Figure 3: Structure and expression of human HEG1.

(a) Structure of open reading frame of human HEG1 gene. HEG1 gene sequences were retrieved from GenBank (accession no. NC_000003, region 124965710–125055958). The positions of SNPs in HEG1 of ACC-MESO-4 are indicated with the reference SNP ID number at the bottom. The domain structure of HEG1 was predicted as follows: signal peptides (exon 1), residues 1–29; proline rich domain (Pro rich) (exon 1), 30–106; serine/threonine rich region (Ser/Thr rich) (exon 2–5), 107–530; Ser/Thr rich (exon 6), 531–630; Ser/Thr rich (exon 7), 631–1086, including poly-serine sequences (polySer), 759–772; EGF domain 1 (typical EGF motif) (exon 8), 1087–1125; EGF domain 2 (Ca²⁺-binding EGF motif) (exon 9), 1126–1165; unknown region (UN) (this domain may contain a potential proteolytic cleavage site) (exon 10–12), 1166–1273; EGF domain 3 (this domain may form a laminin-type EGF-like domain with a following extracellular juxtamembrane region (EJM) 1) (exon 13), 1274–1318; EJM 1 (EJMs consist of a short loop formed by a disulfide-bridge) (exon 14), 1319–1332; EJM 2 (exon 15), 1333–1345; transmembrane domain (TM) (exon 16), 1346–1370; cytoplasmic domain (exon 16–18), 1371–1481. (b) Representative image of RT-PCR (n = 2) of HEG1 with or without exon6. PCR products were resolved at the length of 600 bp or 900 bp. (c) Real-time quantitative PCR of HEG1 in human organs. Values were normalized to the amount of β-actin, and are reported as means ± S.D. of tetraplicate determinations. Similar results were obtained in 2 independent experiments. (d) Glycosylation analysis of HEG1. Purified HEG1 (Fig. 2h) was used as a sample. The signal level was estimated by subtracting the fluorescence intensity of the control without antigen from that of the sample containing SKM9–2 antigen. Original images are shown in Supplementary Fig. 5. (e) Predicted structure of O-glycan attaching to HEG1.