Low Resolution Structural Study of Two Human HSP40 Chaperonesin SolutionDJA1 FROM SUBFAMILY A AND DJB4 FROM SUBFAMILY B HAVE DIFFERENT QUATERNARY STRUCTURES*

Received for publication, July 23, 2004, and in revised form, January 18, 2005Published, JBC Papers in Press, January 20, 2005, DOI 10.1074/jbc.M408349200

Julio C. Borges‡§, Hannes Fischer¶, Aldo F. Craievich¶, and Carlos H. I. Ramos‡�

From the ‡Centro de Biologia Molecular Estrutural, Laboratorio Nacional de Luz Sıncrotron, CP 6192, 13084-971,Campinas, SP, Brazil, the §Departamento de Bioquımica, Instituto de Biologia, Universidade Estadual de Campinas,13084-971, Campinas, SP, Brazil, and the ¶Departamento de Fısica Aplicada, Instituto de Fısica, Universidadede Sao Paulo, 05508-900, Sao Paulo, SP, Brazil

Proteins that belong to the heat shock protein (Hsp)40 family assist Hsp70 in many cellular functions andare important for maintaining cell viability. A knowl-edge of the structural and functional characteristics ofthe Hsp40 family is therefore essential for understand-ing the role of the Hsp70 chaperone system in cells. Inthis work, we used small angle x-ray scattering andanalytical ultracentrifugation to study two represen-tatives of human Hsp40, namely, DjA1 (Hdj2/dj2/HSDJ/Rdj1) from subfamily A and DjB4 (Hlj1/DnaJW) fromsubfamily B, and to determine their quaternary struc-ture. We also constructed low resolution models for thestructure of DjA1-(1–332), a C-terminal-deleted mutantof DjA1 in which dimer formation is prevented. Ourresults, together with the current structural informa-tion of the Hsp40 C-terminal and J-domains, were usedto generate models of the internal structural organiza-tion of DjA1 and DjB4. The characteristics of thesemodels indicated that DjA1 and DjB4 were bothdimers, but with substantial differences in their qua-ternary structures: whereas DjA1 consisted of a com-pact dimer in which the N and C termini of the twomonomers faced each other, DjB4 formed a dimer inwhich only the C termini of the two monomers were incontact. The two proteins also differed in their abilityto bind unfolded luciferase. Overall, our results indi-cate that these representatives of subfamilies A and Bof human Hsp40 have different quaternary structuresand chaperone functions.

The Hsp701 chaperone system is formed by Hsp70 (DnaK-related) and its co-chaperones Hsp40 (DnaJ-related) andGrpE. This system assists many cellular processes involvingproteins, including folding, transport through membranes,degradation, and escape from aggregation (1–6). The affinityof Hsp70 for unfolded proteins is regulated by the binding ofeither ADP (high affinity) or ATP (low affinity) to its nucle-

otide binding domain (NBD), and by interaction with itsco-chaperones (5, 7). Hsp40 assists the folding of nascentproteins, and prevents aggregation and the refolding of ag-gregates by presenting nascent proteins to Hsp70 and stim-ulating the hydrolysis of ATP (8–12).

Hsp40, which can also act as a chaperone by itself (13–15),consists of four conserved functional regions, as determined bygenetic and mutational studies in vivo (16–19) and by biophys-ical methods (11, 20–22) (Fig. 1A). The highly conserved �-hel-ical N-terminal domain, referred to as the J-domain, is char-acteristic of proteins in this family, and the binding of thisdomain to Hsp70 stimulates the ATPase activity of Hsp70 (11,12, 14, 21, 23). Adjacent to the J-domain is a glycine/pheny-lalanine-rich region (G/F-rich; Fig. 1A) that is disordered andlikely to be responsible for flexibility (11, 24). The centralregion consists of a cysteine-rich domain (Cys_rich; Fig. 1A)that includes four repeats of the motif CXXCXGXG (where X isany amino acid) and folds in a zinc-dependent fashion with tworepeats bound to one zinc ion (20, 25, 26). The C-terminal domain(Fig. 1A) forms a �-sheet structure involved in the dimerization ofHsp40 (27). The Cys-rich and C-terminal domains are involved insubstrate binding and presentation (22, 25, 28).

Hsp40 proteins occur throughout the cell and show a highdiversity in eukaryotic genomes (29, 30), with at least 44 genespresent in the human genome (31). Based on their architectureand cellular location (4, 32, 33), Hsp40 proteins are classified inthree main subfamilies (A–C, also referred to as types I-III; Fig.1A) (4, 33, 34): subfamily A consists of proteins with the fourdomains described above, subfamily B contains proteins thatlack the Cys-rich domain, and subfamily C has only the J-domain that is not necessarily located at the N terminus (4, 33).Hsp40 proteins of subfamily A have autonomous chaperoneactivity and may therefore work in an Hsp70-dependent or-independent manner. In contrast, Hsp40 proteins of subfamilyB have no autonomous chaperone activity and depend onHsp70 for full activity (18, 33, 35, 36).

Although high resolution structures of the C-terminal, Cys-rich, and J-domains are available, there is still little informa-tion about how these domains interact with each other, withother domains and also with Hsp70. To obtain further infor-mation about relevant structural and functional characteristicsof human Hsp40 proteins, we used biophysical methods, includ-ing small angle x-ray scattering (SAXS) and analytical ultra-centrifugation (AUC), to study DjA1 (also known as Hdj2/dj2/HSDJ/Rdj1), a representative of subfamily A, its C-terminaldeletion mutant (DjA1-(1–332)), and DjB4 (also known as Hlj1/DnaJW), a representative of subfamily B that shares 65%identity with DjB1/Hdj1.

* This work was supported by fellowships and grants from Fundacaode Amparo a Pesquisa do Estado de Sao Paulo (FAPESP) and ConselhoNacional de Desenvolvimento Cientıfico e Tecnologico (CNPq). Thecosts of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked “adver-tisement” in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

� To whom correspondence should be addressed. Tel.: 55-19-3512-1118; Fax: 55-19-3512-1006; E-mail: cramos@lnls.br.

1 The abbreviations used are: Hsp, heat shock protein; LUC, lucifer-ase; SAXS, small angle X-ray scattering; AUC, analytical ultracentrif-ugation; G/F-rich, glycine/phenylalanine-rich region; NBD, nucleotidebinding domain.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 14, Issue of April 8, pp. 13671–13681, 2005© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 13671

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

EXPERIMENTAL PROCEDURES

Cloning, Expression, and Purification—DjA1 was cloned from thehuman gene DNAJA1 (cDNA clone, GenBankTM accession numberAW247277) and DjB4 was cloned from the human gene DNAJB4 (cDNAclone, GenBankTM accession number AA081471). The Hsp40 nomencla-tures DjA1 and DjB4 were used as described by Ohtsuka and Hata (34).Two primers were used to amplify the DjA1 cDNA by PCR and to createrestriction enzyme sites for NdeI and XhoI: a DjA1 5�-primer (5�-CCG-GCAGGCTAGCATGGTGAAAGAAACAAC-3�) containing an NdeI re-striction site and a DjA1 3�-primer (5�-TGAGTGTTATTCTCGAGTCAT-TAAGAGGTCTG-3�) containing an XhoI restriction site. Two primerswere used to amplify the DjB4 cDNA by PCR and to create restrictionenzyme sites for NdeI and BamHI: a DjB4 5�-primer (5�-TCAAGGCAT-TCCATATGGGGAAAGACTATTA-3�) containing an NdeI restrictionsite and a DjB4 3�-primer (5�-GTGTAACAAAGTGGATCCTACTAT-GAGGCAGG-3�) containing a BamHI restriction site. The C-terminaldeletion of DjA1 was constructed by site-directed mutagenesis using aprimer that created a stop codon at residue Phe333 followed by a restric-tion site for XhoI (5�-TTATCAGGCTCGAGTTAGCCATTCTC-3�).

