Figure 1.
KISS1 is secreted and processed into kisspeptins (KP) in multiple cell lines.
A. (Top panel) Conditioned media collected from cultures of C8161.9, MelJuSo, MDA-MB-435, HeLa, COS7 and HEK293 cell lines which exogenously express KISS1. KP were immunoprecipitated with a monoclonal antibody detecting KISS1 (6a4.27), separated by Tris-tricine SDS-PAGE and then immunoblotted with anti-KISS1 antibody (1∶1000). (Bottom panel) Full-length KISS1 was detected in whole cell lysates using anti-KISS1 antibody as above (Mr∼15.9 kDa; Note: lower molecular mass bands were not detected from whole cell lysates as previously published [15]). B. KISS1 processing into KP can be detected using immunoprecipitated KISS1/KP separated by Tris-tricine SDS-PAGE gel followed by Coomassie dye staining. KP bands were excised and analyzed by ESI-MS/MS [Figure S1]. KP sizes (approximate kDa) are predicted based upon the predicted PC cleavage sites in nascent KISS1 (See Figure 2A). KP sizes and sequence were confirmed by ESI-MS/MS and are shown schematically in Figure S1.
Figure 2.
Proprotein convertases are required for KISS1 processing.
A. Amino acid sequence of KISS1 showing three predicted PC cleavage sites − RK, RR and KR in bold. Note: DYKDDDDK at position 68 is the FLAG epitope tag used for several of the studies described in this manuscript. B. Treatment of C8161.9, MelJuSo, HeLa, COS7 and HEK293 cells over-expressing KISS1 with the general PC inhibitor Dec-RVKR-CMK (100 µM) blocked KISS1 processing as detected by immunoprecipitation and immunoblot using anti-KISS1 antibody [See Figure 1 and Materials and Methods for details]. C. Over-expression of KISS1 and the biological PC inhibitor V5-tagged α1-PDX also resulted in loss of KISS1 processing into KP. Polypeptide sizes correspond to those shown in Figure 1.
Table 1.
RT-PCR analysis of the six members of PC family involved in the secretory pathway.
Figure 3.
Furin is required for the generation of KP from KISS1.
A, C. RT-PCR and immunoblot data showing reduced mRNA and protein levels of furin, PCSK5 and PCSK7 in HEK293 cells transduced with shRNA directed against each PC either individually or in combination. Bars in Panel C (by group; left to right) = furin, PCSK5 and PCSK7. B. shRNA-mediated knockdown of furin, PCSK5 and/or PCSK7 was in HEK293 cells over-expressing KISS1 show that cleavage of KISS1 is lost whenever furin is knocked down, but processing can still occur when PCSK5 or PCSK7 are knocked down. D. C8161.9, MelJuSo, HeLa and COS7 in which furin expression was knocked down by shRNA showed reduced or complete loss of KISS1 processing. KISS1 processing was observed in all cell lines when a scrambled control shRNA (Scr) was used. E. Restoration of furin activity in LoVo cells (which have no furin activity) by transduction of adenoviral particles with V5-tagged furin restored KISS1 processing to KP. LoVo cells transduced with an empty vector still do not show KISS1 cleavage into KP. Loading controls for V5-furin and KISS1 are shown using immunoblots with V5 and KISS1 antibodies. KISS1 and KP polypeptide sizes correspond to those shown in Figure 1.
Figure 4.
Mutation of putative PC sites in KISS1 blocks KP generation but does not abrogate metastasis suppressor function.
A. Immunoblot showing KISS1 processing-deficient mutant (C1C2C3) which was mutated at the three putative furin processing sites (KISS1K57A/R67A/R124A). C1C2C3 was not processed into KP [Compare to Figure 1A and Figure 1B]. * indicates C1C2C3 lane which was over-exposed to highlight lack of laddering below the full-length KISS1 band. B. Relative lung metastases (mean ± SEM) compared to C8161.9 (647±149 metastases). KISS1 and C1C2C3 transfected cells were significantly less metastatic.