HLA-E and HLA-F Are Overexpressed in Glioblastoma and HLA-E Increased After Exposure to Ionizing Radiation

Deep Learning Glioma Grading with the Tumor Microenvironment Analysis Protocol for Comprehensive Learning, Discovering, and Quantifying Microenvironmental Features

Gliomas are primary brain tumors that arise from neural stem cells, or glial precursors. Diagnosis of glioma is based on histological evaluation of pathological cell features and molecular markers. Gliomas are infiltrated by myeloid cells that accumulate preferentially in malignant tumors, and their abundance inversely correlates with survival, which is of interest for cancer immunotherapies. To avoid time-consuming and laborious manual examination of images, a deep learning approach for automatic multiclass classification of tumor grades was proposed. As an alternative way of investigating characteristics of brain tumor grades, we implemented a protocol for learning, discovering, and quantifying tumor microenvironment elements on our glioma dataset. Using only single-stained biopsies we derived characteristic differentiating tumor microenvironment phenotypic neighborhoods. The study was complicated by the small size of the available human leukocyte antigen stained on glioma tissue microarray dataset - 206 images of 5 classes - as well as imbalanced data distribution. This challenge was addressed by image augmentation for underrepresented classes. In practice, we considered two scenarios, a whole slide supervised learning classification, and an unsupervised cell-to-cell analysis looking for patterns of the microenvironment. In the supervised learning investigation, we evaluated 6 distinct model architectures. Experiments revealed that a DenseNet121 architecture surpasses the baseline's accuracy by a significant margin of 9% for the test set, achieving a score of 69%, increasing accuracy in discerning challenging WHO grade 2 and 3 cases. All experiments have been carried out in a cross-validation manner. The tumor microenvironment analysis suggested an important role for myeloid cells and their accumulation in the context of characterizing glioma grades. Those promising approaches can be used as an additional diagnostic tool to improve assessment during intraoperative examination or subtyping tissues for treatment selection, potentially easing the workflow of pathologists and oncologists.

Diversity in the HLA-I Recognition of HLA-F Monoclonal Antibodies: HLA-F or HLA-Ib Monospecific, HLA-E or HLA-G Bispecific Antibodies with or without HLA-Ia Reactivity

Previous investigators have used various anti-HLA-F monoclonal antibodies (mAbs) to demonstrate that the tissue distribution of HLA-F is highly restricted. Notably, these mAbs differed in their immunodiagnostic capabilities. Specifically, mAbs Fpep1.1 and FG1 detected HLA-F intracellularly in B cells but not on the cell surface, whereas mAb 3D11 detected HLA-F on the cell surface. The presence of HLA-F on T cells was recognized by mAb FG1 but not by mAb Fpep1.1. mAb 3D11 detected HLA-F on the cell surface of activated B cells and on peripheral blood lymphocytes, but not on the normal cells. Importantly, mAb 3D11 revealed that HLA-F exists as a heavy chain (HC) monomer, rather than as an HC associated with B2m. Although these mAbs are believed to be specific to HLA-F, their monospecificity has not been formally established, which is critical for immunodiagnostic and therapeutic purposes. Previously, we investigated the diversity of HLA class I reactivities of anti-HLA-E mAbs using HLA-I coated multiplex bead assays on a Luminex platform. We reported that more than 80% of the HLA-E mAbs were cross-reactive with other HLA-I molecules, with exceptionally few truly HLA-E-monospecific mAbs. In the present investigation, we generated IgG mAbs against HCs of HLA-F in Balb/C mice and examined the cross-reactivity of anti-HLA-F mAbs with other HLA-I alleles using a multiplex bead assay on the Luminex platform. Beads coated with an array of HLA homo- and heterodimers of different HLA-Ia (HLA-A, HLA-B, and HLA-C) and Ib (HLA-E, HLA-F, and HLA-G) alleles were used to examine the binding of the anti-HLA-F mAbs. Only two mAbs were HLA-F monospecific, and five were HLA-Ib restricted. Several anti-HLA-F mAbs cross-reacted with HLA-E (n = 4), HLA-G (n = 3), HLA-Ia alleles (n = 9), HLA-G and HLA-Ia (n = 2), and HLA-Ib and HLA-Ia (n = 6). This monospecificity and polyreactivity were corroborated by the presence of HLA-F monospecific and HLA-I-shared sequences. This study emphasizes the need to monitor the mono-specificity of HLA-F for reliable immunodiagnostics and passive immunotherapy.

