Activating and dominant inactivating c-KIT catalytic domain mutations in distinct clinical forms of human mastocytosis

Abstract

Human mastocytosis is characterized by increased mast cells. It usually occurs as a sporadic disease that is often transient and limited in children and persistent or progressive in adults. The c-KIT protooncogene encodes KIT, a tyrosine kinase that is the receptor for mast cell growth factor. Because mutated KIT can transform cells, we examined c-KIT in skin lesions of 22 patients with sporadic mastocytosis and 3 patients with familial mastocytosis. All patients with adult sporadic mastocytosis had somatic c-KIT mutations in codon 816 causing substitution of valine for aspartate and spontaneous activation of mast cell growth factor receptor (P = 0.0001). A subset of four pediatric onset cases with clinically unusual disease also had codon 816 activating mutations substituting valine, tyrosine, or phenylalanine for aspartate. Typical pediatric patients lacked 816 mutations, but limited sequencing showed three of six had a novel dominant inactivating mutation substituting lysine for glutamic acid in position 839, the site of a potential salt bridge that is highly conserved in receptor tyrosine kinases. No c-KIT mutations were found in the entire coding region of three patients with familial mastocytosis. We conclude that c-KIT somatic mutations substituting valine in position 816 of KIT are characteristic of sporadic adult mastocytosis and may cause this disease. Similar mutations causing activation of the mast cell growth factor receptor are found in children apparently at risk for extensive or persistent disease. In contrast, typical pediatric mastocytosis patients lack these mutations and may express inactivating c-KIT mutations. Familial mastocytosis, however, may occur in the absence of c-KIT coding mutations.

Figures

Figure 1

KITWT and KIT mutant phosphorylation in COS-7 cells. Blotting of immunoprecipitated KIT with an antiphosphotyrosine antibody (anti-PY, lanes 1–8) confirms in our system the previously reported results of other researchers (23, 24, 32) showing minimal spontaneous tyrosine phosphorylation of KIT in COS-7 cells expressing KITWT (lane 1) and high levels of spontaneous tyrosine phosphorylation of mutant KITs with substitution in position 816 (lanes 3, 5, and 7). These results serve as controls for Fig. 2. Increased tyrosine phosphorylation of KITWT is seen after exposure to exogenous SCF (+, lane 2), but tyrosine phosphorylation of codon 816 mutants is already maximal, so there is no increase in response to SCF (compare lanes 4, 6, and 8 with lanes 3, 5, and 7). After stripping, reprobing of the blot with anti-KIT antibody demonstrates the relative amounts of protein in each lane (lanes 9–16). Note that much greater amounts of KITWT are needed to demonstrate spontaneous tyrosine phosphorylation than are needed with codon 816 mutants (compare lane 9 with lanes 11, 13, and 15) and that maximal KIT tyrosine phosphorylation is higher in the mutant KITs than in KITWT (compare lane pairs 2 and 10 with 4 and 12, 6 and 14, and 8 and 16).

Figure 2

Codon 839 mutation inactivates KIT. Blotting of immunoprecipitated KIT with an antiphosphotyrosine antibody (anti-Py, lanes 1–6) shows a low level of spontaneous KIT tyrosine phosphorylation in COS-7 cells expressing high levels of KITWT (lane 1) but no spontaneous phosphorylation of KITE839K (lane 3). KITWT is phosphorylated in response to exogenous SCF, but KITE839K is not (lanes 2 and 4, respectively). A double mutant, KITD816V/E839K, which incorporates both the activating D816V mutation and E839K loss-of-function mutation, is not phosphorylated spontaneously or in response to exogenous SCF, indicating that the E839K has an intramolecular dominant–inactivating effect (lanes 5 and 6). Blotting with anti-KIT antibody after stripping the antiphosphotyrosine antibody (anti-KIT, lanes 7–12) shows KIT protein in all transfectants. Note also that KITWT is present as both a 125-kDa form and as a fully glycosylated, mature, 145-kDa form (lanes 7 and 8) but that KITE839K is present predominately as a lower molecular weight form in the steady state (lanes 9, 10, 11, and 12).

Figure 3

Dominant–negative effect of Glu839Lys mutant on wild-type KIT phosphorylation. Anti-py blot of immunoprecipitated wild-type and mutant KITs expressed in COS-7 cells. Again, KITWT shows minimal spontaneous phosphorylation but is phosphorylated in response to exogenous SCF (lanes 1 and 2). The E839K mutant is not phosphorylated spontaneously or in response to SCF (lanes 3 and 4). SCF-induced KITWT phosphorylation (lane 2) is decreased dramatically when the wild-type receptor plasmid is transfected with an equal amount of the E839K mutant plasmid (lanes 5 and 6). As a control, a double amount of KITWT plasmid (2×) was transfected (lanes 7 and 8). Spontaneous and SCF-induced phosphorylation were both increased compared with the lower amounts of KITWT plasmid (compare lanes 1 and 2 with 7 and 8). Reprobing of the anti-Py blot (after stripping) with anti-KIT Ab shows the amount of protein in each lane. Long exposure documents a minor component of 145-kDa KITE839K (lanes 11 and 12). The amounts of 125- and 145-kDa KIT in lanes 13 and 14 equals approximately the sum of these two forms expressed separately (lanes 9 and 10, KITWT; lanes 11 and 12, KITE839K). The increase in protein in the control 2× transfected KITWT (lanes 15 and 16) over the 1× KITWT transfectant is comparable to the increase in phosphorylation (lanes 15 and 16 vs. lanes 9 and 10). Molecular markers are indicated in kilodaltons on the left.