The PCR products were cloned into the pET28A expression vector(Novagen) for His tag purification methods. The correct cloning wasconfirmed by DNA sequencing using an ABI 377 Prism system(PerkinElmer Life Sciences). These procedures created the vectorspET28aDjA1, pET28aDjB4, and pET28aDjA1-(1–332), which weretransformed into the Escherichia coli strain BL21(DE3) for heterolo-gous protein expression by adding isopropyl thio-�-D-galactoside (0.4mM) at A600 � 0.6. The induced cells were grown for 5 h and harvestedby centrifugation for 10 min at 2,600 � g. The bacterial pellet wasresuspended in lysis buffer (50 mM Tris-HCl, pH 8.0, 500 mM KCl and10 mM EDTA; 15 ml of buffer/liter of medium) in the presence of 5 unitsof DNase (Invitrogen, Life Technologies, Inc.) and 30 �g of lysozyme/ml

(Sigma) for 30 min. The pellet was then disrupted by sonication in an icebath, and centrifuged as described above. The supernatant was frac-tionated by metal affinity chromatography in a HiTrap chelating col-umn (Amersham Biosciences) using an AKTA FPLC system (Amer-sham Biosciences). The proteins were eluted with imidazole (500 mM)and loaded onto a HiLoad Superdex 200-pg column (2.6 cm � 60 cm;Amersham Biosciences) using an AKTA FPLC system. The degree ofpurification was estimated by SDS-PAGE, and the protein concentra-tion was determined spectrophotometrically using a calculated extinc-tion coefficient for denatured proteins (37, 38).

Analytical Molecular Exclusion Chromatography—Analytical molec-ular exclusion chromatography was done using a Superose 12 HR 10/30(Amersham Biosciences) column coupled to an AKTA Purifier system(Amersham Biosciences). The column was equilibrated with 25 mM

Tris-HCl, pH 7.5, containing 500 mM NaCl, 1% glycerol, and �-mercap-toethanol (1–10 mM). The column was washed with two column volumesof this same buffer at a flow rate of 0.5 ml/min, and aliquots of proteinsin 100 �l were loaded onto the column. The elution profile was deter-mined by monitoring the absorbance at 280 nm. The apparent molecu-lar mass was calculated using a plot of ln of the molecular mass (kDa)of standard proteins (thyroglobulin, 669 kDa; �-globulin, 160 kDa;bovine serum albumin, 69 kDa; chicken ovalbumin, 45 kDa; cytochromec, 12 kDa) versus the retention time.

Circular Dichroism Spectroscopy—Circular dichroism (CD) measure-ments were done using a Jasco J-810 spectropolarimeter with thetemperature controlled by a Peltier-type control system PFD 425S.Hsp40 proteins were resuspended in 25 mM Tris-HCl, pH 7.5, contain-ing 500 mM NaCl and 1 mM �-mercaptoethanol at final concentrationsof 20–50 �M. The data were collected at a scanning rate of 50 nm/minwith a spectral bandwidth of 1 nm using a 0.1-mm path length cell.CDNN deconvolution software (Version 2, Bioinformatik.biochemtech.

FIG. 1. Hsp40 chaperones. A, schematic representation of the domain structure of the Hsp40 subfamilies (A, B, and C). B, homology of theamino acid sequences of DjA1 (gene DNAJA1, GenBankTM accession number NM_001539) and DjB4 (gene DNAJB4, GenBankTM accession numberNP_008965). The nomenclature is from Ohtsuka and Hata (34). Gray box, J-domain; gray underline, G/F-rich region; black underline, Cys-richdomain; white box, C-terminal domain. The arrow indicates the last residue in the DjA1-(1–332) deletion mutant. DjA1 shares 47% similarity withDjB4 (81% in the J-domain).

Structural Insights into Human Hsp4013672

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

uni-halle.dee/cdnn) was used for secondary structure prediction. Allbuffers used were of analytical grade and were filtered before use toavoid light scattering by small particles.

Measurement of Chaperone Activity—The method described by Luand Cyr (28, 35) was used to assess the ability of Hsp40 proteins tointeract with unfolded proteins. Briefly, luciferase (Promega) waschemically denatured by diluting in guanidinium-HCl (6 M) for 40 minat room temperature and then diluted 25� in 50 �l of 25 mM Tris-HCl,pH 7.5, containing 500 mM NaCl and 5 mM �-mercaptoethanol in theabsence and presence of His-tagged Hsp40 proteins (see figure captionsfor details of the concentrations). In another protocol, luciferase (2 �M)was thermally denatured by incubation at 42 °C for 10 min in theabsence and presence of His-tagged Hsp40 proteins (see figure captionsfor details of the concentrations) and then cooled to room temperature.The solutions were centrifuged for 10 min at 21,000 � g, and 50 �l of thesupernatant was incubated with 50 �l of a 50% slurry of Talon metalchelate resin (Clontech) in 25 mM Tris-HCl, pH 7.5, containing 500 mM

NaCl and 5 mM �-mercaptoethanol for 1 h at room temperature. Themixture was then centrifuged for 1 min at 10,000 � g and 4 °C, and thepellet containing the resin was washed twice with 75 �l of the Trisbuffer indicated above containing 15 mM imidazole. The protein com-plexes bound to the resin were eluted with this same Tris buffer con-taining 150 mM imidazole and then concentrated by precipitation withacetone 80% (Merck) and visualized by SDS-PAGE.

Analytical Ultracentrifugation—The sedimentation velocity and sed-imentation equilibrium experiments were done with a Beckman Op-tima XL-A analytical ultracentrifuge. The proteins DjA1, DjA1-(1–332),and DjB4 were tested at concentrations from 50 to 1000 �g/ml in 25 mM

Tris-HCl, pH 7.5, containing 500 mM NaCl, 1% glycerol (but not forDjA1-(1–332)) and 1 mM �-mercaptoethanol, with no apparent aggre-gation. The sedimentation velocity experiments were done at 20 °C,25,000 rpm for DjA1 and DjB4, and 30,000 and 40,000 rpm for DjA1-(1–332) (AN-60Ti rotor), and the scan data were acquired at 230 and238 nm for low and high protein concentrations, respectively. Thesedimentation equilibrium experiments were done at 20 °C at speeds of6,000, 8,000, and 10,000 rpm with the AN-60Ti rotor and scan dataacquisition at 238 nm. Analysis of the data involved the fitting of amodel of absorbance versus cell radius data using nonlinear regression.All fittings were done using the Origin software package (MicrocalSoftware) supplied with the instrument. The van Holde-Weischet (39)(sediment coefficient plot) and the sedimentation time derivative (g(s*)integral distribution) (40) methods were used to analyze the sedimen-tation velocity experiments. The methods used to analyze the velocityand equilibrium experiments allowed the calculation of the apparentsedimentation coefficient s, the diffusion coefficient D, and the molec-ular mass M. The ratio of the sedimentation to diffusion coefficientsgave the molecular mass in Equation 1,

M �sRT

D(1 � Vbar�)(Eq. 1)

where R is the gas constant and T is the absolute temperature. TheSednterp software (www.jphilo.mailway.com/download.htm) was usedto estimate the partial specific volume of the proteins at 20 °C(VbarDjA1 � 0.7275 ml/g, VbarDjA1-(1–332) � 0.7309 ml/g, andVbarDjB4 � 0.7302 ml/g), the buffer density (� � 1.02163 g/ml) and thebuffer viscosity (� � 1.0851 � 10�2 poise).

The self-association method was used to analyze the sedimentationequilibrium experiments, and several models of association for DjA1and DjB4 were used to fit the data. The distribution of each proteinthroughout the cell, as determined in the equilibrium sedimentationexperiments, was fitted with Equation 2 (41),

C � C0 exp�M(1 � Vbar�)�2(r2 � r0)2RT � (Eq. 2)

where C is the protein concentration at radial position r, C0 is theprotein concentration at radial position r0, and � is the centrifugalangular velocity. The Sednterp software was used to estimate thestandard sedimentation coefficients (s20,w) at each protein concentra-tion and to calculate s0

20,w by extrapolation to a protein concentration of0 mg/ml. This procedure corrects for effects of temperature, solutionviscosity, and molecular crowding (42).

