Recombinant Human Peroxisomal Targeting Signal Receptor PEX5
1999; Elsevier BV; Volume: 274; Issue: 9 Linguagem: Inglês
10.1074/jbc.274.9.5666
ISSN1083-351X
AutoresWolfgang Schliebs, Jürgen Saidowsky, Bogos Agianian, Gabriele Dodt, Friedrich W. Herberg, W H Kunau,
Tópico(s)RNA Research and Splicing
ResumoImport of matrix proteins into peroxisomes requires two targeting signal-specific import receptors, Pex5p and Pex7p, and their binding partners at the peroxisomal membrane, Pex13p and Pex14p. Several constructs of human PEX5 have been overexpressed and purified by affinity chromatography in order to determine functionally important interactions and provide initial structural information. Sizing chromatography and electron microscopy suggest that the two isoforms of the human PTS1 receptor, PEX5L and PEX5S, form homotetramers. Surface plasmon resonance analysis indicates that PEX5 binds to the N-terminal fragment of PEX14-(1–78) with a very high affinity in the low nanomolar range. Stable complexes between recombinant PEX14-(1–78) and both the full-length and truncated versions of PEX5 were formed in vitro. Analysis of these complexes revealed that PEX5 possesses multiple binding sites for PEX14, which appear to be distributed throughout its N-terminal half. Coincidentally, this part of the molecule is also responsible for oligomerization, whereas the C-terminal half with its seven tetratricopeptide repeats has been reported to bind PTS1-proteins. A pentapeptide motif that is reiterated seven times in PEX5 is proposed as a determinant for the interaction with PEX14. Import of matrix proteins into peroxisomes requires two targeting signal-specific import receptors, Pex5p and Pex7p, and their binding partners at the peroxisomal membrane, Pex13p and Pex14p. Several constructs of human PEX5 have been overexpressed and purified by affinity chromatography in order to determine functionally important interactions and provide initial structural information. Sizing chromatography and electron microscopy suggest that the two isoforms of the human PTS1 receptor, PEX5L and PEX5S, form homotetramers. Surface plasmon resonance analysis indicates that PEX5 binds to the N-terminal fragment of PEX14-(1–78) with a very high affinity in the low nanomolar range. Stable complexes between recombinant PEX14-(1–78) and both the full-length and truncated versions of PEX5 were formed in vitro. Analysis of these complexes revealed that PEX5 possesses multiple binding sites for PEX14, which appear to be distributed throughout its N-terminal half. Coincidentally, this part of the molecule is also responsible for oligomerization, whereas the C-terminal half with its seven tetratricopeptide repeats has been reported to bind PTS1-proteins. A pentapeptide motif that is reiterated seven times in PEX5 is proposed as a determinant for the interaction with PEX14. peroxisomal targeting signal tetratricopeptide repeat glutathioneS-transferase dithiothreitol tobacco etch virus polyacrylamide gel electrophoresis response units nickel-nitrilotriacetic acid N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine Proteins residing inside organelles must be translocated across lipid bilayers to reach their final destination. It has been shown for some compartments that protein translocation across hydrophobic membranes occurs through proteinaceous complexes, which are evolutionary conserved (1Pohlschröder M. Prinz W.A. Hartmann E. Beckwith J. Cell. 1997; 91: 563-566Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar). However, the initial notion that the translocation machineries and their mechanisms are also conserved among different organelles has turned out to be an oversimplification (2Schatz G. Dobberstein B. Science. 1996; 271: 1519-1526Crossref PubMed Scopus (914) Google Scholar). Combined classical and molecular genetic analyses of protein import into peroxisomes in lower and higher eukaryotes have identified a large number of genes (PEX genes), the protein products (peroxins) of which play essential roles in the different steps of transport from the cytoplasm to the peroxisomal matrix (for recent reviews, see Refs.3Subramani S. Physiol. Rev. 1998; 78: 171-188Crossref PubMed Scopus (281) Google Scholar, 4Erdmann R. Veenhuis M. Kunau W.-H. Trends Cell Biol. 1997; 7: 400-407Abstract Full Text PDF PubMed Scopus (115) Google Scholar, 5Elgersma Y. Tabak H.F. Biochim. Biophys. Acta. 1996; 1286: 269-283Crossref PubMed Scopus (70) Google Scholar, 6Waterham H.R. Cregg J.M. Bioessays. 1997; 19: 57-66Crossref PubMed Scopus (53) Google Scholar). In a number of cases it has been possible to link mutations in human orthologues to peroxisomal disorders, most being fatal (7Subramani S. Nat. Genet. 1997; 15: 331-333Crossref PubMed Scopus (82) Google Scholar,8Kunau W.-H. Curr. Opin. Microbiol. 1998; 1: 232-237Crossref PubMed Scopus (46) Google Scholar). Import of peroxisomal matrix proteins (for recent reviews, see Refs.3Subramani S. Physiol. Rev. 1998; 78: 171-188Crossref PubMed Scopus (281) Google Scholar, 4Erdmann R. Veenhuis M. Kunau W.