Structural basis of p63a SAM domain mutants involved in AEC syndrome Aruna Sathyamurthy 1 , Stefan M. V. Freund 2 , Christopher M. Johnson 2 , Mark D. Allen 2 and Mark Bycroft 2 1 MRC Centre for Protein Engineering, Cambridge, UK 2 MRC Laboratory of Molecular Biology, Cambridge, UK Keywords 5-helix bundle; AEC syndrome; mutations; p53; p63; p73; sterile alpha motif Correspondence M. D. Allen, MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 0QH, UK Fax: +44 (0)1223 213556 Tel: +44 (0)1223 402409 E-mail: mda201@mrc-lmb.cam.ac.uk (Received 9 July 2010, revised 11 April 2011, accepted 24 May 2011) doi:10.1111/j.1742-4658.2011.08194.x p63 is a member of the p53 tumour suppressor family that includes p73. The p63 gene encodes a protein comprising an N-terminal transactivation domain, a DNA binding domain and an oligomerization domain, but var- ies in the organization of the C-terminus as a result of complex alternative splicing. p63a contains a C-terminal sterile a motif (SAM) domain that is thought to function as a protein–protein interaction domain. Several mis- sense and heterozygous frame shift mutations, encoded within exon 13 and 14 of the p63 gene, have been identified in the p63a SAM domain in patients suffering from ankyloblepharon–ectodermal dysplasia–clefting syn- drome. Here we report the solution and high resolution crystal structures of the p63a SAM domain and investigate the effect of several mutations (L553F ⁄ V, C562G ⁄ W, G569V, Q575L and I576T) on the stability of the domain. The possible effects of other mutations are also discussed. Database Coordinates are available in the Protein Data Bank database under the accession numbers 2Y9U and 2Y9T. Introduction p63 is a member of the p53 tumour suppressor family that includes p73; however, p63 does not function as a classical tumour suppressor and is rarely mutated in human cancers. Sequence homology and initial studies revealed that p63 could act as a DNA-sequence spe- cific transcription factor for target genes that leads to apoptosis or cell-cycle arrest and suggested a function for p63 as a tumour suppressor [1,2]. Unlike p53- knockout mice that developed spontaneous tumours, p63-knockout mice exhibited developmental defects implying that p63 plays an important role in mamma- lian development [3–5]. Consistent with this, mutations of p63 are primarily associated with developmental dis- orders [6–14]. p63 and p73 genes share significant sequence identity with each other and with p53 in the N-terminal transactivation domain (TAD), the DNA binding domain and the oligomerization domain, but differ in having an extended C-terminal coding region that undergoes complex alternative splicing to form six different isoforms [1]. TAp63a, TAp63b and TAp63c contain the N-terminal TAD, whereas the DNp63a, DNp63b and DNp63c are transcribed from an alterna- tive internal promoter and lack the full-length TAD. Only splice variants a in both p63 and p73 have a five helical bundle domain at the C-terminus called a sterile a motif (SAM). The p63 gene undergoes mutations in its DNA bind- ing domain, causing ectrodactyly–ectodermal dysplasia– cleft lip ⁄ palate (EEC) syndrome, where it is predicted to lose its DNA binding capacity [15,16]. Several missense Abbreviations AEC, ankyloblepharon–ectodermal dysplasia–clefting; EEC, ectrodactyly–ectodermal dysplasia–cleft lip ⁄ palate; SAM, sterile a motif; TAD, transactivation domain. 2680 FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS mutations with amino acid substitutions and heterozy- gous frame shift mutations, encoded within exon 13 of the p63 gene, have also been identified in the p63a SAM domain in patients suffering from ankyloblepharon– ectodermal dysplasia–clefting (AEC) syndrome [17–19]. The AEC phenotype is similar to that of the EEC patients and yet differs distinctly in its features. p53, TAp63 and TAp73 can all activate p53- response elements through their TADs that share significant homologies. The activation of promoters for MDM2 and p21, and the repression of Hsp70 pro- moter, appears to depend on the isoform of p63. TAp63 proteins with a mutated SAM domain or no SAM domain (TAp63b and TAp63c) all showed a dis- tinct activation of MDM2 and p21, and a decreased repression of Hsp70 [20], compared with wild-type TAp63a. As such, the SAM domain of p63 has been implicated in the inhibition of p63-mediated transacti- vation of both MDM2 and p21. This result is further supported by mutations occurring in the p63 SAM domain resulting in the activation of Hsp70. Mutations can inactivate a protein either by altering functionally important residues or by the global desta- bilization of the overall fold. It is important to know which mechanism is occurring in order to be able to correctly identify functional sites in the protein. The solution structure of p63a SAM domain 1RG6 pro- vided an insight into how several of the AEC muta- tions observed may cause destabilization of the domain, and indeed two of the mutations studied by the group could only be analysed using molecular dynamics. We have determined high resolution crystal and solution structures of the wild-type p63a SAM domain in order to understand the molecular conse- quences of AEC mutations observed in