Blind Spots in HLA Class II Detection Generate Challenges for Antibody and RNA-based Assays

HLA class II molecules are fundamental to alloimmune responses, presenting peptides to CD4⁺ T cells and orchestrating both cellular and humoral immunity.1 In transplantation, mismatches in HLA class II strongly associate with the development of donor-specific antibodies (DSAs) and allograft rejection.2,3 Therefore, accurately detecting the expression of HLA class II molecules is foundational to our understanding of alloimmunity and improving transplant outcomes. Unfortunately, the same allelic diversity that enables HLA molecules to shape immune recognition also creates a major barrier to reliably detecting and quantifying HLA class II expression.4,5 In this Expert Insight, we examine the specificity and coverage of commonly used antibodies and RNA hybridization probes for HLA class II detection. We highlight how incomplete recognition skews data interpretation, and we provide strategies to improve the accuracy of class II detection for both basic immunology studies and translational applications in transplantation. Accurate measurement of HLA class II protein expression depends critically on antibody specificity. The importance of this has only grown with the rapid adoption of high-dimensional technologies such as spectral flow cytometry and mass cytometry by time-of-flight, where the reliability of each antibody directly determines the accuracy of downstream analyses.6,7 In these platforms, the so-called “pan-class II” antibodies are often trusted to detect all HLA alleles equally. Among these, the Tü39 clone is widely regarded as the gold standard for detecting class II molecules.8,9 Surprisingly, when we tested the specificity of Tü39 using a single antigen bead (SAB) assay, in which each bead is coated with a defined HLA class II allele to directly measure antibody binding, we found that although Tü39 efficiently recognized all HLA-DR and HLA-DP alleles, its coverage of HLA-DQ was incomplete. Specifically, the Tü39 antibody reacted strongly with DQ2 and DQ4 antigens but showed minimal or no reactivity with alleles in the DQ5-9 serogroups (Figure 1A). Given the critical role of HLA-DQ in de novo DSA formation and allograft rejection,10,11 this gap has serious implications for the interpretation of transplant immunology data.FIGURE 1.: Antibody specificity reveals blind spots in HLA class II protein detection. A, LABScreen SAB (One Lambda, ThermoFisher Scientific) assays revealed that the “pan-class II” clone Tü39 efficiently detected all HLA-DR, HLA-DP, and HLA-DQ2/4 molecules but failed to recognize HLA-DQ5–9 antigens. B, Flow cytometry of engineered BLCLs expressing individual HLA-DQ heterodimers confirmed that Tü39 bound HLA-DQ2-expressing cells but showed no detectable signal on HLA-DQ7-expressing cells despite similar expression of HLA-DQ measured by SPVL3. C, SAB assays defined the locus specificity of antibody clones L243 (HLA-DR), SPVL3 (HLA-DQ), and B7/21 (HLA-DP). D, Flow cytometry of BLCLs expressing individual class II molecules validated these findings with L243 binding exclusively to HLA-DR, SPVL3 to HLA-DQ, and B7/21 to HLA-DP. BLCL, B-lymphoblastoid cell line; SAB, single antigen bead.One limitation of SAB assays is that HLA molecules are presented as solubilized, bead-coupled proteins, which may alter their confirmation and antibody recognition. In contrast, cell lines express HLA class II molecules in their native state, enabling proper folding and peptide loading and providing a more physiologic representation of antibody binding. However, the major challenge in assessing the specificity of reagents in cell-based systems is that HLA class II molecules are normally coexpressed in antigen-presenting cells, because class II transactivator activation drives the simultaneous induction of HLA-DR, -DQ, and -DP.12 This makes it nearly impossible to disentangle signals from individual loci in conventional systems. To overcome this, we used B-lymphoblastoid cell lines (BLCLs) generated by collaborators (Crivello and Fleischhauer), in which all classical HLA class II β-chain genes were disrupted using clustered regularly interspaced short palindromic repeats / CRISPR-associated protein 9.13 These knockout BLCLs abolish endogenous class II expression while leaving the antigen processing machinery intact, providing a clean background for reintroducing individual class II B-genes by lentiviral transduction. Using this approach, we generated a panel of BLCLs in which each cell line expresses only a single HLA class II locus at a time, thereby enabling us to assess the locus specificity of antibodies in a physiological setting. Using these engineered BLCLs, we validated the gaps in HLA-DQ coverage of Tü39 observed with the SABs. Tü39 readily detected BLCLs expressing DQ2 (DQB1*02:01), producing a robust signal under standard staining conditions for flow cytometry (Figure 1B). In stark contrast, BLCLs expressing DQ7 (DQB1*03:01) showed no detectable signal despite being assessed under identical conditions (Figure 1B). However, both DQ2 and DQ7 cell lines expressed HLA-DQ as seen by positive SPVL3 staining (Figure 1B). This demonstrates that a reagent long assumed to be a “pan-class II” antibody in fact fails to recognize key HLA-DQ alleles, underscoring both the power of our BLCL system and the urgent need for rigorous validation of reagents across the full allelic spectrum for HLA. Our findings with Tü39 point to a broader problem in current practice: HLA class II is often treated as a single entity, despite HLA-DR, -DQ, and -DP being distinct molecules that each merit independent study. To move beyond this limitation, we systematically evaluated the specificity of other HLA class II antibody clones. SAB assays showed that SPVL3, an anti-HLA-DQ clone, reacted exclusively with DQ alleles with no detectable cross-reactivity to HLA-DR or HLA-DP SABs (Figure 1C). Similarly, L243 (anti-HLA-DR clone) and B7/21 (anti-HLA-DP clone) bound only to their respective loci (Figure 1C), consistent with previously reported SAB results demonstrating the specificity of L243 for HLA-DR and B7/21 for HLA-DP.14 These results were confirmed in our engineered BLCL system, where each cell line expresses a single HLA class II locus. Flow cytometric analyses