-H. Trends Cell Biol. 1997; 7: 400-407Abstract Full Text PDF PubMed Scopus (115) Google Scholar, 5Elgersma Y. Tabak H.F. Biochim. Biophys. Acta. 1996; 1286: 269-283Crossref PubMed Scopus (70) Google Scholar, 6Waterham H.R. Cregg J.M. Bioessays. 1997; 19: 57-66Crossref PubMed Scopus (53) Google Scholar) depends on two well defined targeting signals, termed PTS1 and PTS2.1 Two import receptors, Pex5p and Pex7p, have been identified which specifically bind PTS1 and PTS2, respectively. Both receptor proteins contain repetitive sequence motifs, each belonging to established structural families. Pex5p possesses seven tetratricopeptide repeats (TPRs) (9Hirano T. Kinoshita N. Morikawa K. Yanagida M. Cell. 1990; 60: 319-328Abstract Full Text PDF PubMed Scopus (237) Google Scholar, 10Lamb J.R. Tugendreich S. Hieter P. Trends Biochem. Sci. 1995; 20: 257-259Abstract Full Text PDF PubMed Scopus (545) Google Scholar, 11Goebl M. Yanagida M. Trends Biochem. Sci. 1991; 16: 173-177Abstract Full Text PDF PubMed Scopus (372) Google Scholar), while Pex7p has six WD-40 motifs (12Neer E.J. Schmidt C.J. Nambudripad R. Smith T.F. Nature. 1994; 371: 297-300Crossref PubMed Scopus (1280) Google Scholar). Pex13p and Pex14p, membrane-bound peroxins, have been demonstrated to bind the two PTS receptors. Both are components of a recently reported complex network of interacting peroxins (4Erdmann R. Veenhuis M. Kunau W.-H. Trends Cell Biol. 1997; 7: 400-407Abstract Full Text PDF PubMed Scopus (115) Google Scholar, 8Kunau W.-H. Curr. Opin. Microbiol. 1998; 1: 232-237Crossref PubMed Scopus (46) Google Scholar, 13Huhse B. Rehling P. Albertini M. Blank L. Meller K. Kunau W.H. J. Cell Biol. 1998; 140: 49-60Crossref PubMed Scopus (124) Google Scholar, 14Albertini M. Rehling P. Erdmann R. Girzalsky W. Kiel J.A.K.W. Veenhuis M. Kunau W.-H. Cell. 1997; 89: 83-92Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar). Pex13p binds the PTS1 receptor Pex5p with its cytoplasmic SH3 domain (15Gould S.J. Kalish J.E. Morrell J.C. Bjorkman J. Urquhart A.J. Crane D.I. J. Cell Biol. 1996; 135: 85-95Crossref PubMed Scopus (209) Google Scholar, 16Erdmann R. Blobel G. J. Cell Biol. 1996; 135: 111-121Crossref PubMed Scopus (184) Google Scholar, 17Elgersma Y. Kwast L. Klein A. Voorn-Brouwer T. van den Berg M. Metzig B. America T. Tabak H.F. Distel B. J. Cell Biol. 1996; 135: 97-109Crossref PubMed Scopus (183) Google Scholar). Pex14p interacts with both PTS-dependent receptors (14Albertini M. Rehling P. Erdmann R. Girzalsky W. Kiel J.A.K.W. Veenhuis M. Kunau W.-H. Cell. 1997; 89: 83-92Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar, 18Brocard C. Lametschwandtner G. Koudelka R. Hartig A. EMBO J. 1997; 16: 5491-5500Crossref PubMed Scopus (107) Google Scholar, 19Fransen M. Terlecky S.R. Subramani S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 8087-8092Crossref PubMed Scopus (135) Google Scholar). Therefore, it was proposed that Pex14p may represent the point of convergence of both PTS-dependent import pathways (14Albertini M. Rehling P. Erdmann R. Girzalsky W. Kiel J.A.K.W. Veenhuis M. Kunau W.-H. Cell. 1997; 89: 83-92Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar). The functional importance of Pex14p for peroxisome biogenesis is further supported by its interaction with Pex13p and with an additional membrane bound peroxin Pex17p (13Huhse B. Rehling P. Albertini M. Blank L. Meller K. Kunau W.H. J. Cell Biol. 1998; 140: 49-60Crossref PubMed Scopus (124) Google Scholar). An understanding of peroxisomal protein import at the molecular level requires knowledge about the structure of the peroxins and conformational changes resulting from their interactions. Herein, we report initial biochemical and biophysical studies of the human PTS1 receptor PEX5, overexpressed in Escherichia coli, purified to homogeneity, and its interaction with PEX14. Cloning experiments were performed withE. coli strain DH5α. The cDNA fragments coding for PEX5L, PEX5S, and the truncated versions PEX5L-(1–251) and PEX5L-(323–639) were from pcDNA3 (Invitrogen) derived plasmids. 2G. Dodt, D. Warren, T. Yahraus, M. Soukupova, E. Becker, P. Rehling, and S. J. Gould, manuscript in preparation. A polymerase chain reaction product corresponding to PEX5L-(214–639) was amplified from pGD106 (20Braverman N. Dodt G. Gould S.J. Valle D. Hum. Mol. Genet. 1998; 7: 1195-1205Crossref PubMed Scopus (148) Google Scholar) using the sense primer 5′-ATTGTCGACCATGGAGTTCCTGAAATTC-3′ containing aNcoI site (underlined) and a vector-specific antisense primer corresponding to the Sp6 promotor region (5′-TATTTAGGTGACACTATAG-3′). DNA fragments encoding full-length PEX5L, PEX5S, and the C-terminal fragments PEX5L-(214–639) and PEX5L-(323–639) were digested withNcoI/BglII. The resulting fragments, which also contained additional 218 base pairs from the 3′-noncoding region and anNcoI/NotI fragment corresponding to PEX5L-(1–251), were subcloned into expression plasmids kindly provided by G. Stier (EMBL, Heidelberg). These plasmids were derived from pET9d (Novagen) by replacing the unique NcoI site with a DNA fragment encoding a hexahistidinyl (His6) tag and a TEV (tobacco etch virus) protease cleavage site and containing several unique endonuclease recognition sites. The PEX5 coding fragments were ligated with NcoI/BamHI orNcoI/NotI digested vectors, thus, fusing a peptide with the sequence MKHHHHHHPMSDYDIPTTENLYFQGAM to the N termini of the PEX5 proteins. A DNA fragment encoding GST-PEX14-(1–78) was amplified by polymerase chain reaction using pGEX-PEX14-(1–134) 3Will, G., Soukupova, M., Hong, X., Erdmann, K. S., Kiel, J. A. K. W., Dodt, G., Kunau, W.