human p63a. We have characterized the stability of several addi- tional mutant p63a SAM domains using chemical denaturation to investigate the contribution of each mutation to domain stability and function. The major- ity of the mutations so far observed in human p63a SAM domains that lead to AEC syndrome appear to cause destabilization of the domain. Results and Discussion Domain boundary selection Attempts to crystallize the p63a SAM domain as a construct comprising residues 543–622 were unsuccess- ful. The precise domain boundaries of the p63a SAM domain were subsequently determined by NMR relax- ation experiments (Fig. S1) and a solution structure of the domain was obtained. The structure revealed that only residues 545–611 are involved in secondary structure elements. A truncated construct comprising these residues readily crystallized under a number of conditions. NMR sample preparation and structure determination The initial p63a SAM domain construct (residues 543– 622) was used to determine the NMR solution struc- ture. Complete 1 H ⁄ 13 C ⁄ 15 N assignments and structure determination were carried out using standard methods as described in the Materials and methods. A summary of all conformational constraints and statistics is pre- sented in Table 1. Comparison of our solution struc- ture with a previously published solution structure of the p63a SAM domain [21] (PDB 1RG6) revealed that the global fold and arrangement of helices is essentially the same (Fig. 1) with an rmsd of 1.36 A ˚ . Of interest is a short b-sheet that is occasionally absent in the 1RG6 structure, which brings together the N-terminal region and the third 3 10 -helix, and is defined by the Table 1. Summary of conformational constraints and statistics for the 20 accepted NMR structures of p63a SAM domain. Structural constraints Intra-residue 609 Sequential 303 Medium range (2 £ |i ) j | £ 4) 195 Long range (|i – j | > 4) 219 Dihedral angle constraints 15 TALOS constraints 108 Distance constraints for 37 hydrogen bonds 74 Total 1523 Statistics for accepted structures Statistics parameter (± SD) RMS deviation for distance constraints (A ˚ ) 0.0068 ± 0.0003 RMS deviation for dihedral constraints (°) 0.232 ± 0.020 Mean CNS energy term (kcalÆmol )1 ± SD) E (overall) 66.84 ± 2.56 E (van der Waals) 20.44 ± 1.48 E (distance constraints) 4.92 ± 0.42 E (dihedral and TALOS constraints) 0.81 ± 0.14 RMS deviations from the ideal geometry (± SD) Bond lengths (A ˚ ) 0.0012 ± 0.00004 Bond angles (°) 0.310 ± 0.0041 Improper angles (°) 0.197 ± 0.0068 Average atomic rmsd from the mean structure (± SD) Residues 546–607 (N, C a , C atoms) (A ˚ ) 0.259 ± 0.068 Residues 546–607 (all heavy atoms) (A ˚ ) 0.706 ± 0.067 A. Sathyamurthy et al. Mutants involved in AEC syndrome FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS 2681 presence of an HA-HA NOE (548–572) and two HA-HN NOEs (548–573, 572–549) between the two regions. A corresponding b-sheet region is absent in the p73a SAM domain [22] (PDB 1DXS). Crystal structure The crystal structure of the truncated construct (545–611) of p63a SAM was solved by molecular replacement using the NMR structure as a search model. The crystal structure was refined to 1.6 A ˚ and is consistent with the NMR structure (rmsd of 0.97 A ˚ ). The crystallographic data are summarized in Table 2. The crystal structure enabled analysis of the side-chain atoms of surface residues implicated in AEC muta- tions, and in particular confirmed the presence of salt bridges and hydrogen bonds which could only have been suggested using the NMR structures. Comparing p63a with p73a and other SAM domains The main feature of the p63a SAM domain is a five helical bundle comprising four a-helices and a short 3 10 -helix. Unlike the other structurally homologous SAM domains, there is a short and distinct b-sheet, which brings together the N-terminus and the third 3 10 -helix. Helices 1 and 5 are antiparallel and form a compact hydrophobic core with the other three helices. A sequence alignment of p63- and p73-like SAM domains in different organisms (Fig. 2) highlights the importance of several residues. The aliphatic isoleucine and leucine residues (I549, L553, L556, I573, I576, L584, L587, I589, I597 and I601) that are part of a compact hydrophobic core are highly conserved in all SAM domains. G557 is conserved throughout suggest- ing an important role in forming a turn before the C-X-X-C motif. Interestingly, this sequence is not always present and can be replaced by an L-Q ⁄ G-A-Y motif. Several surface-exposed aspartates, lysines, argi- nines and serines are also highly conserved. The two highly conserved residues F552 and F565 participate in the formation of the hydrophobic core. F593 is par- tially solvent exposed and can be substituted for either histidine or tyrosine in other SAM domains. The fully conserved tryptophan at position 598 in both p63a and p73a SAM is solvent exposed, whereas in homolo- gous SAM domains it participates in hydrophobic core formation. Most of the conserved hydrophobic resi- dues appear to be involved in stabilizing the fold, although some of the solvent-exposed residues may have a functional role. The structural alignments of the p63a (crystal) and p73a [22] SAM domains are