Sequential phosphorylation and multisite interactions characterize specific target recognition by the FHA domain of Ki67

The forkhead-associated (FHA) domain of human Ki67 interacts with the human nucleolar protein hNIFK, recognizing a 44-residue fragment, hNIFK226-269, phosphorylated at Thr234. Here we show that high-affinity binding requires sequential phosphorylation by two kinases, CDK1 and GSK3, yielding pThr238, pThr234 and pSer230. We have determined the solution structure of Ki67FHA in complex with the triply phosphorylated peptide hNIFK226-269(3P), revealing not only local recognition of pThr234 but also the extension of the beta-sheet of the FHA domain by the addition of a beta-strand of hNIFK. The structure of an FHA domain in complex with a biologically relevant binding partner provides insights into ligand specificity and potentially links the cancer marker protein Ki67 to a signaling pathway associated with cell fate specification.     

Structure of the FHA1 domain of yeast Rad53 and identification of binding sites for both FHA1 and its target protein Rad9

Forkhead-associated (FHA) domains have been shown to recognize both pThr and pTyr-peptides. The solution structures of the FHA2 domain of Rad53 from Saccharomyces cerevisiae, and its complex with a pTyr peptide, have been reported recently. We now report the solution structure of the other FHA domain of Rad53, FHA1 (residues 14-164), and identification of binding sites of FHA1 and its target protein Rad9. The FHA1 structure consists of 11 beta-strands, which form two large twisted anti-parallel beta-sheets folding into a beta-sandwich. Three short alpha-helices were also identified. The beta-strands are linked by several loops and turns. These structural features of free FHA1 are similar to those of free FHA2, but there are significant differences in the loops. Screening of a peptide library [XXX(pT)XXX] against FHA1 revealed an absolute requirement for Asp at the +3 position and a preference for Ala at the +2 position. These two criteria are met by a pThr motif (192)TEAD(195) in Rad9. Surface plasmon resonance analysis showed that a pThr peptide containing this motif, (188)SLEV(pT)EADATFVQ(200) from Rad9, binds to FHA1 with a K(d) value of 0.36 microM. Other peptides containing pTXXD sequences also bound to FHA1, but less tightly (K(d)=4-70 microM). These results suggest that Thr192 of Rad9 is the likely phosphorylation site recognized by the FHA1 domain of Rad53. The tight-binding peptide was then used to identify residues of FHA1 involved in the interaction with the pThr peptide. The results are compared with the interactions between the FHA2 domain and a pTyr peptide derived from Rad9 reported previously.     

The ligand specificity of yeast Rad53 FHA domains at the +3 position is determined by nonconserved residues

On the basis of the results from our laboratory and others, we recently suggested that the ligand specificity of forkhead-associated (FHA) domains is controlled by variations in three major factors: (i) residues interacting with pThr, (ii) residues recognizing the +1 to +3 residues from pThr, and (iii) an extended binding surface. While the first factor has been well established by several solution and crystal structures of FHA-phosphopeptide complexes, the structural bases of the second and third factors are not well understood and are likely to vary greatly between different FHA domains. In this work, we proposed and tested the hypothesis that nonconserved residues G133 and G135 of FHA1 and I681 and D683 of FHA2, located outside of the core FHA region of yeast Rad53 FHA domains, contribute to the specific recognition of the +3 position of different phosphopeptides. By rational mutagenesis of these residues, the specificity of FHA1 has been changed from predominantly pTXXD to be equally acceptable for pTXXD, pTXXL, and pYXL, which are similar to the specificities of the FHA2 domain of Rad53. Conversely, the +3 position specificity of FHA2 has been engineered to be more like FHA1 with the I681A mutation. These results were based on library screening as well as binding analyses of specific phosphopeptides. Furthermore, results of structural analyses by NMR indicate that some of these residues are also important for the structural integrity of the loops.     

Structure and function of a new phosphopeptide-binding domain containing the FHA2 of Rad53

The forkhead-associated (FHA) domain is a 55-75 amino acid residue module found in >20 proteins from yeast to human. It has been suggested to participate in signal transduction pathways, perhaps via protein-protein interactions involving recognition of phosphopeptides. Neither the structure nor the ligand of FHA is known. Yeast Rad53, a checkpoint protein involved in DNA damage response, contains two FHA domains, FHA1 (residues 66-116) and FHA2 (residues 601-664), the second of which recognizes phosphorylated Rad9. We herein report the solution structure of an "FHA2-containing domain" of Rad53 (residues 573-730). The structure consists of a beta-sandwich containing two antiparallel beta-sheets and a short, C-terminal alpha-helix. Binding experiments suggested that the FHA2-containing domain specifically recognizes pTyr and a pTyr-containing peptide from Rad9, and that the binding site involves residues highly conserved across FHA domains. The results, along with other recent reports, suggest that FHA domains could have pTyr and pSer/Thr dual specificity.     