-H., and Erdmann, R., in press. as a template and the primers 5′-GCAGTGGTCTCTCATGTCCCCTATACTAGGTT-3′ (sense,BsaI recognition site underlined) and 5′-CCAAGCTTAGTCGACCGAAGGCTCATCGGCAGC-3′ (antisense,HindIII recognition site underlined), digested withBsaI (creating a NcoI-compatible 5′-overhang) andHindIII and subcloned intoNcoI/HindIII digested pET21d plasmid (Novagen). The DNA constructs were verified by restriction analysis and partial DNA sequence analysis using an ABI automated sequencer (Applied Biosystems). A nucleotide exchange was found in the coding region for PEX5L-(1–251), resulting in an amino acid substitution of glutamic acid to aspartic acid at position 33. Expression of His6-tagged PEX5 forms and GST-PEX14-(1–78) was carried out in E. coli strain BL21(DE3). Fresh transformants were grown in Luria Broth medium supplemented with 30 mg/liter kanamycin or 100 mg/liter ampicillin. Cells were induced in the mid-log phase with 0.4 mmisopropyl-β-d-thiogalactopyranoside and grown for another 4–6 h at a temperature of 37 °C. Cells were harvested by centrifugation and were stored at −80 °C. Cell pellets were thawed in buffer A containing 50 mm Tris-HCl, pH 8.0, 150 mm NaCl, and 1 mm dithiothreitol (DTT) and broken with a French pressure cell (Aminco) in one or two passages. Cell debris and other insoluble material were removed by centrifugation (44,000 × g, 45 min). The supernatant containing the soluble proteins was loaded directly on either a Ni-NTA-agarose (Qiagen) or a glutathione-Sepharose 4B (Pharmacia) column equilibrated with buffer A. PEX5L-(323–639) was purified from inclusion bodies on Ni-NTA-agarose columns under denaturing conditions as described in the QIAexpress system instructions (Qiagen). PEX5 proteins were eluted from Ni-NTA-agarose with 60 mm imidazole. For GST-PEX14-(1–78) bound to glutathione-Sepharose a single-step elution with 10 mm glutathione was performed. All affinity purification steps were carried out at 4 °C and monitored by SDS-PAGE analysis. Protease cleavage was performed at 37 °C for 4 h. Thrombin (Serva) was used at a concentration of 3 NIH units/mg of purified GST-PEX14-(1–78), whereas 0.1 NIH units of TEV protease (Life Technologies, Inc.)/mg of PEX5 proteins effectively cleaved off the His6 tag. The thrombin activity was inactivated with 1 mm phenylmethylsulfonyl fluoride for 15 min at 37 °C. A Mono Q ion exchange column (HR5/5, Pharmacia) equilibrated with 20 mm Tris-HCl, pH 8.0, 1 mm DTT was used to remove proteases and fusion parts and to concentrate the recombinant proteins. For that purpose PEX14-(1–78) was eluted with a 20-ml linear gradient between 0 and 120 mm NaCl, whereas for all recombinant PEX5 proteins, 20-ml linear gradients from 150 to 800 mm NaCl were applied. Size exclusion chromatography was performed on a Superose 6 or a Superose 12 column HR 10/30 (Pharmacia) with buffer A at a flow rate of 0.5 ml/min at 20 °C. Molecular mass standards were thyroglobulin (669 kDa), apoferritin (443 kDa), amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), and glutathione S-transferase (GST, 52 kDa), respectively. Protein concentrations were determined according to Bradford (21Bradford M.M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (211983) Google Scholar) using bovine serum albumin as a standard. SDS-denatured proteins were separated by standard PAGE (22Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (205531) Google Scholar) or Tricine-PAGE according to the method of Schägger and von Jagow (23Schägger H. von Jagow G. Anal. Biochem. 1987; 166: 368-379Crossref PubMed Scopus (10410) Google Scholar) and visualized by Coomassie Blue or silver stain (24Heukeshoven J. Dernick R. Electrophoresis. 