in good agreement with an overall rmsd for the backbone atoms of 0.88 A ˚ . The p63a SAM domain differs from p73a by contain- A C B D Fig. 1. (A) Superimposition of the 20 energy-minimized conformers and (B) a MOLSCRIPT [39] ribbon representation of the p63a SAM domain solution structure. (C) The crystal structure of p63a SAM domain, and (D) the crystal structure of p73a SAM domain. Mutants involved in AEC syndrome A. Sathyamurthy et al. 2682 FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS ing a free cysteine (C547) instead of a proline, possibly helping in the formation of a small b-sheet region, and the C-terminus of p63a SAM is significantly longer. The two conserved cysteines (C558 and C561) in p63a SAM domain are also reduced as in the p73a SAM domain. The inclusion of a reducing agent in the crys- tallization of p63a SAM domain would potentially have contributed to the reduced state of the cysteine residues, although the functional significance of this conserved motif is still unclear. The C-S-S-C region in p63a forms the beginning of the second a helix and is consistent in both solution and crystal structure. In the case of p73a, the C-P-N-C motif forms a loop rather than a helix, possibly reflecting the effect of a proline on helix termination. The p63a and p73a SAM domains form a distinct subset of the SAM domain family. SAM domains, in general, were initially thought to be involved in pro- tein–protein interactions by forming homo- or hetero- oligomers with other SAM domains or by interacting with other proteins [23]. The SAM domain of the pro- tein Smaug, however, has been shown to participate in RNA binding [24,25]. The exact role of p63a and p73a SAM domains is still unclear, but it has been speculated that the SAM domain of both p63a and p73a have lipid binding properties [26]. The p63a and p73a SAM domains do not form homo-oligomers or associate with each other. Analysis of mutations involved in the AEC syndrome In the AEC syndrome, the SAM domain of p63a is found to undergo a number of mutations. A number of point mutations have been identified in exons 13 and 14 of the p63 gene. These mutations can be classified into groups according to their sequence position. I549T, F552S, L553F, L553V, C561G, C561W, F565L, I576T, L584P and I597T are present in the hydrophobic core whereas G557V, G569V, T572P, Q575L, S580F, S580P, S580Y, D583C, D583Y, P590L, F593S, R594P, G600V and G600D all occur on the solvent-exposed surface. The positions of these mutations can be seen in Fig.3. To analyse the effect of these mutations, the p63a SAM domain containing some of the point mutations above were cloned and expressed. The expression of point mutations found in the AEC syndrome varied significantly. The wild-type protein and the mutants L553V and C561W were found to be over-expressed in the soluble fraction. In contrast, mutants L553F, C561G, G569V, Q575L and I576T were present only as inclusion bodies, but could be refolded easily during purification. The mutants G569V, Q575L and I576T partially aggregated at the gel filtration stage but suffi- cient material was purified to allow analysis. The sta- bilities of the expressed mutants were determined by analysing their respective denaturation curves (Fig. S2) using the intrinsic fluorescence of W598 as the probe of measurement. The measure of stability of these pro- teins was determined by the amount of the denaturant required to get 50% of the protein in a denatured state ([D] 50% ). The m-value, a constant related to the change in the surface-exposed surface area upon denaturation, gives a very good idea of the degree of unfolding and was calculated for all mutants. The energetics of some of the mutations in the p63a SAM domain are shown in Table 3. The denaturation curves obtained for mutants were substantially different from that of wild- type with respect to both the pre-denaturation slope and the m-value of denaturation. A lower value of m D–N may indicate denaturation from an already par- tially unfolded structure, as a result of a mutation. However, it could also reflect unfolding to an incom- pletely denatured state which would also reduce the change in surface-exposed surface area upon denatur- Table 2. Crystallographic summary of p63a SAM domain. Data set Native Symmetry P2 1 2 1 2 1 Wavelength (A ˚ ) 0.9795 Resolution range (A ˚ ) 27.4–1.6 Unique reflections 8205 Completeness (%) a 98.9 (96.0) R merge b 0.108 (0.302) Multiplicity c 5.3 (5.5) I ⁄ rI a 14.0 (4.7) a = 34.681 A ˚ , b = 38.336 A ˚ , c = 44.471 A ˚ Model refinement Resolution range (A ˚ ) 22.0–1.6 No. of residues A: 545–611 No. of water, ligand molecules 81, 1 sulphate R work ⁄ R free (%) d 0.185, 0.206 B average e 20.29 A ˚ 2 Geometry bonds ⁄ angles f 0.009 A ˚ , 1.192° Ramachandran g 95.1%, 0.0% PDB ID h 2Y9U a Signal to noise ratio of intensities, highest resolution bin in brack- ets. b R m : RhRi |I(h, i ) ) I(h)| ⁄ RhRiI(h, i ) where I(h, i ) are symmetry- related intensities and I(h) is the mean intensity of the reflection with unique index h. c Multiplicity for unique reflections. d 5% of reflections were randomly selected for determination of the free R factor, prior to any refinement. e Temperature factors averaged for all atoms. f RMS