 75 (14.9%) were matched at HLA-DQ but mismatched at HLA-DR. These findings indicate that compatibility expected with strong linkage disequilibrium may not always be present.

DO DSAs AGAINST HLA-C, HLA-DP and HLA-DQ A HAVE DELETRIOUS EFFECTS ON RENAL ALLOGRAFT? WHERE DOES THE EVIDENCE STAND?

Since identification of HLA-C, HLA-DP and HLA-DQ loci researchers have done work on the role of these loci in initiating alloimmune responses and subsequent graft injuries. Multiple case reports and observational studies and translational research has been done. We summarize all these publications in chronological order. Most of the work on HLAs and DSAs against these antigens has been done on different themes and it is difficult to compare studies.

HLA-C and DSA against HLA-C

Initially, DSAs against HLA-C were considered “innocent bystanders”. Albrechtsen et al[75], in their retrospective analysis, found that routine typing for HLA-C had no beneficial effect on renal allograft outcomes. A similar trend was observed in other study that also found no benefit from HLA-C matching[30]. Duquesnoy and Marrari[29] conducted a comprehensive study on antibodies against HLA-C epitopes in patients with rejected kidney transplants who had HLA-C mismatches[29]. Approximately 60% of these patients had DSAs against HLA-C. The triggering of HLA-C antibody production was associated with eplet loads of the HLA-C mismatches. However, the frequency and reactivity of HLA-C antibodies were lower than those induced by donor HLA-A and HLA-B mismatches. An analysis of sensitized patents found a lower frequency of DSAs against HLA-Cw (42%), significantly less than sensitization to HLA-A (80%) and HLA-B (83%) antibodies[9]. Ling et al[5] found similar results, identifying that the prevalence and strength of anti-HLA-Cw and HLA-DP were low compared to other antibodies and had no deleterious effect on graft outcomes.

Over time, substantial evidence emerged contradicting the initial notion that DSAs against HLA-C are benign. Early observations came from case reports. Since the first case report of hyperacute rejection in 1986[30], many case reports have described AMR due to DSAs against HLA-[32,33,76,77]. A review of these case reports showed that sensitizing events like retransplantation[32,76] and pregnancy[77] were present in a few cases. However, no prior sensitizing events were identified in other cases[30,32,33,78]. Contrary to previous findings that HLA-C typing does not affect renal allograft outcomes, Frohn et al[79] using allele-specific PCR, found that HLA-C mismatch was significantly associated with acute transplant rejection, particularly when combined with an additional mismatch on the B locus. In the first retrospective analysis by Gilbert et al[80], a high incidence of AMR and graft loss due to immunological reasons was reported in sensitized patients with additional DSAs against HLA-C and HLA-DP. The combination of DSAs against Cw/DP was found to be as deleterious as DSAs against HLA-A, -B, -DR and -DQ, as demonstrated in another comparative retrospective analysis[10]. Positive flow crossmatch and 2 years biopsy proven rejection were significantly higher in patients with DSAs against Cw/DP and against HLA-A, -B, -DR and -DQ as compared to no DSAs. Cw/DP DSA reduced graft survival similarly to HLA-A, -B, -DR and -DQ DSAs[10]. These studies suggest that the combination of anti-Cw/DP DSAs may synergize to provoke rejection. The impact of anti-HLA-Cw and HLA-DP was studied individually in a recent analysis[81]. Two years post-transplant, AMR occurred in 12% of patients with anti-HLA-Cw DSAs, compared to 28% in patients with anti-HLA-DP DSAs[81]. The difference in rejection was significant below a mean florescence intensity (MFI) of 3000, indicating that the pathogenicity of Cw DSAs is MFI-dependent, manifesting only when MFI exceeds 3000. Several retrospective studies and numerous case reports have established the independent ability of anti-HLA-C DSAs to cause rejection[32,34,76-78]. In a well-planned retrospective case–control study, Aubert et al[9] found that the incidence of AMR in recipients with anti-HLA-C DSAs was 27.3%, significantly higher than the patients without DSAs. Patients with AMR had a higher median MFI of 4966 on day 0 compared to the non-AMR group that had a median MFI of 981[9]. Another retrospective cohort study reported similar risks of rejection and graft survival in patients with anti-HLA-C DSA as in those with anti-HLA-A and HLA-B DSAs[34].

While single-bead assays have played an important role in identifying DSAs against HLA-C, HLA-DP and HLA-DQ, it is essential to note that they may also detect denatured DSAs, which are clinically irrelevant. This important fact should be kept in mind. Visentin et al[82] investigated this issue, comparing anti-denatured DSAs with anti-native DSA against HLA-Cw to study their effects on crossmatches and rejection. They used two types of single-antigen flow beads: (1) Classical beads; and (2) Beads with diminished expression of denatured HLA. Among 135 crossmatches performed, the correlation between classical beads fluorescence and crossmatch ratio was low (P = 0.178), while the correlation was higher with beads with diminished expression of denatured HLA (P = 0.289). Of the 52 patients with DSAs against HLA-Cw, those with anti-native HLA-Cw antibodies had higher rates of AMR and chronic AMR (P = 0.006 and P = 0.03, respectively), along with lower graft survival (P = 0.04). Unfortunately, there is a lack of prospective observational and randomized controlled trials (RCTs) on this topic. More prospective studies, preferably RCTs, are needed to elaborate the significance of DSAs against HLA-C. Table 1 summarizes the evidence for and against anti HLA-C DSAs in chronological order[5,6,9,10,29-34,75-82].

Table 1 Summaries of reported literature on HLA-C typing and anti-HLA-C donor-specific antibody.