1988; 1: 28-32Crossref Scopus (640) Google Scholar). Protein solutions were adjusted to concentrations of 0.1–1 mg/ml and applied onto glow discharged 400 mesh carbon-coated grids for 1 min. Grids were flashed with 150 μl of buffer and stained with a drop of 1% (w/v) aqueous uranyl acetate for 30 s. Excess stain was removed by flash washing. The blotted and air-dried grids were examined in a Philips 400T Transmission Electron Microscope operating at 80 kV. Studies on the interaction between the PEX5 proteins and GST-PEX14-(1–78) were performed by surface plasmon resonance spectroscopy using a BIAcore 2000 instrument (BIAcore AB). Here one binding partner, referred to as the ligand, is immobilized on a sensor chip, and the interaction with an interactant in free solution, the analyte, is detected. Changes in the mass concentration on the sensor surface are proportional to changes in the refractive index of the sensor surface. This refractive index change is monitored by using the physical phenomenon of surface plasmon resonance and expressed in response units (RU), whereas 1000 RU correspond to a change in surface concentration of about 1 ng of protein/mm2 (25Stenberg E. Persson B. Roos H. Urbaniczky C. J. Colloid Interface Sci. 1991; 143: 513-526Crossref Scopus (999) Google Scholar). Anti-GST antibodies (Code number BR-1002-23, BIAcore AB) were coupled to the surface to a total response of 3000 RU by standard amine coupling following the manufacturers' instructions. Purified GST-PEX14-(1–78) at a concentration of 0.09 mg/ml was captured to an immobilization level of about 400 RU to a CM5 sensor chips via anti-GST antibodies. All interaction experiments were carried out in a buffer containing 50 mm Tris-HCl, pH 8.0, and 150 mm NaCl at a constant flow rate of 30 μl/min. PEX5L, PEX5S, PEX5L-(1–251),and PEX5L-(214–639) were injected in varying concentrations from 500 to 1 nm. Unspecific binding was subtracted using an unmodified sensor surface for each concentration individually. Bulk refractive index changes due to buffer variations were subtracted by using runs without protein. After each interaction cycle the surfaces were regenerated by either two injections (1 min each) with 10 mm glycine-HCl, pH 2.2, or one injection with the glycine solution followed by one injection with 0.05% SDS. Association and dissociation phase were both monitored for 5 min. The resulting data were evaluated with the BIAevaluation software version 3.0 (BIAcore AB). The dissociation rate constants were determined separately from the binding curves obtained for the highest concentration of the respective PEX5 protein. The data from 5 to 60 s after injection stop were fitted to Equation 1. R(t)=R0×exp(−kd×(t−t0))Equation 1 Association rate constants were calculated with the previously obtained dissociation rate constants and respective concentrations of analytes according to the pseudo first order model A + B = AB using Equation 2 in global fit analysis. R(t)=kaCRmax/(kaC+kd)×(1−exp[−(ka+kd)×t])Equation 2 The equilibrium binding constants were calculated from the respective rate constants according to Equation 3. KD=kd/kaEquation 3 In these equations the following abbreviations are used:R 0 is the response at the initial timet 0, R(t) is the relative response at a given time t, R max is the maximum analyte binding capacity, C is the molar concentration of the analyte, k a andk d are the association and dissociation rate constants, respectively, and K D is the equilibrium binding constant. It has been reported previously that the PTS1 receptor Pex5p of yeasts and higher eukaryotes can bind to the two peroxins Pex13p (15Gould S.J. Kalish J.E. Morrell J.C. Bjorkman J. Urquhart A.J. Crane D.I. J. Cell Biol. 1996; 135: 85-95Crossref PubMed Scopus (209) Google Scholar, 16Erdmann R. Blobel G. J. Cell Biol. 1996; 135: 111-121Crossref PubMed Scopus (184) Google Scholar, 17Elgersma Y. Kwast L. Klein A. Voorn-Brouwer T. van den Berg M. Metzig B. America T. Tabak H.F. Distel B. J. Cell Biol. 1996; 135: 97-109Crossref PubMed Scopus (183) Google Scholar) and/or Pex14p (14Albertini M. Rehling P. Erdmann R. Girzalsky W. Kiel J.A.K.W. Veenhuis M. Kunau W.-H. Cell. 1997; 89: 83-92Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar, 18Brocard C. Lametschwandtner G. Koudelka R. Hartig A. EMBO J. 1997; 16: 5491-5500Crossref PubMed Scopus (107) Google Scholar, 19Fransen M. Terlecky S.R. Subramani S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 8087-8092Crossref PubMed Scopus (135) Google Scholar) at the cytoplasmic side of the peroxisomal membrane. In higher eukaryotes two forms of PEX5 have been reported which differ in an insertion of 37 amino acids as a result of alternative splicing (20Braverman N. Dodt G. Gould S.J. Valle D. Hum. Mol. Genet. 1998; 7: 1195-1205Crossref PubMed Scopus (148) Google Scholar,26Otera H. Okumoto K. Tateishi K. Ikoma Y. Matsuda E. Nishimura M. Tsukamoto T. Osumi T. Ohashi K. Higuchi O. Fujiki Y. Mol. Cell. Biol. 1998; 18: 388-399Crossref PubMed Google Scholar, 27Dodt G. Braverman N. Wong C. Moser A. Moser H.W. Watkins P. Valle D. Gould S.J. Nat. Genet. 1995; 9: 115-125Crossref PubMed Scopus (379) Google Scholar). To gain insight into the structural basis of the interactions of Pex5p with other peroxins, large quantities of homogeneous protein were required. For this purpose we created His6-tagged versions of both, the long and the short form of human PEX5 (Fig.1, PEX5L and PEX5S), and expressed them in E. coli by using a T7 promotor-based expression system. A TEV protease cleavage site was introduced between the His6tag and the PEX5 sequence, allowing the removal of the affinity tag. Expression and purity were analyzed by SDS-PAGE. With both recombinant PEX5 forms we obtained high levels of expression (Fig. 2 A, lanes 1and 2). During SDS-PAGE both proteins did not migrate with their calculated molecular masses of 67 kDa (PEX5S) and 71 kDa (PEX5L), respectively, but with an apparent molecular mass of 80 kDa, as had been described previously for PEX5S (27Dodt G. Braverman N. Wong C. Moser A. Moser H.W. Watkins P. Valle D. Gould S.J. Nat. Genet. 