deviations from ideal geometry for bond lengths and restraint angles (Engh and Huber). g Percentage of residues in the ‘most favoured region’ of the Ramachandran plot and per- centage of outliers ( PROCHECK). h Protein Data Bank identifiers for coordinates. A. Sathyamurthy et al. Mutants involved in AEC syndrome FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS 2683 ation. The pronounced slopes of the pre-transition phase could be consistent with the former explanation, indicating non-cooperative partial unfolding over this concentration of guanidinium hydrochloride or reflect- ing a more open and dynamic structure as a starting point. Compared with the wild-type, mutations L553F, C561G, C561W, G569V, Q575 and I576T are all signifi- cantly destabilized, as judged by the reduction in [D] 50% and DG H 2 O DÀN . A possible explanation for the instability of each of these point mutations was obtained by looking at the structural features of the wild-type protein. Muta- tion L553F, which is close to F552, would probably cause a severe steric clash between the two phenylala- nine rings and result in overcrowding in the hydropho- bic core. The tryptophan ring in C561W would possible result in a steric clash with the aromatic ring of F593. ** ** T Y PT ** * * ** TPWVFVL * α4 α5 α1 α2 α3 β1 β2 NN P PYP T D CSI V S FLA R L GCSS C LDYF T TQGL T TIYQ P PYP T D CSI V S FLA R L GCSS C LDYF T TQGL T TIYQ P PYP T D CSI V S FLA R L GCSS C VDYF T TQGL T TIYH P PYP M D NSI S S FLL R L GCSA C LDYF T AQGL T NIYQ P PYH A D PSL V S FLT G L GCPN C IEYF T SQGL Q SIYH P PYH A D PSL V S FLT G L GCPN C IECF T SQGL Q SIYH K CEP T E NTI A Q WLT K L GLQA Y IDNF Q QKGL H NMFQ Q NDM Q D NSV S T WLN A L GLGA Y IDGF H EQNL Y SLLQ N GEM T D ISV A A WLN H L GLGA Y IDSF H EHNL Y SVIQ DGA DL LSIS R WLSN IMEKY TQE F I KHG F K VCGH AV M IEH Y S MDD L A SLK I P EQFR H AIWK G ILDH R QL IEH Y S MDD L A SLK I P EQFR H AIWK G ILDH R QL IEH Y S MDD L V SLK I P EQFR H AIWK G ILDH R QL IEN Y N LED L S RLK I P TEFQ H IIWK G IMEY R QT LQN L T IED L G ALK I P EQYR M TIWR G LQDL K QG LQN L T IED L G ALK V P DQYR M TIWR G LQDL K QS LDE F T LED L Q SMR I G TGHR N KIWK S LLDY R RL LDD F S LDD L A KMK I G NSHR N KIWK S LLEL R NQ H EF H DF H DF M EF H DY H DC L SS G FT LDD F S LDD L A KMK I G NAHR N KIWK S VLEL R NE G LT L N S Y SD KKII K NMED C K KIS A Y LLE S N FS S GN Q9H3D4 O88898 Q9DEC7 Q8JHZ6 O15350 Q9JJP2 Q27937 Q9NGC7 Q8T7V3 B3RZS6 HUMAN P63 576 MOUSE P63 576 GAL G P63 478 DAN R P63 482 HUMAN P73 520 MOUSE P73 514 LOL F P53 486 MYA A P73 520 SPI S P53 492 TRI A P53 501 Q9H3D4 O88898 Q9DEC7 Q8JHZ6 O15350 Q9JJP2 Q27937 Q9NGC7 Q8T7V3 B3RZS6 HUMAN P63 541 MOUSE P63 541 GAL G P63 443 DAN R P63 447 HUMAN P73 485 MOUSE P73 479 LOL F P53 451 MYA A P73 485 SPI S P53 457 TRI A P53 466 610 610 512 516 554 548 520 554 526 535 575 575 477 481 519 513 485 519 491 500 * * * VP L * * V L * S F P * S V G D Y Fig. 2. CLUSTALW sequence alignment of SAM domains. Residues that are absolutely conserved are shown in red. Conserved and partially conserved hydrophobic residues are shown in black and grey, respectively. Con- served and partially conserved charged resi- dues are shown in blue and light blue, respectively. The positions and type of AEC mutations are shown above the sequence alignment. A diagrammatic representation of the domain is shown below the alignment. I549 L553 G557 C561 G569 T572 Q575 S580 I576 R594 I597 F565 D581 L584 F593 P590 F552 Fig. 3. Ribbon representation of the p63a SAM domain showing the position of mutations that are associated with AEC syndrome. Table 3. Thermodynamic values of p63a SAM domain AEC mutants. DG H 2 O DÀN is an estimate of the stability of the proteins in buffer calcu- lated from DG H 2 O DÀN ¼ m½D 50% , where m D–N is the m value for denaturation of the protein. The mutants G534V and I541T were highly destabi- lized and their denaturation curves could not be fitted reliably. These mutants proved problematic to express and purify (see Results and Discussion) consistent with this destabilization. Protein Location of mutation [D] 50% (M) m D–N (kcalÆmol )1 ) DG H 2 O DÀN (kcalÆmol )1 ) DDG H 2 O DÀN (kcalÆmol )1 ) Wild-type – 3.50 (± 0.02) 2.31 (± 0.1) 8.11 (± 0.35) – L553V Core 3.38 (± 0.09) 1.31 (± 0.17) 4.43 (± 0.59) 3.68 L553F Core 1.79 (± 0.16) 0.96 (± 0.06) 1.72 (± 0.19) 6.39 C561G Core 1.61 (± 0.17) 1.11 (± 0.08) 1.79 (± 0.23) 6.32 C561W Core 2.59 (± 0.12) 1.12 (± 0.13) 2.90 (± 0.36) 5.21 G569V Surface Highly destabilized Highly destabilized Highly destabilized Highly destabilized Q575L Surface 2.66 (± 0.04) 1.85 (± 0.17) 4.92 (± 0.46) 3.19 I576T Core Highly destabilized Highly destabilized Highly destabilized Highly destabilized Mutants involved in AEC syndrome A. Sathyamurthy et al. 2684 FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS The instability of C561G is most probably caused by the formation of a hydrophobic cavity resulting from the loss of a bulky thiol group. The cavity created would potentially cause some rearrangement of the hydropho- bic core and the associated instability. Introduction of a polar residue into the hydrophobic core, as happens with mutation I576T, would potentially lower the stabil- ity of the domain. G569 adopts a positive angle of phi in the wild-type