Next Article in Journal
Explainable Boosting Machine Predicting Length of Stay After Liver Surgery in Patients with Colorectal Liver Metastases
Previous Article in Journal
Chimeric Antigen Receptor–Immune Cell-Based Therapies for Clear Cell Renal Cell Carcinoma: Latest Advancements and Directions
Previous Article in Special Issue
Prognostic Scores and Risk Stratification of Myeloproliferative Neoplasms: 2026 Updates
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Cancers
Volume 18
Issue 13
10.3390/cancers18132052
cancers-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Immunogenetic and Transcriptomic Evidence Implicating the NKG2D-MICA/MICB Axis in CALR-Mutated Myeloproliferative Neoplasms

by
Velizar Shivarov
1,2,*,
Gergana Tsvetkova
3,
Ilina Micheva
4,
Evgueniy Hadjiev
3,
Jasmina Petkova
5,
Galia Madjarova
5 and
Milena Ivanova
6,*
1
Department of Experimental Research, Medical University Pleven, 5800 Pleven, Bulgaria
2
Department of Clinical Hematology, St. Sophia General Hospital, 1618 Sofia, Bulgaria
3
Department of Clinical Hematology, Alexandrovska University Hospital, Medical University Sofia, 1431 Sofia, Bulgaria
4
Department of Clinical Hematology, Saint Marina University Hospital, Medical University Varna, 9010 Varna, Bulgaria
5
Department of Physical Chemistry, Faculty of Chemistry and Pharmacy, Sofia University “St. Kl. Ohridski”, 1164 Sofia, Bulgaria
6
Department of Clinical Immunology, Alexandrovska University Hospital, Medical University Sofia, 1431 Sofia, Bulgaria
*
Authors to whom correspondence should be addressed.
Cancers 2026, 18(13), 2052; https://doi.org/10.3390/cancers18132052
Submission received: 26 April 2026 / Revised: 17 June 2026 / Accepted: 21 June 2026 / Published: 24 June 2026
(This article belongs to the Special Issue Novel Advances in the Molecular Understanding of Myeloproliferative Neoplasms (MPNs))
Browse Figures
Review Reports Versions Notes

Simple Summary

Myeloproliferative neoplasms are blood cancers in which blood-forming cells grow abnormally. In some patients, these diseases are driven by mutations in the CALR gene. The immune system may help control early abnormal blood-cell clones, but the mechanisms involved are still not fully understood. In this study, we investigated whether genetic variation in the NKG2D–MICA/MICB immune-recognition pathway is linked to CALR-mutated myeloproliferative neoplasms. We found that one MICA allele, MICA*004:001, was more frequent in patients with CALR-mutated disease than in healthy controls. Additional exploratory analyses suggested possible associations involving MICB and KLRK1/NKG2D haplotypes. We also found transcriptomic evidence that CALR-mutant stem and progenitor cells may show altered MICA/MICB expression. Together, these findings suggest that the NKG2D–MICA/MICB axis may contribute to immune surveillance in CALR-mutated myeloproliferative neoplasms. However, larger studies and direct functional experiments are needed before firm biological or therapeutic conclusions can be made.

Abstract

Background/Objectives: Immune surveillance is increasingly recognized as a modifier of myeloproliferative neoplasm (MPN) initiation and evolution, yet the contribution of the NKG2D receptor and its ligands MICA/MICB to CALR-mutated disease remains unclear. Methods: We performed high-resolution next-generation sequencing genotyping of MICA and MICB in 43 patients with CALR-mutated MPN (WHO 2022 criteria) and compared the allele and haplotype distributions with those of 156 healthy Bulgarian controls and 85 patients with JAK2 V617F-positive MPN. Associations were tested using age- and sex-adjusted additive generalized linear models; bi-locus haplotypes were evaluated using haplotype score methods. In a genotyped subgroup (35 CALR-mutated MPN patients and 105 controls), functional KLRK1 (NKG2D) polymorphisms were analyzed for haplotype-level associations. We also performed 700 ns molecular dynamics simulations of selected MICA variants in complex with NKG2D and reanalyzed publicly available single-cell RNA-sequencing data (GSE117826) and RNA-sequencing data from CRISPR/Cas9-edited CALR-mutant iPSC-derived megakaryocytes to evaluate MICA/MICB expression. Results: MICA*004:001 was significantly associated with CALR-mutated MPN versus controls (p = 0.004; Bonferroni-adjusted p = 0.047), while MICB*008:001 showed only nominal association. Exploratory haplotype analyses identified a MICA*009:01-MICB*004:001 haplotype associated with CALR-mutated status (p = 0.008) and a KLRK1 G-A-G-T haplotype (rs1049174-rs2617160-rs2246809-rs2617170) associated with increased CALR-mutated MPN risk (OR = 3.61; p = 0.029). Transcriptomic reanalysis indicated a higher fraction of CALR-mutant stem and progenitor cells expressing detectable MICA/MICB transcripts, and heterozygous CALR-mutant megakaryocytes exhibited higher MICA expression than the wild type. Conclusions: Together, these data support an exploratory immunogenetic and transcriptomic link between the NKG2D-MICA/MICB axis and CALR-mutated MPN, but direct protein-level and functional studies are required before mechanistic or therapeutic conclusions can be drawn.

1. Introduction

Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs) are paradigmatic multistep clonal disorders arising from hematopoietic stem cells (HSCs) and shaped by both cell-intrinsic lesions and the microenvironment [1]. The three major entities—essential thrombocythemia (ET), polycythemia vera (PV), and primary myelofibrosis (PMF)—share recurrent driver mutations in JAK2, CALR, and MPL, yet present with distinct morphology and clinical phenotypes [1]. Beyond oncogenic signaling, MPNs are characterized by a chronic inflammatory milieu that sustains myeloproliferation, contributes to symptom burden, and has therapeutic implications [2]. Growing evidence suggests that immune surveillance operates early in MPN pathogenesis and may constrain the expansion of nascent mutant clones [1]. Accordingly, immunogenetic variation in HLA genes [3,4,5,6], MICA/MICB [7], cytokine loci [8,9], and immune checkpoint genes [10] has been investigated as a determinant of disease susceptibility and progression.
HLA class I (HLA-I) alleles modulate immune control of MPN by determining which neoepitopes can be presented to cytotoxic T cells. Several HLA-I alleles are under-represented among patients with JAK2 V617F-mutated MPNs, consistent with a protective effect via T-cell immunosurveillance [4]. For example, HLA-B*35:01 (as well as HLA-A*02:01 and HLA-C*15:02) has been reported at a lower frequency in JAK2-mutated disease and can bind to a JAK2 V617F-derived peptide, supporting a mechanistic link between antigen presentation and clone elimination [4]. Conversely, MPN stem/progenitor cells may downregulate HLA-I expression, impairing neoantigen display and facilitating immune escape; interferon-alpha and JAK1/2 inhibition have been reported to restore HLA-I expression in this compartment, potentially increasing tumor visibility to the immune system [4].
Recent work further suggests that the host HLA repertoire can influence which driver mutation successfully establishes disease. In a recent immunogenetic analysis, HLA-C*06:02 was under-represented among CALR-mutated MPNs, consistent with efficient presentation of mutant CALR-derived peptides in carriers [6]. More broadly, the predicted capacity of an individual’s HLA-I alleles to present JAK2 V617F-derived versus CALR-mutant peptides showed an inverse relationship, supporting an HLA-mediated immunoediting model that may contribute to the mutual exclusivity of these driver mutations [6]. CALR-mutant hematopoietic cells also displayed lower expression of antigen-processing genes (e.g., TAP1 and CIITA), consistent with downregulation of antigen presentation pathways [6].
The MHC class I-related molecules A and B (MICA and MICB) are stress-inducible ligands for the activating receptor NKG2D expressed on natural killer (NK) cells and subsets of T cells and are, therefore, poised to shape immunosurveillance in hematologic malignancies [11]. In JAK2 V617F-positive MPN, specific MICA alleles (e.g., MICA*008:01 and MICA*016) were reported at a lower frequency than in controls, and circulating soluble MICA was increased, consistent with ligand shedding as an immune-evasion mechanism [7]. While mature myeloid cells may upregulate MICB, MPN CD34+ stem/progenitor cells can downregulate MICA/MICB expression, potentially limiting NK cell recognition during disease establishment [7].
Here, we investigated whether genetic variation in MICA/MICB and the NKG2D receptor gene (KLRK1), as well as transcriptional regulation of MICA/MICB across CALR-mutant hematopoietic compartments, contributes to the pathogenesis of CALR-mutated MPNs.

