Biochemical and Biophysical Research Communications 291, 1180–1186 (2002) doi:10.1006/bbrc.2002.6568, available online at http://www.idealibrary.com on Two Splice Variants of Human PEX19 Exhibit Distinct Functions in Peroxisomal Assembly Peter U. Mayerhofer, Tanja Kattenfeld, Adelbert A. Roscher, and Ania C. Muntau1 Dr. v. Hauner Children’s Hospital, Department of Clinical Chemistry and Biochemical Genetics, Ludwig-Maximilians-University Munich, Lindwurmstrasse 4, D-80337 Munich, Germany Received January 31, 2002 cific biological functions of the different predicted do- PEX19 has been shown to play a central role in the mains of the PEX19 protein. © 2002 Elsevier Science (USA) early steps of peroxisomal membrane synthesis. Com- Key Words: peroxisome biogenesis disorders; splice putational database analysis of the PEX19 sequence variants; farnesylation; peroxin; PEX; ABC half trans- revealed three different conserved domains: D1 (aa porter; peroxisomal membrane proteins. 1–87), D2 (aa 88–272), and D3 (aa 273–299). However, these domains have not yet been linked to specific biological functions. We elected to functionally char- The biogenesis of peroxisomes (reviewed by 1–3) is a acterize the proteins derived from two naturally oc- curring PEX19 splice variants: PEX19⌬E2 lacking the complex process including the biosynthesis of the per- N-terminal domain D1 and PEX19⌬E8 lacking the do- oxisomal membrane, the targeting and insertion of per- main D3. Both interact with peroxisomal ABC trans- oxisomal membrane proteins (PMPs) into this mem- porters (ALDP, ALDRP, PMP70) and with full-length brane, and the import of peroxisomal matrix enzymes PEX3 as shown by in vitro protein interaction studies. across the bilayer into the peroxisomal lumen. The PEX19⌬E8 also interacts with a PEX3 protein lack- proteins involved are termed “peroxins” and are en- ing the peroxisomal targeting region located at the coded by at least 23 different PEX genes. In most cases, N-terminus (⌬66aaPEX3), whereas PEX19⌬E2 does human PEX mutants display defects in the import of not. Functional complementation studies in PEX19- peroxisomal matrix proteins whereas the synthesis of deficient human fibroblasts revealed that transfection some peroxisomal membrane vesicles and the insertion of PEX19⌬E8-cDNA leads to restoration of both perox- of PMPs into these membrane remnants (“peroxisomal isomal membranes and of functional peroxisomes, ghosts”) remain unaffected. Only mutants in PEX3, whereas transfection of PEX19⌬E2-cDNA does not re- PEX16, and PEX19 appear to lack even such peroxiso- store peroxisomal biogenesis. Human PEX19 is partly mal ghosts (4–6), implicating a crucial role of these farnesylated in vitro and in vivo. The farnesylation proteins in the very early stage of peroxisomal mem- consensus motif CLIM is located in the PEX19 domain brane synthesis. D3. The finding that the protein derived from the Among these PEX19 is the only peroxin known to splice variant lacking D3 is able to interact with sev- own a C-terminal CAAX-box, thereby harboring the eral peroxisomal membrane proteins and to restore intrinsic property of being modified by posttransla- peroxisomal biogenesis challenges the previous as- tional farnesylation. In human, PEX19 is in part far- sumption that farnesylation of PEX19 is essential for nesylated in vivo and it has been suggested that far- its biological functionality. The data presented dem- nesylation is required for the peroxisomal localization onstrate a considerable functional diversity of the pro- of PEX19 (6). Whether farnesylation is essential for the teins encoded by two PEX19 splice variants and function of PEX19 in peroxisomal biogenesis (6) or has thereby provide first experimental evidence for spe- only an ancillary effect (7) remains unclear to date. PEX19 appears to be bimodally distributed between Abbreviations used: ABC, half transporter ATP binding cassette transporter; ALDP, adrenoleukodystrophy protein; ALDRP, adreno- the cytoplasm and the peroxisomes, with most of the leukodystrophy related protein; GFP, green fluorescent protein; protein in the cytoplasm (7). PEX19 protein has been GST, glutathione-S-transferase; PEX, peroxisomal assembly protein; shown to bind a wide variety of peroxisomal membrane PMP70, 70-kDa peroxisomal membrane protein; PMP(s), peroxiso- proteins including peroxins and peroxisomal ABC mal membrane protein(s). 