domain, something rarely seen for ali- phatic residues. As such, mutation G569V may cause instability due to the residues in the loops adopting a more unfavourable conformation. Mutation Q575L occurs in a solvent-exposed position and the side-chain carbonyl of Q575 forms a hydrogen bond with the main-chain amides of both T571 and T572 (Fig. S4). As such, any mutation, with the exception of Q575E, would potentially destabilize the domain. Mutations G569V and T572P were analysed by another group using molec- ular simulations due to protein instability and degrada- tion during purification [21]. The side-chain hydroxyl of T572 forms an N-cap hydrogen bond with the main- chain amide of Q575 and thus a mutation to proline would abolish the potential to form this hydrogen bond. L553V was the only conservative AEC-inducing muta- tion analysed but it also produced a significant change in DG H 2 O DÀN at 25 °C. The location of the L553V mutation in the core of the domain precludes this being a func- tional mutation and hence the mutation also appears to be structural, possibly by creating a small hydrophobic cavity in the hydrophobic core. The different pre-denaturation slope and m D–N val- ues for all the mutants may indicate that they were partially denatured or unfolded even under non-dena- turing conditions. As such any attempt to crystallize the domains would be almost impossible, and indeed we were unable to crystallize any of the mutant domains. The change in stability at 25 °C may be enough to result in sufficient instability at 37 °Cto cause complete unfolding of the domain and hence a loss of domain function in vivo. Similar effects have been observed with the human p53 core domain that has a stability of 7.5 kcalÆmol )1 at 25 °C and 3.0 kcalÆ- mol )1 at 37 °C [27]. Structural stability mutants were found with p53 core domain with changes in stability in a similar range to those for p63a SAM domains. Potential effects of other mutations Mutations I549T, F552S, G557V, F565L, S580F, S580P, S580Y D583C, D583Y, L584P, P590L, F593S, R594P, I597T, G600V and G600D were not tested experimen- tally, but analysis of the wild-type crystal structure potentially explains how some of the mutations might cause significant loss of stability. Mutations I549T and I597T would also introduce a polar residue into the hydrophobic core, and presumably these mutants would be destabilized in a similar manner to mutation I576T. Mutations F552S and F565L would both result in the loss of a large aromatic residue from the hydrophobic core. G557 adopts a positive angle of phi in the wild-type domain, and hence mutation G557V may cause instabil- ity due to the residues in the loops adopting a more unfa- vourable conformation in a similar manner to that observed for mutation G569V. Mutations S580F, S580P and S580Y would all result in the loss of the N-cap hydrogen bond from the main-chain amide of D583 to the side-chain hydroxyl of S580, whilst mutations D583C and D583Y would disrupt a hydrogen bond from the main-chain amide of S580 to the side-chain car- bonyl oxygen of D583 (see Fig. S3). Mutation R594P would disrupt a standard i–i+4 hydrogen bond in helix 5. Mutations G600V and G600D would probably result in a steric clash between a-helix 5 and L556. Finally, the in-frame 3 bp insert (573–574 inserting TTC) encoding an additional phenylalanine residue would be expected to be destabilizing as it probably disrupts the packing of the 3 10 -helix to the rest of the protein. Of all the muta- tions only P590L and F593S could not easily be explained as mutations that would either disrupt the hydrophobic core or result in a loss of a stabilizing hydrogen bond or salt bridge. Indeed F593 is a solvent- exposed aromatic residue and a mutation to a polar resi- due might even be expected to increase stability. Several charged and hydrophilic residues are conserved amongst the p63a and p73a SAM domains (Fig. 2). Some of these charge residues may contribute to the stability of the domain by forming hydrogen bonds or salt bridges (D546, E577, R594, R605, H608), whilst others (D563, Q568, D582, K588, Q592, K599 and D603) are solvent exposed and could poten- tially be involved in domain function. It is interesting to note that the two mutations that appear not to be structural mutations, P590L and F593S, are close in sequence to two of the highly conserved hydrophilic residues (K588 and Q592) and may indicate that this interface plays some role in domain function. It is becoming clear that the SAM domain is a very versatile protein module and more work is clearly needed to elucidate its function in p63. The develop- mental malformations in AEC syndrome are primarily caused by mutations in the p63a SAM domain that appear to cause a significant destabilization of the SAM domain. So far, no equivalent mutations have been identified in the p73a isoform. The residues involved in p63 mutations are conserved in p73a SAM domains, and yet only the p63a gene is found to undergo hetero- A. Sathyamurthy et al. Mutants involved in AEC syndrome FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS 2685 zygous in-frame insertion and point mutations in nat- ure. The results presented here provide an insight into the role of domain