A novel nucleolar protein, NIFK, interacts with the forkhead associated domain of Ki-67 antigen in mitosis

In a previous study, we demonstrated that the forkhead associated (FHA) domain of pKi-67 interacts with the novel kinesin-like protein, Hklp2 (Sueishi, M., Takagi, M., and Yoneda, Y. (2000) J. Biol. Chem. 275, 28888-28892). In this study, we report on the identification of a putative RNA-binding protein of 293 residues as another binding partner of the FHA domain of pKi-67 (referred to as NIFK for nucleolar protein interacting with the FHA domain of pKi-67). Human NIFK (hNIFK) interacted with the FHA domain of pKi-67 (Ki-FHA) efficiently in vitro when hNIFK was derived from mitotically arrested cells. In addition, a moiety of hNIFK was co-localized with pKi-67 at the peripheral region of mitotic chromosomes. The hNIFK domain that interacts with Ki-FHA was mapped in the yeast two-hybrid system to a portion encompassed by residues 226-269. In a binding assay utilizing Xenopus egg extracts, it was found that the mitosis-specific environment and two threonine residues within this portion of hNIFK (Thr-234 and Thr-238) were crucial for the efficient interaction of hNIFK and Ki-FHA, suggesting that hNIFK interacts with Ki-FHA in a mitosis-specific and phosphorylation-dependent manner. These findings provide a new clue to our understanding of the cellular function of pKi-67.     

Structural and functional analysis of phosphothreonine-dependent FHA domain interactions

FHA domains are well established as phospho-dependent binding modules mediating signal transduction in Ser/Thr kinase signaling networks in both eukaryotic and prokaryotic species. Although they are unique in binding exclusively to phosphothreonine, the basis for this discrimination over phosphoserine has remained elusive. Here, we attempt to dissect overall binding specificity at the molecular level. We first determined the optimal peptide sequence for Rv0020c FHA domain binding by oriented peptide library screening. This served as a basis for systematic mutagenic and binding analyses, allowing us to derive relative thermodynamic contributions of conserved protein and peptide residues to binding and specificity. Structures of phosphopeptide-bound and uncomplexed Rv0020c FHA domain then directed molecular dynamics simulations which show how the extraordinary discrimination in favor of phosphothreonine occurs through formation of additional hydrogen-bonding networks that are ultimately stabilized by van der Waals interactions of the phosphothreonine γ-methyl group with a conserved pocket on the FHA domain surface.     

The FHA domain is a modular phosphopeptide recognition motif.

FHA domains are conserved sequences of 65-100 amino acid residues found principally within eukaryotic nuclear proteins, but which also exist in certain prokaryotes. The FHA domain is thought to mediate protein-protein interactions, but its mode of action has yet to be elucidated. Here, we show that the two highly divergent FHA domains of Saccharomyces cerevisiae Rad53p, a protein kinase involved in cell cycle checkpoint control, possess phosphopeptide-binding specificity. We also demonstrate that other FHA domains bind peptides in a phospho-dependent manner. These findings indicate that the FHA domain is a phospho-specific protein-protein interaction motif and have important implications for mechanisms of intracellular signaling in both eukaryotes and prokaryotes.     

II. Structure and specificity of the interaction between the FHA2 domain of Rad53 and phosphotyrosyl peptides

The forkhead-associated (FHA) domain is a protein module found in many proteins involved in cell signaling in response to DNA damage. It has been suggested to bind to pThr sites of its target protein. Recently we have determined the first structure of an FHA domain, FHA2 from the yeast protein Rad53, and demonstrated that FHA2 binds to a pTyr-containing peptide (826)EDI(pY)YLD(832) from Rad9, with a moderate affinity (K(d) ca. 100 microM). We now report the solution structure of the complex of FHA2 bound with this pTyr peptide. The structure shows that the phosphate group of pTyr interacts directly with three arginine residues (605, 617, and 620), and that the leucine residue at the +2 position from the pTyr interacts with a hydrophobic surface on FHA2. The sequence specificity of FHA2 was determined by screening a combinatorial pTyr library. The results clearly show that FHA2 recognizes specific sequences C-terminal to pTyr with the following consensus: XX(pY)N(1)N(2)N(3), where N(1)=Leu, Met, Phe, or Ile, N(2)=Tyr, Phe, Leu, or Met, and N(3)=Phe, Leu, or Met. Two of the selected peptides, GF(pY)LYFIR and DV(pY)FYMIR, bind FHA2 with K(d) values of 1.1 and 5.0 microM, respectively. The results, along with other recent reports, demonstrate that the FHA domain is a new class of phosphoprotein-binding domain, capable of binding both pTyr and pThr sequences.     