SAXS—The SAXS experiments were done at the SAS beamline of theLNLS synchrotron radiation facility in Campinas, Brazil (43) under thesame conditions as those described by Borges et al. (44). The measure-ments were done using a monochromatic x-ray beam with a wavelength� � 1.488 Å. For a sample-to-detector distance of 840 mm selected for

the experiments, the modulus of the photon momentum transfer (q �4sin/�, 2 being the scattering angle) covered a range from q � 0.01Å�1 to q � 0.44 Å�1. The scattering intensity curves, I(q), were recordedusing a gas 1 dimensional position sensitive x-ray detector. To monitorpossible effects from radiation damage and synchrotron beam instabil-ities, the SAXS curves were determined in many short frames (90 seach). The SAXS data obtained were normalized to account for thenatural decay in intensity of the synchrotron incident beam and werecorrected for non-homogeneous detector responses. Finally, the scatter-ing intensity produced by the buffer was subtracted, and the differencecurves were scaled to give equivalent protein concentrations. Threetypes of samples with different concentrations were studied by SAXS: 1)3.3 mg/ml DjA1 in 25 mM Tris-HCl, pH 7.5, containing 500 mM NaCl, 1%glycerol and 1 mM dithiothreitol, 2) 2.9 and 6.5 mg/ml of DjA1-(1–332)in 25 mM Tris-HCl, pH 7.5, containing 500 mM NaCl and 1 mM �-mer-captoethanol, and 3) 2.5, 4.4, 6.2, and 8.3 mg/ml DjB4 in 25 mM Tris-HCl, pH 7.5, containing 500 mM NaCl and 1 mM �-mercaptoethanol.

Reliable structural information about the low resolution structure ofproteins in solution can be derived from SAXS data provided that allproteins are in the same (monomeric or oligomeric) state and that thesolution is “dilute,” i.e. interference effects in scattering amplitudesproduced by different proteins are negligible. Under these assumptions,the total scattering amplitude is proportional to the form factor of anisolated (monomeric or oligomeric) protein averaged for all orientations.When samples with a high concentration were available, data mergingwas done using low and high concentration samples for the high andlow q ranges of the scattering intensity, respectively. This procedurereduced the overall statistical error in the scattering curves withoutintroducing unwanted interferences or spatial correlation effects in thesmall q range.

To establish the molecular masses of DjA1, DjA1-(1–332), and DjB4,the SAXS intensity produced by a solution of bovine serum albumin (69kDa, 5 and 10 mg/ml) was also determined. Since, for dilute solutions,the normalized intensity extrapolated to q � 0, I(0), is proportional tothe molecular mass, the molecular masses (M) of the proteins investi-gated here were determined from the ratios between the I(0) valuescorresponding to the different samples and that of the standard bovineserum albumin.

Computer Programs—By applying the indirect Fourier transformprogram GNOM (45) to the normalized SAXS curves, the distancedistribution function p(r) and the radius of gyration, Rg, of the proteinsstudied were evaluated (46). Prior to GNOM analysis, a constant inten-sity background was subtracted from the SAXS data to ensure that theintensity at higher angles decayed as q�4, according to Porod’s law fora two-electron density model (46). The low resolution shapes (or molec-ular envelopes) of the proteins were restored from the experimentalSAXS curves using an ab initio method named DAMMIN (47) andrecently described in Borges et al. (44). Briefly, in this method, the lowresolution shape of the protein was simulated by a set of small spheresthat initially filled another sphere with a diameter equal to the maxi-mum diameter, Dmax, of the protein being studied (previously deter-mined using GNOM). DAMMIN yielded a structural model containinga fraction of the initial number of dummy atoms whose associatedscattering intensity gave the best fit to the SAXS data. The configura-tion of the different protein domains could be refined by manuallyrotating or displacing the domains until the discrepancy between thecalculated and experimental SAXS curves was minimized. The proteinmodels were displayed using the program WebLab ViewerLite software(www.accelrys.com).

The HydroPro software (48) was used to estimate the translationaldiffusion coefficient Dt, the radii of gyration Rg, the sedimentationcoefficient s, and the maximum distance (Dmax) starting from the abinitio models generated by DAMMIN using SAXS data obtained at20 °C. The HydroPro software was configured with the radius of theatomic elements from the ab initio analysis, with sigma factors from 5to 8 (as indicated by the supplier) and a minibeads radius from 6 to 2 Å(SIGMIN and SIGMAX), after an initial evaluation of the two extremes.The parameters Vbar, �, and � were estimated using the softwareSednterp as described above. The translational fraction ratio or Perrinfactor P, which indicates the relationship between the frictional coeffi-cient of the Hsp40 particles and a sphere of the same molecular mass(f/f0), was estimated using Solpro software (49).

RESULTS AND DISCUSSION

Purification of DjA1 and DjB4 as Folded Dimers and ofDjA1-(1–332) as a Folded Monomer—The correct cloning andsequencing of DjA1, DjB4, and DjA1-(1–332) were confirmed by

Structural Insights into Human Hsp40 13673

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

DNA sequencing. All of the proteins were expressed in largequantities and were more than 95% pure as confirmed bySDS-PAGE (Fig. 2A). The proteins were unstable at low ionicstrength but were soluble in the presence of 500 mM NaCl.DjA1 and DjB4 were purified as dimers and the mutant DjA1-(1–332) was purified as a monomer, as confirmed by analyticalmolecular exclusion chromatography (Fig. 2B). The foldingstate of the proteins was investigated by CD, and the resultingspectra corresponded to folded proteins with a minimum at 208nm (Fig. 2C). The analysis using the CDNN deconvolutionsoftware indicated that the proteins had similar amounts ofsecondary structure: 35% �-sheet structure and 10% �-helices.The high content of �-sheet structure agreed with x-ray crys-tallography and nuclear magnetic resonance data indicatingthe presence of �-sheet structures in the C terminus (22, 27)and in the Cys-rich domain (22, 26) of Hsp40.

Although the deletion of the last 54 residues of DjA1 pre-vented the dimerization of this protein, it had no marked effecton the global conformation (Fig. 2B). However, the thermalstability of the deleted mutant was lower than that of the wholeprotein. DjA1 and DjB4 were heated to 50 °C without loss ofstructure whereas the amount of secondary structure in DjA1-(1–332) decreased when the protein was heated above 38 °C(data not shown). These results indicated that the dimeric statewas important for the stability of the protein at temperatureswell above the body temperature, an important characteristicfor proper cell functioning.

Efficiency of DjA1 in Binding Unfolded Luciferase Comparedwith That of DjB4—Because the chaperone action of Hsp40involves hydrophobic interaction with unfolded or partiallyfolded proteins (15, 18, 19, 22), we investigated the ability ofHsp40 proteins to bind unfolded luciferase. His-tagged chaper-ones have a high affinity for the Talon metal chelate resin andbind to it in the absence of imidazole. Any unfolded luciferasethat binds to a His-tagged chaperone co-purifies with the chap-erone when the resin is washed with 150 mM imidazole. Fig. 3shows the SDS-PAGE profiles of washed samples of chemically(Fig. 3A) or thermally (Fig. 3B) denatured luciferase in theabsence or presence of increasing concentrations of DjA1, DjA1-(1–332), or DjB4 (see figure caption for details). At 22 °C, foldedluciferase did not bind to DjA1 or DjB4, whereas at 42 °C

unfolded luciferase bound to DjA1 but not to DjB4 (Fig. 3B).Similar results were obtained with chemically unfolded lucif-erase (Fig. 3A), with this protein binding to DjA1 but having avery low affinity for DjB4. The deletion of the C terminus in themutant DjA1-(1–332) had a major effect on the efficiency ofbinding (Fig. 3A).

These results indicated that the chaperone activities of DjA1and DjB4 were not the same. A similar result was obtainedwhen the ability of yeast Hsp40 proteins to bind unfoldedluciferase was measured: Ydj1 (subfamily A) was 2.5 timesmore active than Sis1 (subfamily B) (19, 28). Others haveshown that DjA1 is more efficient in refolding luciferase thanDjB1 (subfamily B, homologous to DjB4) (50). DjA1-(1–332) hada lower chaperone activity than intact DjA1 and was not used

FIG. 3. Chaperone activity measurements. SDS-PAGE in 10% gelshowing samples of chemically (A) and thermally (B) unfolded lucifer-ase (LUC) in the absence and presence of increasing concentrations ofHsp40 proteins (5–15 �M) (see “Experimental Procedures”). Samples offree luciferase were run for comparison.