Ref.	Journal/year	Study type and sample size	Objectives	Findings
Albrechtsen et al[75]	Transplantation Proceedings/1977	Retrospective: 142 living related 311 cadaveric transplants	To study influence of HLA on the outcome of kidney transplantation	No beneficial of HLA-C matching was found
Bryan et al[6]	Clinical Transplantation/2010	Retrospective: 60 sensitized patients	To describe frequency of HLA class I DSA	There was 42% positivity to HLA-Cw, which was significantly lesser than sensitization to HLA-A (80%) and HLA-B (83%)
Duquesnoy and Marrari[29]	Transplant Immunology/2011	Retrospective: Sera from 45 HLA-C mismatch after allograft nephrectomy	To detect antibodies against HLA-C epitopes in patients with rejected kidney transplants	HLA-C antibody frequencies and reactivities were lower than those induced by donor HLA-A and HLA-B mismatches
Ling et al[5]	Human Immunology/2012	Prospective observational: 1069 patients on waiting list	To determine prevalence and the strength of anti-HLA-Cw and -DP and effect on clinical outcome	Low prevalence and the strength of anti-HLA-Cw and HLA-DP as compared to others and patient and graft survival was 100% without rejection
Frohn et al[79]	Nephrology, Dialysis, Transplantation/2001	Retrospective analysis of 104 pairs analysis for HLA-C matching	To analyze if acute graft rejection is influenced by HLA-C matching	After exclusion of linkage disequilibrium, HLA-C mismatch was significantly associated with rejection (P = 0.004)
Chapman et al[30]	Transplantation/1986	Case report	—	Case report of hyperacute rejection in renal allograft recipient having HLA-Cw5 antibody
Baan et al[32]	Transplantation/1993	Case report of rejection in an unsensitized recipient of deceased kidney	—	Rejection was reported in HLA-C mismatched along with sublocus HLA-Bw22
Bachelet et al[31]	American Journal of Transplantation /2011	Case report of AMR in patient with immunoglobulin A nephropathy undergoing 3rd transplant	—	Anti-HLA-Cw DSA led to positive flow cytometry crossmatch and irreversible acute AMR
Gilbert et al[80]	Transplantation Proceedings/2011	Retrospective analysis of immunized recipients, (n = 176) had antibodies against only classical HLA antigens (A, B, DR and DQ) and 13.3% (n = 27) antibodies against HLA-C and/or HLA-DP	To determine whether the presence of specific HLA-C and HLA-DP antibodies before transplantation influenced graft outcomes in immunized recipients	DSA against HLA-C and HLA-DP along with DSAs against HLA-A, -AB, -DR and -DQ led to significant increase in the number of acute rejection episodes and graft loss due to immunological reasons
Suneja and Kuppachi[33]	Clinical Kidney Journal/2012	Case report of 21-yr-old transplant recipient who developed AMR after 21 mo of transplantation	—	DSA against HLA-Cw17 led to AMR
Aubert et al[9]	American Journal of Transplantation /2014	Retrospective case–control study of 608 renal transplant patients	To evaluate the clinical relevance of the presence of anti-HLA-C DSA at d 0	Incidence of AMR was significantly higher (27.3%) in patients with anti-HLA-C DSA and median MFI was 4966, which was significantly higher in AMR group
Bosch et al[76]	Human Immunology/2014	Case report of AMR in the second transplant	—	Patient had low MFI < 1000 in 1st transplant and has HLA-C mismatched second kidney and developed AMR on d 7
Bachelet et al[10]	Transplantation/2016	Comparative retrospective analysis of 199 who were divided in three groups for comparison	To analyze clinical impact of HLA-Cw/DP DSA by comparing with HLA-sensitized kidney transplant recipients with no DSA at d 0 and recipient with recipients with preformed HLA-A, HLA-B, HLA-DR and HLA-DQ DSAs	Positive flow cross match, 2-yr biopsy proven rejection were more in HLA-Cw/DP DSA and preformed HLA-A, -B, -DR and -DQ DSAs as compared to group with no DSAs and similarly less graft survival in HLA-Cw/DP DSAs and preformed HLA-A, -B, -DR and -DQ DSAs as compared to no DSAs
Santos et al[34]	World Journal of Transplantation/2016	Retrospective cohort study: 12 patients with anti HLA-Cw DSA and 23 with anti-HLA-A or HLA-B	To analyze the clinical impact of preformed anti-HLA-Cw vs anti-HLA-A and/or HLA-B DSAs in kidney transplantation	Similar risk of AMR (P = 1) and impact on graft function (P = 0.528) as compared to anti-HLA-A and/or HLA-B DSA
Persaud et al[77]	Human Immunology/2017	Case report of AMR in a living related transplant	—	Sensitization to HLA-Bw6 via exposure to paternal HLA-C14 during pregnancy likely predisposed this patient to AMR
Abuzeineh et al[78]	Clinical Nephrology Case Studies/2020	Case report of AMR in 39 yr old unsensitized patients	—	De novo DSAs against HLA-C led to AMR. Unexplained Fabry-like zebra bodies were also seen in biopsy
Visentin et al[82]	American Journal of Transplantation /2020	Retrospective observational study of 135 patients	To compare the pathogenicity of preformed anti-denatured and anti-native HLA-Cw antibodies in kidney transplant recipient	Anti-native HLA DSA had more acute and chronic AMR (P = 0.006 and P = 0.03, respectively), and had lower graft survival (P = 0.04)
Laboux et al[81]	Transplant International/2023	Retrospective multicentral observational study of 183 patients	To determinate risk factors of AMR in recipients transplanted with preformed isolated Cw-DSA or DP-DSA	The 12% in the HLA-Cw-DSA group, vs 28% in the HLA-DP DSA group had AMR. The increased risk associated with HLA-DP DSA compared with HLA-Cw DSA, was significant only for MFI < 3000

AMR: Antibody-mediated rejection; DSA: Donor specific antibody; HLA: Human leukocyte antigen; MFI: Mean florescence intensity.

