| Objective:Corin is a cardiac type II transmembrane serine protease. The N-terminaltransmembrane domain anchors corin on the cell surface. Corin is synthesized as aninactive single-chain proenzyme (zymogen) in the ribosome, and modified in theendoplasmic reticula and Golgi apparatus. Corin is not activated until it is expressed onthe cell surface.It has been reported that specific motifs in the membrane-proximal region play animportant role in cell surface targeting and zymogen activation for some transmembraneproteases. Corin has a short cytoplasmic tail at its N-terminus. Recently, it was reportedthat a DDNN motif at residues26-29was critical for the cell surface targeting of humancorin. Interestingly, mouse corin cytoplasmic tail differs from the human corincytoplasmic tail in both length and amino acid sequence. The apparent differencesuggests that other sequences may exit in mouse corin cytoplasmic tail that can regulatecorin cell surface targeting and zymogen activation. To date, no studies have been doneto test such a hypothesis.To goal of this study is to examine the molecular basis underlying corin cell surfacetargeting and zymogen activation. In particular, we studied the role of specific aminoacids in the cytoplasmic tail and near the transmembrane domain in regulating corin cellsurface targeting and zymogen activation.Methods:Plasmids expressing corin mutants were constructed by PCR-based site-directedmutagenesis using wild-type mouse corin expressing plasmid as a template. HEK293cells were transiently transfected with the plasmids. Western blotting were performed toanalyze corin expression level and zymogen activation in cell lysate. Flow cytometricanalysis and biotin labeling of cell surface proteins were used to detect corin expression onthe cell surface. Pro-ANP processing was analyzed by immunoprecipitation, SDS-PAGE and Western blotting to assess corin activity.Results：(1) We constructed and expressed mouse corin truncation mutants, mD50and mD99,deleting50and99amino acids from the cytoplasmic tail, respectively. In transfectedHEK293cells, mD50, mD99and mE1a (the full-length mouse corin) were expressed atsimilar levels. However, an~40-kDa band (corin-p), representing the activated corinprotease domain fragment, was more abundant in samples from mD99mutant (29.00±0.32%) than that in mE1a (12.48±1.89%)(p=0.0001, n=4). It is known that corinactivation depends on its cell surface expression. We therefore examined mD99mutantexpression on the cell surface by biotin labeling cell surface proteins and flow cytometry.Our date showed that mD99expression on the cell surface was more than that in mE1a(69.02±4.56%vs.40.04±5.10%, p=0.0055, n=4), suggesting that there may exist aninhibitory sequence between amino acids50and99.(2) We constructed mouse corin truncation mutant mD60, mD70and mD75, deleting60,70and75amino acids from the cytoplasmic tail, respectively. Western analysis of thelysate from cells expressing these mutants showed that the level of the~40-kDa band inmD75mutant, but not mD60and mD70mutants, was higher than that in mE1a corin(28.73±0.01%vs.16.63±0.04%, p=0.02, n=3). We also found that corin mD75expression on the cell surface was more abundant than that in mE1a corin (76.34±4.43%vs.40.04±5.10%, p=0.0017, n=4).(3) We then constructed mouse corin truncation mutants, mD71, mD72, mD73andmD74, deleting71,72,73and74amino acids from the cytoplasmic tail, respectively.Western analysis showed similar corin expression levels for these mutants in cell lysate.However, the level of the~40-kDa band in mD72mutant, but not mD70and mD71mutants, was higher than that in mE1a corin (26.37±2.65%vs.16.44±2.11%, p=0.02,n=3), suggesting that amino acids around71-74may alter corin cell surface targeting andactivation.We then made mutants, K71A, F72A, Q73A and K71A/F72A/Q73A, using wild-typemouse corin (mE1a) as a template. In Western analysis, corin expression and activationwere similar in mutants K71A, F72A and Q73A. However, the level of the~40-kDa bandin the K71A/F72A/Q73A mutant was higher than that in mE1a corin (28.69±1.29%vs.19.58±1.42%, p=0.009, n=3), indicating that the KFQ motif may inhibit corin cell surface targeting and activation.(4) We also tested the minimal length in the cytoplasmic tail that is required for mousecorin expression the cell surface and activation. We constructed mouse corin truncationmutants, mD106, mD111and mD112, deleting106,111and112amino acids from thecytoplasmic tail, respectively. In transfected HEK293cells, similar expression levelswere found in cell lysate from these mutants. However, we found that mD112mutantwas not expressed on the cell surface. No zymogen activation was detected in this mutant.In pro-ANP processing assay, the mutant had no biological activity.(5) We further examined the effect of charged amino acids next to the transmembranedomain on corin cell surface expression and activation. We made mutants D-R112A,D-R112D and D-R112K to test whether arginine or other positively charged amino acids inthe membrane-proximal region play a role in corin cell surface targeting. Unlike mutantD-R112K, mutants D-R112A and D-R112D did not express on the cell surface and werenot activated. Similar results were observed when mutations of these residues were madein the context of full-length corin. Thus, our results show that positively charged aminoacids in the membrane-proximal region have a critical role in regulating corin cell surfaceexpression and zymogen activation.Conclusion:We have identified a new KFQ motif in the cytoplasmic tail that may inhibit corinintracellular trafficking, cell surface expression and zymogen activation.We have found that positively charged amino acids in the membrane-proximal regionare critical in regulating corin cell surface expression and activation.Our results provide new insights into the molecular basis underlying corinbiosynthesis and activity, and may help to understand the membrane topology and functionin other type II transmembrane proteases, that are important in a variety of biologicalprocesses. |