FIG. 2. Expression, purification,and secondary and quaternary struc-tures of DjA1, DjA1-(1–332), and DjB4.A, SDS-PAGE (12% gel). Lane 1, molecu-lar mass markers in kDa (left); lanes 2, 6,and 10, bacterial pellet before induction;lanes 3, 7, and 11, bacterial pellet afterinduction with isopropyl-1-thio-�-D-galac-topyranoside; lanes 4, 8, and 12, superna-tant of lysed cells; lanes 5, 9, and 13,purified proteins. B, analytical molecularexclusion chromatography. Bottom panel,retention times of standard proteins (see“Experimental Procedures”). Top panel,retention times for DjA1, DjA1-(1–332),and DjB4. The estimated apparent molec-ular masses were 120, 100, and 40 kDa,respectively. C, circular dichroism. Theresidual molar ellipticity of DjA1, DjA1-(1–332) and DjB4 was measured from 195to 260 nm in 25 mM Tris-HCl, pH 7.5,containing 500 mM NaCl and 1 mM �-mer-captoethanol at 20 °C. The amount of sec-ondary structure was estimated usingCDNN deconvolution software and wassimilar among the two proteins: 35%�-sheet, 10% �-helix, 20% turn, and 35%random coil (experimental error �10%).

Structural Insights into Human Hsp4013674

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

in the luciferase thermal unfolding experiments because itunfolded at temperatures �38 °C. The lower chaperone activityof DjA1-(1–332) indicated that the quaternary structure ofHsp40 was related to its chaperone activity, as also shown foryeast Hsp40 Sis1, in which a C-terminal deletion abolished theability to assist Hsp70 (27).

Envelope Models Derived from SAXS—The experimental(corrected and normalized) SAXS curves for DjA1, DjA1-(1–332), and DjB4 are shown in Fig. 4A. Molecular masses of theproteins studied were obtained from the quotient between theirintensity extrapolated to q � 0 and the I(0) value correspondingto bovine serum albumin, as described above. The molecularmasses of DjA1 and DjB4 were determined to be 95 and 90 kDa,respectively, in good agreement with the values calculated

from their amino acid sequences (Table I), assuming that theyare dimers. In contrast, the SAXS data for DjA1-(1–332) indi-cated the existence of a monodisperse set of molecules with amolecular mass of 40 kDa, in good agreement with the valuecalculated from the amino acid sequence (Table I) of the mon-omer. The GNOM program was subsequently used to deter-mine the distance distribution function and related structuralparameters, such as the radius of gyration (Rg) and the maxi-mum diameter (Dmax) of the proteins. The distance distributionfunctions, p(r), of the three proteins studied are shown in Fig.4B. Since these functions showed no negative values, we con-cluded that all of the proteins were in a “dilute” state, asrequired for further analysis of molecular shape. The values ofDmax and Rg determined by GNOM are shown in Table II.

FIG. 4. Experimental SAXS curvesfor DjA1-(1–332), DjA1, and DjB4 insolution and the results of the fittingprocedures. A, log I versus q experimen-tal curve. The open squares (DjA1-(1–332)), circles (DjA1), and triangles (DjB4)represent the experimental curve, andthe solid line represents the scatteringintensity from DAM (obtained from DAM-MIN). B, distance distribution functionsof DjA1, DjB4, and DjA1-(1–332). The ar-rows indicate the maximum diameters(Dmax).

Structural Insights into Human Hsp40 13675

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

DjA1 and its C-terminal deletion mutant, DjA1-(1–332), hadsimilar maximum diameters (Dmax � 140 Å) (Fig. 4B), thusimplying that they had a similar asymmetry. The DjB4 chap-erone had a Dmax of 200 Å (Fig. 4B), despite having a smallermolecular mass than DjA1 (Table I), which suggested a moreelongated conformation than for DjA1. Based on the overalltrend of the p(r) functions (Fig. 4B), we concluded that DjB4had a rather elongated shape (51), and that the shapes of DjA1and DjB4 were clearly different. Indeed, whereas DjA1 had ap(r) function that showed only a smooth, well-defined peak, thep(r) function for DjB4 had two clear shoulders, one above andanother below the main peak (Fig. 4B). This complex profile ofthe p(r) function suggested that DjB4 consisted of well-sepa-rated subunits.

The third step of data analysis was to apply the DAMMINprogram to the SAXS data to obtain the low resolution molec-ular shape of the proteins. The dummy atom model (DAM) ofthe structure was derived from the experimental data by as-suming a 2-fold symmetry for the dimers. This symmetricalrestriction significantly reduced the number of free parametersof the models. The starting volume for the DjA1 dimer corre-sponded to a filled sphere with a Dmax � 140 Å (provided byGNOM) containing 4,196 dummy atoms with a radius (ra) � 4.5Å. By using the DAMMIN program, 600 � 10 dummy atomswere assigned to the resulting final model of the DjA1 dimer. Incontrast, the starting volume for DjA1-(1–332) was filled with5,594 dummy atoms with ra � 3.7 Å within a sphere of Dmax �140 Å. Of these, 395 � 5 dummy atoms were assigned byDAMMIN to the final model of the DjA1-(1–332) monomer. Thestarting volume for DjB4 was filled with 6,699 dummy atomswith ra � 4.8 Å within a sphere of Dmax � 200 Å. Of these,280 � 10 dummy atoms remained in the final model of theDjB4 dimer. Twenty independent ab initio DAMMIN simula-tions were done for each protein.

To determine the uniqueness of the resulting shapes, therestorations were done with different starting conditions, all ofwhich yielded similar results. The DAM models obtained weresuperimposed and averaged using the DAMAVER package

software (52) (Fig. 5). The relevant parameters associated withthe use of DAMMIN and the values of Dmax and Rg derivedfrom it are reported in Table II. These values agreed well withthe same parameters determined by GNOM for the three mol-ecules studied. The DAM-derived hydrodynamic parameters(estimated by HydroPro) are shown in Table III. The maximumconcentrations of protein used for the SAXS studies of the DjA1and DjB4 dimers allowed 30 Å as the maximum resolution ofthe final models (Table II), even when the data were measureddown to 15 Å. This resolution did not permit an unambiguousdetermination of the spatial positions of the secondary struc-tural elements of DjA1, but allowed us to determine the overallshape of the molecule and the relative positions of their indi-vidual domains. Table II shows different resolutions for theGNOM and DAMMIN calculations. For DjA1, the whole SAXScurve determined up to qmax � 0.4 Å�1 was used for GNOM sothat the resolution 2/qmax was equal to � 15 Å. Within thehigh q range from q � 0.2 Å�1 to 0.4 Å�1 (Fig. 4A, curve 2), thestatistical errors were very high, and therefore no significantstructural information was obtained from this high q range. Tospeed up the rather long calculations of the DAMMIN programwhile avoiding a significant loss of structural information, onlySAXS data up to qmax � 0.2 Å�1 were used for this analysis andyielded a resolution of 30 Å. The inherent limitations of shapedetermination methods such as DAMMIN prevent higher res-olution protein envelopes from being obtained from statisticallygood SAXS data at higher qmax values, mainly because thesemethods assume a uniform electron density that limits theresolution to 20–30 Å.

GNOM analysis yielded the same Dmax for DjA1-(1–332) andDjA1, despite the dimer arrangement of the latter. This was aclear indication of the highly asymmetrical shape of the DjA1-(1–332) monomer, a characteristic also deduced from the profileof the p(r) function. Fig. 5A shows the envelope models for DjA1(left) and DjA1-(1–332) (two monomers, center). The two mono-mers of DjA1-(1–332) are shown facing each other to illustratehow they may form a dimer, and the monomer envelope placedwithin the dimer envelope fitted one half of the envelope verywell (Fig. 5A, right). The hydrodynamic properties of the twocombined monomers of DjA1-(1–332), calculated by the Hydro-Pro software were similar to those determined for the DjA1dimer model (data not shown). These results confirmed that theDAM model determined for DjA1 was a good description of itslow resolution structure. Two perpendicular views of the enve-lope model of DjB4 are shown in Fig. 5D. Comparison of theshapes of the DjA1 and DjB4 envelopes strongly indicated thatthese proteins clearly had different quaternary structures.

