Fig. 5. Evaluation of how mutations in the juxtamembrane domain affect Kitl shedding. (A) Mutants carrying deletions in the juxtamembrane domain of Kitl1 or Kitl2 were generated as described in the Materials and Methods in order to identify the cleavage site(s) for ADAM17. The designations of individual deletion mutants are listed on the left. All expression constructs used in this study contain an N-terminal AP-tag and a C-terminal Myc-His tag. Exon 6, which contains the putative major cleavage site, is spliced out in Kitl2. TM, transmembrane domain. Two different kinds of vertical lines indicate the boundary for exons, and the boundary between the extracellular and transmembrane domain, respectively. (B) COS-7 cells were transfected with AP-tagged wild-type Kitl1, Kitl2 or various mutant plasmids, labeled with [³⁵S]-methionine for 30 minutes and chased for 6 hours. Cell lysates were collected, immunoprecipitated with anti-Kitl antibody and treated with Endo H (E, lane 2) or PNGase F (P, lane 3) at 37°C overnight, as described in the Materials and Methods. Untreated sample was loaded in lane 1. Asterisks, black arrowheads and white arrowheads indicate the mature form (125 KDa), intermediate form II (112.5 KDa) and intermediate form I (100 KDa) of Kitl, respectively. The mature forms of Kitl in cells transfected with Kitl2 mutants were resistant to treatment with Endo H to a similar degree as wild-type Kitl2. Since Endo H resistance is acquired during passage through the medial Golgi compartment, these results indicate that similar amounts of the mutant and wild-type proteins were released from the ER and migrated through the medial Golgi compartment. This strongly suggests that the mutants giving rise to Endo H-resistant mature forms are properly folded, and that the lack of shedding of KL2SSTL/KAAK is not caused by retention of this mutant in the ER. Since the KL1DM mutant did not acquire resistance to Endo H it was not used for further shedding analysis. (C) Effects of deletions on Kitl shedding. AP-tagged deletion variants were transfected in wild-type (lanes 1 and 2) or Adam17^–/– mEFs (lanes 3 and 4). Shedding was performed as described above, using 100 mM PV as a stimulus. Supernatants (lanes 1-4) were concentrated by ConA-Sepharose and analyzed on 8% SDS-PAGE as described in the Materials and Methods. An AP-analysis of cell lysates shows similar expression of mutant forms of Kitl2 in wild-type and Adam17^–/– cells in cases in which shedding is abolished. Lower levels of Kitl2 mutants were seen in wild-type cell lysates following PV or PMA stimulation compared with Adam17^–/– cells, presumably because the mutants accumulate in the absence of ADAM17.