found that SPVL3 bound robustly to BLCLs expressing HLA-DQ but showed no reactivity to BLCLs expressing HLA-DR or HLA-DP (Figure 1D). Likewise, L243 exclusively stained HLA-DR expressing BLCLs and B7/21 precisely labeled HLA-DP expressing cells (Figure 1D). Together, these data demonstrate that locus-specific antibodies provide highly accurate detection of individual class II molecules, highlighting the feasibility of resolving HLA class II expression with validated reagents. HLA polymorphism complicates far more than antibody recognition. The genetic variation that diversifies HLA class II molecules also produces diverse allele-specific mRNA transcripts. This creates a fundamental challenge for probe-based transcriptomic platforms such as the molecular microscope diagnostic system15,16 and, more recently, in situ hybridization spatial transcriptomic platforms including MERFISH, Nanostring’s CosMx, and 10X Genomics’ Xenium.17 These technologies promise to not only significantly advance transplantation research but also transform clinical practice by providing tools to improve the diagnoses of T cell–mediated rejection and antibody-mediated rejection from patient biopsies.18 However, the accuracy of these assays hinges on the precise recognition of RNA sequences by a predesigned panel of hybridization probes. Therefore, just as the so-called pan-class II antibodies failed to uniformly detect all HLA class II alleles, the extreme polymorphism of HLA transcripts threatens the reliability and accuracy of probe-based gene expression assays. We encountered this limitation firsthand while investigating how inflammatory cues drive HLA class II expression. Interferon-γ (IFN-γ), acting through the transactivator class II transactivator, is a potent inducer of class II genes in endothelial cells (ECs).19,20 To further analyze this response, we stimulated primary ECs from 11 donors and profiled them with Nanostring’s nCounter Human Organ Transplant Panel.18 At first glance, the data looked deceptively variable with HLA class II transcript expression differing dramatically across donors (Figure 2A). Yet, corresponding measurements told a different story. IFN-γ-stimulated ECs robustly upregulated HLA-DR, -DQ, and -DP at the protein level by flow cytometry and at the mRNA level by RNA sequencing (data not shown). Our observations suggested that the variability detected by the Human Organ Transplantation Panel was not due to biological differences in IFN-γ responses but instead reflected incomplete recognition of hybridization probes across HLA alleles, echoing the limitations of the Tü39 pan-class II antibody. Indeed, closer inspection revealed that the commercial probe set, designed to detect all DR, DQ, and DP alleles, lacked complementarity for a surprisingly large number of common HLA variants. Consequently, the donor-to-donor variability in Figure 2A was largely an artifact of probe design, highlighting a significant limitation of current RNA hybridization probe-based approaches.FIGURE 2.: Hybridization probe specificity reveals blind spots in HLA class II RNA detection. A, Original Nanostring’s nCounter HOT panel probes exhibited inconsistent coverage of HLA class II transcripts across primary EC donors. B, Custom-designed probes targeting HLA-DRB1, HLA-DQA1, and HLA-DQB1 produced allele-specific hybridization patterns that reflected each donor’s HLA genotype. C, Flow cytometry confirmed that HLA-DR, HLA-DQ, and HLA-DP were robustly expressed after IFN-γ stimulation using validated locus-specific antibodies (Figure 1). D, Inclusion of the custom probes into our personalized HOT panel enabled comprehensive detection of HLA class I and class II transcripts across all 11 genetically distinct EC donors. EC, endothelial cell; HOT, Human Organ Transplant; IFN-γ, Interferon-gamma.To overcome this limitation, we engineered custom probes in collaboration with Nanostring to target different allele sequences for HLA-DRB1, HLA-DQA1, and HLA-DQB1 (Figure 2B). When applied to IFN-γ-stimulated ECs, these probes produced distinct hybridization patterns dependent on each donor’s HLA genotype. By retaining the probe with the strongest hybridization signal for each allele, we achieved transcript quantification that closely aligned with protein expression (Figure 2B and C). Incorporating these probes into our customized Human Organ Transplant Panel enabled comprehensive measurement of HLA class I and class II expression across all 11 genetically distinct EC donors (Figure 2D). Our findings reveal that probe-based transcriptomics, although powerful, must be carefully tailored to account for HLA diversity if they are to yield biologically and clinically reliable insights. CONCLUDING REMARKS As high-dimensional technologies become increasingly integrated in transplantation research and clinical practice, it is critical to recognize their blind spots. Antibodies and RNA hybridization probes are only as reliable as their ability to recognize specific peptide or nucleotide sequences, a strength that becomes a liability when faced with the extraordinary polymorphism of HLA. Without accounting for this blind spot, these powerful platforms risk incomplete detection of HLA molecules, ultimately leading to the mischaracterization of the very molecules that lead to DSA formation and alloimmunity. This problem is especially pressing in the analysis of current publicly available data sets, where HLA detection is often incomplete or inconsistent across alleles, leading to inaccurate impressions of HLA class II expression during transplantation. Here, we experimentally validate the specificity of key class II antibodies (L243 for HLA-DR, SPVL3 for HLA-DQ, and B7/21 for HLA-DP) and demonstrate that hybridization probes must also be carefully engineered to capture HLA diversity. Collectively, our findings deliver a clear warning: emerging technologies that overlook HLA allele diversity jeopardize mischaracterizing these critical alloimmune targets and undermining their clinical relevance, whereas approaches designed with this diversity in mind can deliver more precise and meaningful insights. ACKNOWLEDGMENTS The authors thank the laboratory of Dr Katharina Fleischhauer and Dr Pietro Crivello for providing the CRISPR-engineered HLA class II knockout cells and for their expertise.