HLA-DP antigen and DSA against HLA-DP

Like DSA against HLA-C anti-HLA-DP DSA was initially thought to be benign in nature. A retrospective analysis in 1997 reported that isolated HLA-DP mismatches did not affect graft outcome[49]. Similarly, no impact on graft function was reported in recipients having anti-HLA-DP DSAs[48,83]. However, over time, numerous case reports were published, alerting transplant clinicians, as patients developed rejection[50,84-87]. A critical analysis of these case reports revealed that anti-HLA-DP DSAs can cause positive B cell crossmatch[85,88]. These antibodies are able to trigger acute rejection[89-92]. The type of rejections varies. These DSAs can cause AMR[84,91,93,94], mixed rejection with elements of both acute cellular rejection and AMR[89,91,92] and chronic AMR[90]. Most acute rejections occurred early, within 2 d to 2 mo[84-86,89,94], although some occurred later, up to 15 months[50,92]. Sensitizing events in most cases included retransplantation[84,85,89,91,94], blood transfusion and pregnancy[89]. However, some rejections occurred even in first transplant cases[50,85,92]. Anti-HLA-DP DSAs can result in positive B cell crossmatch[85,88,94]. The presence of anti-HLA-DP antibodies with a positive B cell crossmatch should be considered a high immunological risk, putting these patients at an extreme risk of developing AMR. In addition to case reports, multiple studies were conducted to analyze the effects of HLA-DP typing and anti-HLA-DP DSAs, most of which are retrospective in nature. There is a strong need for prospective studies.

Several studies have investigated the impact of HLA-DP matching, finding that better matching improves graft outcomes. For instance, a report of the collaborative transplant study retrospectively DNA-typed was done and was then repeated in 1300 deceased kidney retransplant to assess the effect of HLA-DP mismatches on kidney graft outcome. No beneficial effect was observed for first-year transplant survival in first transplants. However, in second transplants, the 1-year survival rate of grafts with no HLA-DPB mismatch was 83% ± 2%, which was significantly higher than those with one mismatch (76% ± 2%, P = 0.02). The effect was even more pronounced with two mismatches, where graft survival was further reduced (73% ± 3%, P = 0.003)[95]. This suggests that priming of lymphocytes in the first transplant, followed by re-stimulation during the retransplant, leads to immunological sequelae, resulting in reduced graft survival. Laux et al[96], in their novel study, analyzed the impact of HLA-DPB epitope-based matching on graft outcome. They found that fewer than two epitope mismatches resulted in better survival than three epitope mismatches (at 2 years: 77.8% vs 65.8%, P = 0.0112).

Other studies have examined various aspects of anti-HLA-DP antibodies, including its deleterious effects. Comparing these studies can be challenging due to differences in designs and objectives. Qiu et al[97] used a single DP bead antigen and compared anti-DP antibodies in rejected kidneys versus functioning allografts. They found a significantly higher percentage (19.5%) of anti-DP antibodies in rejected kidneys than in functioning allografts (5.1%). Additionally, 42.9% of patients with functioning allografts and 63.9% of those with rejected grafts who had anti-DP antibodies also had DR/DQ antibodies. Around 13% of rejected kidney patients who did not have class I or DQ/DR antibodies had anti-DP antibodies, compared to 3.5% in functioning allografts (P = 0.005)[97]. This suggests that HLA-DP antibodies may develop independently of other HLA antibodies. In another novel study, 14% of a cohort of 338 recipient had DSAs against HLA-DP[98]. Among these, 23% had DSAs before transplantation, while 77% developed them afterward. Among these 77% patients, 30 had a first transplant and seven a second transplant. The study also mapped epitopes and compared the amino acid sequencing of the hypervariable region of HLA-DP. A single hypervariable mismatch was present in 80% of DSAs against HLA-DP[98]. Thus, matching for HLA-DP epitopes may be more practical than allele matching. HLA-DP antibodies can also exist independently of HLA-DR antibodies[7]. For instance, among 271 (42%) of patients with HLA-DP antibodies, 58 (21%) were negative for reactivity with crossreactive HLA-DR, and 16 (5.9%) had no class II antibodies other than anti-HLA-DP. This indicates that anti-HLA-DP can develop on its own. Although significant number of this cohort (62%) who had DSAs against HLA-DP had a previous transplant, others (38%) were nontransplanted patients. Similarly, Hormann et al[99] found that 81/195 (49%) of their consecutive transplant cohort had anti-HLA-DP antibodies. Around 64% of these antibodies were present in pretransplant sera and 36% in post-transplant sera. The presence of these antibodies led to significantly higher number of rejections as compared to those without these antibodies. These studies show that typing for HLA-DP and screening for DSAs against this antigen should be strongly considered even in the first transplant. Seitz et al[100] recently did a well-planned retrospective comparative study. Patients with isolated HLA-DP DSAs were compared with three control groups. The first group included 46 patients with standard immunological risk (calculated reaction frequency < 85%, no current or historical DSAs, and no repeat mismatched antigens with previous transplants). The second group included highly sensitized 27 patients (calculated reaction frequency > 85%). The third group had 18 patients with HLA-DP antibodies that were not donor-specific. Pre-existing HLA-DP DSAs were a significant predictor of AMR (hazard ratio = 9.57, P = 0.012). Moreover, these patients had significantly increased microvascular scores and worse transplant glomerulopathy on biopsy when compared with the standard immunological risk group. Although anti-HLA-DP DSAs are deleterious on their own, it is important to mention that combination of Cw/DP DSAs may synergize each other effects as discussed earlier[10]. Lastly but most importantly, a meta-analysis was recently published on HLA-DP DSAs[101]. The meta-analysis included five studies with 1166 patients. The objective was to see outcome in term of graft loss and acute rejection. This meta-analysis found interesting trends. De novo anti-HLA-DP DSAs showed an increased risk of graft loss or acute rejection (odds ratio = 3.6, 95%CI: 1.6–8.10, P = 0.002, I² = 52%). Table 2[7,48-50,83-101] summarizes the evidence in favor and against anti HLA-DP DSAs in chronological order.

