Characterisation of the Nucleic‐Acid‐Binding Activity of KH Domains Different Properties of Different Domains

Abstract

The KH module is a sequence motif recently identified in a number of diversified RNA-binding proteins and suggested to be the functional element responsible for RNA binding. So far, however, this hypothesis has not received direct experimental support. We have expressed the three KH-domains from heterogeneous nuclear ribonucleoprotein K (hnRNP-K), the poly(C)-binding proteins PCBP-1 and PCBP-2, the first three to four domains from the high-density binding protein HBP, the one and a half domain from the archaeon Halobacterium halobium ORF139 and one and a half domain of the fragile-X protein FMRl in Escherichia coli and analysed their nucleic-acid-binding properties in vitro. The results showed that the in vitro poly(rC)-binding activity of hnRNP-K can be assigned to KH-domain 3, whereas both domains 1 and 3 in the PCBPs bind poly(rC). In addition, all these domains exhibit binding activity towards other nucleic acids, albeit at a significantly lower level. The first KH domain from the FMRl protein binds poly(rG) and single-stranded and double-stranded DNA. The N-terminal three or four domains from HBP bind poly(rG) and, at a much lower level, single-stranded and double-stranded DNA. Thus, single KH domains are discrete and independent nucleic-acid-binding units. Moreover, different KH domains bind different nucleic acids, suggesting that KH domains are composed of a conserved, weakly nucleic-acid-binding, structure that is fine tuned, by sequence variation, resulting in sequencespecific nucleic-acid-binding entities.

The poly(C)-binding proteins: A multiplicity of functions and a search for mechanisms

“…The KH domain has been identified in a wide spectrum of RNA-binding proteins (Gibson et al+, 1993;Siomi et al+, 1993)+ Although this domain was initially identified as a repeated 45-amino-acid motif in hnRNP K (Siomi et al+, 1993), subsequent alignments and structural studies have defined a more extensive 68-72 residue KH "maxi domain" (Musco et al+, 1996)+ NMR spectroscopic and X-ray crystallographic analyses of a number of KH domains (Musco et al+, 1996(Musco et al+, , 1997Baber et al+, 1999;Lewis et al+, 1999;Wimberly et al+, 2000) have revealed that the initially described 45-residue core is configured as a baab unit+ The data further reveal two KH domain subtypes based on the more extensive structure: The type I KH domain (e+g+, KH-3 of hnRNP K) includes a C-terminal ba extension and the type II KH domain (e+g+, ribosomal protein S3) contains an N-terminal ab extension (Grishin, 2001)+ All three KH repeats within the PCBPs belong to the type I KH domain comprising a three-stranded antiparallel b-sheet packed against three a-helices ( baabba; Fig+ 2A)+ Two unstructured surface loops extend from this structure, one containing an invariant GXXG and a second of variable length and sequence+ These loops may be of central importance in determining nucleic acid binding specificity (Adinolfi et al+, 1999)+ An extended survey of PCBP gene organization including the inferred intron-exon organization of the genes encoding human FMR-1 (Eichler et al+, 1993), mouse quaking (Kondo et al+, 1999), Nova-1 and Nova-2, and KOC (known also as IGF-II mRNA-binding protein 3; Fig+ 2A) reveals that introns are restricted to positions between secondary structural segments of the KH domain+ This positioning suggests that the presentday KH domain may have arisen from a combination and assortment of mini-exons encoding individual structural elements+ This model would suggest that prokaryotic and eukaryotic KH domains might have evolved by convergent evolution of nucleic acid-binding modules (Graumann & Marahiel, 1996)+ The presence of multiple KH domains in each of the PCBPs brings up the question of which domain, or combination of domains, dictates RNA-binding specificity and affinity+ At present, there is no simple answer to this question+ Although individual KH domains from PCBPs can bind single-stranded nucleic acids (Dejgaard & Leffers, 1996), it is clear that the binding affinity and specificity of the intact protein reflects a 270…”

