Characterization of the canine mda-7 gene, transcripts and expression patterns

Abstract

Human melanoma differentiation associated gene-7/interleukin-24 (mda-7/IL-24) displays potent growth suppressing and cell killing activity against a wide variety of human and rodent cancer cells. In this study, we identified a canine ortholog of the human mda-7/IL-24 gene located within a cluster of IL-10 family members on chromosome 7. The full-length mRNA sequence of canine mda-7 was determined, which encodes a 186-amino acid protein that has 66% similarity to human MDA-7/IL-24. Canine MDA-7 is constitutively expressed in cultured normal canine epidermal keratinocytes (NCEKs), and its expression levels are increased after lipopolysaccharide stimulation. In cultured NCEKs, the canine mda-7 pre-mRNA is differentially spliced, via exon skipping and alternate 5'-splice donor sites, to yield five splice variants (canine mda-7sv1, canine mda-7sv2, canine mda-7sv3, canine mda-7sv4 and canine mda-7sv5) that encode four protein isoforms of the canine MDA-7 protein. These protein isoforms have a conserved N-terminus (signal peptide sequence) and are dissimilar in amino acid sequences at their C-terminus. Canine MDA-7 is not expressed in primary canine tumor samples, and most tumor derived cancer cell lines tested, like its human counterpart. Unlike human MDA-7/IL-24, canine mda-7 mRNA is not expressed in unstimulated or lipopolysaccharide (LPS), concanavalin A (ConA) or phytohemagglutinin (PHA) stimulated canine peripheral blood mononuclear cells (PBMCs). Furthermore, in-silico analysis revealed that canonical canine MDA-7 has a potential 28 amino acid signal peptide sequence that can target it for active secretion. This data suggests that canine mda-7 is indeed an ortholog of human mda-7/IL-24, its protein product has high amino acid similarity to human MDA-7/IL-24 protein and it may possess similar biological properties to human MDA-7/IL-24, but its expression pattern is more restricted than its human ortholog.

Keywords: Interleukin-24 (IL-24); Melanoma differentiation associated gene-7 (mda-7); cancer cell lines; canine.

Figure 1

Identification of splice variants of canine mda-7. Canine mda-7 splice variants were amplified by nested PCR and cloned into the pCDNA3.1+/Hygro plasmid vector. Recombinant clones were digested with HindIII and XhoI restriction enzymes to release the canine mda-7 inserts for mda-7sv3 (lane 2), mda-7sv1 (lane 3), mda-7sv2 (lane 4), mda-7sv4 (lane 5) and mda-7sv5 (lane 6). Empty vector (pCDNA3.1+/Hygro) was also digested with HindIII and XhoI as a control (lane 1). The samples were visualized by agarose gel electrophoresis. Lane M1 and M2 are 100 bp and 1kbp DNA ladders respectively.

Figure 2

Genomic structure of canine mda-7. The genomic structure of canine mda-7 was determined by comparing mRNA sequence with genomic sequences. The canine mda-7 locus is composed of eight exons and seven introns. The TATA box and transcription start site are labeled. Splice donor and acceptor sites are shown in bold. Alternative splice donor sites are located within exons 2 and 3. The start codon, stop codons and Poly-A signals are shown in boxes. The size of each intron is given in base pairs. Exon numbers are based on relative position and not on the corresponding human exon numbering. AU-rich elements (ARE) are underlined.

Figure 3

Alternative splicing at the canine mda-7 locus results in production of five splice variants. Splice variants range in size from 901- to 1179-bps. 5′ alternate acceptor sites for the second and third exon are used by sv2 and sv5. The fourth exon is skipped in the third splice variant, while exon five is only present in the sv4 and sv5 transcripts. Lines below the introns denote commonly used splices, while the lines above the introns denote unique splices. Curved arrow indicates the position of start codon.

Figure 4

Quantitative RT-PCR probe locations. TaqMan probes were designed against transcript specific sequences to detect sv2, sv3, sv4 and sv5. The copy number of canine mda-7sv1 was deduced by subtracting the sum of the copy number of sv2, sv3, sv4 and sv5 from the total copy number obtained from all transcripts PCR. Similarly, the copy number of sv4 was determined by subtracting copy number of sv5 transcript from copy number of sv4 transcript. Arrows indicate TaqMan PCR primers and bars indicate TaqMan probe positions.

Figure 4

Quantitative RT-PCR probe locations. TaqMan probes were designed against transcript specific sequences to detect sv2, sv3, sv4 and sv5. The copy number of canine mda-7sv1 was deduced by subtracting the sum of the copy number of sv2, sv3, sv4 and sv5 from the total copy number obtained from all transcripts PCR. Similarly, the copy number of sv4 was determined by subtracting copy number of sv5 transcript from copy number of sv4 transcript. Arrows indicate TaqMan PCR primers and bars indicate TaqMan probe positions.

Figure 5

Protein similarities between in human, canine and murine MDA-7. The in-silico predicted canine MDA-7 amino acid sequence was aligned with the amino acid sequences of c49a (rat), FISP (mouse) and human MDA-7/IL-24 using vector NTI 10.0. A maximum likelihood phylogenetic tree was drawn showing that canonical canine MDA-7 has high sequence similarity to human MDA-7/IL-24 protein.

Figure 6

Prediction of a signal peptide sequence for canine MDA-7 protein. Full-length mRNA sequence of canine mda-7sv1 was translated in-silico using vector NTI 10.0 software. The in-silico translated protein sequence representing the canonical canine MDA-7 protein was then analyzed by signal peptide predicting software (signalP3.0 NN). The canonical MDA-7 protein has a 28 amino acid long signal peptide with a predicted cleavage site between the 28^th and 29^th amino acids.

Figure 7

Alignment of canine MDA-7 protein isoforms with Human MDA-7/IL-24. Canine MDA-7 isoforms have a 28 amino-acid long predicted signal peptide sequence. The interleukin-10 signature motif is shown within the box. (*Conserved N-glycosylation site, ^Conserved cysteine amino acid).