Solution structure of the yeast Rad53 FHA2 complexed with a phosphothreonine peptide pTXXL: comparison with the structures of FHA2-pYXL and FHA1-pTXXD complexes

It was proposed previously that the FHA2 domain of the yeast protein kinase Rad53 has dual specificity toward pY and pT peptides. The consensus sequences of pY peptides for binding to FHA2, as well as the solution structures of free FHA2 and FHA2 complex with a pY peptide derived from Rad9, have been obtained previously. We now report the use of a pT library to screen for binding of pT peptides with the FHA2 domain. The results show that FHA2 binds favorably to pT peptides with Ile at the +3 position. We then searched the Rad9 sequences with a pTXXI/L motif, and tested the binding affinity of FHA2 toward ten pT peptides derived from Rad9. One of the peptides, (599)EVEL(pT)QELP(607), displayed the best binding affinity (K(d)=12.9 microM) and the greatest chemical shift changes. The structure of the FHA2 complex with this peptide was then determined by solution NMR and the structure of the complex between FHA2 and the pY peptide (826)EDI(pY)YLD(832) was further refined. Structural comparison of these two complexes indicates that the Leu residue at the +3 position in the pT peptide and that at the +2 position in the pY peptide occupy a very similar position relative to the binding site residues from FHA2. This can explain why FHA2 is able to bind both pT and pY peptides. This position change from +3 to +2 could be the consequence of the size difference between Thr and Tyr. Further insight into the structural basis of ligand specificity of FHA domains was obtained by comparing the structures of the FHA2-pTXXL complex obtained in this work and the FHA1-pTXXD complex reported in the accompanying paper.     

PhosphoThr peptide binding globally rigidifies much of the FHA domain from Arabidopsis receptor kinase-associated protein phosphatase

A net increase in the backbone rigidity of the kinase-interacting FHA domain (KI-FHA) from the Arabidopsis receptor kinase-associated protein phosphatase (KAPP) accompanies the binding of a phosphoThr peptide from its CLV1 receptor-like kinase partner, according to (15)N NMR relaxation at 11.7 and 14.1 T. All of the loops of free KI-FHA display evidence of nanosecond-scale motions. Many of these same residues have residual dipolar couplings that deviate from structural predictions. Binding of the CLV1 pT868 peptide seems to reduce nanosecond-scale fluctuations of all loops, including half of the residues of recognition loops. Residues important for affinity are found to be rigid, i.e., conserved residues and residues of the subsite for the key pT+3 peptide position. This behavior parallels SH2 and PTB domain recognition of pTyr peptides. PhosphoThr peptide binding increases KI-FHA backbone rigidity (S(2)) of three recognition loops, a loop nearby, seven strands from the beta-sandwich, and a distal loop. Compensating the trend of increased rigidity, binding enhances fast mobility at a few sites in four loops on the periphery of the recognition surface and in two loops on the far side of the beta-sandwich. Line broadening evidence of microsecond- to millisecond-scale fluctuations occurs across the six-stranded beta-sheet and nearby edges of the beta-sandwich; this forms a network connected by packing of interior side chains and H-bonding. A patch of the slowly fluctuating residues coincides with the site of segment-swapped dimerization in crystals of the FHA domain of human Chfr. Phosphopeptide binding introduces microsecond- to millisecond-scale fluctuations to more residues of the long 8/9 recognition loop of KI-FHA. The rigidity of this FHA domain appears to couple as a whole to pThr peptide binding.     

Direct observations of shifts in the β-sheet register of a protein-peptide complex using explicit solvent simulations

Using explicit solvent molecular dynamics simulations, we were able to obtain direct observations of shifts in the hydrogen-bonding register of an intermolecular β-sheet protein-peptide complex. The β-sheet is formed between the FHA domain of cancer marker protein Ki67 (Ki67FHA) and a peptide fragment of the hNIFK signaling protein. Potential encounter complexes of the Ki67FHA receptor and hNIFK peptide are misregistered states of the β-sheet. Rearrangements of one of these misregistered states to the native state were captured in three independent simulations. All three rearrangements occurred by a common mechanism: an aromatic residue of the peptide (F263) anchors into a transient hydrophobic pocket of the receptor to facilitate the formation of native hydrogen bonds. To our knowledge, these simulations provide the first atomically detailed visualizations of a mechanism by which nature might correct for errors in the alignment of intermolecular β-sheets.     