1995; 9: 115-125Crossref PubMed Scopus (379) Google Scholar, 28Wiemer E.A.C. Nuttley W.M. Bertolaet B.L. Li X. Francke U. Wheelock M.J. Anné U.K. Johnson K.R. Subramani S. J. Cell Biol. 1995; 130: 51-65Crossref PubMed Scopus (164) Google Scholar). In our studies truncated versions of PEX5 were also used. A similar difference between empirical (45 kDa) and calculated molecular mass (28 kDa) was observed for an N-terminal fragment but not for C-terminal fragments (Fig. 2), suggesting that the aberrant migration is due to the high density of negatively charged residues (about 20%) within the N-terminal half. Both proteins, PEX5L and PEX5S, were virtually completely soluble and could be purified to close homogeneity in one step using Ni-NTA affinity chromatography (Fig. 2 A, lanes 6 and7). The His6 tag was efficiently cleaved off by TEV protease. The protease, degradation products, and minor contaminants were removed by anion exchange chromatography. Typically 20 mg of pure protein could be obtained from one liter bacterial culture. Crystallization experiments have been initiated. Size exclusion chromatography indicated a molecular mass of about 270 kDa for both PEX5 forms (Fig. 3), suggesting a tetramer of identical subunits. The tetrameric assembly of PEX5 was supported by results from electron microscopy. Electron micrographs of negatively stained recombinant PEX5S preparations show numerous particles with a diameter of about 13 nm (Fig.4 A). These particles have the shape of a square with a central stain filled cavity and appear to consist of four subunits. It is important to note that the recombinant PEX5 also tends to form filamentous aggregates, which were detected in the same micrographs. It is noteworthy that their number and size is increased in low salt environment (Fig. 4 B), suggesting that their formation is governed by electrostatic interactions. Similar aggregates consisting of more than 100 molecules have been described for another TPR containing protein (9Hirano T. Kinoshita N. Morikawa K. Yanagida M. Cell. 1990; 60: 319-328Abstract Full Text PDF PubMed Scopus (237) Google Scholar).Figure 4Electron microscopy of recombinant human PEX5. Survey view of negatively stained PEX5S preparations.A, purified PEX5S (0.9 mg/ml) in 20 mm Tris, pH 8.0, 150 mm NaCl, and 1 mm DTT. Typical tetramers are marked by arrowheads. Two of them are shown asinsets at higher magnification. B, purified PEX5S after dialysis against the same buffer but without NaCl. Whereas the number of tetramers is decreased, the number and size of filamentous aggregates is increased. The bar represents 100 nm.View Large Image Figure ViewerDownload (PPT) TPR domains are known to be involved in protein-protein interactions. For example the binding of the PTS1 to its receptor Pex5p is mediated by the TPRs (27Dodt G. Braverman N. Wong C. Moser A. Moser H.W. Watkins P. Valle D. Gould S.J. Nat. Genet. 1995; 9: 115-125Crossref PubMed Scopus (379) Google Scholar, 29Brocard C. Kragler F. Simon M.M. Schuster T. Hartig A. Biochem. Biophys. Res. Commun. 1994; 204: 1016-1022Crossref PubMed Scopus (126) Google Scholar, 30Terlecky S.R. Nuttley W.M. McCollum D. Sock E. Subramani S. EMBO J. 1995; 14: 3627-3634Crossref PubMed Scopus (155) Google Scholar). TPRs are also believed to interact with each other forming inter- and/or intramolecular contacts (10Lamb J.R. Tugendreich S. Hieter P. Trends Biochem. Sci. 1995; 20: 257-259Abstract Full Text PDF PubMed Scopus (545) Google Scholar, 31Gale M.J. Tan S.-L. Wambach M. Katze M.G. Mol. Cell. Biol. 1996; 16: 4172-4181Crossref PubMed Scopus (92) Google Scholar). In order to map the determinants responsible for tetramerization and aggregation of PEX5, three truncated His6-tagged versions, PEX5L-(1–251), PEX5L-(214–639), and PEX5L-(323–639) (Fig. 1), were expressed inE. coli (Fig. 2 A, lanes 3–5), purified by Ni-NTA affinity chromatography (Fig. 2 A,lanes 8–10), and subjected to size exclusion chromatography. While the first two proteins were soluble the third one comprising little more than the seven TPR motifs was insoluble and could only be purified under denaturing conditions. Thus, approximately 100 amino acids, which distinguish PEX5L-(214–639) and PEX5L-(323–639), are sufficient to prevent aggregation and render the former protein soluble. Gel filtration analyses suggested an apparent molecular mass of 105 kDa for PEX5L-(1–251), whereas PEX5L-(214–639) behaved like a protein of 45 kDa (Fig. 3). These results are consistent with a tetrameric structure for PEX5L-(1–251), which contains only the N-terminal half of PEX5, and a monomeric structure for PEX5L-(214–639) comprising the C-terminal