“…Makeyev and S.A. Liebhaber complex sum of its parts+ For example, the high affinity poly(C) binding by hnRNP K can be abolished or significantly reduced by modifying or deleting any of the three individual KH domains (Siomi et al+, 1994)+ The first and/or the second KH domains of hnRNP K exhibit minimal binding activity towards poly(rC) on their own (Dejgaard & Leffers, 1996) and the isolated third KH domain of hnRNP K exhibits a much lower affinity to CT-rich DNA sequence than does the protein as a whole (Baber et al+, 2000)+ In the case of aCP-1 and aCP-2, the first and second KH domains can independently bind poly(rC) with a high affinity and specificity whereas the third KH domain has no such activity (Dejgaard & Leffers, 1996)+ However, when a more complex native target of aCP-2 (poliovirus 59-UTR; see below) was studied, only the first KH domain demonstrated the target RNA-binding activity (Silvera et al+, 1999)+ Such observations lead to the conclusion that the binding specificity of each protein can reflect collaboration of the multiple KH repeats and/or may reflect contributions from the inter-KH regions as well (Dejgaard & Leffers, 1996)+ Structural studies of the intact PCBP proteins with their RNA targets will be needed to gain further insight into this problem+ The presence of multiple KH repeats in each of PCBPs, each of which may theoretically bind independently to RNA, and the ability of PCBPs to directly interact with themselves and with other proteins (see below) suggests that RNP complexes involving PCBPs may differ significantly in their stoichiometry+ To address this issue, the stoichiometry of several of the PCBP-containing RNP complexes has been explored+ NMR measurements demonstrate a 1:1 interaction between the isolated third KH domain of hnRNP K and its targets (Baber et al+, 2000)+ Analysis of the complex formed at the C-rich segment of the human a-globin mRNA revealed that this "a-complex" is comprised of a single molecule of aCP bound to the mRNA target site (Chkheidze et al+, 1999)+ This binary interaction of the aCP with the a-globin 39 UTR (39 untranslated region) target could reflect interaction between the RNA target and a single KH domain of aCP or by interactions between each of the three KH domains and the target sequence+ In this regard, comparison of optimized RNA targets generated by in vitro SELEX revealed that aCP-2 and hnRNP K have distinct requirements for highaffinity RNA binding: The optimal target sequence for hnRNP K was found to be a single short "C-patch," whereas that of aCP-2 encompassed three such short C-patches within a highly exposed single-stranded conformation (Thisted et al+, 2001)+ These data suggest that whereas a single C-patch may mediate a high affinity interaction with a single KH domain in hnRNP K, a tandem array of three C-patches maximizes aCP-2 binding to its RNA target+ Thus, the binding of aCP to its optimized target might reflect individual interactions by each of the three KH domains+ The interesting questions of how these multiple contacts are organized and why the closely related aCP and hnRNP K differ in this seemingly fundamental aspect of their biology must await structural analyses of the respective RNA-protein complexes+…”

Control of mRNA translation and stability in haematopoietic cells: The function of hnRNPs K and E1/E2

“…Several laboratories showed that a subfamily of hnRNPs, hnRNP K and hnRNP E1/E2, functions in translational regulation and/or mRNA stabilization in haematopoiesis. These proteins all bear three copies of an RNA binding motif, the hnRNP K homology (KH)-domain (Siomi et al, 1993, Gibson et al, 1993 consisting of 65-70 amino acids (Dejgaard andLeffers, 1996, Musco et al, 1996). In contrast to hnRNP E1/E2, hnRNP K carries additional domains: An N-terminal bipartite nuclearlocalization signal (NLS) and an hnRNP K-specific nuclear shuttling signal located between the second and third KHdomain, which confers the capacity for bi-directional transport across the nuclear envelope (Matunis et al, 1992, Michael et al, 1995.…”

Regulation of α1(I) Collagen Messenger RNA Decay by Interactions with αCP at the 3′-Untranslated Region

“…Some hnRNP proteins have RNA binding activity regulated directly by specific kinase phosphorylation, suggesting kinases may play important roles in regulating gene transcription or translation in response to certain stimuli (21,23). The hnRNP family member ␣CP contains three RNA-binding K domain homology domains and is alternatively spliced to produce several discrete isoforms (24,25). Yeast two-hybrid studies have demonstrated that ␣CP may form homodimers and have identified other nucleic acid-binding proteins with which it may interact, such as poly(A)-binding protein (PABP) (13,26).…”