The Autoproteolysis Of SEA Module Within Carboxy-terminal Domain Of RMuc3 Determines Its Functional Composition

Background and objectives:In a previous paper, we presented evidence that the expressed C-terminal domain of the rodent membrane mucin Muc3 (construct p20) undergoes proteolytic cleavage between the glycine and serine within the LSKGSIVV amino acid sequence during an early period of biosynthesis in the endoplasmic reticulum (ER). The 30 kDa N-terminal cleavage fragment is not secreted, but remains associated with the 49 kDa C-terminal membrane tethered fragment by non-covalent disulphide bond-independent interactions. The biological purposes of the cleavage and the association of the fragments are not understood, but are possibly important for the later release of the soluble extracellular domain into the intestinal lumen. Then, we found that cleavage within the SEA module (sea-urchin sperm protein, enterokinase and agrin module) of rat Muc3 requires participation of peptide sequences located C-terminal of and distant from the LSKGSIVV cleavage site, and association of the fragments requires the SEA module, but is independent of N-linked oligosaccharides. Further, we showed that the 49 kDa membrane-anchored fragment undergoes a further cleavage reaction, which decreases its size to 30 kDa. Western blotting, pulse–chase metabolic incubations, immunoprecipitation and deglycosylation with N-glycosidase F were used to detect and identify the proteolytic products. Both the first and second cleavages are presumed to facilitate solubilization of Muc3 at the apical surface of enterocytes and/or enhance the potential for Muc3 to participate in ligand–receptor and signal transduction events for enterocyte function in vivo. But it is still unclear how the proteolytic cleavage in the LSKGSIVV motif occurred, whether it was caused by protease or autoproteolysis. This kind of data is important to the SEA module-containing protein. Recently, the cleavage within the SEA module of human MUC1 was identified as an autocatalytic reaction, and the mechanism for this cleavage was elucidated However, the sequence similarity between the 104 amino acids that constitute the human MUC1 SEA module and the 174 residues of the rat Muc3 SEA module is only 6.9% (12/174). Due to the lack of sequence homology between the MUC1 SEA module and the rat Muc3 SEA module, the finding of the cleavage reaction in the MUC1 SEA module cannot be extrapolated to other mucins. This study was designed to demonstrate the mechanisms related to our previously described proteolysis within the SEA module of rodent Muc3 and decipher its rationale of this kind of posttranslatinal modification.Methods:1.Expression of the interested protein: DNA encoding residus of 381 amino acids within carboxyl terminal domain of rodent Muc3 obtained by PCR templated from p20 and p20G/A which were described previously(ref. 26 ) were inserted into the pQE30 prokaryotic expression vector with a coding for an N-terminal 6×His tag and the construction of prokaryotic expression of the rodent Muc3 carboxy-terminal domain was designated as pQE30-Muc3 and pQE30-Muc3(g/a). M15 cells were transfected by the pQE30-Muc3 or pQE30-Muc3(g/a). Cells were cultured in LB medium with 100mg/L Ampicilin and 50 mg/L kanamycin and induced at OD600=0.5 with 1mM IPTG and continued for 6h at 37℃.The product from pQE30-Muc3(g/a) transfected bacteria was purified by Ni-NTA agarose and the purified protein was incubated in 37℃for 4, 8, 16, 24, 36 and 48 hours in PBS separately.The incubated protein was subject to SDS/PAGE (12%) and detected by anti-V5 or anti-Myc antibodies (anti-V5, 1:2500; anti-Myc, 1:1000).2.Site-directed mutation: Sequence alignment of the SEA modules was performed by ClustalX. The conserved amino-acid residues were 174th serine,201th thioserine,212th tyrosine and 223th tyrosine, then mutated to stop coden by PCR templated with p20SEA to produce a series of new constructions which produced truncated forms of rMuc3. COS-7 cells were transfected transiently with these plasmids to produce different truncated SEA module of rodent Muc3 in order to find new and unidentified polypeptide sequences which located at the 79 amino acids, C-terminal end of SEA module of rMuc3 and were critical to keep its proper conformation and the occurrence of autoproteolysis.3.Molecular Modelling: The molecular modelling of rMuc3 SEA module was performed in Homology Model Block of SGI Insight II software package, SWISS MODEL WORKSPAC based on the solution structures of SEA domain from the murine homologue of ovarian antigen CA125 (MUC16) and membrane-bound MUC1 mucin identified by NMR.4.Cell culture and transfection experiments: Lovo cells (Shanghai Academia Sinica Life Science Research Institute) were cultured in F12K supplemented with 10% (v/v) FBS. cells were seeded into 3.5-cm-diameter tissue-culture dishes at a density of approx. 8×106 cells/dish. At 40-50% confluence, DNA transfections in cells were carried out with 2μg of plasmid and 10μl of LIPOFECT2000 in the presence of F12K without FBS. pSecTag2 transfection served as a control in each experiment. The hygromycin -resistant colonies were isolated by the ring cloning method, expanded, and maintained in medium supplemented with 100ug/mL hygromycin .PCR,Western blotting and membrane-targetting experiments were to identify the successful transfection.5.Cell Cycle Analysis: Lovo cells were transfected with p20, p20G/A, p20S/A or the empty vector, pSec, respectively. The transfected cells were incubated at 37°C, 5%CO2 for 48h. Cells were washed once in PBS, fixed in 90% Methanol, and stained in PI buffer containing Propidium Iodide, RNase A (Invitrogen), 0.1% Triton X-100 in PBS. Analysis was performed on flow cytometry and G2/G1 ratios were calculated.6.Cell migration assay: Lovo cells grown to 90%-100% confluency in 24-well plates were cultured overnight in serum-free medium. The medium was replaced with PBS, and the monolayers were wounded mechanically using a steriled transferpettor tip. After wounding, cells were rinsed twice with PBS and further incubated in F12K medium without serum for 24h at 37℃, 5%CO2. Those cells that had migrated from the wounded edge were counted at 100×, using an inverted light microscope. Five successive fields were counted and averaged within 3 well.The assay was repeated three times.8.Cell invasion assay: Motility and invasion capability in vitro were measured in transwells chambers assay. 