Table 2 Summaries of reported literature on HLA-DP typing and anti-HLA donor-specific antibody.

Ref.	Journal/year	Study type and sample type	Objectives	Findings
Rosenberg et al[49]	Human Immunology/1992	Retrospective study of 37 patients	To study the influence isolated HLA-DP mismatches between donors and recipients	No benefit was found
Pfeiffer et al[48]	Transplant International/1995	Retrospective analysis of sera from 505 patients	To study the frequency and impact on graft function of HLA-DP antibodies	HLA-DP antibodies were found in 7.3%, and those with prior antibodies who had retransplantation had no impact on graft function
Redondo-Pachón et al[83]	Transplant Immunology/2016	Retrospective analysis of 440 kidney transplant patients	To study the effect of antibodies against HLA-DP detected with solid-phase assays on graft survival after kidney transplantation	No effect of survival was found
Mytilineos et al[95]	Transplantation/1997	Retrospective study in which 3600 retrospective DNA typing was performed first and then repeated in 1300 deceased kidney transplant	To assess influence of HLA-DPB mismatches on kidney graft outcome	No effect on 1st transplant. In 2nd transplant, the 1-yr survival rate of transplants with no HLA-DPB mismatch was 83% ± 2%, which was significantly higher than grafts with 1 mismatch (76% ± 2%, P = 0.02) and that of 2 mismatches (73% ± 3%, P = 0.003)
Laux et al[96]	Transplantation/2003	Retrospective analysis of 1478 patients who received a cadaver kidney retransplant	To study the effect of HLA-DPB1 epitopes on graft outcome	< 2 epitope mismatches have better survival than three epitope mismatches (at 2 yr: 77.8% vs 65.8%, P = 0.0112)
Qiu et al[97]	Transplantation/2005	Multicenter brief report of 232 sera from 4 centers	To describe the frequency of HLA-DP antibodies found in 323 patients who had functioning and rejected renal allografts	The 5.1% of 138 patients with functioning grafts, and 19.5% of 185 patients with rejected grafts (P < 0.001) had anti HLA-DP antibodies
Samaniego et al[87]	Clinical Transplants/2006	Case report of AMR due to HLA-DP DSA	—	Anti-HLA-DP antibody led to HLA-C4d-positive AMR
Vaidya et al[88]	Human Immunology/2007	Cas report of positive B cell crossmatch in full match patient	—	A single HLA-DP allele mismatch (DPB1 0601) resulted in positive B cells
Goral et al[89]	Nephrology, Dialysis, Transplantation/2008	2 case reports. 1st case was sensitized due to retransplant, pregnancy and blood transfusion and 2nd case was sensitized by blood transfusion only	—	Both cases has mixed rejections (acute cellular and AMR) needing therapy. First case has rejection in 2 mo and second case on d 12
Thaunat et al[90]	Transplant Immunology/2009	Case report of chronic AMR due to anti HLA-DP and antibodies to nondonor-specific HLA-DP, which has same amino acid sequence	—	Anti HLA-DP antibodies resulted in chronic AMR and author recommended epitope matching instead of antigen matching
Singh et al[91]	Transplantation/2010	Case report of fully matched patient who has 3 failed transplants and rejection due to anti-HLA-DP antibody	—	Patient develop borderline cellular rejection and AMR after 2 wk. There was a mismatch at the HLA-DPA1 locus and pre- and post-transplant sera identified DSA against DPA1 0103
Billen et al[98]	Tissue Antigens/2010	Retrospective analysis of pre- and post-transplant sera for HLA-DP antibodies	To analyze the incidence of HLA-DP antibodies in renal patients	The 14% (48/338) had anti HLA-DP DSAs. The 23% of these had DSAs pretransplant and 77 had DSAs after transplant. All DSAs had a single mismatch at a hypervariable region in 80% of cases
Jolly et al[85]	American Journal of Transplantation/2012	2 case reports. Case 1 was sensitized and had 3rd transplant with mismatches only for HLA-C 15 and HLA-DPB1 01 with negative crossmatch. Case 2 was live unrelated transplant from wife with 1-2-1 HLA-A, HLA-B, HLA-DR mismatched graft, with an additional single mismatch at the DP locus with positive B cell crossmatch	—	Patient 1 had acute cellular rejection on d 3 and AMR at wk 4. Patient 2 had AMR on d 11
Callender et al[7]	Human Immunology/2012	Retrospective analysis of 650 renal patients on waiting list	To determine the frequency of HLA-DP-specific antibodies in presence and absence of crossreactive HLA-DR antibodies	42% were reactive for HLA-DP antibodies. 58 of these were negative for crossreactive HLA-DR antigens, and 16 had no class II antibodies other than anti-HLA-DP
Mierzejewska et al[84]	Human Immunology/2014	Case report of recipient who had 3rd transplant and was completely matched except at HLA-DPA1 and -DPB1	—	AMR at d 13 due to presence of C1q binding IgG1 DSA against donor HLA-DPA1 and -DPB1
Cippà et al[50]	Human Immunology/2014	Case report of de novo donor HLA-DP-specific antibodies in a nonsensitized patient	—	Late AMR due to anti-HLA-DPS
Hörmann et al[99]	Clinical Transplantation/2016	Retrospective analysis of 195 consecutive kidney transplant patients	To study incidence and impact of anti-HLA-DP antibodies in renal transplantation	81 (49%) patients had anti-HLA-DP antibodies. Around 64% (n = 52) of patients were positive in the pretransplant samples and 36% (n = 29) were positive post-transplant. Anti-HLA-DP antibody-positive patients had a higher rate of rejection (P = 0.01)
Thammanichanond et al[92]	Transplantation Proceedings/2018	Case report of acute AMR by de novo Anti-HLA-DPβ and -DPα antibodies after kidney transplantation	—	Developed acute cellular and AMR after 15 months
Nikaein et al[93]	Transplant Immunology/2018	Brief communication of 2 cases who received transplants, 1 from living unrelated and the other from deceased donors. Both cases had DSAs to HLA-DPB with MFI > 15000	—	Both developed AMR (acute humoral rejection)
Marie et al[94]	Transplantation Report/2021	Case report of kidney transplants in 3 highly sensitized individuals with significant sensitization with donor-directed HLA-DP antibody and had kidney transplants from donors after brain death with positive B cell flow cytometry crossmatch	—	Case 1 (retransplant) had transplant glomerulopathy after 2 yr. Case 2 had no events. Case 3 had retransplant and AMR on d 10
Thammanichanond et al[86]	BMC Nephrology/2022	Case report of acute AMR associated with preformed HLA-DPα and HLA-DPβ DSAs that were not detected before transplantation	—	AMR at d 15 (this was first transplant)
Seitz et al[100]	Kidney International Reports/2022	Retrospective case–control study of 23 patients	To study the effect of pre-existing isolated HLA-DP-DSAs on renal allograft outcomes	Pre-existing HLA-DP DSAs was risk factor for AMR on multivariate analysis (HR = 9.578, P = 0.012). Patients with HLA-DP DSAs had increased microvascular scores (P = 0.0346) and worse transplant glomerulopathy (P = 0.015) compared with the standard immunological risk group
Pan et al[101]	HLA/2023	Meta-analysis of 5 studies with 1166 kidney transplant patients	To study the impact of preformed and de novo HLA-DP antibodies after renal transplantation on graft loss and rejection	De novo HLA-DP antibodies after transplantation showed an increased risk of graft loss or acute rejection (OR = 3.6, 95%CI: 1.6–8.10, P = 0.002, I2 = 52%). Preformed anti-HLA-DP antibodies did not show any effect