2. Materials and Methods

2.1. Patients

We analyzed peripheral blood DNA from 43 patients with CALR-mutated, Philadelphia chromosome-negative MPN diagnosed according to the 2022 WHO criteria [12]. The MICA and MICB allele and haplotype distributions were compared with those of 85 patients with JAK2 V617F-positive MPNs from our prior cohort [4] and with those of 156 healthy Bulgarian controls from our immunogenetics of aging studies [13]. The demographic and clinical characteristics are summarized in Table 1. Because controls were younger than patients, age was included as a covariate in association models; sex was also included. All participants provided written informed consent. This study was approved by the Institutional Ethics Committee at the Medical University—Sofia under protocol 05/20 March 2025. All subjects provided informed consent in accordance with the Declaration of Helsinki.

2.2. DNA Extraction

Genomic DNA was extracted from sodium citrate peripheral blood (PB) using the IPrep (Life Technologies, Paisley, UK) automated system, as described previously [13]. Genomic DNA samples were stored at −20 °C until being genotyped.

2.3. Genotyping for MPN-Associated Mutations

Genotyping for JAK2 (including V617F), MPL, and CALR mutations was performed using direct sequencing and/or a bead-based liquid assay, as previously described [14,15,16].

2.4. MICA and MICB Genotyping and Association Analyses

MICA and MICB genotyping were performed on 43 CALR-mutant MPNs, 85 MPNs from JAK2 V617F-positive patients and 156 healthy controls by NGS whole-gene amplification using NGSgoR AmpX MICA with the MICB amplification kit (GenDX, Utrecht, The Netherlands). MICA and MICB genotypes were assigned by the NGSengine analysis software (GenDX, Utrecht, The Netherlands) and the IPD-IMGT/HLA database, as described previously [17]. Considering the difference in terms of age between patients and controls in order to avoid data analysis bias, the analysis of the association of MICA and MICB alleles with MPN was performed by fitting additive generalized linear models with age and gender as covariates [7,17]. Analyses were initially performed independently for each locus using two-field resolution. The analysis was executed through the HIBAG package (version 1.44.0) for Bioconductor (release 3.21) for R for Windows (version 4.5.0) [18]. Associations between MPN and bi-locus haplotypes were analyzed using the haplo.stats package (version 1.9.7) (https://cran.r-project.org/web/packages/haplo.stats/index.html) (accessed on 01 March 2026) through implementation of imputed haplotype score calculation, as described previously [7]. p-values below 0.05 were considered significant.

2.5. KLRK1 Genotyping and Association Analysis

Five functional single-nucleotide polymorphisms (SNPs) in KLRK1 (rs2255336, rs1049174, rs2617170, rs2246809, and rs2617160) were genotyped in a subgroup of 35 CALR-mutated MPN patients and 105 healthy controls using a real-time PCR-based allelic discrimination assay. Genotypes were called using standard instrument software. Haplotype associations were evaluated using SNPStats (Institut Català d’Oncologia, Barcelona, Spain) [19], with adjustment for age and sex. Because rs2255336 deviated from the Hardy–Weinberg equilibrium in controls, haplotype analyses were restricted to rs1049174, rs2617160, rs2246809, and rs2617170.

2.6. In Silico Modelling

Molecular dynamics simulations (MDSs) of MICA*001:01:01, MICA*004:01:01, and MICA*112:02 proteins in complex with the NKG2D dimer were performed using the AMBER03 force field in GROMACS 2023 [20]. The initial crystal structure was obtained from the Protein Data Bank (PDB ID: 1HYR) [21]. The systems were constructed by modifying the MICA chain (chain C in 1HYR) to match the MICA*004:01:01 and MICA*112:02 sequences retrieved from the IPD-IMGT/HLA Database (release 3.61) [22]. Amino acid substitutions were introduced in PyMOL 2.5.1 [23].
Each complex was solvated with TIP3P water and neutralized by addition of Na+ and Cl− counterions. Protonation states at pH 7.0 were estimated using PROPKA [24], and 150 mM NaCl was added to mimic physiological ionic strength. After energy minimization, systems were heated to 310 K in the constant-temperature, constant-volume (NVT) ensemble for 400 ps, followed by production simulations in the isothermal–isobaric (NPT) ensemble at 310 K and 1 atm.
For all simulations, the production phase lasted for 700 ns, and structural snapshots were saved every 20 ps. The root-mean-square deviations (RMSDs) of atomic coordinates for all atoms of the complex were calculated relative to the initial production structure using built-in GROMACS tools. To characterize the effect of the structural changes in MICA variations on the stability of the NKG2D–MICA complex, a detailed analysis of the ligand–receptor interfaces was performed using LigPlot+ v2.3.1 [25,26].

2.7. Gene Expression Analyses

We reanalyzed a publicly available single-cell RNA sequencing (scRNA-seq) dataset with genotype-resolved transcriptomes from bone marrow cells of five CALR-mutated essential thrombocythemia (ET) and five CALR-mutated myelofibrosis (MF) patients (GSE117826) [27]. Processed, filtered, and integrated data were handled as a Seurat object, as described previously [6]. Dimensionality reduction was performed using t-SNE [28] in Seurat version 5.4.0 [29]. Cell-type annotation was performed using the celldex (version 1.22.0) and SingleR (version 2.14.0) packages with a hematopoietic reference dataset [30], as described previously [6]. MICA and MICB expression in CALR-mutant versus wild-type HSCs, common myeloid progenitors (CMPs), and megakaryocyte-erythroid progenitors (MEPs) was visualized using dot plots implemented through the SCpubr (version 3.0.1) package [31].
We additionally analyzed an RNA-sequencing dataset of human iPSC-derived megakaryocytes (MKs) edited by CRISPR/Cas9 to harbor heterozygous or homozygous CALR frameshift mutations, alongside an isogenic CRISPR control line (GSE182479) [32]. Processed gene counts were downloaded from GEO and normalized to log2-transformed transcripts per million (TPM). MICA and MICB expression was compared between genotypes using two-sided t-tests and visualized with the ggpubr package (v. 0.6.1). Gene set enrichment analysis (GSEA; v4.4.0) was performed on normalized expression data using the standard workflow [33,34] and custom gene sets for antigen processing and presentation pathways, as described previously [4,5,6].

3. Results

3.1. MICA*004:001 Allele Is Associated with CALR-Mutated MPNs

When we compared the frequencies of MICA alleles in CALR-mutated MPNs vs. healthy controls after adjusting for age and sex, we did not identify any protective MICA allele (Supplementary Tables S1–S3). However, the MICA*004:001 allele was significantly associated with CALR-mutant status (p = 0.004; adjusted p = 0.047) (Figure 1A,B and Supplementary Table S1). When we tested for any associations with alleles defined by exon 5 or codon 129 polymorphisms, we did not identify any significant associations (Figure 1C,D and Supplementary Tables S2 and S3). When we performed analogous analyses comparing CALR-mutated MPNs vs. JAK2 V617F-positive MPNs, we did not identify any significant associations (Figure 2) (Supplementary Tables S4–S6). Interestingly, we identified a novel MICA allele in a CALR-mutant patient which was annotated as MICA*112:02 (Supplementary Table S1) [35].

3.2. MICB*008:001 Allele Shows Nominal Association with CALR-Mutated MPNs

When we searched for association of MICB alleles with CALR-mutated MPNs vs. healthy controls (Figure 3A and Supplementary Table S7), one allele showed a nominal positive association, MICB*008:001 (p = 0.037), which was lost after Bonferroni adjustment (p = 0.184) (Figure 3B). When we compared CALR-mutated MPNs vs. JAK2 V617F-positive MPNs, no significant associations were found (Figure 4A,B and Supplementary Table S8). Of note, however, we identified two novel MICB alleles in CALR-mutant patients, annotated as MICB*055 and MICB*005:15 (Figure 3A and Figure 4A; Supplementary Table S7) [36].

