Molecular Basis for SUMOylation-dependent Regulation of DNA Binding Activity of Heat Shock Factor 2

Abstract

Heat shock factor 2 (HSF2) is a member of a vertebrate transcription factor family for genes of heat shock proteins and is involved in the regulation of development and cellular differentiation. The DNA binding property of HSF2 is modulated by the post-translational modification of a specific lysine residue in its DNA binding domain by small ubiquitin-like modifier (SUMO), but the consequences of SUMOylation and its underlying molecular mechanism remain unclear. Here we show the inhibitory effect of SUMOylation on the interaction between HSF2 and DNA based on biochemical analysis using isolated recombinant HSF2. NMR study of the SUMOylated DNA binding domain of HSF2 indicates that the SUMO moiety is flexible with respect to the DNA binding domain and has neither a noncovalent interface with nor a structural effect on the domain. Combined with data from double electron-electron resonance and paramagnetic NMR relaxation enhancement experiments, these results suggest that SUMO attachment negatively modulates the formation of the protein-DNA complex through a randomly distributed steric interference. Heat shock factor 2 (HSF2) is a member of a vertebrate transcription factor family for genes of heat shock proteins and is involved in the regulation of development and cellular differentiation. The DNA binding property of HSF2 is modulated by the post-translational modification of a specific lysine residue in its DNA binding domain by small ubiquitin-like modifier (SUMO), but the consequences of SUMOylation and its underlying molecular mechanism remain unclear. Here we show the inhibitory effect of SUMOylation on the interaction between HSF2 and DNA based on biochemical analysis using isolated recombinant HSF2. NMR study of the SUMOylated DNA binding domain of HSF2 indicates that the SUMO moiety is flexible with respect to the DNA binding domain and has neither a noncovalent interface with nor a structural effect on the domain. Combined with data from double electron-electron resonance and paramagnetic NMR relaxation enhancement experiments, these results suggest that SUMO attachment negatively modulates the formation of the protein-DNA complex through a randomly distributed steric interference. The heat shock response is a cellular defense against environmental and physiological stresses, such as heat, heavy metals, oxygen radicals, and viral or bacterial infections. Eukaryotic cells maintain homeostasis through elevating the expression of heat shock proteins (HSPs), 4The abbreviations used are: HSP, heat shock protein; DBD, DNA binding domain; SUMO, small ubiquitin-like modifier; HSF, heat shock factor; HSE, heat shock element; EMSA, electrophoretic mobility shift assay; GST, glutathione S-transferase; hHSF2FL, full-length human HSF2β; TDG, thymine DNA glycosylase; SPR, surface plasmon resonance; PDB, Protein Data Bank; MTSL, (1-oxyl-2,2,5,5-tetramethyl-d3-pyrroline-3-methyl) methane thiosulfonate; SDSL, site-directed spin labeling; HSQC, heteronuclear single quantum coherence; SIM, SUMO interaction motif; HR, hydrophobic repeat; DEER, double electron-electron resonance; PRE, paramagnetic NMR relaxation enhancement; HtH, head-to-head; TtT, tail-to-tail. which function as molecular chaperones, in response to various stress stimuli. The heat shock response is also evoked during nonstress-induced processes, such as early development and cellular differentiation in mammals (1Akerfelt M. Trouillet D. Mezger V. Sistonen L. Ann. N. Y. Acad. Sci.. 2007; 1113: 15-27Google Scholar). The HSP genes are up-regulated by heat shock elements (HSEs) comprising multiple, and at least three, inverted repeats of the pentanucleotide 5′-nGAAn-3′, where n denotes any nucleotide. The proteins within the heat shock factor (HSF) family, evolutionarily conserved from yeast to human, bind to HSEs and regulate the transcription level of a set of the HSP genes. One of the mammalian members, HSF2, is involved in development and differentiation-related processes, whereas HSF1 is responsive to classical stress stimuli (2Loones M.T. Rallu M. Mezger V. Morange M. Cell. Mol. Life Sci.. 1997; 53: 179-190Google Scholar, 3Sistonen L. Sarge K.D. Phillips B. Abravaya K. Morimoto R.I. Mol. Cell. Biol.. 1992; 12: 4104-4111Google Scholar). Recent genetic and biological studies have revealed that interplays between HSF1 and HSF2 gain functional versatility in the stress response and developmental signaling pathways. For example, HSF1 and HSF2 form heterocomplexes on the promoters of chaperone genes, clusterin and hsp70, and are cooperatively involved in the regulation of these genes (4Loison F. Debure L. Nizard P. le Goff P. Michel D. le Drean Y. Biochem. J.. 2006; 395: 223-231Google Scholar, 5Ostling P. Bjork J.K. Roos-Mattjus P. Mezger V. Sistonen