AMR: Antibody-mediated rejection; DSA: Donor specific antibody; HLA: Human leukocyte antigen.

HLA-DQ and DSA against HLA-DQ

Initially, DSA against HLA-DQ was considered less important due to its underexpression[61] and its masking by the HLA-DR locus[71,72,102,103]. Case reports on pathogenicity of these DSAs have shown mixed findings. An earlier case report of B-cell-positive crossmatch due to IgG antibody reported a successful transplant, with the graft maintaining good function up to a 1-year follow-up. This was despite immunoperoxidase techniques showing normal expression of HLA-DQ in biopsies during follow-up[104]. In another similar case, a patient with anti-HLA-DQ5 underwent a successful second transplant without any sequalae at a 2-year follow-up[105]. However, later on, there were case reports of chronic AMR due to de novo DSAs against HLA-DQ, occurring after 2 years in a retransplant patient, and mixed rejection in a first transplant due to DSAs against HLA-DQA1 after 2 years[106,107]. A brief communication compared renal transplant recipients and sensitized candidates with a failed graft, showing that 10% of newly transplanted patients and 10.8% of the control group developed DSAs against HLA-DQ within the first 6 mo[108]. The presence of these DSAs did not cause any graft rejection or deterioration in graft function. However, it is important to note that no protocol biopsies were done to detect subclinical rejection, and the short follow-up could not account for late or chronic AMR. A study from Japan examined the etiology of chronic AMR using ELISA and Luminex bead assays, classifying positive sera into high (> 20% of positive control), moderate (10%–20%) and low (2%–10%) categories. Among high levels of class II antibodies, 78% were DSAs against HLA-DQB, and 44% were against HLA-DRB. Although DSAs against HLA-DQB were high prevalent, they were not found to be associated with chronic AMR in any sera category (low, moderate or high). DSAs against HLA-DRB were more frequently associated with chronic AMR in the high and moderate groups[109].

Various studies were conducted in clinical settings to explore HLA-DQ, its characteristics, and its implications. Recipient HLA-DQ-directed antibodies may also target the patient's own HLA-DQ beta chain if paired with a non-self-DQ alpha chain. This phenomenon was studied using Luminex-based HLA class II beads for antibody identification and Luminex PCR-sequence-specific oligo prob hybridization for HLA-DQA1/DQB1 typing. Around 71% of patients had antibodies directed to their own DQα-chain or β-chain components, detected by test beads coated with the patient's own DQα-chain or β-chain components. The 34% had antibodies directed to their DQ β-chain in combination with non-self DQα-chain, and 62% had antibodies against their own DQα-chain in combination with non-self-DQ β-chain[110]. It is important to keep this in mind when analyzing virtual crossmatch results. Another advancement in understanding HLA is at the eplet level. Eplets are small patches of polymorphic amino acids on the surface of HLAs, which are the targets of HLA antibodies. Like other HLAs, HLA-DQ has been studied at eplet level. Using HLA matchmaker and Cn3D software, around 10 DQA eplets or eplet combinations and 13 DQB eplets or combinations were identified in one study[110]. The impact of HLA typing was also examined in a couple of studies. Earlier studies on HLA-DQ typing showed a negative correlation with graft dysfunction, reinforcing the idea that anti-HLA-DQ DSAs are benign in nature. One study on HLA-DQ mismatches reported that HLA-DQ mismatches between donor and recipient, in the presence of HLA-DR compatibility, did not influence the function or outcome of renal transplants[54]. Another study analyzed the effect of HLA-DQ matching in first kidney transplant recipients. A Cox regression model showed nonsignificant reduction of 3% only when adjusted for donors and recipients’ race, age and sex, cold ischemia time, body mass index, cyclosporine use, year of transplant, diabetes, HLA-A, HLA-B and HLA-DR match[111]. However, contrary to these negative findings, a few studies have reported the deleterious effects of HLA-DQ mismatches. Efforts to reduce HLA-DR and DQ mismatches have been shown to reduce the odds of developing transplant glomerulopathy[112]. Another study examined the effect of HLA-DQ mismatches on rejection. Compared to zero mismatch, one or two HLA-DQ mismatches led to significant increase in rejection events, late rejections, and AMR[74].