The Hydrodynamic Parameters Calculated by AUC Corrob-orated the Models Generated from the SAXS Data—The hydro-dynamic properties of DjA1 and DjB4 were established bysedimentation equilibrium (Fig. 6, A and B) and sedimentationvelocity (Fig. 7A). In these experiments, the proteins behavedas a monodisperse system and showed no aggregation. Theprofiles of the sedimentation equilibrium for DjA1 (Fig. 6A) andDjB4 (Fig. 6B) fitted well to a single exponential, as indicated

TABLE IMolecular masses of DjA1, DjA1-(1–332), and DjB4 predicted from

their amino acid sequences and determined by SAXS and AUCThe SAXS data were determined by comparing the values of the

extrapolated scattering intensity, I(0), with that of bovine serum albu-min (69 kDa). The sedimentation equilibrium technique was used tocalculate the molecular mass for DjA1 and DjB4. For DjA1-(1–332), themolecular mass was calculated from the ratio s/D (Equation 1), with thesedimentation coefficient (s0

20,w) estimated from sedimentation velocityanalysis and the diffusion coefficient (D0

20,w) estimated from dynamiclight scattering experiments (data not shown).

MethodMolecular mass

DjA1 dimer DjA1-(1–332) DjB4 dimer

kDa

Amino acid sequence 94.0 39.5 80.0SAXS 95 � 10 40 � 5 90 � 10AUC 93.0 � 1.0 42.0 � 2.0 80.0 � 1.0

TABLE IIStructural parameters of the proteins DjA1, DjA1-(1–332), and DjB4 in solution

These were studied by SAXS, and mathematical parameters were calculated using the programs GNOM and DAMMIN. In all cases, theexperimental error was less than 10%.

Mathematical and structural parametersDjA1 dimer DjA1-(1–332) DjB4 dimer

GNOM DAMMIN GNOM DAMMIN GNOM DAMMIN

Free parameter 19 2098 19 5594 26 3350Discrepancy � 0.85 1.26 1.32Resolution (Å) 15 30 15 21 15 30Dmax (Å) 140 145 140 140 200 195Rg (Å) 46 46 41 41 56 55

Structural Insights into Human Hsp4013676

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

by the analysis of their residuals. The molecular masses ofDjA1 and DjB4 derived from the sedimentation equilibriummeasurements were 93.0 � 1.0 kDa and 80.0 � 1.0 kDa, re-spectively (Table I). For DjA1-(1–332), the molecular mass was42.0 � 2.0 kDa, which was estimated from Equation 1 using thesedimentation coefficient (s0

20,w) calculated from the sedimen-tation velocity analysis (see below) and the diffusion coefficient(D0

20,w) calculated from dynamic light scattering experiments(data not shown). The molecular masses determined here forthe human Hsp40 proteins agreed well (within experimentalerror) with the molecular masses calculated from the aminoacid sequences and from the SAXS results (Table I). To deter-mine the sedimentation coefficients s of the proteins studied,the sedimentation velocity experiments were done at differentconcentrations and velocities (Fig. 7). The standard sedimen-tation coefficients s20,w obtained were plotted as a function ofthe protein concentration used to calculate the standard sedi-mentation coefficient at 0 mg/ml, s0

20,w, by extrapolation. Thisapproach was used to correct for effects of temperature, solu-tion viscosity and molecular crowding (42). For DjA1, we ob-tained s0

20,w � 4.63 S, and for DjB4 s020,w � 3.78 S (Fig. 7B). The

sedimentation coefficient for DjA1-(1–332) was velocity-dependent, with s0

20,w � 2.85 S at 30,000 rpm and 3.40 S at40,000 rpm, both of which were larger than the value predictedby the DjA1-(1–332) DAM model (s0

20,w � 2.46 S; see Table III).This effect may be caused by aggregation or, more likely in thiscase, by a highly asymmetrical conformation (53). Some sub-domains of Hsp40 proteins may become more flexible in themonomer form, which would reduce the frictional ratio whencompared with a rigid body and would distort the measured svalues (54). Since the sedimentation velocity measures thetime average s of the particles in solution, it is not possible todistinguish a molecule with a high frictional ratio from anotherwith flexible segments (55). In addition, the Perrin factor andthe elongated shape of DjA1-(1–332) determined from theSAXS data and its domain structure suggested that this pro-tein was flexible, thereby reinforcing the hypothesis that theapparently high s0

20,w was caused by molecular flexibility.The hydrodynamic parameters determined from the AUC

experiments for DjA1, DjA1-(1–332), and DjB4 were similar tothose derived from the DAM models of DjA1 and DjB4 by theHydroPro software (Table III). This finding confirmed the re-liability of the structural models obtained from the SAXS re-sults. Together, the SAXS and the AUC data indicated thatDjA1 and DjB4 were dimers with different structural features,and that DjA1-(1–332) was a monomer with flexible segments(Tables I and III). The structural parameters obtained from theAUC measurements strongly suggested that the DAM modelsderived from the SAXS data represented a good, low resolutionstructure of the real conformation of the proteins studied here.

The results presented in this and in the preceding sectionsindicate that DjA1, DjA1-(1–332), and DjB4 were produced asfolded proteins with the conformational properties of membersof the Hsp40 chaperone family. The next step was to use thelow resolution models of the proteins determined from theSAXS data, together with information about the crystallo-graphic and NMR structures of Hsp40 domains available in theProtein Data Bank, to obtain a more detailed insight into thestructural conformation of DjA1 (dimer and monomer) andDjB4.

Quaternary Structure of DjA1—The DjA1-(1–332) mutantwas a monomer with an elongated shape, as determined byanalytical molecular exclusion chromatography, AUC andSAXS. The NMR structure of the J-domain (56) displayed inred in Fig. 5B, and the crystallographic structure of the Cys-rich and C-terminal domains (22) shown in green in Fig. 5B,

FIG. 5. Low resolution DAM models generated from the SAXSdata for human Hsp40 and molecular fitting. A, DjA1 dimer isshown in white (left, dimensions: 140 � 92 � 52 Å) and the DjA1-(1–332) monomers are shown in red and yellow (center, dimensions:140 � 57 � 42 Å each). DjA1-(1–332) is shown as a dimer modelconstructed from monomers to allow comparison with DjA1. The dimerDjA1-(1–332) has hydrodynamic properties very similar to those ofDjA1, as predicted by HydroPro software (data not shown). One view ofthe superposition of DjA1 and DjA1-(1–332) is represented on the right.B, DjA1-(1–332) ab initio model proved to be a good envelope for theavailable structure of the Hsp40 domains. Red, the NMR structure ofthe J-domain (residues 2–77, 1HDJ, Ref. 56), and green, the crystallo-graphic structure of the Cys-rich and C-terminal domains (residues103–350, 1NLT, Ref. 22), connected by a dotted line (G/F-rich region,structure not available). The subdomains (I-III, Ref. 22) of the Cys-richand C-terminal domains are shown. C, profile of the domain arrange-ments obtained in B, was used to construct a similar profile for the DjA1SAXS model. Red, the NMR structure of the J-domain (residues 2–77,1HDJ, Ref. 56), and green, the crystallographic structure of the Cys-richand C-terminal domains (residues 103–350, 1NLT, Ref. 22), connectedby a dotted line (G/F-rich region, structure not available). D, ab initiomodel for DjB4 (dimensions: 200 � 93 � 48 Å). E, ab initio model forDjB4 proved to be a good envelope for the available structure of theHsp40 domains. Red, the NMR structure of the J-domain (residues2–77, 1HDJ, Ref. 56), and green, the crystallographic structure of theCys-rich and C-terminal domains (residues 180–349, 1C3G, Ref. 27),connected by a dotted line (G/F-rich region, structure not available). Therelative position between the DjB4 and DjA1 J-domains is a strongindication of differences in the quaternary structures of these proteins.The superposition was done using the WebLab ViewerLite software.Additional illustrations can be found at www.lnls.br/saxstructure.

Structural Insights into Human Hsp40 13677

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

were fitted into the DjA1-(1–332) envelope. The final arrange-ment is shown in Fig. 5B, where some important features can beseen, namely, the location of the C terminus within the thinnestpart of the protein envelope and that of the J-domain in thethickest part of the envelope (Fig. 5B). Some portions of thedomains in this model were highly flexible (arrows in Fig. 5B)because of the presence of the G/F-rich region (see below). As

argued before, this could explain why the hydrodynamic proper-ties predicted by the HydroPro software for the DjA1-(1–332)envelope were not in perfect agreement with the experimentalAUC data. The apparently high sedimentation coefficient forDjA1-(1–332) may be justified by the presence of the G/F-richregion, which is believed to function as a flexible spacer betweenthe J-domain and the remaining structure of the protein (11, 24).