1 To whom correspondence and reprint requests should be ad- transporters (7–13). Among these, PEX19 has been dressed. Fax: ϩ49 89 5160 3343. E-mail: [email protected]. shown to interact with PEX3, another key component uni-muenchen.de. in early peroxisomal membrane synthesis, in yeast (8) 0006-291X/02 $35.00 1180 © 2002 Elsevier Science (USA) All rights reserved. Vol. 291, No. 5, 2002 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS and human (4, 5, 14). In yeast, this particular protein- System (Promega) following standard protocols. For the genera- protein interplay appears to be a prerequisite for the tion of glutathione-S-transferase (GST) fusion proteins, different cDNAs of human PEX19-splice variants were subcloned into the generation of preperoxisomal structures that can ac- prokaryotic expression plasmid pGEX-3X (Pharmacia) as described quire and stabilize newly synthesized PMPs (15). elsewhere (16). Expression and affinity purification using gluta- PEX19 harbors distinct binding sites for PEX3 and thione-Sepharose 4B (Pharmacia) of the fusion proteins were per- PEX16 (13). However, these sites have not yet been formed as previously described (22). The various purified GST-fusion linked to a specific protein domain organization of proteins were bound to glutathione-Sepharose beads (Pharmacia) in PBSfor1hat4°C and washed three times in PBS. For the in vitro PEX19. interaction assay the GST-fusion proteins (0.5–1.5 ␮g protein bound We have previously cloned and characterized the to 1 ␮l bead suspension) were incubated with 5 ␮lofin vitro trans- human PEX19 gene, which is composed of eight exons lated, [35S]-labeled protein in 200 ␮l binding buffer (100 mM NaCl, 50 (16). On the transcriptional level, we identified differ- mM potassium phosphate pH 7.4, 1 mM MgCl2, 10% glycerol, 0.1% ent PEX19 splice variants (16). The unspliced variant Tween 20, 1.5% BSA) for2hat4°C. The beads were then pelleted and washed four times in 1 ml binding buffer not containing BSA. PEX19all encodes for the full-length PEX19 protein, Pellets were resuspended in SDS-sample buffer and boiled for 5 min whereas the variants PEX19⌬E2 and PEX19p⌬E8 are before being analyzed by SDS–PAGE according to standard proto- expected to lack N- or C-terminal parts of the protein. cols. To verify stability of the GST-fusion proteins and equal loading, Tissue- or cell-specific expression of splice variants gels were stained with Coomassie blue. The gels were dried and has been reported for many genes, e.g., for genes en- exposed to X-ray film for 24 h. coding cytokines or tyrosine kinases. Functional diver- Transfection, antibodies, and indirect immunofluorescence analy- sities of these variants, like distinct protein-protein sis. Plasmids were transfected into COS7 cells or into human PEX19-deficient fibroblasts using Lipofectamine (Life Technologies) interactions, are known (17–19). For PEX19, the splice following protocols provided by the manufacturer. To create stable variant PEX19all has been described to be the predom- transfectants, the fibroblasts were incubated with 500 ␮g/ml neomy- inant transcript in mRNA of several cells and tissues cin (SIGMA). For indirect immunofluorescence analysis, stable in human, although the variant PEX19⌬E2 has also transfectants were fixed with 3% formalin/PBS for 30 min and per- meabilized with 1% Triton X-100 for 5 min. Thereafter, cells were been reported to account for a significant level of total incubated for 1 h with a polyclonal rabbit antibody against catalase human PEX19 mRNA. Interestingly, the ratio between (Biodesign; 1:100 dilution) and/or a monoclonal mouse antibody PEX19all and PEX19⌬E2 expression was completely against ALDP (ALD-1D6, Euromedex; 1:100 dilution). The cells were reversed in uterine mRNA prepared from total human then washed 10 times with PBS and incubated for 1 h with secondary uterine tissue (16), but a possible biological impact of antibodies: Rhodamine (TRITC)-conjugated goat anti-mouse IgG and/or fluorescein (FITC)-conjugated goat anti-rabbit IgG (all Jack- these findings remained still unclear. Therefore, the son Immunoresearch Laboratory Inc.; 1:50 dilution). After 10 wash characterization of PEX19 splice variants as to their steps in PBS, the cells were embedded in an antifading reagent ability to interact with other peroxisomal proteins and (Vectashield, Vector Laboratories Inc.). Fluorescence microscopy was to initiate early peroxisomal membrane synthesis performed using an Axiovert 135 inverted fluorescence microscope could provide insights into the potential biological rel- (Zeiss). evance of these PEX19 splice variants. Furthermore, the definition of functional regions of PEX19 by the RESULTS characterization of truncated PEX19 splice variants Domain Structure of the Human PEX19 Protein could contribute to the knowledge on specific protein in Relation to PEX19 Splice Variants domains within PEX19. Computational database analysis of the full-length PEX19 protein sequence using the NCBI-BLASTP pro- MATERIALS AND METHODS gram and the ProDom database revealed three differ- Computational analysis. Database searches for the prediction of ent conserved domains in