Killer instincts: natural killer cells as multifactorial cancer immunotherapy

Natural killer (NK) cells integrate heterogeneous signals for activation and inhibition using germline-encoded receptors. These receptors are stochastically co-expressed, and their concurrent engagement and signaling can adjust the sensitivity of individual cells to putative targets. Against cancers, which mutate and evolve under therapeutic and immunologic pressure, the diversity for recognition provided by NK cells may be key to comprehensive cancer control. NK cells are already being trialled as adoptive cell therapy and targets for immunotherapeutic agents. However, strategies to leverage their naturally occurring diversity and agility have not yet been developed. In this review, we discuss the receptors and signaling pathways through which signals for activation or inhibition are generated in NK cells, focusing on their roles in cancer and potential as targets for immunotherapies. Finally, we consider the impacts of receptor co-expression and the potential to engage multiple pathways of NK cell reactivity to maximize the scope and strength of antitumor activities.

HLA-F transcriptional and protein differential expression according to its genetic polymorphisms

Many specificities single out HLA-F: its structure, expression regulation at cell membrane and function. HLA-F mRNA is detected in the most cell types and the protein is localized in the ER and Golgi apparatus. When expressed at cell surface, HLA-F may be associated to β2-microglobulin and peptide or expressed as an open-conformer molecule. HLA-F reaches the membrane upon activation of different primary cell types and cell-lines. HLA-F has its highest affinity for the KIR3DS1-activating NK receptor, but also binds inhibitory immune receptors. Some studies reported that HLA-F expression is associated with its genotype. Higher HLA-F mRNA expression associated with F*01:01:02, and 3 noncoding SNPs, rs1362126, rs2523405, and rs2523393, located in HLA-F-AS1 or upstream the HLA-F sequence were associated with HLA-F mRNA expression. Given the implication of HLA-F in many clinical setting, and the undisclosed process of its expression regulation, we aim to confirm the effect of the aforementioned SNPs with HLA-F transcriptional and protein expression. We analyzed the distribution, frequency and linkage disequilibrium of these SNPs at worldwide scale in the 1000 Genomes Project samples. Influence on the genotype of each SNP on HLA-F expression was explored using RNAseq data from the 1000 Genomes Project, and using Q-PCR and intracellular cytometry in PBMC from healthy individuals. Our results show that the SNPs under studied displayed remarkably different allelic proportion according to geography and confirm that rs1362126, rs2523405, and rs2523393 displayed the most concordant results, with the highest effect size and a double-dose effect.

Polymorphisms of Killer Ig-like Receptors and the Risk of Glioblastoma

PURPOSE

The immune responses of natural killer (NK) cells against cancer cells vary by patient. Killer Ig-like receptors (KIRs), which are some of the major receptors involved in regulating NK cell activity for killing cancer cells, have significant genetic variation. Numerous studies have suggested a potential association between the genetic variation of KIR genes and the risk of development or prognosis of various cancer types. However, an association between genetic variations of KIR genes and glioblastoma (GB) remains uncertain. We sought to evaluate the association of genetic variations of KIRs and their ligand genes with the risk of GB development in Koreans.

METHODS

A case-control study was performed to identify the odds ratios (ORs) of KIR genes and Classes A, B, and, C of the human leukocyte antigen (HLA) for GB. The GB group was comprised of 77 patients with newly diagnosed IDH-wildtype GB at our institution, and the control group consisted of 200 healthy Korean volunteers.