L. J. Biol. Chem.. 2007; 282: 7077-7086Google Scholar). The HSF members are characterized by a conserved domain structure that consists of a winged helix-turn-helix motif near the N terminus followed by an extended hydrophobic heptad repeat (HR-A/B). By itself, the helix-turn-helix DNA binding domain (DBD) of HSFs is capable of binding to at least two inverted HSEs, regardless of the repeat directions, that is head-to-head (nGAAnnTTCn) and tail-to-tail (nTTCnnGAAn) repeats (6Perisic O. Xiao H. Lis J.T. Cell.. 1989; 59: 797-806Google Scholar). However, it has been reported that Drosophila and yeast HSFs preferentially bind to the tail-to-tail elements rather than the head-to-head ones (7Bonner J.J. Ballou C. Fackenthal D.L. Mol. Cell. Biol.. 1994; 14: 501-508Google Scholar, 8Xiao H. Perisic O. Lis J.T. Cell.. 1991; 64: 585-593Google Scholar). The HR-A/B domain facilitates trimerization through the formation of a coiled-coil bundle upon receiving stress signals. The trimer formation enables HSFs to bind the HSE repeats with high affinity, and hence transcription of the downstream HSP genes is activated. Nevertheless, the isolated DBDs of Kluyveromyces lactis have been shown to bind to HSEs in vitro (9Littlefield O. Nelson H.C. Nat. Struct. Biol.. 1999; 6: 464-470Google Scholar); thus it is suggested that isolated DBDs of HSFs retain a binding affinity for HSE repeats and that the trimer formation of HSF1 and HSF2 has a regulative role in DNA binding. In addition to these functional domains, HSF1 and HSF2 possess a C-terminal hydrophobic repeat (HR-C) that suppresses the trimer formation to maintain an inactive state in the absence of stimuli (10Rabindran S.K. Haroun R.I. Clos J. Wisniewski J. Wu C. Science.. 1993; 259: 230-234Google Scholar). Functional properties of mammalian HSFs were found to be modulated by post-translational modification with small ubiquitin-like modifier (SUMO). Modification with SUMO (SUMOylation) has been recognized as an important regulatory mechanism in diverse cellular processes such as nuclear trafficking, transcription, and DNA repair (11Seeler J.S. Dejean A. Nat. Rev.. 2003; 4: 690-699Google Scholar). SUMO shares structural similarity with ubiquitin and can be covalently attached to lysine residues within a ψKXE consensus sequence (ψ indicates hydrophobic residue and X is any amino acid) in target proteins via an isopeptide bond formed through an enzymatic pathway that is similar to but distinct from that in ubiquitination (11Seeler J.S. Dejean A. Nat. Rev.. 2003; 4: 690-699Google Scholar, 12Gill G. Genes Dev.. 2004; 18: 2046-2059Google Scholar, 13Rodriguez C. J. Biol. Chem.. Scholar). SUMO and have been in mammals to H. J. J. Biol. Chem.. Scholar). The of SUMOylation thus are nuclear (11Seeler J.S. Dejean A. Nat. Rev.. 2003; 4: 690-699Google Scholar, G. Dev.. Scholar). transcription the of SUMO a interaction surface the DNA binding to the of distinct for or J. Biol. Chem.. Scholar, G. Mol. Cell. Biol.. 2004; Scholar). In SUMOylation expression by the of G. J. Genes Dev.. 2003; Scholar, L. H. A. F. Genes Dev.. Scholar). The molecular underlying the regulation of protein-DNA by SUMO to be but has been by biochemical and structural studies on a DNA repair thymine DNA D. N. Y. H. K. F. H. H. M. Scholar, J. P. J.. Scholar). The modification with is suggested to a in that is with noncovalent between the covalently and and it to a of DNA binding. The HSF SUMOylation that complex physiological whereas in HSF1 is by or Y. Sarge K.D. J. Biol. Chem.. and in HSF2 are by and J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; Scholar, Y. Sarge K.D. J. Biol. Chem.. Scholar). these in HSF2, which is in a of the DBD, is the SUMOylation in the trimerization domain has been to be than J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; Scholar). The of SUMOylation at of HSF2 has been by its in of the in H. Y. Sarge K.D. Science.. Scholar). However, the functional consequences of the SUMOylation are the SUMO modification of HSF2 at has been reported to and its DNA binding J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; Scholar, Y. Sarge K.D. J. Biol. Chem.. Scholar, Sarge K.D. 2004; Scholar). the effect of SUMOylation on the DNA interaction of HSF2, we the biochemical properties of isolated recombinant human HSF2 with to the molecular mechanism by which the DNA binding of HSF2, we structural analysis on the HSF2 using NMR and biochemical and structural data that SUMOylation of HSF2 with the binding to HSEs through a molecular mechanism distinct from that by structural studies of the SUMOylated The attachment of SUMO any in the HSF2 or form a noncovalent interaction with HSF2. Here we a in which the randomly distributed SUMO moiety to the flexible in HSF2 of HSF2 to DNA through a steric in a in for the formation of the Protein and of human by and the bacterial expression with an The