3.3. Haplotype Associations

We further tested for biallelic haplotype associations with CALR-mutant status in comparison to healthy controls (Supplementary Tables S9–S15) and JAK2 V617F-mutated MPNs (Supplementary Tables S16–S22). We defined MICA~MICB and single-locus HLA-A/B/C~MICA/MICB haplotypes. Because several alleles and haplotypes were evaluated in a relatively small CALR-mutated cohort, these analyses should be interpreted as exploratory and require independent replication. Analyses that showed nominal associations are shown in Figure 5. The MICA*009:01-MICB*004:001 haplotype was nominally associated with CALR-mutated MPN versus healthy controls (p = 0.008) (Figure 5A and Supplementary Table S15). The B*35:01-MICA*002:001 haplotype showed a positive nominal association, whereas the B*35:01-MICA*016 haplotype showed an apparent protective nominal association (Figure 5B). The lowest p-value was observed for the HLA-C*15:02-MICA*027:01 haplotype (Figure 5C). Finally, several weaker nominal associations were observed for MICB-containing biallelic haplotypes (Figure 5D).

3.4. NKG2D Polymorphisms

In a genotyped subgroup (35 CALR-mutated MPN patients and 105 controls), we assessed the associations of four KLRK1 SNPs (rs1049174, rs2617160, rs2246809, and rs2617170) with CALR-mutated status. The estimated haplotype frequencies are shown in Table 2 (upper panel). In age- and sex-adjusted haplotype association models, a single haplotype (G-A-G-T; rs1049174-rs2617160-rs2246809-rs2617170) was associated with increased risk of CALR-mutated MPN (Table 2, lower panel).

3.5. Molecular Dynamics Simulations

To obtain structural context for selected MICA variants, we performed molecular dynamics simulations of MICA*004:01:01 and MICA*112:02 in complex with NKG2D, using the published MICA*001:001-NKG2D structure as the initial template [21]. The initial structures were modified according to the conventional amino acid alignment (Figure 6A) and then subjected to 700 ns production MD simulations (Figure 6B–D). These simulations were designed to compare conformational behavior and interface contacts in silico; they were not intended to quantify binding affinity or to infer NKG2D-mediated immune recognition in patients. The final 100 ns of each trajectory were subjected to statistical analysis. Five thousand (5000) frames were used for similarity-based clustering of structures via two cluster analysis methods: Jarvis-Patrick [37] and gromos [38], using a cutoff of 0.16 nm. In the former method, clustering was performed based on the coordinates of the entire complex system, while in the second, it was based on the coordinates of the backbone atoms. Clusters with a population of over 1 500 structures were considered representative. Regardless of the method, the MICA*001:01:01-NKG2D complex clustered into only one average (central) structure, whereas MICA*004:01:01-NKG2D clustered into nine structures (with two representative structures using the Jarvis-Patrick method) or 43 structures (with two representative structures using the gromos method), and the MICA*112:02-NKG2D complex clustered into one structure (with one representative structure using the Jarvis-Patrick method) or 22 structures (with one representative structure using the gromos method).
Figure 7 shows interaction maps of the representative clusters of each studied system. The amino acid substitutions defining MICA*004:01:01 and MICA*112:02 do not directly involve residues described as core contact residues in the crystal structure of the MICA*001:01:01-NKG2D complex. Nevertheless, molecular dynamics simulations under physiological conditions indicated allele-specific changes in the interface architecture and identified alternative contacts involving additional amino acids.

3.6. Single-Cell RNA-Sequencing Analysis

We analyzed publicly available scRNA-seq data with genotype-resolved transcriptomes from five CALR-mutated ET and five CALR-mutated MF patients (GSE117826) [27], enabling direct comparison of MICA and MICB expression between CALR-mutant and wild-type cells within the same microenvironmental context [6]. In ET cases (Figure 8A,B), CALR-mutant HSCs, CMPs, and MEPs showed a higher fraction of cells with detectable MICA and MICB transcripts than the wild type. Similar patterns were observed in MF cases, with higher MICA expression in CALR-mutant HSC and MEP compartments (Figure 8C,D).
We further examined whether CALR gene dosage affects MICA/MICB expression in megakaryocytes using RNA-seq data from CRISPR/Cas9-edited human iPSC-derived megakaryocytes (MKs) (Figure 8E–G) [32]. GSEA showed relative enrichment of MHC-I and MHC-II antigen processing and presentation pathways in CALR-mutant MKs (Figure 8E,F). Heterozygous CALR-mutant MKs showed higher MICA expression relative to controls, which is compatible with a cell-intrinsic effect of mutant CALR on MICA regulation, whereas MICB expression was not materially altered (Figure 8G).