DSAs against HLA-DQ are de novo in nature and usually develop within 6 mo[108]. However, some DSAs may take up to 11 mo to develop[57]. Approximately 55%–77% of all DSAs are anti-HLA-DQ in nature[113]. The variability in the prevalence of anti-HLA-DQ DSAs is due to differences in MFI cutoff values, follow-up duration, and the techniques used to identify these antibodies. Various studies have explored different aspect of anti HLA-DQ antibodies and their pathogenicity. In a study of retransplant patients, HLA-DQB and -DQA mismatches led to the production of anti-HLA-DQB antibodies in 87% of cases, and HLA-DQA antibodies in 64% of cases. This signifies that both alpha and beta chains possess immunological epitopes that can trigger antibody production[114]. The first well-organized study describing the pathogenic effects of anti-HLA-DQ DSAs was conducted by Willicombe et al[56]. They found DSAs in 18.2% of their cohort, with 54.3% of those being anti-HLA-DQ DSAs. Anti-HLA-DQ DSAs were significantly associated with an increased risk of AMR, transplant glomerulopathy, and allograft loss[56]. Another study, published the same year, found that anti-HLA-DQ DSAs were the most prevalent, present in 78% of patients either alone or in combination with other DSAs. Anti-HLA-DQ DSAs led to acute rejection in 21% of cases, although this was not significant compared to patients without DSAs. However, mean serum creatinine and proteinuria were significantly higher in patients with anti-HLA-DQ DSAs compared to those without DSAs. Worse 3-year survival was reported when HLA-DQ antibodies were combined with non-HLA-DQ antibodies (52%), compared with HLA-DQ alone, non-HLA-DQ antibodies alone, or no antibodies[57]. In a novel study, Freitas et al[115] analyzed de novo anti-HLA-DQ DSAs, focusing on IgG subclasses and C1q-binding anti-HLA-DQ DSAs. Anti-HLA-DQ DSAs, in combination with non-HLA-DQ DSAs, significantly associated with acute rejection, increased risk of allograft loss, and lower graft survival. Recipients with acute rejection had significantly higher levels of IgG1/IgG3 combinations and C1q-binding anti-HLA-DQ DSAs. C1q-binding anti-HLA-DQ DSAs also led to significantly lower survival, with a 30% reduction[115]. Table 3[29,54,56,57,74,104-112,115,116] summarizes the evidence for and against anti-HLA-DQ DSAs in chronological order.

Table 3 Summaries of reported literature on HLA-DQ typing and anti-HLA-DQ donor-specific antibody.

Ref.	Journal/year	Study type and sample type	Objectives	Findings
Taylor et al[104]	Tissues Antigens/1987	Case report	—	B cell crossmatch positive due to IgG DSA against HLA-DQ did not affect graft function till 1 yr follow-up
Bushell et al[54]	Human Immunology/1989	Retrospective study of 25 HLA-DQ-mismatched but DR-matched patient	To study effect of HLA-DQ mismatch in HLA-DR-matched patients	No beneficial effect
Freedman et al[111]	Clinical Transplants/1997	Retrospective study of 12050 deceased 1st kidney transplant	To analyze the effect of HLA-DQ phenotype matching on renal allograft survival	Nonsignificant 3.0% reduction in graft failure (P = 0.38) was observed for each level of increasing HLA-DQ match when using the Cox regression model adjusted for recipient and donor race, age and sex, cold ischemia time, body mass index, cyclosporine A use, year of transplant, diabetes mellitus, HLA-A, HLA-B and HLA-DR match
Iniotaki-Theodoraki et al[108]	Transplantation/2003	Case–control study of 142 patients. Group A had 32 immediately post-transplant patients and group B had 110 sensitized patients who had failed grafts	To study humoral immune reactivity against HLA-DQ graft molecules in the early post transplantation period	No rejection or graft dysfunction in first 6 months
Duquesnoy and Marrari[29]	Transplant Immunology/2008	Retrospective analysis of 75 class-II-sensitized patients with different types of failed allografts including 60 kidney, 4 liver, 4 heart, 2 lung, 2 pancreas and 3 small bowel transplants	To describe the effect the donor-specific HLA class II epitope mismatching on antibody reactivity patterns	HLA-DQB and -DQA mismatches led to production of anti-HLA-DQB antibodies in 87% and HLA-DQA antibodies in 64%
Hartono et al[105]	Journal of Medical Case Reports/2009	Case report of a successful retransplant with pre-existing anti-HLA-DQ5 antibodies	—	Graft function good until 2 years without any deleterious effects
Tambur et al[110]	Transplantation/2010	Observational study of the sera of 104 patients	To identify HLA-DQ antibodies directed to patient’s own DQ or DQ chain linked to non-self DQ chains	7% of patients had anti -HLA-DQ antibodies against patient’s own DQ or DQ chain. 21% had antibodies to their own DQβ chain and 62% had antibodies to their own DQα chain
Kobayashi et al[109]	Human Immunology/2011	Observational study of 586 kidney transplant recipients	To know the impact of DSAs against HLA-DRB and -DQB on development of chronic AMR in high, moderate and low risk	Anti-HLA-DQB DSAs were not associated with chronic AMR in all immunological risk categories
Willicombe et al[56]	Transplantation/2012	Retrospective analysis of 505 kidney transplant recipients	The aim of this study was to establish the incidence and outcomes after the development of HLA-DQ DSAs	Patients with anti-HLA-DQ DSAs were at significant risk for AMR, transplant glomerulopathy, and allograft loss (P < 0.0001)
DeVos et al[57]	Kidney International/2012	Prospective analysis of retrospective data of 347 without pretransplant DSAs	To study development of de novo anti-HLA-DQ DSAs and its impact on patient and graft	78% of all DSAs were anti-HLA-DQ. No relation with rejection found. Mean creatinine and proteinuria higher in anti-HLA-DQ DSAs. Anti-HLA-DQ DSAs along with non-DQ DSAs led to reduced 3-yr survival
Freitas et al[115]	Transplantation/2013	Retrospective analysis of 284 transplant recipients	To study complement-binding characteristics of HLA-DQ DSAs	Anti-HLA-DQ DSAs and non-DQ DSAs caused more acute rejection (P = 0.0009), increased graft loss and reduced 5-yr survival. Acute rejection had more IgG1/IgG3 combination and C1q-binding antibodies (51%, P = 0.01; and 63%, P = 0.001)
Tambur et al[116]	Transplantation/2014	Clinical and translational research in 40 transplant recipients	To analyze eplet and epitope of HLA-DQ in immunologically naive patients before failed transplantation	10 HLA-DQA eplets or eplet combinations and 13 HLA-DQB eplets or combinations identified
Sapir-Pichhadze et al[112]	American Journal of Transplantation/2015	Case–control study of 156 kidney transplant recipients. Cases consisted of patients with transplant glomerulopathy and controls without transplant glomerulopathy	To assess risk of transplant glomerulopathy as a function of donor and recipient HLA-DR and HLA-DQ incompatibility at the eplet level	Logistic regression model showed increased odd of transplant glomerulopathy (OR = 2.84, 95%CI: 1.73-7.84) in 27-43 eplet mismatches and (OR = 4.62, 95%CI: 1.51-14.14)
Lim et al[74]	Clinical Journal of the American Society of Nephrology/2016	Retrospective observational data of 788 recipients followed for 2.2 years	To assess impact of HLA-DQ mismatches on rejection	Compared with 0 HLA-DQ mismatched kidneys, those who received 1 or 2 HLA-DQ mismatched had more rejections (P < 0.01), late rejections (P = 0.03), and AMR (P = 0.01)
Chowdhry et al[106]	Asian Journal of Transfusion Science/2019	Case report of de novo DSAs against HLA-DQ in a retransplant	—	Chronic AMR after 2 years
Liu et al[107]	Transplant Immunology/2022	Case report of DSA against HLA-DQA1	—	Mixed rejection after 2 years

