Self-Association of the Single-KH-Domain Family Members Sam68, GRP33, GLD-1, and Qk1: Role of the KH Domain

Abstract

Sam68 is a member of a growing family of proteins that contain a single KH domain embedded in a larger conserved domain of ϳ170 amino acids. Loops 1 and 4 of this KH domain family are longer than the corresponding loops in other KH domains and contain conserved residues. KH domains are protein motifs that are involved in RNA binding and are often present in multiple copies. Here we demonstrate by coimmunoprecipitation studies that Sam68 self-associated and that cellular RNA was required for the association. Deletion studies demonstrated that the Sam68 KH domain loops 1 and 4 were required for self-association. The Sam68 interaction was also observed in Saccharomyces cerevisiae by the two-hybrid system. In situ chemical cross-linking studies in mammalian cells demonstrated that Sam68 oligomerized in vivo. These Sam68 complexes bound homopolymeric RNA and the SH3 domains of p59 fyn and phospholipase C␥1 in vitro, demonstrating that Sam68 associates with RNA and signaling molecules as a multimer. The formation of the Sam68 complex was inhibited by p59 fyn , suggesting that tyrosine phosphorylation regulates Sam68 oligomerization. Other Sam68 family members including Artemia salina GRP33, Caenorhabditis elegans GLD-1, and mouse Qk1 also oligomerized. In addition, Sam68, GRP33, GLD-1, and Qk1 associated with other KH domain proteins such as Bicaudal C. These observations indicate that the single KH domain found in the Sam68 family, in addition to mediating protein-RNA interactions, mediates protein-protein interactions.

The KH domains of Xenopus Vg1RBP mediate RNA binding and self-association

“…In addition to the truncated variants, seven full-length variants were constructed that contained point-mutated KH domains (individually, as didomains, or all four; see Fig+ 2), named so that an ( m ) follows the number of the mutated KH domain (see position of mutations in Fig+ 1A)+ The mutation introduces an asparagine residue instead of a conserved aliphatic residue from the KH domain core, adjacent to the RNA binding platform (Lewis et al+, 2000)+ In FMR1, the gene product mutated in the fragile X mental retardation syndrome, which contains 2 KH domains, this naturally occurring mutation in the second KH domain (I304N) severely impairs RNA binding (Siomi et al+, 1994)+ Equivalent mutations in the KH domains of hnRNP K and Sam68 also abrogate RNA binding (Siomi et al+, 1994;Chen et al+, 1997)+ We noted that some of the I304N-like mutations introduced into KH domains of Vg1RBP have an adverse effect on the solubility of the protein: R-K12-K3 m 4 m is partially soluble whereas the quadruple mutant R-K1 m 2 m -K3 m 4 m is almost completely insoluble and is included in this study only as an in vitro translated product (Fig+ 2)+ An important issue to consider is whether the addition of a tag, even one as short as six His residues, may affect the folding and hence the activity of the fused protein+ Two independent results indicate that the added tag does not affect RNA binding+ First, in vitro-translated native and His-tagged Vg1RBP variants bind poly(C)-agarose to the same extent (Fig+ 3D)+ Similarly, in UV crosslinking analysis, recombinant Vg1RBP added into oocyte extract shows binding to 32 P-labeled VLE RNA that is indistinguishable from its native counterpart (data not shown)+ These experiments suggest that neither the overexpressed nor the in vitro-translated Vg1RBP are adversely affected by the presence of a 6ϫHis tag+…”

A single point mutation in the nuclear localization domain of Sam68 blocks the Rev/RRE-mediated transactivation

“…Unlike the RevM10, which localizes to the nucleus and competes with wild type Rev for binding to RRE, the Sam68 P439?R mutant is in the cytoplasm. However, since this protein can multimerize (Chen et al, 1997), it is conceivable that co-expression of Sam68 P439?R with Rev might form a nonfunctional complex in the cytoplasm. Therefore, its mechanism may be trapping Rev in the cytoplasm by direct protein-protein interaction.…”

“…The 443 amino acid protein is comprised of two putative RNA binding domains, namely, an RGG box (a domain containing several Arg-Gly-Gly motifs) and a KH (for hnRNP-K homology) domain (Fumagalli et al, 1994). The RGG box is dispensable for RNA binding and the multimerization of Sam68 (Chen et al, 1997). The KH domain is essential for self-association, RNA binding as well as cellular localization of wild type Sam68 (Chen et al, 1997;McBride et al, 1998).…”

“…The RGG box is dispensable for RNA binding and the multimerization of Sam68 (Chen et al, 1997). The KH domain is essential for self-association, RNA binding as well as cellular localization of wild type Sam68 (Chen et al, 1997;McBride et al, 1998). The KH domain is highly conserved in several RNA binding proteins such as hnRNP-K (Siomi et al, 1993a), GRP33 (Cruz-Alvarez and Pellicer, 1987), fragile X mental retardation gene product FMR-1 (Siomi et al, 1993b) and the Ceanorhabditis elegans germline-speci®c tumor suppressor GLD-1 (Jones and Schedl, 1995).…”

“…Sam68 binds to single stranded and homopolymeric RNAs in vitro and this binding is dependent on the phosphorylation of tyrosine residues of the protein (Chen et al, 1997;Lin et al, 1997). Recently, we have demonstrated that Sam68 binds to RRE in vitro and in vivo, and functionally replaces as well as synergizes with HIV-1 Rev in RRE-mediated gene expression and virus replication (Reddy et al, 1999).…”