4. Discussion

Immunoediting is increasingly recognized in early MPN pathogenesis [1,39]. Adaptive immune recognition of MPN stem cells can be driven by both driver and passenger mutations and is expected to be HLA-restricted. Consistent with this, several studies have identified HLA class I alleles that are under-represented in JAK2 V617F- or CALR-mutated MPN, suggesting partial protection [3,4,6]. Because such effects are not absolute, additional immunogenetic factors may contribute to immune control. We previously observed associations implicating the NKG2D ligand axis in JAK2 V617F-positive MPN (MICA*008:01 and MICA*016 as potentially protective alleles and MICB*018:01) [7]. In the current study, we evaluated whether the NKG2D-MICA/MICB axis also contributes to susceptibility to CALR-mutated MPN.
We genotyped MICA and MICB in 43 Bulgarian patients with CALR-mutated MPN and compared allele frequencies with those in 85 JAK2 V617F-positive MPN patients and 156 healthy Bulgarian controls [7]. We identified MICA*004:001 as an allele associated with increased CALR-mutated MPN risk versus controls. MICA*004 is among the most common alleles in European populations [40,41,42,43], and despite our modest sample size, we also observed three previously unreported MICA/MICB alleles [35,36]. Because MICA polymorphisms can influence NKG2D binding, we performed molecular dynamics simulations to explore potential structural correlates. Prior work has classified MICA variants into higher- versus lower-affinity types based on linked polymorphic residues [44,45]. In this framework, MICA*004:001 is categorized as a higher-affinity (type I) variant, whereas MICA*001:01 and MICA*112:02 fall within the lower-affinity (type II) group. Consistent with this classification, 700 ns simulations showed similar dynamics for MICA*001:01 and MICA*112:02 that differed from MICA*004:001. These structural observations are hypothesis-generating only. They do not demonstrate altered NKG2D binding affinity, NK-cell or CD8+ T-cell activation, cytokine production, or immune recognition in patients.
The MICA*001:01:01-NKG2D complex demonstrated conformational homogeneity, with 100% of the trajectory classified into a single cluster with a cutoff of 0.16 nm. Cluster analysis revealed greater conformational dynamics in the MICA*004:01:01-NKG2D complex, which formed nine distinct clusters at the same cutoff. The presence of two dominant populations (Cluster 1 and Cluster 9), which were nearly equal in size (~30% and ~29% of the frames, respectively), suggests that the simulated MICA*004:01:01-NKG2D complex sampled two main conformational states. In contrast, the two amino acid differences in the MICA*112:02-NKG2D model were associated with a more stable trajectory that was dominated by one representative conformation. These computational differences should be interpreted as structural observations rather than evidence of altered biological function.
A detailed analysis of the MICA*001:01:01-NKG2D complex shows a specific network of hydrogen bonds, primarily involving acidic residues such as Asp (aspartic acid) on the MICA (Chain C) and basic or polar residues (Lys and Ser) on the NKG2D (Chains A and B). The binding of Chain A to Chain C is primarily stabilized by hydrogen bonds and salt bridges. Notable interactions include the polar coordination between His159(A) and Asp226(C), and the electrostatic coupling of Lys197(A) with Asp163(C); Ser172(A) forms a hydrogen bond with Asp219(C). Interaction analysis of the Chain B sub-unit reveals a distinct binding mode. Hydrogen bonds are observed between Tyr152(B) and Lys71(C), Arg64(C) and Arg74(C). Furthermore, a salt-bridge network is established by Asp115(B) with Arg64(C) and Arg74(C). After 700 ns of simulation, interactions of the two chains of NKG2D with MICA remain asymmetric.
The mutations in the MICA*112:02 allele do not change the binding pattern in the MICA*112:02-NKG2D complex. The loss of the Thr24 hydroxyl group and the reversal of the Lys125 charge cause a conformational adaptation. The interaction between Arg74 and Lys71 of MICA*112:02 forms persistent contacts with Tyr152, Gln185 and Glu201 of NKG2D. Tyr152(A) and Arg74(C) remain the factors in high structural convergence and populate a single conformational cluster.
The substitution of eleven amino acids in the MICA*004:01:01 allele induced significant structural heterogeneity, resulting in two dominant clusters (Cluster 1 and Cluster 9). The representative conformations of both clusters show a significant shift in the bonding network compared with the MICA*001:01:01-NKG2D complex. The loss of native contacts caused by a high mutational rate is compensated by the formation of alternative salt bridges and polar interactions. Key residues, including Arg64 and Asp149 of MICA*004:01:01, interact with Asp115 and Glu116 of NKG2D. The NKG2D interface in this alternative state is stabilized by Thr155 and His156 of MICA through coordinated interactions with Lys197 and Glu201 of NKG2D.
The comparative data indicate that Tyr152 of the NKG2D dimer and the basic residues Arg74 and Lys71 of MICA constitute important recognition elements required for assembly of the complex. The two-residue difference in MICA*112:02 appears structurally well tolerated and preserves key hydrogen-bonding interactions involving Tyr152 (on both Chains A and B) and Arg74 of Chain C. In contrast, the LigPlot analysis suggests that MICA*004:01:01 is associated with a broader reconfiguration of protein–receptor interactions. The transition from a lysine-rich cationic interface in MICA*001:01:01 to a more glutamate-rich anionic interface in MICA*004:01:01, together with loss of specific hydrogen bonds, is compatible with distinct simulated interface geometry. However, this modeling result does not demonstrate altered NKG2D binding affinity, cytotoxicity, cytokine production, or pro-inflammatory signaling.
NKG2D activity is influenced not only by MICA/MICB polymorphism but also by functional variation in KLRK1, which affects NKG2D expression and lymphocyte cytotoxicity [46,47]. In our cohort, one KLRK1 haplotype (G-A-G-T; rs1049174-rs2617160-rs2246809-rs2617170) showed an exploratory association with increased risk of CALR-mutated MPN and has previously been reported to confer higher cytotoxic activity in CD8+ T and NK cells [46,48]. This haplotype has also been linked to susceptibility to rheumatoid arthritis [49], consistent with the broader role of NKG2D in chronic inflammatory and autoimmune disorders [50] and with the inflammatory state in MPNs that contributes to disease progression [51,52]. Notably, functional studies on MPN have demonstrated quantitative and qualitative NK-cell defects and modulation by interferon-alpha or ruxolitinib therapy [53,54]. Together, these observations suggest that inherited NKG2D-related variation may interact with disease-associated NK-cell dysfunction, but direct functional assays are required to determine whether this translates into altered immune surveillance or inflammation in CALR-mutated MPN.
MICA/MICB expression represents an additional regulatory layer for the NKG2D axis. By reanalyzing scRNA-seq data from CALR-mutated ET and MF, we found that only a minority of HSCs, CMPs, and MEPs had detectable MICA or MICB transcripts, but the fraction and expression level were higher in CALR-mutant compared with wild-type cells. Because immune selection may extend beyond the stem/progenitor compartment, we analyzed an RNA-seq dataset of iPSC-derived megakaryocytes with engineered CALR mutations. Heterozygous CALR-mutant megakaryocytes exhibited higher MICA expression than controls, whereas MICB was not appreciably changed. These findings support the possibility that a subset of CALR-mutant megakaryocytes—and potentially platelets—may engage NKG2D-expressing lymphocytes independently of effective MHC-I presentation. This interpretation is consistent with observations of increased platelet-CD8+ T-cell aggregates in MPN and immunoregulatory platelet phenotypes enriched in CALR-mutant disease [55,56], but it remains to be validated at the protein and functional levels.
Overall, our data support a modest and exploratory link between the NKG2D-MICA/MICB axis and the immunogenetic landscape of CALR-mutated MPN. The association of MICA*004:001 and a KLRK1 haplotype, together with increased MICA/MICB transcription in CALR-mutant hematopoietic compartments, is compatible with a role for immune selection, but these findings do not establish a direct mechanistic contribution to disease pathogenesis. From a translational perspective, therapies that increase surface expression of NKG2D ligands or counteract ligand shedding could theoretically enhance innate immune recognition [57,58,59]. However, such approaches would require careful preclinical evaluation because excessive or chronic activation of NK-cell and CD8+ T-cell pathways may also contribute to inflammatory pathology. Thus, therapeutic modulation of this axis should be regarded as a testable hypothesis rather than a conclusion of the current study.
Several limitations should be considered when interpreting the present results. First, the CALR-mutated cohort was small (n = 43), and the KLRK1 analysis was performed in an even smaller subgroup (n = 35). Given the number of alleles, SNPs, and haplotypes tested, some nominal associations may represent false-positive findings. Although the MICA*004:001 allele remained significant after Bonferroni correction in the comparison with healthy controls, the haplotype-level findings should be considered exploratory and require validation in larger, independent cohorts with harmonized ancestry and clinical annotation. Second, protein-level and functional validation of the NKG2D-MICA/MICB axis was not available in the analyzed cohort. Specifically, soluble MICA and MICB concentrations were not measured, and viable patient-derived material was not available for systematic assessment of surface MICA/MICB expression, NK-cell or CD8+ T-cell cytotoxicity, degranulation, or cytokine production, including IL-6, IL-10, IL-2, and IFN-gamma. Therefore, although the genetic, structural, and transcriptomic data are compatible with involvement of this axis in CALR-mutated MPN, they do not directly prove a pro-inflammatory functional interaction between MICA/MICB and NKG2D-expressing lymphocytes. Future studies should integrate MICA/MICB and KLRK1 genotyping with plasma soluble-ligand quantification, flow-cytometric analysis of ligand expression in CD34+ and megakaryocytic compartments, and functional assays of NK-cell and CD8+ T-cell activation.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cancers18132052/s1, Supplementaty Table S1. Association of MICA alleles with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S2. Association of MICA exon 5 alleles with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S3. Association of MICA codon 129 alleles with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S4. Association of MICA alleles with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S5. Association of MICA exon 5 alleles with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S6. Association of MICA codon 129 alleles with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S7. Association of MICB alleles with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S8. Association of MICB alleles with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S9. Association of HLA-A~MICA haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S10. Association of HLA-A~MICB haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S11. Association of HLA-B~MICA haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S12. Association of HLA-B~MICB haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S13. Association of HLA-C~MICA haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S14. Association of HLA-C~MICB haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S15. Association of MICA~MICB haplotypes with CALRmut+ MPNs vs. healthy controls. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S16. Association of HLA-A~MICA haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S17. Association of HLA-A~MICB haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S18. Association of HLA-B~MICA haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S19. Association of HLA-B~MICB haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S20. Association of HLA-C~MICA haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S21. Association of HLA-C~MICB haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariables. Supplementaty Table S22. Association of MICA~MICB haplotypes with CALRmut+ MPNs vs. JAK2 V617F+ MPNs. Additive generalized linear model with age (continuous) and gender (binary) as covariable.

Author Contributions

V.S. proposed research, designed and performed research, provided important reagents, analyzed data and drafted this manuscript; G.T. provided important reagents and revised this manuscript; I.M. provided important reagents and revised this manuscript; E.H. provided important reagents and revised this manuscript; J.P. performed MD simulations, analyzed data and revised this manuscript; G.M. performed MD simulations, analyzed data and revised this manuscript; M.I. provided important reagents, performed experiments, analyzed data and drafted this manuscript. All authors have read and agreed to the published version of this manuscript.

Funding

This work was partly supported by an NSF grant (Bulgaria), project BG-175467353-2024-11-0053/KP-06-ПH83/3. The computational resources were provided by the Project № BG16RFPR002-1.014-0006 “National Centre of Excellence Mechatronics and Clean Technologies” (co-funded by the European Union under the “Research Innovation and Digitization for Smart Transformation” program 2021–2027).

Institutional Review Board Statement

This study was approved by the Institutional Ethics Committee at the Medical University—Sofia under protocol 05/20 March 2025. All subjects provided informed consent in accordance with the Declaration of Helsinki.

Informed Consent Statement

Not applicable. No personal identifying information is provided in this manuscript.

Data Availability Statement

Genotype data for Bulgarian subjects are available from the authors upon reasonable request and after signing a material transfer agreement. All other datasets are publicly available and can be accessed as described in the main text.

