Integrative Analysis of Proteomics and Transcriptomics Reveals Endothelin Receptor B as Novel Single Target and Identifies New Combinatorial Targets for Multiple Myeloma

Abstract

Despite the recent introduction of next-generation immunotherapeutic agents, multiple myeloma (MM) remains incurable. New strategies targeting MM-specific antigens may result in a more effective therapy by preventing antigen escape, clonal evolution, and tumor resistance. In this work, we adapted an algorithm that integrates proteomic and transcriptomic results of myeloma cells to identify new antigens and possible antigen combinations. We performed cell surface proteomics on 6 myeloma cell lines based and combined these results with gene expression studies. Our algorithm identified 209 overexpressed surface proteins from which 23 proteins could be selected for combinatorial pairing. Flow cytometry analysis of 20 primary samples confirmed the expression of FCRL5, BCMA, and ICAM2 in all samples and IL6R, endothelin receptor B (ET_B), and SLCO5A1 in >60% of myeloma cases. Analyzing possible combinations, we found 6 combinatorial pairs that can target myeloma cells and avoid toxicity on other organs. In addition, our studies identified ET_B as a tumor-associated antigen that is overexpressed on myeloma cells. This antigen can be targeted with a new monoclonal antibody RB49 that recognizes an epitope located in a region that becomes highly accessible after activation of ET_B by its ligand. In conclusion, our algorithm identified several candidate antigens that can be used for either single-antigen targeting approaches or for combinatorial targeting in new immunotherapeutic approaches in MM.

Figure 1.

Recently developed combinatorial strategies. These strategies require that 2 antigens are present to activate immune effector cells. (A) They are based on the activation of costimulatory pathways in CAR-T cells; (B) the use of hemibodies with alignment of the CD3 binding heavy and light chain variable domains; and (C) the coadministration of bispecific antibodies that bind to a costimulatory receptor. CAR = chimeric antigen receptor.

Figure 2.

Combinatorial target identification strategy. (A) Analysis workflow. (B) The consensus tissue list. (C) MA plots (mean vs ratio) illustrating the identified DHE genes in different contrasts. DHE genes are genes with higher log ratios of mean + 2 SD and having higher expression than average expression among all patients. (D) Venn diagram representing the overlap of identified DHE genes in different contrasts. A = average expression; BM BC = bone marrow B cell; DHE = differentially highly expressed; M = log ratio; MM = multiple myeloma; PB BC = peripheral blood B cell; PC = plasma cell.

Figure 3.

Tissue expression and pairing of candidate and existing immunotherapeutic targets. (A) Viable pairs between candidate proteins/genes and existing immunotherapeutic targets (left) as well as self-combinations (right). Viable pairs are selected based on having no expression in vital tissues and at most low expression in nonvital tissues as pairs. (B) Tissue expression of candidate proteins/genes (top) and existing immunotherapeutic targets (bottom).

Figure 4.

Selection of ideal combinatorial pairs. (A) The frequencies of expression of candidate genes on MM PC from the bone marrow of MM patients and on normal PC, normal CD34⁺ HSC, normal T and NK cells (T-NK) from the BM of healthy donors. (B) Tissue expression distributions of the genes in the top combinatorial target pairs. (C) The combined expression levels for each possible pair in MM cells. BM = bone marrow; HSC = hematopoietic stem cells; MM = multiple myeloma; PC = plasma cells.

Figure 5.

Median expression (scaled) profiles of genes among transcriptional and cytogenetic patient categories. (A) Median expression profiles of candidate and immunotherapeutic target genes over UAMS categories in the IFM dataset (GSE83503). (B) Median expression profiles of candidate and immunotherapeutic target genes over cytogenetic categories in the COMMPASS dataset. In both datasets, the data is min-max scaled using 99th percentile as the maximum value.

Figure 6.

The expression profiles of ET_B gene at proteomic and transcriptional level as well as in different patient subgroups. (A) Transcriptomic expression of ET_B (EDNRB) at single-cell level in bone marrow shown among myeloma plasma cells (left) obtained from 13 MM subjects, and CD38⁺ immune cells (right) obtained from 13 MM and 5 healthy subjects. (B) Expression profile among different cytogenetic categories in COMMPASS dataset. (C) Flow cytometry analysis conducted on the bone marrow sample from MM patient 10, as a sample. The population in red represents CD38⁺ MM plasma cells. (D) Results of the survival analysis conducted by SurvivalGenie. Patients in the COMMPASS dataset are divided into low- and high-expression categories using the cutp option (left) and survival analysis is performed (right). (E) Histogram illustrating the flow cytometry results with the control IgG1k (in blue) and RB49 (in red), confirming the expression of ET_B. ETB = endothelin receptor B; MM = multiple myeloma.