stability in the developmental mal- formations observed in AEC. Our high resolution crys- tal structure will contribute to an understanding of the potential destabilizing or functional effects of AEC mutations within the p63a SAM domain. Materials and methods Cloning, expression and purification of the p63a SAM domain All reagents were purchased from Sigma (Sigma-Aldrich Corp., St Louis, MO, USA) and were Anal-R grade or higher, with the exception of ultrapure guanidine hydrochloride which was purchased from ICN Biomedicals Inc. (Aurora, OH, USA). Human p63a SAM domain (543– 621) gene was codon optimized and synthesized using overlapping primers and cloned into a modified pRSETa expression vector [28]. The shortened construct (545–611) was subcloned from the full-length gene into the same vec- tor. Seven mutants involved in the AEC syndrome – L553F, L553V, C561G, C561W, G569V, Q575L and I576T – were made using the Quickchange kit. The plasmids were trans- formed into Escherichia coli C-41 host cells and grown in 2YT medium containing 50 lgÆmL )1 ampicillin. These cul- tures were induced with 1 m M isopropyl thio-b-D-galactoside with A 600 = 0.8 and harvested after 4 h at 37 °C by centrifu- gation. Isotopically labelled p63 SAM domain was prepared by growing cells in K-MOPS minimal media [29] containing 15 NH 4 Cl and ⁄ or [ 13 C]-glucose. The resulting protein was purified by Ni 2+ -nitrilotriacetic acid affinity chromatogra- phy, TEV protease digestion and a second Ni 2+ -nitrilotri- acetic acid affinity chromatography to remove the lipoyl domain fusion tag. Final purification was performed with gel filtration using a HiLoad 26 ⁄ 60 Superdex 75 column. SDS ⁄ PAGE and MALDI-TOF mass spectrometry were used to confirm proteins were of the expected mass and to assess their purity. The mutants L553F, C561G, G569V, Q575L and I576T were present as inclusion bodies. The pro- tein was solubilized in 8 M guanidinium hydrochloride and refolded on the Ni 2+ -nitrilotriacetic acid affinity column. The homogeneity of the refolded material was confirmed by gel filtration and fluorescence measurements. Folding analysis All folding experiments were performed at 298 K using a buffer of 50 m M sodium phosphate pH 7.0 and 10 mM dithiothreitol. Guanidium hydrochloride was used as denaturant in the unfolding experiments. Tryptophan fluo- rescence was used as a monitor for protein denaturation of p63a SAM domain upon the addition of guanidinium hydrochloride. The tryptophan residue W598 of p63a SAM domain (and tryptophan W561 in the case of mutant C561W) was used as a probe to monitor folding. Excitation was at 280 nm using a 4 nm bandwidth and emitted light was collected after passage through a 325 nm bandpass. Denaturation experiments were performed on a Hitachi F4500 spectrofluorimeter. Data from denaturation experi- ments were fitted to equations assuming two-state kinetics. Fitting of data was performed using the KALEIDAGRAPH version 3.6 and PRISM version 4.0a softwares. NMR spectroscopy and structure determination Protein samples prepared for NMR spectroscopy experi- ments were typically 1.5 m M in 90% H 2 O, 10% D 2 O, con- taining 50 m M potassium phosphate, pH 6.5, 200 mM NaCl and 5 m M d-dithiothreitol. All spectra were acquired using either a Bruker DRX800 or DRX600 spectrometer equipped with pulsed field gradient triple resonance at 25 °C, and referenced relative to external sodium 2,2-dimethyl-2-silapentane-5-sulfonate for proton and car- bon signals, or liquid ammonium for that of nitrogen. Assignments were obtained using standard NMR methods using 13 C ⁄ 15 N-labelled, 15 N-lablled, 10% 13 C-labelled and unlabelled p63 NMR samples [30,31]. Backbone assign- ments were obtained using the following standard set of 2D and 3D heteronuclear spectra: 1H-15N HSQC (Fig. S4), HNCACB, CBCA(CO)NH, HACACO, HNCO, CCCONH and 1H-13C HSQC. Additional assignments were made using 2D TOCSY and DQF-COSY spectra. A set of dis- tance constraints were derived from 2D NOESY spectra recorded from a 1.5 m M p63 domain sample with a mixing time of 150 ms. Hydrogen bond constraints were included for a number of backbone amide protons whose signals were still detected after 10 min in a 2D 1 H- 15 N HSQC spec- trum recorded in D 2 O at 278 K (pH 5.0). For hydrogen bond partners, two distance constraints were used where the distance (D) H–O (A) corresponded to 1.5–2.5 A ˚ and (D) N– O (A) to 2.5–3.5 A ˚ . Torsional angle constraints were obtained from an analysis of C¢,N,C a ,H a and C b chemical shifts using the program TALOS [32]. The stereospecific assignments of H b resonances determined from DQF-COSY and HNHB spectra were confirmed by analysing the initial ensemble of structures. Stereospecific assignments of H c and H d resonances of Val and Leu residues, respectively, were assigned using a fractionally 13 C-labelled protein sample [33]. Stereospecific assignments were identified for resolved resonances when the side-chain atoms were sufficiently well defined in the ensemble of structures. The 3D structures of the p63a SAM domain were calculated using the standard torsion angle dynamics simulated annealing protocol in the program CNS 1.2 [34]. Structures were accepted where no distance violation was