Acknowledgments

Thanks to A. Ivanova for fruitful discussions regarding MD simulations.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
ETEssential thrombocythemia
PMFPrimary myelofibrosis
PVPolycythemia vera
MPNMyeloproliferative neoplasm
NKNatural killer
HLAHuman leukocyte antigen
SNPSingle-nucleotide polymorphism
iPSCInduced pluripotent stem cell
MKMegakaryocyte
scRNA-seqSingle-cell RNA sequencing
MDSMolecular dynamics simulation

References

  1. Paz, D.L.; Kralovics, R.; Skoda, R.C. Genetic basis and molecular profiling in myeloproliferative neoplasms. Blood 2023, 141, 1909–1921. [Google Scholar] [CrossRef] [PubMed]
  2. Koschmieder, S.; I Mughal, T.; Hasselbalch, H.C.; Barosi, G.; Valent, P.; Kiladjian, J.-J.; Jeryczynski, G.; Gisslinger, H.; Jutzi, J.S.; Pahl, H.L.; et al. Myeloproliferative neoplasms and inflammation: Whether to target the malignant clone or the inflammatory process or both. Leukemia 2016, 30, 1018–1024. [Google Scholar] [CrossRef] [PubMed]
  3. Gigoux, M.; Holmström, M.O.; Zappasodi, R.; Park, J.J.; Pourpe, S.; Bozkus, C.C.; Mangarin, L.M.B.; Redmond, D.; Verma, S.; Schad, S.; et al. Calreticulin mutant myeloproliferative neoplasms induce MHC-I skewing, which can be overcome by an optimized peptide cancer vaccine. Sci. Transl. Med. 2022, 14, eaba4380. [Google Scholar] [CrossRef] [PubMed]
  4. Ivanova, M.; Tsvetkova, G.; Lukanov, T.; Stoimenov, A.; Hadjiev, E.; Shivarov, V. Probable HLA-mediated immunoediting of JAK2 V617F-driven oncogenesis. Exp. Hematol. 2020, 92, 75–88.e10. [Google Scholar] [CrossRef] [PubMed]
  5. Shivarov, V.; Tsvetkova, G.; Hadjiev, E.; Ivanova, M. The relevance of HLA class II genes in JAK2 V617F-positive myeloproliferative neoplasms. Hum. Immunol. 2023, 84, 199–207. [Google Scholar] [CrossRef] [PubMed]
  6. Shivarov, V.; Tsvetkova, G.; Micheva, I.; Hadjiev, E.; Petrova, J.; Ivanova, A.; Madjarova, G.; Ivanova, M. Differential modulation of mutant CALR and JAK2 V617F-driven oncogenesis by HLA genotype in myeloproliferative neoplasms. Front. Immunol. 2024, 15, 1427810. [Google Scholar] [CrossRef] [PubMed]
  7. Ivanova, M.; Tsvetkova, G.; Lessichkova, S.; Gesheva, N.; Hadjiev, E.; Shivarov, V. Exploration of the role of NKG2D ligands MICA and MICB in JAK2 V617F-positive myeloproliferative neoplasms. HLA 2023, 102, 168–178. [Google Scholar] [CrossRef] [PubMed]
  8. Pedersen, K.M.; Çolak, Y.; Ellervik, C.; Hasselbalch, H.C.; Bojesen, S.E.; Nordestgaard, B.G. Loss-of-function polymorphism in IL6R reduces risk of JAK2V617F somatic mutation and myeloproliferative neoplasm: A Mendelian randomization study. eClinicalMedicine 2020, 21, 100280. [Google Scholar] [CrossRef] [PubMed]
  9. Macedo, L.C.; Quintero, F.d.C.; Pagliari-E-Silva, S.; Pagnano, K.B.B.; Rodrigues, C.; de Alencar, J.B.; Sell, A.M.; Visentainer, J.E.L. Association of TNF polymorphisms with JAK2 (V617F) myeloproliferative neoplasms in Brazilian patients. Blood Cells Mol. Dis. 2016, 57, 54–57. [Google Scholar] [CrossRef] [PubMed]
  10. Feenstra, J.D.M.; Jäger, R.; Schischlik, F.; Ivanov, D.; Eisenwort, G.; Rumi, E.; Schuster, M.; Gisslinger, B.; Machherndl-Spandl, S.; Bettelheim, P.; et al. PD-L1 Overexpression Correlates with JAK2-V617F Mutational Burden and Is Associated with 9p Uniparental Disomy in Myeloproliferative Neoplasms. Am. J. Hematol. 2022, 97, 390–400. [Google Scholar] [CrossRef]
  11. Boneva, E.; Shivarov, V.; Ivanova, M. A Concise Review of the Role of the NKG2D Receptor and Its Ligands in Cancer. Immuno 2025, 5, 9. [Google Scholar] [CrossRef]
  12. Khoury, J.D.; Solary, E.; Abla, O.; Akkari, Y.; Alaggio, R.; Apperley, J.F.; Bejar, R.; Berti, E.; Busque, L.; Chan, J.K.C.; et al. The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Myeloid and Histiocytic/Dendritic Neoplasms. Leukemia 2022, 36, 1703–1719. [Google Scholar] [CrossRef] [PubMed]
  13. Ivanova, M.; Creary, L.E.; Al Hadra, B.; Lukanov, T.; Mazzocco, M.; Sacchi, N.; Ameen, R.; Al-Shemmari, S.; Moise, A.; Ursu, L.D.; et al. 17th IHIW component “Immunogenetics of Ageing”—New NGS data. Hum. Immunol. 2019, 80, 703–713. [Google Scholar] [CrossRef] [PubMed]
  14. Shivarov, V.; Ivanova, M.; Hadjiev, E.; Naumova, E. Rapid quantification of JAK2V617F allele burden using a bead-based liquid assay with locked nucleic acid-modified oligonucleotide probes. Leuk. Lymphoma 2011, 52, 2023–2026. [Google Scholar] [CrossRef] [PubMed]
  15. Pavlov, I.; Hadjiev, E.; Alaikov, T.; Spassova, S.; Stoimenov, A.; Naumova, E.; Shivarov, V.; Ivanova, M. Calreticulin Mutations in Bulgarian MPN Patients. Pathol. Oncol. Res. 2017, 24, 171–174. [Google Scholar] [CrossRef] [PubMed]
  16. Ivanova, M.I.; Shivarov, V.S.; Hadjiev, E.A.; Naumova, E.J. Novel multiplex bead-based assay with LNA-modified probes for detection of MPL exon 10 mutations. Leuk. Res. 2011, 35, 1120–1123. [Google Scholar] [CrossRef] [PubMed]
  17. Ivanova, M.; Al Hadra, B.; Yordanov, S.; Lesichkova, S.; Stoyanov, H.; Shivarov, V.; Deliverska, E. Associations of high-resolution-typing-defined MICA and MICB polymorphisms, and the levels of soluble MICA and MICB with Oral Squamous Cell Carcinoma in Bulgarian patients. J. Oral Pathol. Med. 2021, 50, 758–765. [Google Scholar] [CrossRef] [PubMed]
  18. Zheng, X.; Shen, J.; Cox, C.; Wakefield, J.C.; Ehm, M.G.; Nelson, M.R.; Weir, B.S. HIBAG—HLA genotype imputation with attribute bagging. Pharmacogenomics J. 2013, 14, 192–200. [Google Scholar] [CrossRef] [PubMed]
  19. Solé, X.; Guinó, E.; Valls, J.; Iniesta, R.; Moreno, V. SNPStats: A web tool for the analysis of association studies. Bioinformatics 2006, 22, 1928–1929. [Google Scholar] [CrossRef] [PubMed]
  20. Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A.E.; Berendsen, H.J.C. GROMACS: Fast, flexible, and free. J. Comput. Chem. 2005, 26, 1701–1718. [Google Scholar] [CrossRef] [PubMed]
  21. Li, P.; Morris, D.L.; Willcox, B.E.; Steinle, A.; Spies, T.; Strong, R.K. Complex structure of the activating immunoreceptor NKG2D and its MHC class I–like ligand MICA. Nat. Immunol. 2001, 2, 443–451. [Google Scholar] [CrossRef] [PubMed]
  22. Robinson, J.; Barker, D.J.; Marsh, S.G. 25 years of the IPD-IMGT/HLA Database. HLA 2024, 103, e15549. [Google Scholar] [CrossRef] [PubMed]
  23. DeLano, W.L. Pymol: An open-source molecular graphics tool. CCP4 Newsl. Protein Crystallogr. 2002, 40, 82–92. [Google Scholar]
  24. Li, H.; Robertson, A.D.; Jensen, J.H. Very fast empirical prediction and rationalization of protein pKa values. Proteins Struct. Funct. Bioinform. 2005, 61, 704–721. [Google Scholar] [CrossRef] [PubMed]
  25. Wallace, A.C.; Laskowski, R.A.; Thornton, J.M. LIGPLOT: A program to generate schematic diagrams of protein-ligand interactions. Protein Eng. Des. Sel. 1995, 8, 127–134. [Google Scholar] [CrossRef] [PubMed]
  26. Laskowski, R.A.; Swindells, M.B. LigPlot+: Multiple ligand–protein interaction diagrams for drug discovery. J. Chem. Inf. Model. 2011, 51, 2778–2786. [Google Scholar] [CrossRef] [PubMed]
  27. Nam, A.S.; Kim, K.-T.; Chaligne, R.; Izzo, F.; Ang, C.; Taylor, J.; Myers, R.M.; Abu-Zeinah, G.; Brand, R.; Omans, N.D. Gen-otyping of Transcriptomes links somatic mutations and cell identity. Nature 2019, 571, 355. [Google Scholar] [CrossRef] [PubMed]
  28. Van der Maaten, L.; Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res. 2008, 9, 2579–2605. [Google Scholar]
  29. Satija, R.; A Farrell, J.; Gennert, D.; Schier, A.F.; Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 2015, 33, 495–502. [Google Scholar] [CrossRef] [PubMed]
  30. Novershtern, N.; Subramanian, A.; Lawton, L.N.; Mak, R.H.; Haining, W.N.; McConkey, M.E.; Habib, N.; Yosef, N.; Chang, C.Y.; Shay, T.; et al. Densely Interconnected Transcriptional Circuits Control Cell States in Human Hematopoiesis. Cell 2011, 144, 296–309. [Google Scholar] [CrossRef] [PubMed]
  31. Blanco-Carmona, E. Generating publication ready visualizations for Single Cell transcriptomics using SCpubr. bioRxiv 2022. [Google Scholar] [CrossRef]
  32. Olschok, K.; Han, L.; de Toledo, M.A.; Böhnke, J.; Graßhoff, M.; Costa, I.G.; Theocharides, A.; Maurer, A.; Schüler, H.M.; Buhl, E.M.; et al. CALR frameshift mutations in MPN patient-derived iPSCs accelerate maturation of megakaryocytes. Stem Cell Rep. 2021, 16, 2768–2783. [Google Scholar] [CrossRef] [PubMed]
  33. Mootha, V.K.; Lindgren, C.M.; Eriksson, K.-F.; Subramanian, A.; Sihag, S.; Lehar, J.; Puigserver, P.; Carlsson, E.; Ridderstråle, M.; Laurila, E.; et al. PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat. Genet. 2003, 34, 267–273. [Google Scholar] [CrossRef] [PubMed]
  34. Subramanian, A.; Tamayo, P.; Mootha, V.K.; Mukherjee, S.; Ebert, B.L.; Gillette, M.A.; Paulovich, A.; Pomeroy, S.L.; Golub, T.R.; Lander, E.S.; et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA 2005, 102, 15545–15550. [Google Scholar] [CrossRef] [PubMed]