expression were in at to an of at and with for at For of and used of were by in and The the by from the by the with for at The using and affinity with and with than full-length human by using two at the to The with to the followed by as a in cells and by affinity The with on the glutathione in of from the using with a and with a of SUMOylated of SUMOylated a and a and SUMOylation and were cells Y. M. K. M. H. 2004; Scholar). The cells were in at to an of at and were with at for of the in the as the during the and SUMOylated were using an in vitro SUMOylation and were in of a and The were at for SUMOylated proteins were in the as DNA were used in as and with the or SUMOylated in a binding and The at for and to on in in at for at The DNA were by with and of for HSE were by using a and HSE were on a response for with and response for with were with the binding and and were at a of or The the between by and the were and using the and the binding at a of NMR NMR of and SUMOylated were with an with a at in for were by and using a NMR data were using F. G. J. A. J. 6: and of Scholar). were using a the from and residues were by or at in Wu C. A. J. 1994; Scholar). of and were by site-directed as and were and as reported Y. M. K. M. H. 2004; Scholar). and of were also by site-directed and as For and were SUMOylated in vitro with and as For with using the in vitro SUMOylation and using MTSL, in to SUMOylated in a of and the at for The by affinity using a were on of which were in as H. M. J. 2007; Scholar, M. A. G. J. Scholar). data were at on an with a and The were by using the and on For a sequence H. M. J. 2007; Scholar, M. A. G. J. Scholar). The set to the of the The set to which to a The The and the were amino of human and Drosophila HSF were using J. K. J. Mol. Biol.. Scholar). of of with N. D. B. A. Mol. Biol.. using of the of Drosophila HSF A. J. Wu C. A. Nat. Struct. Biol.. 1994; in the Protein Data as a of these were of the in the structure of the K. lactis HSF to the repeat of using the K. 2004; and M. K. J. Mol. 14: Scholar). The were using The Scholar). of structure of the with C. J. 2003; Scholar). of the of the and structure from an NMR were used as the The C-terminal of and the flexible of were as from the by were used in the and for the between of and of and between of and of were set based on the that in between two of and between the two of a has been shown to from to J. Scholar). the isopeptide bond between and SUMO, a between the of of and the of of the with a of of The were to and the with the were and Modification the DNA of with biochemical the structure of the K. lactis HSF in complex with DNA a formation of its upon binding to two inverted HSEs in the tail-to-tail (9Littlefield O. Nelson H.C. Nat. Struct. Biol.. 1999; 6: 464-470Google Scholar). of the HSE repeats is recognized by an of K. lactis HSF, and the specific protein-DNA interaction is by the of the In with results from the that the isolated of human HSF2, amino residues as to two or inverted HSE consensus repeats and but to a single analysis using that as a DNA in the of upon specific binding to the a for the of the HSE we using a two inverted repeats in the or The and the a shift 2 and The affinity of for HSE using the binding of in a the for the and repeats were to be and and to two inverted HSE for the or with recombinant SUMOylated to the of at on the DNA binding in using a in which and an are with SUMOylation and Y. M. K. M. H. 2004; In the using the interaction of with neither nor repeats and of for the and DNA be by the were and a state of In to bind to a DNA HSE data that the DNA binding of regardless of the of the HSE DNA binding were in a The effect of SUMOylation on the DNA binding affinity of also in in the of between HSE DNA and or SUMOylated as by be to small be to small be to small be to small in a SUMO Modification the DNA of effect of SUMO modification on the DNA binding of also using and with recombinant SUMOylated The by an in vitro SUMOylation the of the SUMO to SUMOylation in the HR-A/B to In the shown in of the complex with were for and 2 and the to the HSE repeats than but DNA binding by the SUMO moiety with and and and on the of for and repeats were to be and whereas the of were to be for and for and The modification with also shown to the affinity of for HSEs in the by an these that to the two inverted HSEs in the and with similar and that its DNA binding affinity is by than by SUMOylation at regardless of the HSE repeat a with the were noncovalent between and are covalently through an isopeptide The molecular interaction and by the attachment were using by an in vitro SUMOylation in which the or the moiety with to NMR from the shift were in a The of to with that of shift for the resonance to the of at the C-terminal of residue is involved in the formation of the isopeptide bond with the target residue of the shift is