“…In addition, genetic evidence from yeast and¯ies suggests that a group of kinases which appear to function upstream of MAPKKKs may exist in certain MAPK modules (Widmann et al, 1999;Kyriakis, 1999), despite the lack of direct biochemical evidence that these kinases activate MAPKKK by phosphorylation. Thus, at least 14 di erent but structurally related kinases in this group are suggested to function as mammalian MAPKKKKs, including PAK1, 2, 3 and 4 (Manser et al, 1994;Knaus et al, 1995;Abo, 1998), GCK (Katz et al, 1994), GCKR/KHS (Shi and Kehrl, 1997;Tung and Blenis, 1997), GLK , HPK1 (Kiefer et al, 1996;Hu et al, 1996), NIK (Nck-interacting kinase)/HGK (Su et al 1997;Yao et al 1999), SOK1 (previously called UK-1; Pombo et al, 1996), Krs-1 (Taylor et al, 1996), MST1/Krs-2(Creasy and Cherno , 1995; Taylor et al, 1996), MST3 (Schinkmann and Blenis, 1997) and LOK (Kuramochi et al, 1997). Most of the MAPKKKKs, but not SOK1, Krs-1, MST3 and LOK, can activate JNK by overexpression.…”