two-thirds of PEX5. Electron micrographs of negatively stained PEX5L-(1–251) showed particles very similar in shape but smaller in size to those found in preparations of full-length PEX5 (data not shown). However, in contrast to full-length PEX5 no filamentous aggregates were detected. A comparison of the properties determined for the four soluble PEX5 proteins, PEX5L, PEX5S, PEX5L-(1–251), and PEX5L-(214–639), revealed that the presence of the segment between amino acid 214 and 251 does not correspond to the ability to form tetramers. Thus, this structural property is dependent on a determinant located within the first 213 amino acids. Aggregation in contrast seems to be due to the seven TPR motifs. InSaccharomyces cerevisiae interaction of Pex5p and Pex14p was indicated previously by the two-hybrid method leading to the suggestion that Pex14p could be a candidate for the docking protein of the PTS1 receptor at the peroxisomal membrane (14Albertini M. Rehling P. Erdmann R. Girzalsky W. Kiel J.A.K.W. Veenhuis M. Kunau W.-H. Cell. 1997; 89: 83-92Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar, 18Brocard C. Lametschwandtner G. Koudelka R. Hartig A. EMBO J. 1997; 16: 5491-5500Crossref PubMed Scopus (107) Google Scholar). Deletion analysis mapped the binding site for ScPex5p to the first 58 amino acid residues of ScPex14p. 4K. Niederhoff and W.-H. Kunau, manuscript in preparation. Recently, the human orthologue of ScPex14p was identified and also was shown to interact with PEX5 (19Fransen M. Terlecky S.R. Subramani S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 8087-8092Crossref PubMed Scopus (135) Google Scholar). Sequence comparison of all available Pex14p orthologues suggests that the first 58 amino acid residues of ScPex14p correspond to the first 78 amino acid residues of human PEX14 (Fig.5). Thus, we used a fusion protein composed of GST, a thrombin cleavage site, and the first 78 amino acid residues of human PEX14 to demonstrate the direct binding of PEX5 and PEX14 in vitro. The fusion protein was expressed in E. coli and could be purified in high amounts (Fig. 2 B). Lysates of cells expressing either GST-PEX14-(1–78) or His6-PEX5L in similar amounts were mixed and divided into two parts. One aliquot was run through a glutathione-Sepharose column, while the other was subjected to Ni-NTA chromatography. As seen in Fig.6 A, in both cases the identical complex consisting of His6-PEX5L and GST-PEX14-(1–78) could be specifically eluted by glutathione and imidazole, respectively. The binding of PEX5 to the N-terminal fragment of PEX14 was specific, since expression of GST alone instead of GST-PEX14-(1–78) did not recruit His6-PEX5L to the glutathione column (data not shown). We conclude that PEX5 directly binds PEX14. This conclusion was supported by another in vitro binding experiment with purified GST-PEX14-(1–78) and His6-PEX5L. The two proteins were mixed in a molar ratio of 1:6 (PEX5L: GST-PEX14-(1–78)) and subsequently analyzed by sizing chromatography. Fig. 6 B shows that the mixture of the two proteins contained an early eluting species, which is not detectable when each protein was analyzed individually. This high molecular weight complex consisted of GST-PEX14-(1–78) and His6-PEX5 as confirmed by SDS-PAGE analysis (Fig. 6 A, lane 4).Figure 6Isolation of in vitro formed complexes between His6-PEX5L and GST-PEX14-(1–78). A, SDS-PAGE analysis of complexes isolated by glutathione (lane 2) and Ni-NTA (lane 3) affinity chromatography and by size exclusion chromatography (lane 4). Proteins were separated in 12% polyacrylamide gels and visualized with Coomassie Blue. The formation of complexes was carried out in 20 mm Tris, pH 8.0, 150 mm NaCl, and 1 mm DTT. For the affinity purification of the complexes by glutathione-Sepharose and by Ni-NTA-agarose, total cell lysates of E. coli expressing His6-PEX5L or GST-PEX14-(1–78) were mixed (lane 1) and incubated for 1 h at 4 °C with gentle shaking. Size exclusion chromatography on a Superose 6 column was performed with a mixture of both affinity purified proteins incubated for 5 min at room temperature. B shows the gel filtration chromatograms obtained for purified His6-PEX5L and GST-PEX14-(1–78) and for the mixture of both proteins.View Large Image Figure ViewerDownload (PPT) The formation of stable complexes indicated a high affinity binding between PEX5 and PEX14-(1–78). The kinetics of interaction between PEX5 and the peroxisomal docking protein PEX14 were analyzed using surface plasmon resonance spectroscopy. Both recombinant PEX5 isoforms, PEX5L and PEX5S, bind to the immobilized GST-PEX14-(1–78) with a fast association rate, whereas the dissociation rate is very slow (Fig.7). The calculated equilibrium binding constants are in the low nanomolar range (TableI). Data were also analyzed by steady state analysis (not shown) and found to be in good agreement with those obtained by evaluating the apparent rate constants. Vice versa experiments with His6-tagged PEX5 immobilized on a Ni-NTA sensor chip surface