by a in the properties of the or in the of with the isopeptide bond The absence of shift for that the structure of is upon the to and that has noncovalent with for a near the the of with with that of The shift of the formation of the to are to the and its and of residues shift than and in the and results suggest that have a noncovalent interface with the attached and the that a in the is with the SUMO between and the DNA the mechanism of the effect of SUMOylation on the DNA binding of we the structure of in complex with HSE repeats based on a structure of K. lactis HSF in a complex with DNA (9Littlefield O. Nelson H.C. Nat. Struct. Biol.. 1999; 6: 464-470Google and NMR of Drosophila HSF A. J. Wu C. A. Nat. Struct. Biol.. 1994; for a the SUMOylation sequence to of yeast and Drosophila whereas the of the are conserved In an of NMR of Drosophila DBD, the is than A. J. Wu C. A. Nat. Struct. Biol.. 1994; Scholar). The in yeast HSF is in a form (9Littlefield O. Nelson H.C. Nat. Struct. Biol.. 1999; 6: 464-470Google Scholar). that of be and thus the SUMOylation to be in a specific in In the is in the of the interface the that the formation for DNA binding. However, in the the two SUMOylation are to be the and to be from the interface SUMOylation the binding affinity of HSF2 to HSEs in the and we that the DNA binding of through a mechanism than the of the we between the flexible SUMO moiety and DNA the interaction between and by using that the two inverted HSE repeats with at which are to be to be with the in the to bind to the DNA in the and However, to bind to but form a complex with the of the DNA the HSE which are to with is to the effect of on the protein-DNA interaction in the the of these we that of the interaction by SUMOylation be to steric or between the SUMO moiety and The binding of to also in in a binding also with but with that the SUMOylation of in binding affinity for the repeats than for the repeats of the to the of the of the attached to we a with site-directed spin using The can be between two that are to a residue in a target For attachment of to the of to the be by the at that to the of the we a residue with a the of and were to of these proteins a of spin at in the moiety with or in the moiety of and The of were from of the The of a with the at and a of and a with the at and a of The of and are than than the for of attached to residues at and of a ubiquitin H. M. J. 2007; Scholar). The of of the between the and the spin can the by J. Scholar); the and of are the of is than which from the of the of The for and suggest that the of to a of the of the of in the complex with using paramagnetic NMR relaxation enhancement with using The paramagnetic effect of the of of NMR of within on the between the paramagnetic and the N. J. Scholar, J. 2006; Scholar). In we a of to a single and the on the to the In addition to the we the that a of a or The for the were that molecular of and are in the on an and and are on the negatively surface were for in the of in which with with the of SDSL, of the were by the in of by The for residue of using the of the NMR of the to the In experiments, the residues were distributed on the DNA interaction in the DNA of residues and of were in the of the and in which the attached to the surface of the results suggest the of to has a of the of is the DNA interaction surface of the is distributed and is near the surface than the of the surface of to be to the DNA interaction surface of of a between the of and with the the form in the is of the of which is to the C-terminal of and thus to the residue in the flexible of and were in a as the shift on the from the of and and the of we the of the to based on a using the C. J. 2003; Scholar). upon SUMO were from the results of shift experiments, we of and a structure of from an of NMR as were for using for between of and of and between of and of which were from the of that the were a whereas the structure is with the results of experiments, and thus is from In these the between and is and between and is the the the between and and between and to be and The in between two of and between the two of a has been shown to from to on the of the the and the J. Scholar). the by the are with the results of the the between and of the which isopeptide bond In the of but two the the surface at the of the SUMO is in the of the surface of and the are with the results of the of the of the SUMO moiety in the HSF2 has been shown to be a SUMOylation and its DNA binding to HSE is modulated by of SUMO to the target lysine in the the HSE binding of SUMOylated HSF2 by two and of enhancement and of the have been in using in proteins J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; Scholar, Y. Sarge K.D. J. Biol. Chem.. Scholar). be to of SUMOylation interplays with post-translational or in DNA or used in the binding In a SUMO has been in HSF2 in addition