Crystal structure of the FHA domain of the Chfr mitotic checkpoint protein and its complex with tungstate

The Chfr mitotic checkpoint protein is frequently inactivated in human cancer. We determined the three-dimensional structure of its FHA domain in its native form and in complex with tungstate, an analog of phosphate. The structures revealed a beta sandwich fold similar to the previously determined folds of the Rad53 N- and C-terminal FHA domains, except that the Rad53 domains were monomeric, whereas the Chfr FHA domain crystallized as a segment-swapped dimer. The ability of the Chfr FHA domain to recognize tungstate suggests that it shares the ability with other FHA domains to bind phosphoproteins. Nevertheless, differences in the sequence and structure of the Chfr and Rad53 FHA domains suggest that FHA domains can be divided into families with distinct binding properties.     

Direct Observations of Shifts in the β-Sheet Register of a Protein-Peptide Complex Using Explicit Solvent Simulations

Using explicit solvent molecular dynamics simulations, we were able to obtain direct observations of shifts in the hydrogen-bonding register of an intermolecular β-sheet protein-peptide complex. The β-sheet is formed between the FHA domain of cancer marker protein Ki67 (Ki67FHA) and a peptide fragment of the hNIFK signaling protein. Potential encounter complexes of the Ki67FHA receptor and hNIFK peptide are misregistered states of the β-sheet. Rearrangements of one of these misregistered states to the native state were captured in three independent simulations. All three rearrangements occurred by a common mechanism: an aromatic residue of the peptide (F263) anchors into a transient hydrophobic pocket of the receptor to facilitate the formation of native hydrogen bonds. To our knowledge, these simulations provide the first atomically detailed visualizations of a mechanism by which nature might correct for errors in the alignment of intermolecular β-sheets.     

The molecular basis of FHA domain:phosphopeptide binding specificity and implications for phospho-dependent signaling mechanisms

Forkhead-associated (FHA) domains are a class of ubiquitous signaling modules that appear to function through interactions with phosphorylated target molecules. We have used oriented peptide library screening to determine the optimal phosphopeptide binding motifs recognized by several FHA domains, including those within a number of DNA damage checkpoint kinases, and determined the X-ray structure of Rad53p-FHA1, in complex with a phospho-threonine peptide, at 1.6 A resolution. The structure reveals a striking similarity to the MH2 domains of Smad tumor suppressor proteins and reveals a mode of peptide binding that differs from SH2, 14-3-3, or PTB domain complexes. These results have important implications for DNA damage signaling and CHK2-dependent tumor suppression, and they indicate that FHA domains play important and unsuspected roles in S/T kinase signaling mechanisms in prokaryotes and eukaryotes.     

Identification of Ser/Thr kinase and Forkhead Associated Domains in Mycobacterium ulcerans: Characterization of Novel Association between Protein Kinase Q and MupFHA

Background

Mycobacterium ulcerans, the causative agent of Buruli ulcer in humans, is unique among the members of Mycobacterium genus due to the presence of the virulence determinant megaplasmid pMUM001. This plasmid encodes multiple virulence-associated genes, including mup011, which is an uncharacterized Ser/Thr protein kinase (STPK) PknQ.

Methodology/Principal Findings

In this study, we have characterized PknQ and explored its interaction with MupFHA (Mup018c), a FHA domain containing protein also encoded by pMUM001. MupFHA was found to interact with PknQ and suppress its autophosphorylation. Subsequent protein-protein docking and molecular dynamic simulation analyses showed that this interaction involves the FHA domain of MupFHA and PknQ activation loop residues Ser¹⁷⁰ and Thr¹⁷⁴. FHA domains are known to recognize phosphothreonine residues, and therefore, MupFHA may be acting as one of the few unusual FHA-domain having overlapping specificity. Additionally, we elucidated the PknQ-dependent regulation of MupDivIVA (Mup012c), which is a DivIVA domain containing protein encoded by pMUM001. MupDivIVA interacts with MupFHA and this interaction may also involve phospho-threonine/serine residues of MupDivIVA.

Conclusions/Significance

Together, these results describe novel signaling mechanisms in M. ulcerans and show a three-way regulation of PknQ, MupFHA, and MupDivIVA. FHA domains have been considered to be only pThr specific and our results indicate a novel mechanism of pSer as well as pThr interaction exhibited by MupFHA. These results signify the need of further re-evaluating the FHA domain –pThr/pSer interaction model. MupFHA may serve as the ideal candidate for structural studies on this unique class of modular enzymes.     