greater than 0.25 A ˚ and no dihedral angle violations were greater than 5°. The final coordinates have been deposited in the Protein Data Bank (PDB 2Y9T). Mutants involved in AEC syndrome A. Sathyamurthy et al. 2686 FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS Crystallization and structural determination Crystals were obtained using sitting drops containing 2 lLof 14 mgÆmL )1 wild-type p63a SAM domain (510–576) protein in 100 m M sodium citrate, pH 6.4, 500 mM lithium sulphate, 500 m M ammonium sulphate and 5 mM dithiothreitol. Clus- tered plate-like crystals were obtained in 2 days at 4 °C. Sin- gle crystals were obtained by seeding. The crystal contained one molecule per asymmetric unit and grew in the space group P2 1 2 1 2 1 (a = 34.64 A ˚ , b = 38.30 A ˚ , c = 44.52 A ˚ , a = b = c =90°). Crystal was cryo-protected in reservoir buffer with an additional 20% glycerol prior to stream freez- ing. The data set was collected at the European Synchrotron Radiation Facility on beam line ID14-2 to 1.6 A ˚ . Data pro- cessing and integration was done using CCP4 [35]. The struc- ture was solved by molecular replacement using PHASER [36] with the NMR solution structure as a starting model. Struc- ture calculations were done using PHENIX [37]. The model refinement was done using MAIN [38]. The final coordinates have been deposited in the Protein Data Bank (PDB 2Y9U). Acknowledgements We would like to thank the Nehru Trust and the Cambridge Commonwealth Trust for the scholarship award and financial support. References 1 Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD, Do ¨ tsch V, Andrews NC, Caput D & McKeon F (1998) p63, a p53 homolog at 3q27-29, encodes multiple prod- ucts with transactivating, death-inducing, and domi- nant-negative activities. Mol Cell 2, 305–316. 2 Osada M, Ohba M, Kawahara C, Ishioka C, Kana- maru R, Katoh I, Ikawa Y, Nimura Y, Nakagawara A, Obinata M et al. (1998) Cloning and functional analysis of human p51, which structurally and functionally resembles p53. Nat Med 4, 839–843. 3 Yang A,Schweitzer R,Sun D,Kaghad M, Walker N, Bron- son RT, Tabin C, SharpeA, Caput D, Crum Cet al. (1999) p63 isessential for regenerative proliferation in limb, cra- niofacial and epithelial development. Nature398, 714–718. 4 Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR & Bradley A (1999) p63 is a p53 homologue required for limb and epidermal morphogenesis. Nature 398, 708–713. 5 Keyes WM, Vogel H, Koster MI, Guo X, Qi Y, Pether- bridge KM, Roop DR, Bradley A & Mills AA (2006) p63 heterozygous mutant mice are not prone to sponta- neous or chemically induced tumors. Proc Natl Acad Sci USA 103, 8435–8440. 6 Koga F, Kawakami S, Fujii Y, Saito K, Ohtsuka Y, Iwai A, Ando N, Takizawa T, Kageyama Y & Kihara K (2003) Impaired p63 expression associates with poor prognosis and uroplakin III expression in invasive urothelial carcinoma of the bladder. Clin Cancer Res 9, 5501–5507. 7 Park BJ, Lee SJ, Kim JI, Lee SJ, Lee CH, Chang SG, Park JH & Chi SG (2000) Frequent alteration of p63 expression in human primary bladder carcinomas. Cancer Res 60, 3370–3374. 8 Urist MJ, Di Como CJ, Lu ML, Charytonowicz E, Verbel D, Crum CP, Ince TA, McKeon FD & Cordon-Cardo C (2002) Loss of p63 expression is associated with tumor pro- gression in bladder cancer. Am J Pathol 161, 1199–1206. 9 Wang X, Mori I, Tang W, Nakamura M, Nakamura Y, Sato M, Sakurai T & Kakudo K (2002) p63 expres- sion in normal, hyperplastic and malignant breast tis- sues. Breast Cancer 9, 216–219. 10 Celli J, Duijf P, Hamel BC, Bamshad M, Kramer B, Smits AP, Newbury-Ecob R, Hennekam RC, Van Bug- genhout G, van Haeringen A et al. (1999) Heterozygous germline mutations in the p53 homolog p63 are the cause of EEC syndrome. Cell 99, 143–153. 11 van Bokhoven H, Jung M, Smits AP, van Beersum S, Ru ¨ schendorf F, van Steensel M, Veenstra M, Tuerlings JH, Mariman EC, Brunner HG et al. (1999) Limb mammary syndrome: a new genetic disorder with mam- mary hypoplasia, ectrodactyly, and other hand ⁄ foot anomalies maps to human chromosome 3q27. Am J Hum Genet 64, 538–546. 12 van Bokhoven H, Hamel BC, Bamshad M, Sangiorgi E, Gurrieri F, Duijf PH, Vanmolkot KR, van Beusekom E, van Beersum SE, Celli J et al. (2001) p63 Gene mutations in EEC syndrome, limb–mammary syndrome, and isolated split hand–split foot malformation suggest a genotype– phenotype correlation. Am J Hum Genet 69, 481–492. 13 Ianakiev P, Kilpatrick MW, Toudjarska I, Basel D, Beighton P & Tsipouras P (2000) Split-hand ⁄ split-foot malformation is caused by mutations in the p63 gene on 3q27. Am J Hum Genet 67 , 59–66. 14 McGrath JA, Duijf PH, Doetsch V, Irvine AD, de Waal R, Vanmolkot KR, Wessagowit V, Kelly A, Atherton DJ, Griffiths WA et al. (2001) Hay–Wells syn- drome is caused by heterozygous missense mutations in the SAM domain of p63. Hum Mol Genet 10, 221–229. 15 Rinne T, Hamel B, van Bokhoven H & Brunner HG (2006) Pattern of p63 mutations and their phenotypes – update. Am J Med Genet A 140 , 1396–1406. 16 van Bokhoven H & Brunner HG (2002) Splitting p63. Am J Hum Genet 71, 1–13. 17 Rinne T, Brunner HG & van Bokhoven H (2007) p63- associated disorders. Cell Cycle 6, 262–268. 