100μl diluted Matrigel gel solution was put into upper chambers of the transwell inserts (6.5 mm, 8μm pore size; Costar Inc., USA). Incubated the inserts at 37°C for 4 h for gelling and then pretreated with serum-free medium at 37°C for 1 h before seeding cell at a density of 1×105 per well in 100μl medium with 1% FBS. The lower chambers of the transwells were filled with 500μl medium containing 10% FBS. The transwells were then incubated at 37°C with 5% CO2 for 24 h to allow cells to migrate. At the end of incubation, the cells on the upper side of the insert filter were completely removed by wiping with cotton swab. Cells that had invaded through the Matrigel-coated filter were stained with HE. Cells that had invaded the Matrigel and reached the lower surface of the filter were counted under a inverted microscope of 200×. Five fields of vision were chosen and the numbers of the invaded cells at the lower surface of the filter were counted, and the results from three separate chambers were then averaged. The assay was performed in triplicate.Results:1. The carboxy-terminal domain of rodent Muc3 were expressed in E. coli at 37℃induced by IPTG at 100mmol/l. The products contained the 55kDa, full-length, carboxy-terminal domain of rodent Muc3 recognized by both anti-V5 and anti-Myc antibodies, 49 kDa C-terminal fragment recognized only by anti-Myc antibody, but not by anti-V5 antibody, and 30kDa N-terminal fragment recognized only by anti-V5 antibody, but not by anti-Myc antibody, which indicated that the carboxy-terminal domain of rodent Muc3 was cleaved in bacteria as same as in the eukaryotic cells. Based on the general knowledge, it is impossible for the bacteria to have the specific protease to cleave the carboxy-terminal domain of rodent Muc3, a protein present in rat, not in the bacteria. This data provides a primary evidence to the autoproteolysis of the SEA module of the rMuc3.2. Products from pQE30-Muc3(g/a) was purified and incubated in PBS at 37℃, then deteced by Western blot with anti-V5 antibody. The main products after 4h or 8h incubation were the 55kDa full length one. After 16h incubation, the 55kDa full-length products decreased and the cleaved 30kd N-terminal fragment increased dramatically. With further incubation, the full-length products disappeared, only the 30kd N-terminal fragment existed. This data confirmed that the carboxy-terminal domain of rodent Muc3 was undergone the further cleavage without any other protein influence. The N-terminal 30 kDa fragment was the only cleaved fragment during the different time of incubation, it excluded the possibility of the degradation of the purified carboxy-terminal domain of rodent Muc3 from eukaryotic cells. So we concluded the cleavage in the SEA module in the carboxy-terminal domain of rodent Muc3 was based on the autoproteolysis.3. The conserved amino acids including S174,C201,Y212,Y223 in the SEA module were mutated to stop coden, respectively, and then COS-7 cells were transfected transiently with a series of mutated plasmids which produced a truncated forms of rMuc3 carboxyl terminal domain and secreted into the spent media. The spent media were detected by Western blotting with anti-V5 antibody. The cleaved 30kd N-terminal fragments were detected in 201C,212Y,223Y mutated constructs. So the amino acid sequence between 174 and 201 is critical to keep the proper comformation of the SEA module of rMuc3 and guarantes the occurrence of the autoproteolysis of the SEA module of rMuc3.4. The SEA module consists of a four-stranded antiparallelβ-sheets and fourα-helices occurring in the order ofβ1→α1→α2→β2→β3→α3→α4→β4 and is similar to that of human MUC1. The cleavage site is located in the turn betweenβ2 andβ3.The sequence between 174S and 201C resides in the turn formed byα4 andβ4, and is near to the site of cleavage. It is the possible reason for the sequence between 174S and 201C resides to affect the autoproteolysis. The molecular modelling of the SEA module of rMuc3 provides a structural information to understand the autoproteolysis of the SEA module cleavage.5. The introduction of the carboxy-terminal domain of rodent Muc3 into the Lovo cells (which had a different splicing variant of MUC3 and no cytoplasmic tail of MUC3, and definitely affected the function of MUC3 in Lovo cells) drove more cells into the G2/M phase than the other groups measured by FACS (p<0,05); Cells transfected with p20 showed an evident increase in cell migration and invasion over 24h compared with cells transfected with p20G/A,p20S/A and pSec vecor or non-transfected. The data indicated that the autoproteolysis of the SEA module of rMuc3 controlled its function.Conclusion:1. Our studies indicate that the SEA module within carboxyl terminal domain of rodent Muc3 undergoes autoproteolysis.2. The amino-acid residues between 174S and 201C may be critical to for the autoproteolysis.3. The autoproteolysis of the SEA module of rMuc3 determined its functional composition. The autoproteolytic rMuc3 C-terminal domain may mediate cell proliferation, stimulates cell migration and modulates cell invasion in vitro.