FIG. 6. Sedimentation equilibriumexperiments. The sedimentation equi-librium experiments were done at 6,000,8,000, and 10,000 rpm (AN-60Ti rotor) in25 mM Tris-HCl, pH 7.5, containing 500mM NaCl, 1% glycerol, and 0.5 mM �-mer-captoethanol, with scan data acquisitionat 238 nm. The best fits of the experimen-tal data for 250 �g/ml of either DjA1 orDjB4 at 6,000, 8,000, and 10,000 rpm at20 °C are shown (see “Experimental Pro-cedures”). The random distribution of theresiduals (bottom panel) indicated thatthe fit was satisfactory. The sedimenta-tion equilibrium data for DjA1 (A) andDjB4 (B) agreed with dimer structures of93.0 � 1.0 and 80.0 � 1.0 kDa,respectively.

TABLE IIIHydrodynamic parameters of DjA1, DjA1-(1–332), and DjB4 in solution

These were calculated from the AUC data and the structural parameters derived from the DAM model generated by the SAXS experiments usingthe HydroPro software. The partial specific volume, solvent viscosity, and density were determined using Sednterp software. In all cases, theexperimental error was less than 2%.

Hydrodynamic parameters

Method

DjA1 dimer DjA1-(1–332) DjB4 dimer

AUC DAM by HydroPro AUC DAM by HydroPro AUC DAM by HydroPro

s020,w (S) 4.63 4.44 2.85 2.46 3.78 3.63

D020,w (� 107 cm2/s) 4.5 4.2 6.5 5.6 4.4 4.1

P (f/f0) 1.33 1.40 1.36

Structural Insights into Human Hsp4013678

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

In the Cys-rich domain determined by crystallography (22),the subdomains I and II (Fig. 5B) form an L-shaped structure(90o angle), whereas the same domain determined by NMRforms an angle of about 45o resulting in a V-shaped structure(26). This indicates that subdomain II may have some flexibil-ity and could account for the apparently high sedimentationcoefficient of DjA1-(1–332) calculated in the AUC experiments.In addition, the Hsp40 proteins studied here were highly sus-ceptible to degradation (data not shown) and the high flexibil-ity of the G/F-rich region may be an explanation for thisbehavior.

The results of the SAXS experiments indicated that DjA1and DjA1-(1–332) had an elongated shape and similar Dmax

values, which suggested that they probably shared the sameaxial orientation. The superposition of the two models (Fig. 5A,

right) revealed that a dimer composed of two DjA1-(1–332)monomers (Fig. 5A, center) had nearly the same structure asthe DjA1 dimer. This finding indicated that the region involvedin dimer formation was located mainly in the C terminus ofDjA1. The deletion of this region blocked dimer formation anddecreased the protein stability, but had no effect on the second-ary and tertiary (low resolution) structures of DjA1-(1–332).Based on the observed orientation of the monomers in the DjA1dimer model, we fitted the high resolution structures of the J-,Cys-rich, and C-terminal domains into the monomer envelope(Fig. 5B), and then built a model to show the arrangement ofthe Hsp40 domains in the DjA1 dimer (Fig. 5C). The DAMmodel generated from the SAXS data indicated that DjA1 wasa dimer with a bent horseshoe shape, similar to the structuredetermined for the C-terminal and Cys-rich domains of Ydj1(subfamily A) (22). Fig. 5C shows that the DjA1 protein wasbullet-shaped and that the C termini monomers were located inthe thinnest side of the model and were responsible for thedimerization. The N-terminal monomers (J-domains) were lo-cated in the thickest part of the model where they could act asa clamp for binding to Hsp70 (Fig. 5C). The C-terminal andCys-rich domains of Ydj1 (green in Fig. 5C; 1NLT, Ref. 21) wereplaced on the axial axis, where they fitted very well. TheL-shape of this complex was a consequence of the presence ofthe Cys-rich domain, which was placed in the center to face itscounterpart in the other monomer (Fig. 5B).

Quaternary Structure of DjB4—The data generated by SAXSwere used to create an envelope model of the quaternary struc-ture of DjB4 that agreed with the structural parameters meas-ured (Fig. 5D). The envelope model of DjB4 was compatiblewith an elongated structure, with a Dmax greater than that forDjA1 (200 Å versus 140 Å). This was an unexpected findingbecause DjA1 has a higher molecular mass than DjB4, asshown by the AUC experiments (Table I). The similaritiesbetween the predicted and the calculated hydrodynamic pa-rameters of the envelope model were a strong indication of theaccuracy of the model (Table III). The DjB4 ab initio modelgenerated from the SAXS data using DAMMIN andDAMAVER is shown in Fig. 5D. The high resolution structuresof the Hsp40 domains available, namely, the NMR structure ofthe J-domain (56), which is shown in red in Fig. 5E, and thecrystallographic structure of the C terminus of Sis1 (subfamilyB) (22), which is shown in green in Fig. 5E, were fitted into theDjB4 envelope in the best possible position. Studies of C-ter-minal deletions of Hsp40 (22, 27, this work) have indicated thatthis region is responsible for dimerization, which implies thatthe C-terminal monomers must be located near to each other.The localization of the two C-terminal domains of DjB4 in thecenter of the model as shown in Fig. 5E fulfilled this require-ment. Any other arrangement for these domains would compli-cate the formation of the dimer, and would be contrary to thestrong tendency of Hsp40 proteins to form dimers (22, 27). TheJ-domains were placed at the extremities in the model (Fig. 5E)because they are connected to the C-terminal by a long, flexibleregion formed by the G/F-rich region (residues 68–139) plus asequence of 60 amino acids of unknown function (residues140–199). There is no high resolution structure of this region,which makes it difficult to predict the correct arrangement ofthese regions (residues 68–199) in the model. This uncertaintyis represented by a dotted line (Fig. 5E).

DjA1 and DjB4: Similarities and Differences—Genetic andbiochemical studies have shown that the Hsp40 family is struc-turally and functionally diverse (33). Such diversity means thatthe interaction of Hsp70 with different Hsp40 generates spe-cialized combinations that facilitate specific actions of theHsp70 chaperone machinery within the cell (19, 33, 35). Al-

FIG. 7. Sedimentation velocity experiments. A, sedimentationvelocity experiments were done at 20 °C, with scan data acquisitionat 230 and 238 nm at low and high protein concentrations, respec-tively. Rotor (AN-60Ti) velocities were 25,000 rpm for DjA1 and DjB4and 30,000 rpm for DjA1-(1–332). The figure shows experiments with0.6 mg/ml DjA1, 0.8 mg/ml DjA1-(1–332) and 0.48 mg/ml DjB4 in 25mM Tris-HCl, pH 7.5, containing 500 mM NaCl, 1% glycerol, and 0.5mM �-mercaptoethanol (except for DjA1-(1–332), which was analyzedin the absence of glycerol). The g(s*) distributions were fitted usingOrigin (Microcal Software) with a Gaussian fit giving apparent sed-imentation coefficients (s*) of about 3.8, 3.2, and 2 Svedberg for DjA1,DjB4, and DjA1-(1–332), respectively. B, plots of s20,w versus proteinconcentrations fitted by linear regression to calculate the s0

20,w: 4.63 �0.05, 3.78 � 0.01, and 2.85 � 0.03 for DjA1, DjB4, and DjA1-(1–332),respectively.

Structural Insights into Human Hsp40 13679

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

though Hsp40 from subfamilies A and B differ in their chaper-one activity, the reasons for these differences are still incom-pletely understood. Determination of the tertiary structure ofsome of the domains of Hsp40 proteins from subfamilies A andB (22, 26, 27, 56) has contributed to our understanding of thedifferences in their chaperone activity. The data on the quater-nary structure of the two proteins studied here provides addi-tional information that may help to explain the differences intheir binding to substrates and to Hsp70 proteins.