as ligand and GST-PEX14-(1–78) as analyte confirmed the high affinity binding (data not shown).Table IRate and equilibrium binding constants of the interaction of immobilized GST-PEX14-(1–78) and PEX5 proteinsAnalytek ak dK DM−1s−1s−1mPEX5L1.2 × 1060.8 × 10−30.7 × 10−9PEX5S0.4 × 1061.3 × 10−33.3 × 10−9PEX5L-(1–251)1.2 × 1060.8 × 10−30.7 × 10−9PEX5L-(214–639)0.2 × 1061.5 × 10−37.5 × 10−9Association rate constants, k a, dissociation rate constants, k d, and equilibrium binding constants,K D, were determined from surface plasmon resonance spectroscopy using GST-PEX14-(1–78) immobilized to an anti-GST surface as the ligand and the various PEX5 forms as analytes. Open table in a new tab Association rate constants, k a, dissociation rate constants, k d, and equilibrium binding constants,K D, were determined from surface plasmon resonance spectroscopy using GST-PEX14-(1–78) immobilized to an anti-GST surface as the ligand and the various PEX5 forms as analytes. In order to determine whether the GST part of GST-PEX14-(1–78) affects the binding with PEX5, the fusion protein was cleaved with thrombin to obtain a 10-kDa fragment PEX14-(1–78) (Fig. 2 B). This allowed in vitro binding experiments with four different forms of PEX5: the long form PEX5L, the short form PEX5S, the N-terminal fragment PEX5L-(1–251), and the C-terminal fragment PEX5L-(214–639). For this purpose the components were mixed and analyzed by sizing chromatography. Fig. 8summarizes the findings. In all four binding experiments PEX14-(1–78) was present in excess. Comparison of the elution profiles obtained with the individual PEX5 proteins in the presence or absence of PEX14-(1–78) revealed that the PEX5 proteins eluted earlier, when PEX14-(1–78) was present, indicating the formation of complexes. SDS-PAGE analyses of the corresponding peak fractions confirmed the presence of both PEX14-(1–78) and the respective PEX5 protein, whereas the unbound PEX14-(1–78) was found as a late peak in all elution profiles. This is shown for PEX5L in Fig. 8. One surprising result was that all soluble forms of PEX5 were able to interact with PEX14-(1–78), suggesting more than one binding site. This is consistent with the findings from surface plasmon resonance analysis that both truncated versions PEX5L-(1–251) and PEX5L-(214–639) bind to GST-PEX14-(1–78) with similar affinity as the long and the short form of PEX5 (Fig. 7 and Table I). Using the apparent molecular masses of PEX14-(1–78), of the different PEX5 proteins, and of their complexes, we estimated the subunit composition of the complexes (Table II). This led to interesting and unexpected results. The tetrameric long and short forms of PEX5 seem to bind seven and six PEX14-(1–78) fragments per subunit, respectively, suggesting that the 37-amino acid insertion (amino acids 215–251) of the long form contains one binding site. The calculated number of binding sites was confirmed by analysis of the complexes formed by the two truncated PEX5 proteins. The tetrameric N-terminal fragment PEX5L-(1–251) seems to bind four fragments of PEX14-(1–78) per subunit, whereas the C-terminal fragment PEX5L-(214–639) binds three or four fragments. As the overlapping region of these two forms of PEX5 contained the insertion (amino acids 215–251) with one putative binding site, their presumed binding sites add up to six or seven.Table IISummary of the results of molecular mass analysis of PEX5 proteins in absence and presence of isolated PEX14-(1–78) fragmentProteinm calcm app− PEX14-(1–78)+ PEX14-(1–78)kDakDaPEX5L71270550PEX5S67270500PEX5L-(1–251)28105270PEX5L-(214–639)474580The apparent molecular masses of PEX5 proteins and their complexes with PEX14-(1–78) were estimated by size exclusion chromatography as shown in Fig. 8. m calc, molecular mass, calculated from the amino acid sequence; m app, apparent molecular mass. Open table in a new tab The apparent molecular masses of PEX5 proteins and their complexes with PEX14-(1–78) were estimated by size exclusion chromatography as shown in Fig. 8. m calc, molecular mass, calculated from the amino acid sequence; m app, apparent molecular mass. In order to initiate a structural analysis of the import machinery for peroxisomal matrix proteins, we prepared recombinant human PEX5 (PTS1 receptor) and examined its binding to human PEX14 (binding partner at the peroxisomal membrane) in vitro. In these studies we used an N-terminal fragment of PEX14 comprising only the first 78 amino acid residues rather than the full-length protein as the binding site for the PTS1 receptor of S. cerevisiae had been mapped to the corresponding fragment of ScPex14p. 5K. Niederhoff and W.-H. Kunau, manuscript in preparation. Human PEX5, as all other members of the steadily growing Pex5 protein family, is composed of two distinctly different parts, the highly conserved C-terminal half comprising seven TPR motifs and the N-terminal half, which possesses only a few amino acids that are strictly conserved among all members (Fig. 9). While the TPR region was shown to mediate the binding to PTS1-containing proteins (27Dodt G. Braverman N. Wong C. Moser A. Moser H.W. Watkins P. Valle D. Gould S.J. Nat. Genet. 