to J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; and the has been reported to the SUMOylation of HSF2 J. Y. Sarge K.D. J. Biol. Chem.. Scholar). In the of the SUMO on the interaction using an isolated in a that the of and SUMOylation at biochemical from and have that the interaction of full-length with the HSE DNA is by the of at its In effect of SUMOylation on the DNA binding of the a that have the which that a of DNA binding be to the of the trimer J. V. K. Sistonen L. Mol. Cell. Biol.. 2006; have reported that SUMOylation with HSF2 The of in complex with DNA that the interface of the DBDs in the is by for the However, in experiments, the DNA of were with the and we that the modification with a on the interface between HSF2 and SUMOylation functional regulation have been SUMOylation with post-translational for target lysine of specific lysine residues with ubiquitin and an have been reported in nuclear and a C. B. G. Scholar, N. J. G. C. Dejean A. D. Mol. Cell. Biol.. 2007; Scholar). the SUMO a binding of the for its downstream A. P. F. Nat. Struct. Mol. Biol.. 12: Scholar). as based on the studies of the TDG, of SUMO in and molecular D. N. Y. H. K. F. H. H. M. Scholar). SUMO that is covalently to can as a binding a in the which an molecular interface for with downstream as for B. M. C. Scholar, D. Scholar, Y. J. M. H. J. Biol. Chem.. 2006; Scholar). a the SUMO interaction motif in proteins M. K. P. J. Biol. Chem.. 2006; Scholar, O. 2007; Scholar). The to a interaction that is by these the has been shown to a protein-DNA The high binding affinity of for DNA an is by the SUMOylation of TDG, through a via noncovalent between the SUMO and a sequence in D. N. Y. H. K. F. H. H. M. Scholar). of the of protein-DNA by SUMO has been The DNA binding of a human transcription Drosophila a is by the SUMOylation of its DNA binding domain C. H. J. Biol. Chem.. 2007; 282: Scholar). has the SUMOylation in its we the that the SUMOylation of to the DNA interaction through a of in a similar to the SUMOylation of However, NMR on have that the of the interaction by SUMOylation be to a interaction between the SUMO and HSF2 or to a in HSF2. mechanism that the SUMO attachment the interface of which is for the specific binding to the HSE for the of binding to the HSE repeats but for binding to the the results of the using and that the to the with the of DNA the with that the of the to is but a with the SUMOylation is in a flexible as shown in the of the Nevertheless, the results of the that the and of the are in such a that is near the surface of with its surface on the results of the and experiments, we for a of the of The of in to is by an interaction between the surface of and the surface of Nevertheless, the of the of the from the that the SUMO moiety can from the shown in to J. Scholar). In of the of an interaction between the surface of and the surface of shift of in a that the of and of and and of in addition to residues near the small but shift upon and shift were in the of for and of these residues are on the surface of or the surface of the shift at be by between the of and The of as from the DEER, PRE, and shift experiments, the mechanism for the of DNA binding of between the of of shown in and the complex that in the as to a steric with the of the DNA HSE in a In the of the SUMOylated complex that can to steric The DNA the that with the steric between and DNA to be a of the of DNA binding of upon Nevertheless, we the that the attached also an effect that to the DNA binding affinity of the of that in of the the surface is to be within from the of the attached steric and with the DNA are on the of in the the complex between SUMOylated and DNA is biochemical and structural data a molecular mechanism attachment of negatively modulates formation of the protein-DNA The attached to with the surface of but in a to the and or with DNA binding to in a The of is to the between the negatively surface of and the surface of of proteins are such in of SUMOylated the mechanism for HSF2 be to to of the nuclear DNA binding are by HSF2 a in which is by the pathway A. S.K. Morimoto R.I. Mol. Cell. Biol.. 18: Scholar); on the HSF1 is and in and its is by with proteins R.I. Cell.. Scholar). of HSF1 and HSF2 in nuclear stress which is a by various stress M. A. V. M. J. Sci.. 2003; HSF1 and HSF2 are to be involved in distinct processes, classical heat shock response and developmental processes, In HSF2 has been shown to of genes by HSF1 or negatively P. Bjork J.K. Roos-Mattjus P. Mezger V. Sistonen L. J. Biol. Chem.. 2007; 282: 7077-7086Google Scholar). the HSF2 has been shown to be by in a to H. Y. Sarge K.D. Science.. Scholar). SUMOylation of HSF2 a molecular to its DNA binding and a of its on the are to the DNA binding of HSF2 by SUMOylation is to these cellular are to for expression of SUMOylation and the in SUMOylation and for in the