Search

In this issue of Blood, Ahci and colleagues describe a simple and robust method for identifying the phenomenon of leukemic HLA loss often associated with treatment failure of hematopoietic stem cell transplantation from haploidentical donors (haplo-HSCT).1

Comprehensive knockout of HLA class II (HLA-II) β-chain genes is complicated by their high polymorphism. In this study, we developed CRISPR/Cas9 genome editing to simultaneously target HLA-DRB, -DQB1, and -DPB1 through a single guide RNA recognizing a conserved region in exon 2. Abrogation of HLA-II surface expression was achieved in five different HLA-typed, human EBV-transformed B lymphoblastoid cell lines (BLCLs). Next-generation sequencing–based detection confirmed specific genomic insertion/deletion mutations with 99.5% penetrance in sorted cells for all three loci. No alterations were observed in HLA-I genes, the HLA-II peptide editor HLA-DMB, or its antagonist HLA-DOB, showing high on-target specificity. Transfection of full-length HLA-DPB1 mRNA into knockout BLCLs fully restored HLA-DP surface expression and recognition by alloreactive human CD4 T cells. The possibility to generate single HLA-II–expressing BLCLs by one-shot genome editing opens unprecedented opportunities for mechanistically dissecting the interaction of individual HLA variants with the immune system.

T cell alloreactivity is mediated by a self-human leukocyte antigen (HLA)-restricted T cell receptor (TCR) repertoire able to recognize both structurally similar and dissimilar allogeneic HLA molecules (i.e., differing by a single or several amino acids in their peptide-binding groove). We hypothesized that thymic selection on self-HLA molecules could have an indirect impact on the size and diversity of the alloreactive response. To test this possibility, we used TCR Vß immunophenotyping and immunosequencing technology in a model of alloreactivity between self-HLA selected T cells and allogeneic HLA-DPB1 (DPB1) differing from self-DPB1*04:02 by a single (DPB1*02:01) or several (DPB1*09:01) amino acids in the peptide-binding groove. CD4+ T cells from three different self-DPB1*04:01,*04:02 individuals were stimulated with HeLa cells stably transduced with the relevant peptide processing machinery, co-stimulatory molecules, and HLA-DP. Flow cytometric quantification of the DPB1-specific T cell response measured as upregulation of the activation marker CD137 revealed significantly lower levels of alloreactivity against DPB1*02:01 compared with DPB1*09:01 (mean CD4+CD137+ frequency 35.2 ± 9.9 vs. 61.5 ± 7.7%, respectively, p < 0.0001). These quantitative differences were, however, not reflected by differences in the breadth of the alloreactive response at the Vß level, with both alloantigens eliciting specific responses from all TCR-Vß specificities tested by flow cytometry, albeit with higher levels of reactivity from most Vß specificities against DPB1*09:01. In line with these observations, TCRB-CDR3 immunosequencing showed no significant differences in mean clonality of sorted CD137+CD4+ cells alloreactive against DPB1*02:01 or DPB1*09:01 [0.39 (0.36-0.45) and 0.39 (0.30-0.46), respectively], or in the cumulative frequencies of the 10 most frequent responding clones (55-67 and 58-62%, respectively). Most of the clones alloreactive against DPB1*02:01 (68.3%) or DPB1*09:01 (75.3%) were characterized by low-abundance (i.e., they were not appreciable among the pre-culture T cells). Interestingly, however, their cumulative frequency was lower against DPB1*02:01 compared with DPB1*09:01 (mean cumulative frequency 35.3 vs. 50.6%, respectively). Our data show that, despite lower levels of alloreactivity, a similar clonal diversity can be elicited by structurally similar compared with structurally dissimilar HLA-DPB1 alloantigens and demonstrate the power of TCRB immunosequencing in unraveling subtle qualitative changes not appreciable by conventional methods. OA gold - CA Fleischhauer

Although quantitative PCR (qPCR) has been explored for chimerism monitoring after allogeneic stem cell transplantation (SCT), evidence regarding its clinical utility compared with standard short tandem repeat (STR) is still limited. We retrospectively studied commercial qPCR and STR chimerism with respective positivity thresholds of .1% and 1% in 359 peripheral blood (PB) and 95 bone marrow (BM) samples from 30 adult patients after first HLA-matched SCT for myeloid malignancies or acute lymphatic leukemia. Concordance between the 2 methods was 79.5%, with all discordant samples positive in qPCR but negative in STR. Of the latter, sporadic qPCR positivity without clinical correlates was seen mostly in BM samples early post-transplant. In 7 of 21 patients with available follow-up samples in the first months after transplantation, qPCR but not STR revealed low levels (1%) of sustained host chimerism in PB, reflecting delayed engraftment or persistent mixed chimerism (PMC). These conditions were associated with donor-recipient cytomegalovirus (CMV) serostatus and early CMV reactivation but not with immunosuppressive regimens or clinical outcome. qPCR predicted all 8/8 relapses with samples in the 6 months before onset by sustained positivity in both PB and BM compared with 1/8 relapses predicted by STR mainly in BM. The response kinetics to donor lymphocyte infusions for the treatment of PMC or relapse was shown by qPCR but not STR to be protracted over several months in 3 patients. Our results demonstrate the superior clinical utility of qPCR compared with STR for monitoring subtle changes of host chimerism associated with different clinical conditions, making a case for its use in the clinical follow-up of transplant patients.

Aim Next-generation sequencing (NGS) provides a cost and time efficient approach for routine HLA typing. Analysis software typically presumes a diploid genotype with counts of sequencing reads from homozygous or heterozygous genotypes. However, chimeric samples with more than two alleles per locus and imbalanced read counts may occur, for instance, after partially HLA-mismatched hematopoietic stem cell transplantation. Detection of alleles at even very low allele frequencies can then become diagnostically relevant in the context of post-transplant relapse. Here, we describe and validate a novel algorithm for rare HLA allele detection using NGS data. Methods Sequencing reads are mapped against expected HLA-alleles in a chimeric DNA-sample. Only polymorphic nucleotide positions differentiating the expected allele sequences are considered for further analysis. Reads are classified based on the frequency of discriminatory di-nucleotides. The classification algorithm was validated using in vitro admixtures of DNA samples down to allelic frequencies of 0.5 percent based on NGS data targeting exons 2 and 3 for the HLA genes A, B, C, DQB1, DRB1, and DPB1. Results We found that sensitivity and specificity are allele and locus-specific and depend on the number of sequence differences between the alleles. In each of the analyzed admixture samples a subset of the 12 amplicons harbored sufficient sequence differences to detect genotypes at dilutions as low as 0.5 percent. Fig. 1 shows an example for 3 amplicons. Conclusions The proposed algorithm may support the application of cost effective NGS for sensitive detection of HLA microchimerism in different settings including transplantation and pregnancy. Download : Download high-res image (200KB) Download : Download full-size image