1995; 9: 115-125Crossref PubMed Scopus (379) Google Scholar, 29Brocard C. Kragler F. Simon M.M. Schuster T. Hartig A. Biochem. Biophys. Res. Commun. 1994; 204: 1016-1022Crossref PubMed Scopus (126) Google Scholar, 30Terlecky S.R. Nuttley W.M. McCollum D. Sock E. Subramani S. EMBO J. 1995; 14: 3627-3634Crossref PubMed Scopus (155) Google Scholar), a specific function was not yet assigned to the N-terminal half. In this study we present evidence that the N-terminal half of PEX5 binds PEX14. Recombinant human PEX5 is a soluble protein (Fig. 2) with a tetrameric structure, as judged by sizing chromatography (Fig. 3) and by electron microscopy (Fig. 4). However, it also forms aggregates. As the C-terminal fragment (amino acids 214–639) alone, in contrast to the N-terminal fragment (amino acids 1–251), behaves as a monomer the ability to form tetramers appears to be a property of the first 213 amino acids. Conversely, the present data suggest that aggregation could be caused by the TPR motifs. A C-terminal fragment (amino acids 323–639), which consists virtually of nothing else but the TPR motifs is completely insoluble, while the N-terminal fragment (amino acids 1–251) does not aggregate. A striking property of the recombinant PEX5 is the occurrence of multiple (at least six or seven per subunit) binding sites for the PEX14 fragment which seem to bind with very high affinity. Interestingly, sequence analysis of PEX5L reveals seven pentapeptide repeats within the segment of amino acids 115–315. These repeats are characterized by the consensus WXXXF/Y. Six of them have been reported to exist in the short form of human PEX5 (27Dodt G. Braverman N. Wong C. Moser A. Moser H.W. Watkins P. Valle D. Gould S.J. Nat. Genet. 1995; 9: 115-125Crossref PubMed Scopus (379) Google Scholar). An alignment of the known Pex5 proteins shows that these repeats are found in each but the number and spacings within the N-terminal half of the molecules differ (Fig. 9). For example, Pex5p of water melon (Citrullus lanatus) possesses nine of these motifs whereas Pex5p of S. cerevisiae contains only two. Sequence analysis of these repeats in all of the orthologues revealed that the first position between the conserved aromatic amino acid residues can be any amino acid, whereas there is a strong preference for hydrophilic residues at the second position and the third position is occupied in 35 out of 36 motifs by aspartic acid, glutamic acid, or glutamine. Secondary structure prediction (32Rost B. Methods Enzymol. 1996; 266: 525-539Crossref PubMed Google Scholar) for all known Pex5 proteins revealed that the WXXXF/Y motifs are parts of amphipathic α-helices, suggesting that the strictly conserved aromatic residues of tryptophan, phenylalanine, or tyrosine are positioned to the same side of the helix. As the number and distribution of the repeats in human PEX5 closely matches the estimated number of binding sites for the PEX14 fragment in the various PEX5 forms (Table II), it is tempting to speculate that the repeats might provide the structural basis for this interaction. It seems unlikely that all the binding sites found for the PEX14-(1–78) fragment in vitro in the absence of any other possible interacting protein (peroxins and/or cargoproteins) are occupied in vivo by full-length PEX14 simultaneously. It is very difficult to envisage how a protein of this size, and especially with all its predicted partners, could sterically fit into seven binding sites within a segment of about 200 amino acid residues. This then raises the question which function the observed multiple binding sites of PEX5 for PEX14 could have in vivo. One explanation could be that the number of binding sites increases the probability of binding between the two partners without the necessity that they all have to be occupied. In the case that the individual binding sites turn out to have different binding constants, it is conceivable that PEX14 first interacts with low affinity binding sites and is then transferred to the high affinity one(s). To address this possibility additional fragments and mutant forms of PEX5, as well as full-length PEX14, have to be tested. Taken together, the fact that recombinant PEX5 can be obtained in high amounts and successfully used for in vitro binding studies, including surface plasmon resonance analysis, now opens up new ways to study the structural basis of interactions between essential components of peroxisomal protein import. Studies including ternary complexes are in progress. We thank Uschi Dorpmund for her excellent technical assistance, Gunther Stier for the supply of plasmids, Garnet Will for providing PEX14 cDNA, and Dr. John Williams for critical reading the manuscript. We are most grateful to all members of the Structural Biology Program, EMBL (Heidelberg) for encouragement and helpful discussions throughout the course of this work.
Referência(s)