  35. Ivanova, M.; Shivarov, V. Characterisation of the novel MICA*112:02 allele by next generation sequencing. HLA 2024, 103, e15583. [Google Scholar] [CrossRef] [PubMed]
  36. Shivarov, V.; Ivanova, M. Characterisation of novel MICB alleles by next generation sequencing: MICB*055 and MICB*005:15. HLA 2024, 103, e15581. [Google Scholar] [CrossRef] [PubMed]
  37. Jarvis, R.; Patrick, E. Clustering Using a Similarity Measure Based on Shared Near Neighbors. IEEE Trans. Comput. 1973, C-22, 1025–1034. [Google Scholar] [CrossRef]
  38. Daura, X.; Gademann, K.; Jaun, B.; Seebach, D.; Van Gunsteren, W.F.; Mark, A.E. Peptide folding: When simulation meets experiment. Angew. Chem. Int. Ed. 1999, 38, 236–240. [Google Scholar] [CrossRef]
  39. Holmström, M.O.; Cordua, S.; Skov, V.; Kjær, L.; Pallisgaard, N.; Ellervik, C.; Hasselbalch, H.C.; Andersen, M.H. Evidence of immune elimination, immuno-editing and immune escape in patients with hematological cancer. Cancer Immunol. Immunother. 2020, 69, 315–324. [Google Scholar] [CrossRef] [PubMed]
  40. Muñoz-Saá, I.; Cambra, A.; Pallarés, L.; Espinosa, G.; Juan, A.; Pujalte, F.; Matamoros, N.; Mila, J.; Julià, M. Allelic diversity and affinity variants of MICA are imbalanced in Spanish patients with Behçet’s disease. Scand. J. Immunol. 2006, 64, 77–82. [Google Scholar] [CrossRef] [PubMed]
  41. Ďurmanová, V.; Tirpakova, J.; Stuchlikova, M.; Shawkatova, I.; Kuba, D.; Sapak, M.; Buc, M. Characterization of MICA gene polymorphism of HLA complex in the Slovak population. Ann. Hum. Biol. 2011, 38, 570–576. [Google Scholar] [CrossRef] [PubMed]
  42. Birtsas, V.; Batrinou, A.; Dinou, A.; Routsias, J.; Gennimata, V.; Iniotaki, A.; Spyropoulou, M.; Tsakris, A. Distribution of MICA alleles and haplotypes associated with HLA-B in Greek population. Hum. Immunol. 2021, 82, 588–592. [Google Scholar] [CrossRef] [PubMed]
  43. Koskela, S.; Tammi, S.; Clancy, J.; Lucas, J.A.M.; Turner, T.R.; Hyvärinen, K.; Ritari, J.; Partanen, J. MICA and MICB allele assortment in Finland. HLA 2023, 102, 52–61. [Google Scholar] [CrossRef] [PubMed]
  44. Xiao, C.; Wang, W.; Zhang, Y.; Huang, T.; Chen, C.; Liu, B.; Liu, C.; Yang, Y.; Lu, F.; Zhu, Q. Association Between Amino Acid Polymorphisms in MICA and MICA-NKG2D Interaction Affinity: Implications and Significance for Tumor Immunity. Biomolecules 2025, 16, 47. [Google Scholar] [CrossRef] [PubMed]
  45. Luo, Q.; Yin, X.; Zhu, Q.; Luo, W.; Liu, R.; Wei, L.; Zou, Y. Two major human phenotypes of MICA molecules and their differential activation to NK cells via NKG2D receptor. Front. Immunol. 2025, 16, 1563872. [Google Scholar] [CrossRef] [PubMed]
  46. Hayashi, T.; Imai, K.; Morishita, Y.; Hayashi, I.; Kusunoki, Y.; Nakachi, K. Identification of the NKG2D Haplotypes Associated with Natural Cytotoxic Activity of Peripheral Blood Lymphocytes and Cancer Immunosurveillance. Cancer Res. 2006, 66, 563–570. [Google Scholar] [CrossRef] [PubMed]
  47. Imai, K.; Hayashi, T.; Yamaoka, M.; Kajimura, J.; Yoshida, K.; Kusunoki, Y.; Nakachi, K. Effects of NKG2D haplotypes on the cell-surface expression of NKG2D protein on natural killer and CD8 T cells of peripheral blood among atomic-bomb survivors. Hum. Immunol. 2012, 73, 686–691. [Google Scholar] [CrossRef] [PubMed]
  48. Furue, H.; Matsuo, K.; Kumimoto, H.; Hiraki, A.; Suzuki, T.; Yatabe, Y.; Komori, K.; Kanemitsu, Y.; Hirai, T.; Kato, T.; et al. Decreased risk of colorectal cancer with the high natural killer cell activity NKG2D genotype in Japanese. Carcinogenesis 2008, 29, 316–320. [Google Scholar] [CrossRef] [PubMed]
  49. Mariaselvam, C.M.; Tamouza, R.; Krishnamoorthy, R.; Charron, D.; Misra, D.P.; Jain, V.K.; Negi, V.S. Association of NKG2D gene variants with susceptibility and severity of rheumatoid arthritis. Clin. Exp. Immunol. 2016, 187, 369–375. [Google Scholar] [CrossRef] [PubMed]
  50. Babic, M.; Romagnani, C. The Role of Natural Killer Group 2, Member D in Chronic Inflammation and Autoimmunity. Front. Immunol. 2018, 9, 1219. [Google Scholar] [CrossRef] [PubMed]
  51. Hasselbalch, H.C.; Bjørn, M.E. MPNs as Inflammatory Diseases: The Evidence, Consequences, and Perspectives. Mediat. Inflamm. 2015, 2015, 102476. [Google Scholar] [CrossRef] [PubMed]
  52. Koschmieder, S.; Chatain, N. Role of inflammation in the biology of myeloproliferative neoplasms. Blood Rev. 2020, 42, 100711. [Google Scholar] [CrossRef] [PubMed]
  53. Riley, C.H.; Hansen, M.; Brimnes, M.K.; Hasselbalch, H.C.; Bjerrum, O.W.; Straten, P.T.; Svane, I.M.; Jensen, M.K. Expansion of circulating CD56bright natural killer cells in patients with JAK2-positive chronic myeloproliferative neoplasms during treatment with interferon-α. Eur. J. Haematol. 2014, 94, 227–234. [Google Scholar] [CrossRef] [PubMed]
  54. Schönberg, K.; Rudolph, J.; Vonnahme, M.; Parampalli Yajnanarayana, S.; Cornez, I.; Hejazi, M.; Manser, A.R.; Uhrberg, M.; Verbeek, W.; Koschmieder, S.; et al. JAK Inhibition Impairs NK Cell Function in Myeloproliferative Neoplasms. Cancer Res. 2015, 75, 2187–2199. [Google Scholar] [CrossRef] [PubMed]
  55. Simões, A.M.C.; Holmström, M.O.; Aehnlich, P.; Rahbech, A.; Peeters, M.J.W.; Radziwon-Balicka, A.; Zamora, C.; Klausen, T.W.; Skov, V.; Kjær, L.; et al. Patients With Myeloproliferative Neoplasms Harbor High Frequencies of CD8 T Cell-Platelet Aggregates Associated With T Cell Suppression. Front. Immunol. 2022, 13, 866610. [Google Scholar] [CrossRef] [PubMed]
  56. Holmström, M.O.; Ruders, J.H.; Riley, C.H.; Larsen, M.K.; Grauslund, J.H.; Kjær, L.; Skov, V.; Ellervik, C.; Guo, B.B.; Linden, M.; et al. The CALR mutations enhance the expression of the immunosuppressive proteins GARP and LAP on peripheral blood lymphocytes through increased binding of activated platelets. Br. J. Haematol. 2024, 205, 1417–1429. [Google Scholar] [CrossRef] [PubMed]
  57. Ferrari de Andrade, L.; Tay, R.E.; Pan, D.; Luoma, A.M.; Ito, Y.; Badrinath, S.; Tsoucas, D.; Franz, B.; May, K.F., Jr.; Harvey, C.J. Antibody-mediated inhibition of MICA and MICB shedding promotes NK cell–driven tumor immunity. Science 2018, 359, 1537–1542. [Google Scholar] [CrossRef] [PubMed]
  58. Kumar, R.; Gupta, R. Epigenetic regulation of NKG2D ligand and the rise of NK cell-based immunotherapy for cancer treatment. Front. Oncol. 2024, 14, 1456631. [Google Scholar] [CrossRef] [PubMed]
  59. Wildschut, M.H.E.; Mena, J.; Dördelmann, C.; van Oostrum, M.; Hale, B.D.; Settelmeier, J.; Festl, Y.; Lysenko, V.; Schürch, P.M.; Ring, A.; et al. Proteogenetic drug response profiling elucidates targetable vulnerabilities of myelofibrosis. Nat. Commun. 2023, 14, 6414. [Google Scholar] [CrossRef] [PubMed]
Cancers 18 02052 g001
Figure 1. MICA allele distributions and associations in CALR-mutated MPN versus healthy controls. (A) Frequencies of identified MICA alleles in CALR-mutated MPN patients and healthy controls. (B) Odds ratios (ORs) for association of MICA alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model). (C) ORs for association of MICA exon 5-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model). (D) ORs for association of MICA codon 129-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model).
Figure 1. MICA allele distributions and associations in CALR-mutated MPN versus healthy controls. (A) Frequencies of identified MICA alleles in CALR-mutated MPN patients and healthy controls. (B) Odds ratios (ORs) for association of MICA alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model). (C) ORs for association of MICA exon 5-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model). (D) ORs for association of MICA codon 129-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model).
Cancers 18 02052 g001
Cancers 18 02052 g002
Figure 2. MICA allele distributions and associations in CALR-mutated versus JAK2 V617F-positive MPN. (A) Frequencies of identified MICA alleles in CALR-mutated MPN patients and JAK2 V617F-positive MPN patients. (B) Odds ratios (ORs) for association of MICA alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model). (C) ORs for association of MICA exon 5-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model). (D) ORs for association of MICA codon 129-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model).
Figure 2. MICA allele distributions and associations in CALR-mutated versus JAK2 V617F-positive MPN. (A) Frequencies of identified MICA alleles in CALR-mutated MPN patients and JAK2 V617F-positive MPN patients. (B) Odds ratios (ORs) for association of MICA alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model). (C) ORs for association of MICA exon 5-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model). (D) ORs for association of MICA codon 129-defined groups with CALR-mutated status (age- and sex-adjusted generalized linear model).