AMR: Antibody-mediated rejection; DSA: Donor specific antibodies; HLA: Human leukocyte antigen; IgG: Immunoglobulin G; OR: Odds ratio.

WAY FORWARD IN VIEW OF CRITICAL REVIEW

Historically, there were various reasons for not typing HLA-C, HLA-DP and HLA-DQ. HLA-C[1-3] is the least expressed compared to other class I HLA-A and other HLA-Bs. This is despite equal intracellular synthesis of these antigens. Factors such as increased turnover of the heavy chain mRNA[59], inefficient association of the heavy chain with β₂-microglobulin[117], and retention in the endoplasmic reticulum[60] contribute to reduce HLA-C expression. Initial clinical studies showed that typing for HLA-C had no beneficial effect on graft outcomes[30,75]. Several early studies also indicated that frequencies and reactivities of anti-HLA-C were lower than anti-HLA-A and anti-HLA-B antibodies and that they have no deleterious effect on graft outcomes[5,9,29]. Similarly, among class II antigens, underexpression of HLA-DP/DQ[61], along with the masking of HLA-DQ by DR[69,70,102], reduced enthusiasm for typing these loci. HLA-DR expression is more abundant[61] compared to HLA-DP and HLA-DQ, and similarly expression of HLA-DR in response to IFN- and TNF is even more pronounced[64,65]. Additionally, HLA-DR expression is quicker, whereas HLA-DP and HLA-DQ expression is delayed[66]. Early clinical studies also found no impact of HLA-DP mismatches on graft outcomes[49] and anti-HLA-DP antibodies did not affect graft function[48,83]. Similarly, early research showed that anti-HLA-DQ antibodies did not influence graft outcomes[104,105,108,109]. These findings led to a greater focus on typing HLA-A, HLA-B and HLA-DR.

One might expect DSAs against these less-expressed antigens in cases of retransplantation, due to the priming of lymphocytes during a previous transplant, which could justify typing these loci. How about typing in the absence of sensitizing events, such as retransplantation, blood transfusion, or pregnancy? In-depth analysis of DSAs against each of the above loci identified cases where no sensitizing events were found. Multiple previous publications showed that no prior sensitizing events were identified in patients with rejection due to anti-HLA-C DSAs[30,32-34,78]. One retrospective analysis found similar risk of rejection with anti-HLA-C DSAs as in patients with anti HLA-A and HLA-B DSAs[34]. Similarly, anti-HLA-DP antibodies have been shown to cause rejection in first transplants[50,85,92]. Approximately 23%–64% of patients waiting for their first kidney transplant have anti-HLA-DP antibodies in their sera[7,98,99], indicating that antibodies can develop even in first transplants. Around 13% of the patients with rejected kidneys did not have class I and HLA-DQ/DR antibodies but had anti-HLA-DP antibodies as compared to 3.5% with functioning allografts[97]. This suggests that anti-HLA-DP antibodies are capable of rejection on their own. Similarly, there is much evidence that de novo anti-HLA-DQ DSAs can appear in the first transplant[57,107,109,113], which can have deleterious effect on the graft outcomes. The question is how these least expressed antigens in first transplants could lead to alloimmune response. One theoretical explanation is that inflammation could augment expression of these antigens. The presence of a renal allograft may keep immune response activated, leading to persistent low-grade inflammation[118-120]. Inflammatory markers such as C-reactive protein, IL-6 and terminal C5b–9 complement complex have been identified in kidney transplant recipients and can contribute to graft loss[121,122]. Another well-known cause of inflammation is ischemia–reperfusion, which is common in kidney transplant recipients[123,124]. Ischemia–reperfusion injury induces a proinflammatory Th1 and Th17 response instead of anti-inflammatory Treg cell response. Activated Th17 cells produce cytokines such as IL-17A, IL-17F and IL-21, and chemokines, such as chemokine (C-X-C motif) ligand (CXCL) such as CXCL1, CXCL2 and chemokine (C-C motif) ligand 2[125]. Th1 and Th17 responses override the Treg cell response. This results in recruitment of neutrophils and monocytes that could lead to allograft rejection[126].