The presence of a J-domain defines a protein as a memberof the Hsp40 family. Since the conserved residues in thisdomain are responsible for stabilizing the anti-parallel coiledcoil (56), we suggest that this domain has the same confor-mation in DjA1 and DjB4. However, the J-domains are lo-cated differently in their quaternary structure (because thedimerization sites, at the C terminus, must be in the center ofthe model): the J-domains from each monomer of DjA1 arearranged facing each other, whereas in DjB4, these domainsare placed in the extremities. Since there is still no highresolution structure of the G/F-rich region, which is an im-portant site for the binding of Hsp40 to Hsp70 (20, 21), werepresented this region as a dotted line in our models (Fig. 5,B, C, and E). Based on primary sequence homology, theG/F-rich region is larger in DjB4 than in DjA1 (Fig. 1B) and,in combination with a region of about 60 residues for whichthere is also no available structure, this allowed for the moreextended conformation shown for DJB4 (Fig. 5E), in whichthe J-domains are located in the extremities. The C-terminalregions of both subfamilies A and B are responsible for sub-strate binding and are structurally similar (22), but differ intheir ability to refold luciferase (28, 35).

Lu and Cyr (35) studied the interaction of Ydj1 (subfamily A)and Sis1 (subfamily B) with two Hsp70s and concluded thatthese two Hsp40 proteins differed in their chaperone functions,perhaps because of the zinc finger region in Ydj1 which maystabilize the polypeptide binding groove of Hsp40. Fan et al.(19) engineered two Hsp40 chimeras in which the central partof Ydj1 (residues 101–255) was replaced by the central part ofSis1 (residues 108–257) and vice versa. This exchange alteredthe substrate binding specificity of the proteins and their abil-ity to stimulate the luciferase refolding activity of Hsp70.These results indicated that selective substrate binding sitesare needed for substrate selection in Hsp40 and that the cen-tral portion of this protein is important for specifying chaper-one activity (19). In agreement with these observations, thecentral part of DjA1 and DjB4 appeared to be the main regionresponsible for the differences in the low resolution modelsdescribed here: the central part of DjA1 (G/F-rich region plusCys-rich region, residues 68–207) apparently keeps the J-do-

mains close to each other whereas the central part of DjB4(residues 68–199) keeps the J-domains far apart from eachother. A different global structural conformation rather thanspecific amino acid interactions could be responsible for thedifference in the substrate specificity of the chaperones, assuggested by Lopez et al. (57), who compared the ability ofSis1/Hdj1 (subfamily B) and Ydj1/Hdj2 (subfamily A) Hsp40s tomaintain the Rnq1 prion. Hence, the differences in the quater-nary structures of DjA1 and DjB4 mentioned above could con-tribute to the variations in chaperone activity.

The high diversity of Hsp40 proteins may mediate severaltypes of interactions between Hsp70 and Hsp40, and couldexplain the need for proteins with different quaternary struc-tures in subfamilies A and B. Several lines of evidence suggestthe presence of multiple binding sites among Hsp70 and Hsp40partners: Suh et al. (21) showed the existence of at least twobinding sites for DnaK in DnaJ, both located in the J-domain.Linke et al. (36) suggested that a possible contact between theCys-rich domain may be important for the transfer of substratefrom DnaJ to DnaK. Demand et al. (58) and Qian et al. (59)provided evidence that the C terminus of eukaryotic Hsp70interacts with Hsp40 from subfamily B. Laufen et al. (12)proposed that the interaction between Hsp40 and Hsp70 mayinvolve multiple steps in which the Hsp40 protein “targets”unfolded substrate to Hsp70, as follows: 1) interaction betweenthe J-domain with the Hsp70 NBD-ATP, 2) shift of substratefrom Hsp40 to the Hsp70 substrate binding domain (SBD), 3)transfer of the information regarding substrate binding fromthe Hsp70 SBD to the Hsp70 NBD, resulting in the completehydrolysis of ATP and closing of the Hsp70 SBD lid and, finally,4) Hsp40 dissociation from the Hsp70-ADP-substrate complex.Because of the high diversity of Hsp40 and the existence ofmultiple contacts among different Hsp40-Hsp70 partners (21,36, 58, 59), the functional cycle described above may consist ofadditional steps that depend on the Hsp40-Hsp70 partners.

The fact that DnaJ (subfamily A) is unable to bind to theisolated N- and C-domains of DnaK (21), but that Hsp40 fromsubfamily B does (58, 59), is in good agreement with the qua-ternary models proposed here: the J-domains of DjA1 are ar-ranged in a position that allows binding to both domains ofHsp70 at once, while the J-domains of DjB4 can act independ-ently of each other when binding to Hsp70. The quaternarystructure of DjA1 presented here is compatible with the modelof interaction between DnaJ and DnaK proposed by Suh et al.(Ref. 21; Fig. 5 in their report), and the quaternary structure ofDjB4 presented here is compatible with the model of interac-tion between Sis1 and Hsp70 proposed by Sha et al. (Ref. 27;Fig. 6 in their report). Several authors (22, 27, 59, 60) havesuggested that the peptide binding domain of Hsp70 nestles

FIG. 8. Proposed models for the in-teraction of DjA1 and DjB4 withHsp70. See text for discussion.

Structural Insights into Human Hsp4013680

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

into the groove between the monomers of Hsp40. Despite theirlow resolution, the size and shape of the models presented herebased on SAXS are compatible with these views.

Fig. 8 provides a diagram suggesting the possible interac-tions between DjA1 and DjB4 with Hsp70, based on the qua-ternary structure determined here for DjA1 and DjB4, and onthe current view regarding the interaction of Hsp40 from sub-families A and B with Hsp70 (21, 36, 58, 59, 61). In this model,the interaction between DjA1 and Hsp70 may occur by twobinding sites located in each J-domain, one interacting with theHsp70 NBD and the other with a region close to the Hsp70substrate binding site (21). In the interaction model for DjA1-Hsp70, the Cys-rich domain is in a proper position to interactwith Hsp70 and to transfer substrate to it, as proposed byLinke et al. (36). In the interaction model for DjB4-Hsp70, thebinding site located in one J-domain interacts with the Hsp70NBD, and the C terminus of DjB4 is able to interact with the Cterminus of Hsp70, as previously suggested (58, 59). The size ofDjB4 allows the suggested double interaction between the J-domain and the Hsp70 NBD and between the Hsp40 C termi-nus and the Hsp70 C terminus to occur in one cycle.

The data presented here are in good agreement with thecurrent view of the interaction between Hsp40 and Hsp70described above. The conformational arrangement of the J-domain and G/F-rich region found for DjA1 and DjB4 mayconfer flexibility to these proteins, allowing them to accommo-date and interact with both nucleotide and substrate bindingdomains of Hsp70 in the same cycle of interaction (see discus-sion above). Low or high resolution structural studies for otherHsp40 from subfamilies A and B would provide more informa-tion about the interactions between Hsp70-Hsp40 partners.Finally, our results shed light on several structural aspects ofDjA1 and DjB4 that are crucial to the intracellular chaperonefunction of these proteins.

CONCLUSION

Small angle x-ray scattering experiments yielded low reso-lution structural models (molecular envelopes) for the proteinsDjA1 and DjB4. A low resolution model for the structure of aC-terminal-deleted mutant of DjA1, DjA1-(1–332), in whichthere is no dimer formation, was also determined. The hydro-dynamic properties calculated from these protein modelsagreed well with those determined directly from AUC experi-ments. This agreement supported and confirmed the structuralmodels derived from the SAXS results. The characteristics ofthe low resolution models and the values of their molecularmasses indicated that DjA1-(1–332) was a monomer whereasDjA1 and DjB4 were both dimers. An additional analysis com-bining the envelope functions derived from SAXS data, and theknown information about the high resolution structure of theHsp40 C-terminal and J-domains showed that DjA1 had arather compact structure in which two highly asymmetricalmonomers bound in a bullet shape, while in DjB4 the N terminihad a completely different arrangement in which they pointedaway from the compact C termini core responsible for dimer-ization. These differences may explain why these two Hsp40subfamilies have different chaperone activities and contributeto our understanding of the role of Hsp40 in substrate binding,transport, and interaction with Hsp70 proteins.

Acknowledgments—We thank H. L. M. Guedes for helpful discus-sions, the technical staff of LNLS for assistance, and Dr. S. Hyslop forcritical reading of this manuscript.

REFERENCES

1. Gething, M. J., and Sambrook, J. (1992) Nature 355, 33–452. Bukau, B., and Horwich, A. L. (1998) Cell 92, 351–366

3. Fink, A. L. (1999) Physiol. Rev. 79, 425–4494. Mayer, M. P., Brehmer, D., Gassler C. S., and Bukau, B. (2001) Adv. Protein

Chem. 50, 1–455. Hartl, F. U., and Hayer-Hartl, M. (2002) Science 295, 1852–18586. Borges, J. C., and Ramos, C. H. I. (2005) Protein Pept. Lett. 12, 257–2617. Mayer, M. P., Schroder H., Rudiger, S., Pall, K., Laufen, T., and Bukau, B.