RESULTS

There was no significant difference in the frequency of KIR genes and KIR haplotypes between the GB and control groups. Genetic variations of KIR-2DL1, 3DL1, and 3DS1 with their ligand genes (HLA-C2, HLA-Bw4/6, and Bw4, respectively) had effects on the risk of GB in Korean patients. The frequency of KIR-2DL1 with HLA-C2 (OR 2.05, CI 1.19-3.52, p = 0.009), the frequency of KIR-3DL1 without HLA-Bw4 (80I) (OR 8.36, CI 4.06-17.18, p < 0.001), and the frequency of KIR-3DL1 with Bw6 (OR 4.54, CI 2.55-8.09, p < 0.001) in the GB group were higher than in the control group. In addition, the frequency of KIR-2DL1 without HLA-C2 (OR 0.44, CI 0.26-0.75, p = 0.003), the frequency of KIR-3DL1 with HLA-Bw4 (80T) (OR 0.13, CI 0.06-0.27, p < 0.001), the frequency of KIR-3DL1 without Bw6 (OR 0.27, CI 0.15-0.49, p < 0.001), and the frequency of KIR-3DS1 with Bw4 (80I) (OR 0.03, CI 0.00-0.50, p < 0.001) in the GB group were lower than in the control group.

CONCLUSIONS

This study suggests that genetic variations of KIRs and their ligand genes may affect GB development in the Korean population. Further investigations are needed to demonstrate the different immune responses for GB cells according to genetic variations of KIR genes and their ligand genes.

Disruption of the NKG2A:HLA-E Immune Checkpoint Axis to Enhance NK Cell Activation against Cancer

Ligation of the inhibitory receptor NKG2A by its ligand HLA-E negatively regulates the activation of natural killer (NK) cells, as well as subsets of CD8+ T cells and innate T cell populations. NKG2A has recently become a novel immune checkpoint target for the treatment of cancer and direct antibody mediated blockade of NKG2A function is currently under assessment in two phase 3 clinical trials. In addition to direct targeting, the NKG2A:HLA-E axis can also be disrupted indirectly via multiple different targeted cancer agents that were not previously recognised to possess immunomodulatory properties. Increased understanding of immune cell modulation by targeted cancer therapies will allow for the design of rational and more efficacious drug combination strategies to improve cancer patient outcomes. In this review, we summarise and discuss the various strategies currently in development which either directly or indirectly disrupt the NKG2A:HLA-E interaction to enhance NK cell activation against cancer.

Sequence Variations Within HLA-G and HLA-F Genomic Segments at the Human Leukocyte Antigen Telomeric End Associated With Acute Graft-Versus-Host Disease in Unrelated Bone Marrow Transplantation

Acute graft-versus-host disease (aGVHD) is defined as a syndrome of an immunological response of graft to the host that occurs early after allogeneic hematopoietic stem cell transplantation (HCT). This disease is frequently observed even in HCT matched for human leukocyte antigen (HLA) alleles at multiple gene loci. Although the HLA region represents complex and diverse genomic characteristics, detailed association analysis is required for the identification of uncharacterized variants that are strongly associated with aGVHD. We genotyped three loci, OR2H2, HLA-F-AS1, and HLA-G, that are located in the 460 kb of HLA telomeric region and statistically analyzed the genotypes including HLA-DPB1 with clinical and transplantation outcomes using 338 unrelated bone marrow transplantation (UR-BMT) patient–donor pairs who were matched for HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1 (HLA-10/10). Multivariate analyses demonstrated that HLA-F-AS1 and HLA-DPB1 mismatches were associated with grade II–IV aGVHD (hazard ratio (HR), 1.76; 95% CI, 1.07–2.88; p = 0.026; and HR, 1.59; CI, 1.02–2.49; p = 0.042, respectively). There was no confounding between HLA-F-AS1 and HLA-DPB1 (p = 0.512), suggesting that the HLA-F-AS1 mismatch has a strong effect on aGVHD independently of HLA-DPB1. Moreover, a stratified analysis suggested possible associations of HLA-F-AS1, HLA-DPB1, and/or HLA-G mismatches with grade II–IV aGVHD and the more severe grade III–IV aGVHD. These findings provide new insights into understanding the molecular mechanism of aGVHD caused by HLA-matched UR-BMT.