Phosphoprotein and phosphopeptide interactions with the FHA domain from Arabidopsis kinase-associated protein phosphatase

FHA domains are phosphoThr recognition modules found in diverse signaling proteins, including kinase-associated protein phosphatase (KAPP) from Arabidopsis thaliana. The kinase-interacting FHA domain (KI-FHA) of KAPP targets it to function as a negative regulator of some receptor-like kinase (RLK) signaling pathways important in plant development and environmental responses. To aid in the identification of potential binding sites for the KI-FHA domain, we predicted (i) the structure of a representative KAPP-binding RLK, CLAVATA1, and (ii) the functional surfaces of RLK kinase domains using evolutionary trace analysis. We selected phosphopeptides from KAPP-binding Arabidopsis RLKs for in vitro studies of association with KI-FHA from KAPP. Three phosphoThr peptide fragments from the kinase domain of CLV1 or BAK1 were found to bind KI-FHA with KD values of 8-20 microM, by NMR or titration calorimetry. Their affinity is driven by favorable enthalpy and solvation entropy gain. Mutagenesis of these three threonine sites suggests Thr546 in the C-lobe of the BAK1 kinase domain to be a principal but not sole site of KI-FHA binding in vitro. The brassinosteroid receptor BRI1 and KAPP are shown to associate in vivo and in vitro. Further genetic studies indicate that KAPP may be a negative regulator of the BRI1 signaling transduction pathway. 15N-Labeled KI-FHA was titrated with the GST-BRI1 kinase domain and monitored by NMR. BRI1 interacts with the same 3/4, 4/5, 6/7, 8/9, and 10/11 recognition loops of KI-FHA, with similar affinity as the phosphoThr peptides.     

Structural delineation of MDC1-FHA domain binding with CHK2-pThr68

Mammalian MDC1 interacts with CHK2 in the regulation of DNA damage-induced S-phase checkpoint and apoptosis, which is directed by the association of MDC1-FHA and CHK2-pThr68. However, different ligand specificities of MDC1-FHA have been reported, and no structure is available. Here we report the crystal structures of MDC1-FHA and its complex with a CHK2 peptide containing pThr68. Unlike other FHA domains, MDC1-FHA exists as an intrinsic dimer in solution and in crystals. Structural and binding analyses support pThr+3 ligand specificity and provide structural insight into MDC1-CHK2 interaction.     

Structural and functional versatility of the FHA domain in DNA-damage signaling by the tumor suppressor kinase Chk2

The Chk2 Ser/Thr kinase plays crucial, evolutionarily conserved roles in cellular responses to DNA damage. Identification of two pro-oncogenic mutations within the Chk2 FHA domain has highlighted its importance for Chk2 function in checkpoint activation. The X-ray structure of the Chk2 FHA domain in complex with an in vitro selected phosphopeptide motif reveals the determinants of binding specificity and shows that both mutations are remote from the peptide binding site. We show that the Chk2 FHA domain mediates ATM-dependent Chk2 phosphorylation and targeting of Chk2 to in vivo binding partners such as BRCA1 through either or both of two structurally distinct mechanisms. Although phospho-dependent binding is important for Chk2 activity, previously uncharacterized phospho-independent FHA domain interactions appear to be the primary target of oncogenic lesions.     

NMR structure of the N-terminal domain of Saccharomyces cerevisiae RNase HI reveals a fold with a strong resemblance to the N-terminal domain of ribosomal protein L9.

In addition to the conserved and well-defined RNase H domain, eukaryotic RNases HI possess either one or two copies of a small N-terminal domain. The solution structure of one of the N-terminal domains from Saccharomyces cerevisiae RNase HI, determined using NMR spectroscopy, is presented. The 46 residue motif comprises a three-stranded antiparallel beta-sheet and two short alpha-helices which pack onto opposite faces of the beta-sheet. Conserved residues involved in packing the alpha-helices onto the beta-sheet form the hydrophobic core of the domain. Three highly conserved and solvent exposed residues are implicated in RNA binding, W22, K38 and K39. The beta-beta-alpha-beta-alpha topology of the domain differs from the structures of known RNA binding domains such as the double-stranded RNA binding domain (dsRBD), the hnRNP K homology (KH) domain and the RNP motif. However, structural similarities exist between this domain and the N-terminal domain of ribosomal protein L9 which binds to 23 S ribosomal RNA.