18 Rinne T, Bolat E, Meijer R, Scheffer H & van Bokhoven H (2009) Spectrum of p63 mutations in a selected patient cohort affected with ankyloblepharon– ectodermal defects–cleft lip ⁄ palate syndrome (AEC). Am J Med Genet A 149, 1948–1951. A. Sathyamurthy et al. Mutants involved in AEC syndrome FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS 2687 19 Berk DR, Crone K & Bayliss SJ (2009) AEC syndrome caused by a novel p63 mutation and demonstrating erythroderma followed by extensive depigmentation. Pediatr Dermatol 26, 617–618. 20 Ghioni P, Bolognese F, Duijf PH, Van Bokhoven H, Mantovani R & Guerrini L (2002) Complex transcrip- tional effects of p63 isoforms: identification of novel activation and repression domains. Mol Cell Biol 22, 8659–8668. 21 Cicero DO, Falconi M, Candi E, Mele S, Cadot B, Di Venere A, Rufini S, Melino G & Desideri A (2006) NMR structure of the p63 SAM domain and dynamical properties of G534V and T537P pathological mutants, identified in the AEC syndrome. Cell Biochem Biophys 44, 475–489. 22 Wang WK, Bycroft M, Foster NW, Buckle AM, Fersht AR & Chen YW (2001) Structure of the C-ter- minal sterile alpha-motif (SAM) domain of human p73 alpha. Acta Crystallogr D Biol Crystallogr 57, 545–551. 23 Leone M, Cellitti J & Pellecchia M (2008) NMR studies of a heterotypic Sam-Sam domain association: the inter- action between the lipid phosphatase Ship2 and the EphA2 receptor. Biochemistry 47, 12721–12728. 24 Green JB, Gardner CD, Wharton RP & Aggarwal AK (2003) RNA recognition via the SAM domain of Smaug. Mol Cell 11, 1537–1548. 25 Aviv T, Lin Z, Lau S, Rendl LM, Sicheri F & Smibert CA (2003) The RNA-binding SAM domain of Smaug defines a new family of post-transcriptional regulators. Nat Struct Biol 10, 614–621. 26 Barrera FN, Poveda JA, Gonza ´ lez-Ros JM & Neira JL (2003) Binding of the C-terminal sterile motif (SAM) domain of human p73 to lipid membranes. J Biol Chem 278, 46878–46885. 27 Bullock AN & Fersht AR (2001) Rescuing the function of mutant p53. Nat Rev Cancer 1, 68–76. 28 Dodd RB, AllenMD, Brown SE,Sanderson CM, Duncan LM,Lehner PJ,BycroftM & ReadRJ (2004) Solution structure of theKaposi’ssarcoma-associated herpesvirus K3 N-terminal domainrevealsa novelE2-bindingC4HC3- typeRING domain.JBiol Chem279,53840–53847. 29 Neidhardt FC, Bloch PL & Smith DF (1974) Culture medium for enterobacteria. J Bacteriol 119, 736–747. 30 Bax A, Ikura M, Kay LE, Barbato G & Spera S (1991) Multidimensional triple resonance NMR spectroscopy of isotopically uniformly enriched proteins: a powerful new strategy for structure determination. Ciba Found Symp 161, 108–119; discussion 119–135. 31 Englander SW & Wand AJ (1987) Main-chain-directed strategy for the assignment of 1H NMR spectra of proteins. Biochemistry 26, 5953–5958. 32 Cornilescu G, Delaglio F & Bax A (1999) Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J Biomol NMR 13, 289–302. 33 Neri D, Szyperski T, Otting G, Senn H & Wu ¨ thrich K (1989) Stereospecific nuclear magnetic resonance assign- ments of the methyl groups of valine and leucine in the DNA-binding domain of the 434 repressor by biosyn- thetically directed fractional 13C labeling. Biochemistry 28, 7510–7516. 34 Brunger AT (2007) Version 1.2 of the crystallography and NMR system. Nat Protoc 2, 2728–2733. 35 CCP4 (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr D Biol Crystallogr 50, 760–763. 36 McCoy AJ (2007) Solving structures of protein com- plexes by molecular replacement with Phaser. Acta Crystallogr D Biol Crystallogr 63, 32–41. 37 Adams PD, Grosse-Kunstleve RW, Hung LW, Ioerger TR, McCoy AJ, Moriarty NW, Read RJ, Sacchettini JC, Sauter NK & Terwilliger TC (2002) PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr D Biol Crys- tallogr 58, 1948–1954. 38 Turk D (1996) MAIN 96: An interactive software for density modifications, model building, structure refine- ment and analysis. Proceedings from the 1996 Meeting of the International Union of Crystallography Macromo- lecular Computing School, 1996. 39 Kraulis PJ, Domaille PJ, Campbell-Burk SL, Van Aken T & Laue ED (1994) Solution structure and dynamics of ras p21.GDP determined by heteronuclear three- and four-dimensional NMR spectroscopy. Biochemistry 33, 3515–3531. Supporting information The following supplementary material is available: Fig. S1. Comparison of the B-factors obtained for the backbone amide atoms of the p63a crystal struc- ture with the NMR dynamics of the same domain in solution. Fig. S2. KALEIDOGRAPH plots of the denaturation curves of wild-type protein and mutants L553V, L553F, C561G, C561W and Q575L. Fig. S3. Close up views of several of the residues involved in side-chain hydrogen bonds and salt bridges. Fig. S4. 15 N 1 H HSQC spectrum of p63a SAM domain. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be reorganized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Mutants involved in AEC syndrome A. Sathyamurthy et al. 2688 FEBS Journal 278 (2011) 2680–2688 ª 2011 The Authors Journal compilation ª 2011 FEBS . representation of the p63a SAM domain solution structure. (C) The crystal structure of p63a SAM domain, and (D) the crystal structure of p73a SAM domain. Mutants involved. reflecting the effect of a proline on helix termination. The p63a and p73a SAM domains form a distinct subset of the SAM domain family. SAM domains, in general,