A novel unconventional antigen MPD5 elicits anti-tumor humoral immune responses in a subset of patients with polycythemia vera

Abstract

In an effort to define the antigenic mechanism that contributes to beneficial therapeutic outcome in patients with polycythemia vera (PV), we screened a human testis cDNA library with serological cloning derived from sera of three PV patients who had undergone therapeutic-induced remission. As a result, we identified a novel antigen, MPD5, which belongs to the group of cryptic antigens with unconventional genomic intron/exon structure. Moreover, MPD5 elicited IgG antibody responses in a subset of PV patients who had benefited from a variety of therapies--including IFN-alpha, Hydroxyurea, Imatinib mesylate, Anagrelide, and phlebotomy--but not in untreated PV patients or healthy donors, suggesting that MPD5 is a PV-associated, therapy-related antigen. In the granulocytes of PV patients who are responsive to therapy, upregulated MPD5 expression may serve to enhance immune responses. These findings provide new insight into the mechanism underlying regulation of the self-antigen repertoire that elicits anti-tumor immune responses in patients with myeloproliferative diseases, indicating the potential of these self-antigens as targets of novel immunotherapy.

Fig. 1. The novel PV-associated SEREX antigen, MPD5, is an unconventional antigen

A. A schematic representation of the location of the unconventional antigen, MPD5, in the complementary (i.e., antisense) strand within the intron 1 region of the NEK6 (NIMA-6) gene (GenBank accession number: NT_00847016). B. The structure of MPD5 promoter and the structure of MPD5 mRNA transcript. MPD5 has an ORF of 144 bp (GenBank accession number: AY567967). Four interferon regulatory factor 1 (IRF-1) binding sites—including NF-KB, AP1, and other transcription factor binding sites—were found in the 2kb region of the MPD5 promoter. These results suggest that MPD5 transcription may be regulated by interferon-α stimulation. C. The in vitro transcription and translation product of MPD5. D. The protein sequence of MPD5. E. The hydrophilicity index plot of MPD5, analyzed by the Kyte-Doolittle method, suggests that there are hydrophilic regions within the N-terminal and C-terminal regions of MPD5.

Fig. 2. The expression of MPD5 transcripts is elevated in tumors and PV

A. The expression of MPD5 transcripts in normal tissues. Lanes N1 to N10 indicate various normal tissues in the brain (N1), liver (N2), placenta (N3), small intestine (N4), colon (N5), thymus (N6), spleen (N7), prostate (N8), testis (N9), and ovary (N10), respectively. The transcript sizes are indicated as kilobases (kb). B. Expression of MPD5 transcripts in tumor cells, as detected by Northern blot. Lanes T1 to T10 indicate various tumor cells in acute T cell leukemia (Jurkat cells) (T1), Burkitt’s lymphoma (CA46) (T2), breast cancer (MDA-MD-453) (T3), Burkitt’s lymphoma (Namalwa) (T4), epidermal carcinoma (A-431) (T5), uterine carcinoma (MES-SA) (T6), Burkitt’s lymphoma (Raji) (T7), osteosarcoma (MG-63) (T8), histiocytic lymphoma (U-937) (T9), and cervical adenocarcinoma (Hela S3) (T10), respectively. In the lower panel, the specificity of the NEK6 intron 1 probe (outside of MPD5) was confirmed by Southern blot of an agarose gel analyzing a NEK6 intron 1 fragment that was located outside the MPD5 region and had no overlap with MPD5. C. The expression of MPD5 transcripts in the granulocytes from PV patients and healthy donors, as detected by quantitative RT-PCR. The expression levels of MPD5 transcripts are expressed as the ΔCT (MPD5-18S). Low ΔCT values indicate higher expression of the specific gene. The groups, whose expression of MPD5 transcripts is statistically different from that of healthy donors (p<0.05), are marked with an *. D. The MPD5 expression in K562 cells stimulated with IFN-α, as detected by semi-quantitative RT-PCR (the left panel). RT-PCR of β-actin served as a housekeeping control for IFN-α stimulation. The RT-PCR of ISG15 was used as a positive control for IFN-α stimulation. In the right panel, the densitometric units were calculated by normalizing the densities of the PCR products of MPD5 and ISG15 with those of β-actin in the same sample.

Fig. 3. MPD5 peptide specifically reacts to sera from PV patients responding to therapies

IgG antibody responses to the C-terminal antigenic epitope (from aa 30 to aa 47) of MPD5 were detected by peptide ELISA. These experiments were repeated three times, and representative results are shown. The mean plus three standard deviations (SD) of the OD405 peptide ratios over the coating control (derived from 23 healthy donors) was calculated as the upper limit of the normal range of MPD5 peptide antibody responses (the mean + 3SD = 1.38). The groups with detection rates of IgG antibody responses to MPD5 peptide that are statistically higher than those of healthy donors (the chi-square goodness-of-fit test; p<0.05) are marked with an *.