(2000) Nat. Struct. Biol. 7, 586–5938. Martin, J., and Hartl, F. U. (1997) Curr. Opin. Struct. Biol. 7, 41–529. Hendrick, J. P., Langer, T., Davis, T. A., Hartl, F. U., and Wiedmann, M.

(1993) Proc. Natl. Acad. Sci. U. S. A. 90, 10216–1022010. Liberek, K., Wall, D., and Georgopoulos, C. (1996) Proc. Natl. Acad. Sci.

U. S. A. 92, 6224–622811. Karzai, A. W., and MaMacken, R. (1996) J. Biol. Chem. 271, 11236–1124612. Laufen, T., Mayer, M. P., Beisel, C., Klostermeier, D., Mogk, A., Reistein, J.,

and Bukau, B. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 5452–545713. Langer, T., Lu, C., Echols, H., Flanagan, J., Hayer, M. K., and Hartl, F. U.

(1992) Nature 356, 683–68914. Szabo, A., Langer, T., Schroder, H., Flanagan, J., Bukau, B., and Hartl, F. U.

(1994) Proc. Natl. Acad. Sci. U. S. A. 91, 10345–1034915. Rudiger, S., Scheneider-Mergener, J., and Bukau, B. (2001) EMBO J. 20,

1042–105016. Yan, W., and Craig, E. A. (1999) Mol. Cell. Biol. 19, 7751–775817. Johnson, J. L., and Craig, E. A. (2001) J. Cell Biol. 152, 851–85618. Lee, S., Fan, C. Y., Younger, J. M., Ren, H., and Cyr, D. M. (2002) J. Biol.

Chem. 277, 21675–2168219. Fan, C. Y., Lee, S., Ren, H. Y., and Cyr, D. M. (2004) Mol. Biol. Cell 15,

761–77320. Szabo, A., Korszun, R., Hartl, F. U., and Flanagan, J. (1996) EMBO J. 15,

408–41721. Suh, W-C., Lu, C. Z., and Gross, C. A. (1999) J. Biol. Chem. 274, 30534–3053922. Li, J., Qian, X., and Sha, B. (2003) Structure 11, 1475–148323. Greene, M. K., Maskos, K., and Landry, S. J. (1998) Proc. Natl. Acad. Sci.

U. S. A. 95, 6108–611324. Szyperski, T., Pellechia, M., Wall, D., Georgopoulus, C., and Wuthrich, K.

(1994) Proc. Natl. Acad. Sci. U. S. A. 91, 11343–1134725. Banecki, B., Liberek, K., Wall, D., Wawrzynow, A., Georgopoulos, C., Bertoli,

E., Tanfani, F., and Zylicz, M. (1996) J. Biol. Chem. 271, 14840–1484826. Martinez-Yamout, M., Legge, G. B., Zhang, O., Wright, P. E., and Dyson, H. J.

(2000) J. Mol. Biol. 300, 805–81827. Sha, B., Lee, S., and Cyr, D. M. (2000) Structure 8, 799–80728. Lu, Z., and Cyr, D. M. (1998) J. Biol. Chem. 273, 5970–597829. Miernyk, J. A. (2001) Cell Stress Chaperones 6, 209–21830. Borges, J. C., Peroto, M. C., and Ramos, C. H. I. (2001) Genet. Mol. Biol. 24,

85–9231. Venter, J. C., Adams, M. D., Myers, E. W., Li, P. W., Mural, R. J., Sutton, G.

G., Smith, H. O., Yandell, M., Evans, C. A., Holt, R. A. (2001) Science 291,1304–1351

32. Cyr, D. M., Langer, T., and Douglas, M. G. (1994) Trends Biochem. Sci. 19,176–181

33. Cheetham, M. E., and Caplan, A. J. (1998) Cell Stress Chaper. 3, 28–3634. Ohtsuka, K., and Hata, M. (2000) Cell Stress Chaper. 5, 98–11235. Lu, Z., and Cyr, D. M. (1998) J. Biol. Chem. 273, 27824–2783036. Linke, K., Wolfram, T., Bussemer, J., and Jakob, U. (2003) J. Biol. Chem. 278,

44457–4446637. Edelhock, H. (1967) Biochemistry 6, 1948–195438. Gill, S. C., and von Hippel, P. H. (1989) Anal. Biochem. 182, 319–32639. Van Holde, K. E., and Weischet, W. O. (1978) Biopolymers 17, 1397–140340. Stafford, W.F. (1994) Methods Enzymol. 240, 478–50141. Johnson, M. L., Correia, J. J., Yphantis, D. A., and Halvorson, H. R. (1981)

Biophys. J. 36, 575–58842. Laue, T. M. (2001) Curr. Opin. Struct. Biol. 11, 579–58343. Kellermann, G., Vicentin, F., Tamura E., Rocha, M., Tolentino, H., Barbosa,

A., Craievich, A., and Torriani, I. (1997) J. Appl. Crystallogr. 30, 880–88344. Borges, J. C., Fischer, H., Craievich, A. F., Hansen, L. D., and Ramos, C. H. I.

(2003) J. Biol. Chem. 278, 35337–3534445. Svergun, D. I. (1992) J. Appl. Crystallogr. 25, 495–50346. Porod, G. (1982) in Small Angle X-ray Scattering (Glatter, O., and Kratky, O.,

eds) pp. 17–51, Academic Press, London47. Svergun, D. I. (1999) Biophys. J. 76, 2879–288648. Garcia de la Torre, J., Huertas, M. L., and Carrasco, B. (2000) Biophys. J. 78,

719–73049. Garcia de la Torre, J., Carrasco, B., and Harding, S. E. (1997) Eur. Biophys. J.

25, 361–37250. Terada, K., and Mori, M. (2000) J. Biol. Chem. 275, 24728–2473451. Feigin, L. A., and Svergun, D. I. (1987) Structure Analysis by Small Angle

X-ray and Neutron Scattering, pp. 83–87, Plenum Press, New York52. Volkov, V. V., and Svergun, D. I. (2003) J. Appl. Crystallogr. 36, 860–86453. Zimm, B. H. (1974) Biophys. Chem. 1, 279–29154. Laue, T. M., Shal, B. D., Ridgeway, T. M., and Pelletier, S. M. (1992) (Harding,

S., Rowe, A., and Horton, J. C., eds) pp. 90–125, Royal Society of Chemistry,London

55. Waxman, E., Laws, W. R., Laue, T. M., and Ross, J. B. A. (1994) (Schuster,T. M., and Laue, T. M., eds) pp. 189–208, Birkhauser, Boston

56. Qian, Y. Q., Patel, D., Hartl, F. U., and McColl, D. J. (1996) J. Mol. Biol. 260,224–235

57. Lopez, N., Aron, R., and Craig, E. A. (2003) Mol. Cell. Biol. 14, 1172–118158. Demand, J., Luders, L., and Hohfeld, J. (1998) Mol. Cell. Biol. 18, 2023–202859. Qian, X., Hou, W., Zhengang, L., and Sha, B. (2002) Biochem. J. 361, 27–3460. Landry, S. J. (2003) Structure 11, 1465–146661. Fan, C.-Y., Ren, H.-Y., Lee, P., Caplan, A. J., and Cyr, D. M. (2005) J. Biol.

Chem. 280, 695–702

Structural Insights into Human Hsp40 13681

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from

Júlio C. Borges, Hannes Fischer, Aldo F. Craievich and Carlos H. I. RamosDIFFERENT QUATERNARY STRUCTURES

DJA1 FROM SUBFAMILY A AND DJB4 FROM SUBFAMILY B HAVE Low Resolution Structural Study of Two Human HSP40 Chaperones in Solution:

doi: 10.1074/jbc.M408349200 originally published online January 20, 20052005, 280:13671-13681.J. Biol. Chem. 

  10.1074/jbc.M408349200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/280/14/13671.full.html#ref-list-1

This article cites 57 references, 22 of which can be accessed free at

by guest on February 22, 2020http://w

ww

.jbc.org/D

ownloaded from