Cancers 18 02052 g002
Cancers 18 02052 g003
Figure 3. MICB allele distributions and associations in CALR-mutated MPN versus healthy controls. (A) Frequencies of identified MICB alleles in CALR-mutated MPN patients and healthy controls. (B) Odds ratios (ORs) for association of MICB alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model).
Figure 3. MICB allele distributions and associations in CALR-mutated MPN versus healthy controls. (A) Frequencies of identified MICB alleles in CALR-mutated MPN patients and healthy controls. (B) Odds ratios (ORs) for association of MICB alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model).
Cancers 18 02052 g003
Cancers 18 02052 g004
Figure 4. MICB allele distributions and associations in CALR-mutated versus JAK2 V617F-positive MPN. (A) Frequencies of identified MICB alleles in CALR-mutated MPN patients and JAK2 V617F-positive MPN patients. (B) Odds ratios (ORs) for association of MICB alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model).
Figure 4. MICB allele distributions and associations in CALR-mutated versus JAK2 V617F-positive MPN. (A) Frequencies of identified MICB alleles in CALR-mutated MPN patients and JAK2 V617F-positive MPN patients. (B) Odds ratios (ORs) for association of MICB alleles with CALR-mutated status (age- and sex-adjusted additive generalized linear model).
Cancers 18 02052 g004
Cancers 18 02052 g005
Figure 5. Volcano plots of haplotype associations with CALR-mutated MPN versus healthy controls. (A) MICA~MICB haplotypes. (B) HLA-B~MICA haplotypes. (C) HLA-C~MICA haplotypes. (D) HLA-C~MICB haplotypes.
Figure 5. Volcano plots of haplotype associations with CALR-mutated MPN versus healthy controls. (A) MICA~MICB haplotypes. (B) HLA-B~MICA haplotypes. (C) HLA-C~MICA haplotypes. (D) HLA-C~MICB haplotypes.
Cancers 18 02052 g005
Cancers 18 02052 g006
Figure 6. Molecular dynamics simulations of selected MICA variant molecules in complex with NKG2D. (A) Clustal-based alignment of protein sequences of selected MICA alleles (MICA*001:01:01, MICA*004:01:01, and MICA*112:02) from the IPD-IMGT/HLA database. The red arrows mark the key amino acid residues defining the two main subgroups of MICA molecules, as discussed in the main text. (B) RMSD values from molecular dynamics simulation of the MICA*001:01:01-NKG2D complex. (C) RMSD values from molecular dynamics simulation of the MICA*004:01:01-NKG2D complex. (D) RMSD values from molecular dynamics simulation of the MICA*112:02-NKG2D complex.
Figure 6. Molecular dynamics simulations of selected MICA variant molecules in complex with NKG2D. (A) Clustal-based alignment of protein sequences of selected MICA alleles (MICA*001:01:01, MICA*004:01:01, and MICA*112:02) from the IPD-IMGT/HLA database. The red arrows mark the key amino acid residues defining the two main subgroups of MICA molecules, as discussed in the main text. (B) RMSD values from molecular dynamics simulation of the MICA*001:01:01-NKG2D complex. (C) RMSD values from molecular dynamics simulation of the MICA*004:01:01-NKG2D complex. (D) RMSD values from molecular dynamics simulation of the MICA*112:02-NKG2D complex.
Cancers 18 02052 g006
Cancers 18 02052 g007
Figure 7. Representative MICA-NKG2D interface contacts observed during last 100 ns of productive molecular dynamics simulations. Hydrogen bonds and salt bridges were identified across the trajectories for MICA*001:01:01, MICA*004:01:01, and MICA*112:02 complexes.
Figure 7. Representative MICA-NKG2D interface contacts observed during last 100 ns of productive molecular dynamics simulations. Hydrogen bonds and salt bridges were identified across the trajectories for MICA*001:01:01, MICA*004:01:01, and MICA*112:02 complexes.
Cancers 18 02052 g007
Cancers 18 02052 g008
Figure 8. Expression of MICA and MICB in CALR-mutant hematopoietic compartments. (A) Two-dimensional scatter plot after t-SNE transformation of integrated scRNA-seq data from five CALR-mutated ET cases. (B) Dot plot summarizing MICA and MICB expression in HSCs, CMPs, and MEPs from ET cases stratified by CALR genotype. (C) Two-dimensional scatter plot after t-SNE transformation of integrated scRNA-seq data from five CALR-mutated MF cases. (D) Dot plot summarizing MICA and MICB expression in HSCs, CMPs, and MEPs from MF cases stratified by CALR genotype. (E) GSEA plots of enrichment of MHC-I antigen processing and presentation pathway genes in CALR-mutant versus CALR-wild type MKs. (F) GSEA plots of enrichment of MHC-I antigen processing and presentation pathway genes in CALR-mutant versus CALR-wild type MKs. (G) MICA and MICB expression in iPSC-derived megakaryocytes (MKs) (control, heterozygous CALR-mutant, and homozygous CALR-mutant). p-values are from two-sided t-tests. Statistical notation: ns, p ≥ 0.05; *, p < 0.05. Abbreviations: HSCs, hematopoietic stem cells; CMPs, common myeloid progenitors; MEPs, megakaryocyte–erythroid progenitors.
Figure 8. Expression of MICA and MICB in CALR-mutant hematopoietic compartments. (A) Two-dimensional scatter plot after t-SNE transformation of integrated scRNA-seq data from five CALR-mutated ET cases. (B) Dot plot summarizing MICA and MICB expression in HSCs, CMPs, and MEPs from ET cases stratified by CALR genotype. (C) Two-dimensional scatter plot after t-SNE transformation of integrated scRNA-seq data from five CALR-mutated MF cases. (D) Dot plot summarizing MICA and MICB expression in HSCs, CMPs, and MEPs from MF cases stratified by CALR genotype. (E) GSEA plots of enrichment of MHC-I antigen processing and presentation pathway genes in CALR-mutant versus CALR-wild type MKs. (F) GSEA plots of enrichment of MHC-I antigen processing and presentation pathway genes in CALR-mutant versus CALR-wild type MKs. (G) MICA and MICB expression in iPSC-derived megakaryocytes (MKs) (control, heterozygous CALR-mutant, and homozygous CALR-mutant). p-values are from two-sided t-tests. Statistical notation: ns, p ≥ 0.05; *, p < 0.05. Abbreviations: HSCs, hematopoietic stem cells; CMPs, common myeloid progenitors; MEPs, megakaryocyte–erythroid progenitors.
Cancers 18 02052 g008
Table 1. Demographic and clinical characteristics of study participants.
Table 1. Demographic and clinical characteristics of study participants.
ParameterGroupp-Value
CALR-Mutated MPNsJAK2 V617F-Positive MPNsHealthy Controls
Sex 0.2233 *
Male233861
Female204795
Total4385156
Age <2.2 × 10−16 †
Median667132
Range30–8732–9620–99
Diagnosis
ET1630NA<2.2 × 10−16 ‡
MF post-ET21NA
PMF1411NA
PV037NA
MF post-PV02NA
MPN; NOS114NA
Total4385NA
Abbreviations: ET, essential thrombocythemia; PMF, primary myelofibrosis; PV, polycythemia vera; MPN, NOS, myeloproliferative neoplasm, not otherwise specified; NA, not applicable. p-values: * chi-squared test across three groups; † Kruskal–Wallis test across three groups; ‡ chi-squared test across two groups.
Table 2. KLRK1 (NKG2D) haplotype frequencies and association with CALR-mutated MPN using SNPStats web-server analysis.
Table 2. KLRK1 (NKG2D) haplotype frequencies and association with CALR-mutated MPN using SNPStats web-server analysis.
Haplotype Frequency Estimation (n = 140)
rs1049174
(NKC-3)		rs2617160
(NKC-7)		rs2246809
(NKC-9)		rs2617170
(NKC-11)		TotalHealthy ControlsCALR-Mutated MPNsCumulative Frequency
1CTGC0.5630.58990.48480.563
2GAAT0.1850.20060.12780.7482
3GAGT0.1720.15640.22730.9197
4CTGT0.0440.02890.08470.9635
5GTGT0.0120.00750.03030.9757
6CAGC0.0080.004900.9836
7GTGC0.0070.00490.01450.991
8GTAT0.0060.006800.9969
9CAGT0.003NA0.01551
10CAAC0NA0.0151
Haplotype association with response (n = 140, adjusted by sex and age)
rs1049174
(NKC-3)		rs2617160
(NKC-7)		rs2246809
(NKC-9)		rs2617170
(NKC-11)		FrequencyOR (95% CI)p-Value
1CTGC0.5631---
2GAAT0.1860.85 (0.29–2.53)0.78
3GAGT0.1713.61 (1.16–11.26)0.029
4CTGT0.0443.75 (0.60–23.42)0.16
5GTGT0.0133.65 (0.15–91.62)0.43
Rare****0.0249.08 (0.96–86.05)0.057
The upper panel represents comparison of the frequencies in healthy controls and CALR-mutated MPN patients using chi-squared testing without adjustment. The lower panel represents comparison of the frequencies in healthy controls and CALR-mutated MPN patients using logistic regression adjusted for sex and age. p-values below 0.05 were considered statistically significant. Abbreviations: OR, odds ratio; NA, not applicable. Abbreviations: * any nucleotide.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Shivarov, V.; Tsvetkova, G.; Micheva, I.; Hadjiev, E.; Petkova, J.; Madjarova, G.; Ivanova, M. Immunogenetic and Transcriptomic Evidence Implicating the NKG2D-MICA/MICB Axis in CALR-Mutated Myeloproliferative Neoplasms. Cancers 2026, 18, 2052. https://doi.org/10.3390/cancers18132052