Another important point to consider is that anti-HLA-C and anti-HLA-DP DSAs can augment the pathogenicity of HLA-A, HLA-B, HLA-DR and HLA-DQ DSAs, leading to more rejection, graft survival and graft loss[10,80]. Likewise, anti-HLA-DQ DSAs in combination with the other non-HLA-DQ DSAs leads to more rejection, reduced graft survival, and graft loss[57,113]. These facts underscore the need for typing and screening for DSAs against these loci. Finally, typing for HLA-C[79], HLA-DP[95,96] and HLA-DQ[74,112] has been shown to impact graft outcomes significantly. Therefore, it is crucial to type these loci to minimize mismatches, and to ensure better graft outcome or in case of mismatches to do surveillance for development of DSAs against these loci. Universal adoption of typing and screening for DSAs against these loci will enable clinicians to identify well match pairs, plan immunosuppression, minimize rejection, and improve graft and patient survival, making it one of the cost-effective strategies to be adopted. Looking at these facts, transplant organizations and societies have changed their approach to HLA typing. Australian New Zealand Pair Exchange requires typing for HLA-C, HLA-DPA1, HLA-DPB1, HLA-DQA1 and HLA-DQB1[127]. Eurotransplant recommendations initially suggest that every recipient must be typed for HLA-A, HLA-B, HLA-C, HLA-DR and HLA-DQ using DNA-based typing. However, they further suggest that every recipient should be typed for 11 loci (HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DRB3/4/5, DQB1, HLA-DQA1, HLA-DPB1 and HLA-DPA1) by DNA typing[128]. The British Transplantation Society and The Renal Association in 2018 recommended that HLA typing of the recipients and all potential living donors should be performed using DNA-based methods to at least two-digit (low) resolution for HLA-A, HLA-B, HLA-C, HLA-DR and HLA-DQ. However, they recommend HLA-DP typing only if the recipient is known to have HLA-DP-specific alloantibodies[129]. The Organ Procurement and Transplantation Network initially recommended typing for HLA-A, -B, -Bw4, -Bw6, -C, -DR, -DR51, -DR52, -DR53, -DQA1, -DQB1 and -DPB1 antigens. In August/September 2021, HLA-DPA1 was added to the list[130]. It is clear that there is variability in recommendations for HLA typing among transplant organizations and there is a need for consensus to bring uniformity. The consensus for typing these loci is possible through close liaison among transplant clinicians across the world. Regional transplant societies can sit together to formulate recommendations. International societies such as the World Health Organization, International Society of Nephrology or The Transplantation Society can act as bridge to provide a common platform to all clinicians worldwide to bring together regional societies for uniform recommendation to type these loci. This will ensure uniform practice across the globe, improve graft and patient survival, and help in future research.

RECOMMENDATIONS

We suggest the following recommendations: (1) HLA-C and HLA-DP can express themselves even in the first transplant and should be routinely typed universally; (2) Masking of HLA-DQ by HLA-DR via linkage disequilibrium is not adequate and HLA-DQ can evade this phenomenon. Therefore, typing for HLA-DQ should also be done routinely; (3) We suggest to look further into interaction of HLA-C molecules, with dendritic cells, lymphocytes and KIR and its potential role in rejection; (4) HLA-C, HLA-DP and HLA-DQ can induce alloimmune responses and antibody production. Therefore, patients with any mismatch should be screened and monitored for DSAs against these antigens; (5) We suggest single bead antigen for better identification of DSAs against HLA-C, HLA-DP and HLA-DQ antigens; (6) We suggest that eplet matching may determine unacceptable antigens more accurately and should be used; (7) Combination of anti HLA-Cw and HLA-DP DSAs along with DSAs against other antigens has more deleterious effects; therefore, their presence pre- or post-transplant should warrant clinicians to plan induction and maintenance therapy appropriately; (8) Anti-HLA-DQ DSA with non-HLA-DQ DSAs also synergize and clinicians must carefully monitor and optimize immunosuppression for better outcome; (9) There is a need for greater consensus amongst regional societies to call for unified position statements to include HLA-C, HLA-DP and HLA-C typing prior to transplantation universally across the globe; and (10) There is need to capture more data about these antigens and DSAs against them along with more prospective randomized control trial to know about these antigens and antibodies and their pathogenic effects.

CONCLUSION

HLA-C, HLA-DP and HLA-DQ, along with DSAs against them, are not benign. They can contribute to AMR, chronic AMR, mixed cellular rejection, and reduced graft survival. Although their harmful effects are more pronounced in retransplantation, they can also be detrimental in first transplantation. They alone can cause harmful effects and can augment the harmful effects of DSAs against HLA-A, HLA-B and HLA-DR. There is a clear need for more prospective studies and universal screening of these antigens and the DSAs they elicit.

ACKNOWLEDGEMENTS

We acknowledge our colleagues including consultants, registrars, fellows and nurses who help us while writing this manuscript.