Ecto-calreticulin expression in multiple myeloma correlates with a failed anti-tumoral immune response and bad prognosis

Abstract

Immunogenic cell death (ICD) has been proposed to be a crucial process for antitumor immunosurveillance. ICD is characterized by the exposure and emission of Damage Associated Molecular Patterns (DAMP), including calreticulin (CRT). A positive correlation between CRT exposure or total expression and improved anticancer immunosurveillance has been found in certain cancers, usually accompanied by favorable patient prognosis. In the present study, we sought to evaluate CRT levels in the plasma membrane of CD38⁺ bone marrow mononuclear cells (BMMCs) isolated from 71 patients with varying degrees of multiple myeloma (MM) disease and examine the possible relationship between basal CRT exposure and the bone marrow immune microenvironment, as well as its connection with different clinical markers. Data show that increased levels of cell surface-CRT were associated with more aggressive clinical features and with worse clinical prognosis in MM. High CRT expression in MM cells was associated with increased infiltration of NK cells, CD8⁺ T lymphocytes and dendritic cells (DC), indicative of an active anti-tumoral immune response, but also with a significantly higher presence of immunosuppressive Treg cells and increased expression of PD-L1 in myeloma cells.

Keywords: Calreticulin; immunogenic cell death; multiple myeloma.

Figure 1.

Analysis of CRT exposure in patients with plasma cell dyscrasias. (a) Ecto-CRT levels were determined in CD38⁺ BMMCs isolated from clinically diagnosed patients with varying degrees of MM disease (MGUS, SMM, MM). Control BM (HD) was obtained from individuals with other pathologies that held an uncompromised BM evidenced by histological or flow cytometry examination. Statistical analysis was performed using one-way ANOVA test with Tukey post test, where *p < .05, **p < .01, ***p < .001, ns, non-significant. (b) Time to progression (TTP) was compared between CRT^high and CRT^low subgroups. Data are shown as median with interquartile range. Statistical analysis was performed using unpaired t test with Welch’s correction (*p < .05). (c) The expression of ecto-CRT in patients showing or not extramedullary plasmacytomas was analyzed. Statistical analysis was performed using Fisher’s exact test, where *p < .05, **p < .01, ***p < .001, ns, non-significant. (d) Kaplan–Meier curve comparing the survival of patients in the CRT^high and CRT^low groups. Overall survival of patients with overt MM was calculated from the time of sample retrieval and CRT analysis. HR was obtained by the univariate log-rank (Mantel-Cox) test. (e) Multivariate Cox regression analysis of the indicated variables. (f) Patients were stratified in two groups (treated and untreated) depending on whether they have previously received any chemotherapeutic regimen or not at the moment when BM sample was collected. Ecto-CRT levels were compared between these two groups. Statistical analysis was performed using two-way t test, where *p < .05, **p < .01, ***p < .001, ns, non-significant. (g) Samples were stratified in two groups based on their median ecto-CRT levels and number of total lines of therapy patients received was compared between CRT^high and CRT^low subgroups. Statistical analysis was performed using Chi-square test for trend (*p < .05, **p < .01, ***p < .001) Left: global analysis; Right: samples from overt MM patients. (h) Individuals with MM or MGUS/SMM diagnosis were stratified depending on their cytogenetic profile and ecto-CRT levels were compared between these two groups. Bars indicated median with interquartile range. Statistical analysis was performed using Mann-Whitney test. (i) Relationship between CRT exposure and the presence of different cytogenetic abnormalities with varying degrees of risk. Statistical analysis was performed by comparing each of the different groups harboring cytogenetic abnormalities with the normal group, using Kruskal-Wallis test where *p < .05.

Figure 2.

CRT exposure and the BM immune landscape in MM. Patients were stratified in two different cohorts based on their median ecto-CRT levels. Individuals with an equal or inferior value to the median CRT exposure cutoff of the entire population were deemed CRT^low, while individuals with an ecto-CRT expression level over the median cutoff were classified under CRT^high group. (a) BM frequencies of CD4⁺CD3⁺ T cells, CD8⁺CD3⁺ T cells, CD3⁺ T lymphocytes, as well as the percentage of CD4⁺ and CD8⁺ T cells within CD3⁺ population and the corresponding CD4⁺/CD8⁺ ratios were analyzed by flow cytometry. (b) The percentage of CD56⁺CD45⁺ NK cells in BM was determined by flow cytometry in the CRT^high and CRT^low groups. (c) Plasmacytoid lineage-negative/HLA-DR⁺/CD123⁺ DCs (pDCs) and myeloid lineage-negative/HLA-DR⁺/CD11c⁺ DCs (mDCs) subpopulations were also analyzed. Lineage-negative cells (Lin-) were determined by staining with FITC-conjugated antibodies against the following markers: CD3, CD14, CD16, CD19, CD20 and CD56. Statistical analysis was performed using two-way t test (*p < .05, **p < .01, ***p < .001). Data plotted in the figure represent the frequency of each cell type in individual samples. The mean and SD values in each patient cohort are also illustrated.

Figure 3.

CRT exposure and immunosuppressive markers in MM. Patients were stratified in two different cohorts based on their median ecto-CRT levels. (a) The BM frequencies of Tregs were analyzed in CRT^high and CRT^low samples. (b) PD-L1 expression in CD38⁺ BM cells was analyzed and compared between CRT^high and CRT^low groups. (c-f) PD-1 expression was analyzed in the indicated immune subpopulations and compared in patient’s subgroups based on their ecto-CRT expression levels. Statistical analysis was performed using two-tail t test (*p < .05, **p < .01, ***p < .001). Data plotted in the figure represent the frequency of each cell type in individual samples. Global mean and SD in each cohort are also illustrated.