AMA Style

Shivarov V, Tsvetkova G, Micheva I, Hadjiev E, Petkova J, Madjarova G, Ivanova M. Immunogenetic and Transcriptomic Evidence Implicating the NKG2D-MICA/MICB Axis in CALR-Mutated Myeloproliferative Neoplasms. Cancers. 2026; 18(13):2052. https://doi.org/10.3390/cancers18132052

Chicago/Turabian Style

Shivarov, Velizar, Gergana Tsvetkova, Ilina Micheva, Evgueniy Hadjiev, Jasmina Petkova, Galia Madjarova, and Milena Ivanova. 2026. "Immunogenetic and Transcriptomic Evidence Implicating the NKG2D-MICA/MICB Axis in CALR-Mutated Myeloproliferative Neoplasms" Cancers 18, no. 13: 2052. https://doi.org/10.3390/cancers18132052

APA Style

Shivarov, V., Tsvetkova, G., Micheva, I., Hadjiev, E., Petkova, J., Madjarova, G., & Ivanova, M. (2026). Immunogenetic and Transcriptomic Evidence Implicating the NKG2D-MICA/MICB Axis in CALR-Mutated Myeloproliferative Neoplasms. Cancers, 18(13), 2052. https://doi.org/10.3390/cancers18132052

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop