Polypeptides differentially expressed in ... - Wiley Online Library

Eur. J. Biochem. 267, 487±497 (2000) q FEBS 2000

Polypeptides differentially expressed in imaginal discs define the peroxiredoxin family of genes in Drosophila Javier Rodriguez1, Marta Agudo1, Jozef Van Damme2, Joel Vandekerckhove2 and Juan F. SantareÂn1 1

Centro de BiologõÂa Molecular `Severo Ochoa', Universidad AutoÂnoma de Madrid, Spain; 2Flanders Interuniversity Institute of Biotechnology, Department of Biochemistry, Faculty of Medicine, Universiteit Gent, Belgium

2D gel electrophoresis followed by microsequencing has been used to purify and identify a protein (catalogued in the database as SSP5111) from Drosophila wing imaginal discs of third instar larvae that showed significant differences in their level of expression when compared with other imaginal discs of the same age. The microsequence data showed identity with amino acids encoded by the human proliferation association gene, pag, which is a thiol-specific antioxidant. By virtue of this homology we have cloned and sequenced two cDNAs that appear to define the peroxiredoxin family of Drosophila. One of them, Jafrac1, encodes the SSP5111 protein searched, had 194 amino acids and mapped in the region 11E in the X chromosome. The other, Jafrac2, encodes a protein of 242 amino acids and mapped in the region 62F in the 3 L chromosome. Both new peroxidases contain two conserved cysteines and share homology with other peroxidases that extends over the entire sequence and ranges between 47% and 76%. An antiserum raised against the SSP5111 protein showed significant changes in the amount of protein in different stages of Drosophila development, being a major product in early embryos. In 2D gels the antibody not only recognizes the SSP5111 polypeptide but also a related one (catalogued in the database as SSP6107) that exhibits identical amino-acid sequence over at least 85% of its sequence. The data also suggest that the SSP5111 polypeptide could be a maternal-effect product. Keywords: Drosophila; peroxidase; thioredoxin; 2D PAGE.

We are using high resolution 2D gel electrophoresis [1,2] to analyse some aspects of the development, morphogenesis, and genetic variation in the wing imaginal disc of the mature larvae of Drosophila melanogaster as a model system. Improvements in the separation made it possible to resolve 1718 labelled polypeptides from as little as one imaginal disc. From this we have constructed a database [3] that acquired a quantitative character when compounded with computer analysis [4]. One application of the database is the monitoring of systemic changes in the pattern of protein synthesis caused by mutations or involved in different aspects of development, differentiation and morphogenesis in Drosophila. In fact, using this database as a reference we have detected changes in gene activity that are dependent on the tissue type [5] or are associated with the different stages of development [6] or with mutations [7]. In a previous study we used the database for a comparison of the standard wing pattern with those of the haltere, leg 1, leg 2, leg 3 and eye-antenna imaginal discs of the same developmental stage. Then we defined a set of 17 polypeptides whose levels of expression vary quantitatively between the different imaginal discs [8]. We have tried in this study to purify and to identify one of those proteins, specifically an acidic protein that was named IEF39g in the first report [3] and later on catalogued Correspondence to J. F. SantareÂn, Centro de BiologõÂa Molecular `Severo Ochoa', Facultad de Ciencias. Universidad AutoÂnoma, Cantoblanco, 28049-Madrid. Spain. Fax: + 34 1 3974799, Tel.: + 34 1 3978447, E-mail: [email protected] Abbreviations: pag, proliferation associated gene; Tpx, thioredoxin peroxidase; Trx, thioredoxin; Prx, peroxiredoxin; MALDI-MS, matrix-assisted laser desorption/ionization MS; ORF, open reading frame; TOF, time of flight. (Received 27 July 1999, revised 20 October 1999, accepted 16 November 1999)

as SSP5111 when the computerized analysis was introduced [4]. By using preparative 2D electrophoresis and microsequencing we have obtained two amino-acid sequences with a close homology with the human proliferating associated protein, a member of the Tpx family [9]. The use of the pag probe allowed us to clone and obtain the cDNA sequences of two Drosophila genes, one of them encoding the SSP5111 polypeptide searched and another one that encodes a slightly larger polypeptide; they share 57% overall identity with each other and have significant streches of similarity with other peroxidases from different organisms, defining in this way the peroxiredoxin family in Drosophila. Initially identified in yeast, thioredoxin peroxidase (Tpx) is the immediate enzyme that reduces H2O2 with the use of electrons provided by thioredoxin (Trx) in a redox chain in which Trx reductase (Tr) is also involved [10]. More than 40 proteins from a wide variety of species, ranging from prokaryotes to mammals, show sequence similarity to yeast Tpx [11]. These homologous proteins were named the peroxiredoxin (Prx) family. They were not termed the Tpx family because not all members use Trx as the hydrogen donor [12]. Although a primary function of Tpx appears to be to prevent or limit cellular damage caused by reactive oxygen species, the physiological function of Tpx in cells is not yet known. Several lines of evidence suggest that the regulation of intracellular redox, a process highly conserved in organisms ranging from bacteria to humans, is a versatile control mechanism in signal transduction and gene expression [13]. In mammalian cells, intracellular redox status has been linked with cellular differentiation, inmune response, growth control, tumour promotion, and apoptosis, as well as activation of viruses, notably HIV, from latency [14,15]. The Prx family includes 12 mammalian proteins that were identified without reference to peroxidase

488 J. Rodriguez et al. (Eur. J. Biochem. 267)

activity but rather by association with a variety of diverse cellular functions including proliferation, differentiation, and immune response [16]. In this paper we report the isolation and characterization of two full-length cDNAs encoding Drosophila Tpxs with their chromosomal localizations. The hypothetical role of this new family of genes in Drosophila is discussed.

M AT E R I A L S A N D M E T H O D S Strains Laboratory stocks of wild-type D. melanogaster [strain Vallecas (Spain)] and D. virilis were used. Cell culture We have used an established cell line from wing imaginal discs of mature larvae of D. melanogaster (strain Oregon R) designated CME W2. Cells (3 105) were plated out in multiwell plates (2 cm2) according to the original protocol [17]. Preparation of samples and 2D gel electrophoresis Dissection and labelling of wing imaginal discs was carried out as described previously [3]. Analytical 2D gel electrophoresis was performed as described by O'Farrell [1] with some modification [3]. Gels were processed for fluorography [18], dried, and exposed at 270 8C for various periods of time. Approximately 106 trichloroacetic acid (TCA)-precipitable c.p.m. [35S]methionine + [35S]cysteine was routinely applied to each gel. For preparative gels used for microsequencing the cell contents of 20 confluent 65-cm2 plates of CME W2 cells were used to prepare proteins. The cell monolayer was washed with Hanks' solution and scraped off, with a rubber policeman, in to 4 mL of the same solution. After centrifugation, cells were sonicated, treated with DNase and RNase [19] and passed several times through a narrow needle. The sample was lyophilized and resuspended in 1 mL lysis buffer. Then 107 c.p.m. [35S]methionine + [35S]cysteine-labelled proteins from the wing imaginal discs were added and identified by autoradiography. Gels were immediately dried and exposed for 5 days. Polypeptides of interest were located by autoradiography and excised from the gels. Protein recovery from 2D gels and microsequencing Protein spots from several (up to 20) dried 2D gels were cut using a scalpel and collected. Details of gel rehydratation, protein electrotransfer, poly(vinylidene difluoride) membrane in situ protease cleavage, peptide separation by reversed-phase HPLC and amino-acid sequencing has been published [20,21]. cDNA cloning and sequence analysis JAFRAC1 and JAFRAC2 cDNAs were isolated from a D. melanogaster embryo lambda gt11 cDNA library (Stratagene) by DNA hybridization using a human thiol-specific antioxidant protein, pag [9], cDNA as probe and following standard procedures. Inserts from positive clones were excised and cloned into pUC119. Plasmid DNA was prepared by alkaline lysis, and DNA sequencing reactions were carried out with the ABI PRISM dye terminator cycle sequencing ready reaction kit with Ampli Taq DNA polymerase, and subjected to sequencing analysis using a model 377 ABI PRISM DNA

q FEBS 2000

sequencer. Positive clones were sequenced fully in both strands. Nucleotide and peptide sequences were analysed with the Wisconsin software package (version 9.0, Genetics Computer Group, Inc.). Multiple alignments of protein sequences were generated with a progressive pair-wise algorithm [22]. MALDI-MS and Q-time of flight (TOF)-MS analysis SSP5111 and SSP6107 polypetides were electroblotted onto poly(vinylidene difluoride) membranes and digested with trypsin for 20 h as described above, except that in this case resulting peptides were partially labelled at their C-terminal carboxyl moiety by the incorporation of 18O during the enzymatic hydrolysis in digestion buffer containing 50% H18 2 O. A small aliquot was processed by the bead-peptide concentration procedure [23] and subjected to matrix-assisted laser desorption/ionization (MALDI)-MS analysis. The spectra were recorded with a Bruker Reflex II Instrument (BrukerFranzen Analytik GmbH, Bremen, Germany). Ionization was achieved using a conventional nitrogen laser (337 nm beam, 3 ns pulse width, 5 Hz) set at an attenuation between 20 and 30. The instrument was calibrated using the known masses of synthetic peptides. The remained product of the enzymatic digestions were fractionated by reverse-phase HPLC as described above and analysed in a Q-TOF-MS instrument from Micromass. Approximately 5% of the HPLC outlet flow was directed into the Q-TOF-MS. Scans were acquired between 250 and 1500 m/z in 2.5 s with a resolution of 5000. The standard z-spray electrospray interface was use with standard conditions: capillary voltage 3000 V; cone voltage 45 V; source block and desolvation temperature 80 8C and 120 8C, respectively. Desolvation and nebulizer flow were set at 250 L´h21 and 45 L´h21, respectively. [Glu1]fibrinopeptide B (Sigma F-3261) was use to calibrate the instrument according to the manufacturer's instruction. Antibody production The cell content of 20 confluent 500-cm2 trays (NUNC) of CME W2 cells was used to prepare SSP5111 and SSP6107 polypeptides for immunization. The cell monolayers were washed with Hanks' solution and scraped off with a rubber policeman in to 10 mL of Hanks'. After centrifugation cells were sonicated, treated with DNase and RNase and passed several times through a narrow gauge needle. Then the sample was lyophilized and resuspended in 5 mL of lysis buffer. [35S]Methionine + [35S]cysteine labelled proteins from wing imaginal discs were added to identify the polypeptides by autoradiography. Two-dimensional gels (100 gels) were prepared as described above. Gels were immediately dried (without fixation) and exposed for 4 days. Protein SSP5111 or SSP6107 was located using the autoradiographs and cut out from the gels. The gel pieces were rehydrated and the protein was recovered by electroelution. The antigens (approximately 100 mg) were injected into rabbits and the rabbit sera were tested by immunobloting against SSP5111 or SSP6107 proteins as described below. Immunoblotting Proteins resolved by 2D gels were transferred onto nitrocellulose filters at 130 mA for 6 h. The filters were incubated in NaCl/Pi containing 10% bovine serum and anti-SSP5111 or anti-SSP6107 serum diluted to 1 : 100 for 120 min at room temperature. The filters were then incubated with peroxidaseconjugated anti-(rabbit IgG) IgG for 120 min in the same

q FEBS 2000

Thioredoxin peroxidase of Drosophila (Eur. J. Biochem. 267) 489

buffer. The immunocomplexes were visualized using diaminobenzidine as a substrate. Fluorescence in situ hybridization to polytene chromosomes Polytene chromosomes were prepared as described previously [24]. Clone Jafrac1 was labelled with biotin-16-UTP, and clone Jafrac2 was labelled with digoxigenin-11-UTP using, respectively, Biotin-Nick Translation Mix and DIG-Nick Translation Mix (Boheringer Manheim). Hybridization and detection were performed by following protocols described previously [25]. Biotinylated probe was detected with FluoroLink CY3-Avidin (Amersham) and digoxigenin-labelled probe was detected by incubation with antidigoxigenin±fluoresceine isothiocyanate (Boheringer Manheim). Chromosomes were counter-stained with 4 0 ,6-diamino-2-phenylindole. Slides were analysed using a Zeiss Axioplan epifluorescence microscope equipped with a cooled charge-coupled device camera (Photometrics). The fluorescent signals from the CY3, fluoresceine isothiocyanate and 4 0 ,6-diamino-2-phenylindole staining were recorded separately as grey scale digital images and then pseudocolored and merged using the Adobe Photoshop program.

R E S U LT S The general pattern of protein synthesis Figure 1A shows a representative fluorogram of a 2D gel of acidic polypeptides from wing imaginal discs of mature larvae of D. melanogaster (wild-type, strain Vallecas). The gels were analysed with a computer using the pdquest system and each polypeptide was assigned a number in the database. In this way we obtained a database for which a total of 1718 [35S]methionine + [35S]cysteine-labelled polypeptides (1352 acidic, IEF and 366 basic, NEPHGE) were separated, numbered, quantified, and catalogued. In a previous paper [8] we compared the pattern of protein synthesis among six different imaginal discs. There, we observed some quantitatively significant changes in 17 polypeptides: one of them was catalogued as IEF 39 g in our `manual' database [3] corresponding to SSP5111 in the catalogue obtained with the computer [4]. Figure 1B shows detail of the IEF gel that contains this polypeptide and the quantification of the level of expression of such a spot in the six imaginal discs analysed. Taking the wing imaginal disc as a reference, the polypeptide increased its level of expression in haltere, leg 1, leg 2 and leg 3 whereas the eye-antenna disc show a level similar to that in the wing. In this paper we have focused on the purification, identification and characterization of this SSP5111 polypeptide. Purification of polypeptide SSP5111 for microsequencing As the amount of protein derived from wing imaginal discs is limited, we used material derived from a permanent cell line, CME W2, obtained from the wing imaginal disc of Drosophila; we have shown previously that it is possible to use proteins from the CME W2 cell line, purified from a reasonable number of preparative 2D gels, to isolate and identify proteins in wing imaginal discs [21]. We ran 20 preparative 2D gels by loading a precise amount of polypeptide from the CME W2 cells mixed with a small amount of [35 S]methionine + [35 S]cysteine-labelled polypeptide from wing imaginal discs. The spots corresponding to SSP5111 were recovered from dried gels, concentrated in a special gel, transferred to nitrocellulose and digested with trypsin for 4 h. The tryptic peptides were separated by HPLC and two well-resolved peaks were used for microsequencing

Fig. 1. 2D gel electrophoresis map of total [35S]methionine + [35S]cysteine polypeptides from wing imaginal disc of late third instar larvae of D. melanogaster. (A) The pH ranges from 7.0 (left) to 4.5 (right). One area of interest is boxed. (B) Close-up of the boxed region in (A); polypeptide SSP5111 is indicated by an arrow. Inset, each bar represents the quantification of the SSP5111 polypeptide in each gel of the matchset; from left to right the bars represent wing, haltere, leg 1, leg 2, leg 3 and eyeantenna imaginal discs. The number at the upper right of the histogram is the quantification of the maximum bar in the graph. The other bars are drawn in proportion to the highest bar.

enabling us to obtain two sequences of 11 and 20 amino acids, respectively. The results of the database searches show these sequences to have no homology with any known Drosophila protein. However, both peptides share strong homology with different Tpx sequences, such as the human proliferation associated gene (pag, a gene constitutively expressed in most human cells and which is induced to higher levels upon serum stimulation in untransformed and Ras-transformed HBL cells [9]), and a murine gene product (MER5) that is preferentially expressed in erythroleukemia cells during the early period of cell differentiation [26]. cDNA sequence of Jafrac1 and Jafrac2 genes, and predicted amino-acid composition A Lambda gt11 cDNA library of Drosophila embryos was screened for plaques which cross-reacted with the human pag


q FEBS 2000

Fig. 2. Nucleotide and deduced amino-acid sequences of the Jafrac1 gene of Drosophila. The sequence contains a single ORF that encodes a protein (SSP5111 polypeptide) of 194 amino acids as shown below the nucleotide sequence.

probe. Two positive clones (Jafrac1 and Jafrac2) were isolated from 105 plaques. Both of them were recovered and processed for sequencing. The most intense of the positive clones (Jafrac1) enabled the determination of a 955 nucleotide sequence, including a poly(A) tail, that contained one open reading frame (ORF) that was deduced to encode a novel 194 amino-acid protein (calculated Mr = 21737 and pI = 5.43); this is shown in Fig. 2 aligned with the complete sequence of SSP5111. This polypeptide has the two critical motifs consistent with a Prx: i.e. the cysteine-containing segments surrounding Cys47 (FYPLDFTFVCPTEI) and Cys168 (GEVCPA). These cysteine motifs have been considered to be important for catalysis of peroxides [27]. The nucleotide sequence of the 1.1-kB cDNA insert of the other positive clone (Jafrac2) revealed an ORF that encodes a polypeptide of 242 amino acids with a calculated molecular mass of 26 726 Da and an apparent pI of 6.78. In the protein encoded by Jafrac2 the motifs mentioned above are found in positions corresponding to Cys95 and Cys216 (Fig. 3). These sequences were found in neither the nbrf nor the swissprot protein databases. Comparison of amino-acid sequences from the available databases The nonredundant protein database was searched using the blastp algorithm with the deduced amino-acid sequences codified by Jafrac1 and Jafrac2. These sequences were not found in the database, but significant homology was found with

different Tpxs from many species. An alignment of the deduced amino-acid sequences encoded by Jafrac1 and Jafrac2 compared to those of 10 known Prxs is presented in Fig. 4. Several positions along the sequence are perfectly conserved. The protein sequence coded by Jafrac1 (protein SSP5111) has homology ranging from 47% to 76% with known Prxs including Tpx 1 from mouse, rat and human (67% identity and 75% similarity in all three), Tpx 2 from mouse (66% identity and 75% similarity), rat (68% identity and 76% similarity) and human (68% identity and 76% similarity). Conservations are especially high around the conserved cysteine residues. These conserved regions are widely separated from another, but they retain the same relative order in most homologous proteins, suggesting that these cysteines might be catalytic residues. At the same time some residues in SSP5111 polypeptide (e.g. Tyr164) differ from the consensus found for the other Prxs. The sequence of the polypeptide codified by the Jafrac2 gene shows homology ranging from 50% to 68% with the 10 Prxs selected in Fig. 4. In addition to the similarity already mentioned the products codified by Jafrac1 and Jafrac2 revealed significant similarities with amoebial and bacterial proteins, for example a 29-kDa antigen of pathogenic Entamoeba histolytica, located on the surface of trophozoites [28]. Chromosomal mapping of Jafrac1 and Jafrac2 genes We mapped the Drosophila Jafrac1 and Jafrac2 genes by in situ hybridization to polytene chromosomes (Fig. 5). The

q FEBS 2000


Fig. 3. Nucleotide and deduced amino-acid sequences of the Jafrac2 gene of Drosophila. The sequence contains a single open reading frame that encodes a protein of 242 amino acids as shown below the nucleotide sequence.

biotin-labelled cDNA probe revealed a single hybridizing band for Jafrac1 located on the X chromosome (Fig. 5A) Close examination suggested that the Jafrac1 gene maps at region 11E of this chromosome (Fig. 5C). An analysis on the FlyBase Cyto Search shows that the wary (wy), congested (cgd ), strawberry notch (sno) and glutamate pyruvate transaminase (Gpt) genes mapped around this region. Chromosomal localization of the Jafrac2 gene demonstrated that it resides on band 62F on the 3 L chromosome (Fig. 5A,C). Antibodies against polypeptide SSP5111 A rabbit antibody was raised against purified SSP5111 protein recovered from 2D gels. The specificity of the antibody was determined by immunoblotting of 1D gels containing total cellular extracts from wing imaginal discs. Only one band with an apparent molecular mass of 22 kDa was obtained (data not shown). This suggests that the predicted size of the SSP5111 polypeptide is correct. Several studies have suggested that thiol antioxidants could be bridged through interchain disulfides [29]. These observations led us to consider whether Drosophila Prx might form homo- and heterodimers and whether this type of protein±protein complex might contribute mechanistically to the regulation of functional specificity. To test this hypothesis

25 wing imaginal discs were heated at 95 8C for 5 min in SDS sample buffer, in the absence or presence of dithiothreitol, analysed by SDS/PAGE, transferred to nitrocellulose filters and probed with the anti-SSP5111 serum. In the absence of dithiothreitol, cysteine residues would be expected to be oxidized during heating. However, in both cases only one band was observed with a mobility identical to that in reducing gels regardless of the presence or absence of the reducing agent (data not shown). This result suggests that SSP5111 does not form intermolecular disulfide linkages upon oxidation. Presence of polypeptide SSP5111 in different stages of Drosophila development We used the antibody against the SSP5111 polypeptide in Western blotting experiments to assess the levels of expression of this Prx throughout Drosophila development. To do this, 100 mg of total protein obtained at different stages of development (from early embryos to adult) were resolved by SDS/PAGE, transferred to nitrocellulose filters and probed with anti-SSP5111 serum. The result is presented in Fig. 6 and clearly shows how SSP5111 is abundant in early embryos (0± 14 h). From this time the protein decreases gradually until third

Fig. 4. Alignment of the predicted amino-acid sequence of the Drosophila Jafrac1 and Jafrac2 gene products with different Trxs. Identities are boxed in black and similarities in grey. Gaps have been introduced to optimize alignment, and residue numbers are indicated on the right. Abbreviations on left margin: TDX1_Mouse: Trx-dependent peroxide reductase 1 (TSA); TDX1_Rat: Trx-dependent peroxide reductase 1 (TSA); TDX1_Human: Trx-dependent peroxide reductase 1 (TSA) (PRP) (NKEF-B); TDX2_Rat: Trx-dependent peroxide reductase 2 (HBP23); TDX2_Mouse: Trx-dependent peroxide reductase 2 (OSF-3); TDX2_Human: Trx-dependent peroxide reductase 2 (PAG) (NKEF-A); TDXM_Human: mitochondrial Trx-dependent peroxide reductase precursor; TDXM_Mouse: mitochondrial Trx-dependent peroxide reductase precursor (MER5); TDXN_Human: antioxidant enzyme AOE372; TDXN_Mouse: antioxidant enzyme AOE372.

492 J. Rodriguez et al. (Eur. J. Biochem. 267) q FEBS 2000

q FEBS 2000


Fig. 5. Chromosomal localization of Jafrac1 and Jafrac2 genes. (A) Double in situ hybridization to a D. melanogaster polytene chromosome (blue) with digoxigenin-labelled Jafrac1 (green) and biotin-labelled Jafrac2 (red). This general view shows that each gene hybridizes on a single band. (B) Magnification of the 3L chromosome arm. Gene Jafrac2 maps at polytene chromosome band 62F. (C) Magnification of the X chromosome arm. Gene Jafrac1 hybridizes on polytene chromosome band 11E.

instar larvae. Then, the level of SSP5111 polypeptide increases from young pupae until adult. No differences were found between adult males and females (data not shown).

Specificity of the Anti-SSP5111 antibody in 2D gels The specificity of the antibody was also determined by immunoblotting of 2D gels containing total cellular extracts from wing imaginal discs. As shown in Fig. 7A the antibody recognized two spots in a 2D blot. The exact position of SSP5111 was determined by autoradiography of the blot as the cell extracts were labelled with [35S]methionine + [35S]cysteine (Fig. 7B). The other spot recognized by the antibody was the spot SSP6107 (see Fig. 1B), a spot with identical molecular mass but slightly more acidic than SSP5111. To corroborate the cross-hybridization we purified the spot SSP6107 from 80 Fig. 6. Presence of SSP5111 polypeptide at different stages of Drosophila development. Total cellular extracts resolved by 15% SDS/PAGE, and transferred onto nitrocellulose filters were analysed for the presence of SSP5111. The same amount of protein (100 mg) was applied to each line. EE (Embryos 0±12 h), EL(Embryos 12±22 h), LI(First instar Larvae), LII(Second instar Larvae), LIII(Third instar Larvae), PE (Pupae), PL (Pupae), A (Adult).

Fig. 7. Specificity of polyclonal antibodies against SSP5111 and SSP6107 polypeptides. Twenty wing imaginal discs of late third instar larvae of D. melanogaster were labelled with [35S]methionine + [35S]cysteine, resolved by 2D electrophoresis and transferred to nitrocellulose filters. The filters were analysed with a polyclonal antibody against SSP5111 polypeptide (A) or with a polyclonal antibody against SSP6107 polypeptide (C). (B) and (D) are the autoradiographs of (A) and (C), respectively.

preparative gels and raised an antibody against this polypeptide. The specificity of the new antibody was again determined by immunoblotting of 2D gels with the result shown in Fig. 7C. Again two spots were seen, one corresponding to polypeptide


q FEBS 2000

Fig. 8. Ìn vitro' translation products of polyA+ mRNA from Drosophila embryos (A) or Drosophila adults (B). In each case 1 mg of polyA+ mRNA was translated in a rabbit reticulocyte system. The product of the ìn vitro' translation was mixed with 20 wing imaginal discs, resolved by 2D electrophoresis and transferred to nitrocellulose filters. The filters were assayed with the antiserum against SSP5111 polypeptide and exposed for autoradiography. (A) and (B) are the autoradiographs of (C) and (D) respectively. Arrows indicate polypeptides SSP5111 (left) and SSP6107 (right).

SSP6107 and other corresponding to SSP5111. These results suggest that the polypeptides could be related. Taking into account their relative positions on the gel, the more acidic polypeptide (SSP6107) could be a post-translational modification of polypeptide SSP5111. Experiments in which wing imaginal discs were labelled with 32P or with [14C]glucosamine were negative, in the sense that in neither case did we obtain any kind of label in the polypeptide SSP6107. Polypeptides SSP5111 and SSP6107 are the products of two different genes or the result of alternative splicing: ìn vitro' translation experiments To determine the nature of the two spots, SSP5111 and SSP6107, recognized by the antiserum in 2D gels, we designed an experiment with the aim of determining whether these two spots arise from a modification of the polypeptide SSP5111, different from phosphorylation or glycosylation, or if they are products of two different genes. Therefore, we performed an ìn vitro' translation experiment using rabbit reticulocyte lysate in the presence of 1 mg Drosophila embryo polyA+ mRNA. The final product of the ìn vitro' translation was mixed with 20 wing imaginal discs, resolved by 2D electrophoresis and transferred to a nitrocellulose filter. The filter was then assayed with the anti-SSP5111 serum and exposed for autoradiography (Fig. 8). Fig. 8C shows the two spots logically expected in the immunoblotting. When the autoradiography (Fig. 8A) was superimposed over the filter, no radioactive spots were detected

over the two positions corresponding to the peptides. We conclude that the polyA+ RNA from embryos does not translate the polypeptides searched. Similar negative results were obtained when the ìn vitro' translation was performed in the presence of dog pancreas microsomes (data not shown). This negative result could indicate ineffective translation. We then tested the hypothesis that these two polypeptides correspond to maternal-effect genes transferred by the mother to the embryos, and that the embryo does not produce mRNA for such products: the ìn vitro' translation experiment was repeated using 1 mg polyA+ mRNA isolated from adult Drosophila. The results obtained are summarized in Fig. 8B,D. The nitrocellulose filter (Fig. 8D) shows the two expected spots, but in this case it was possible to superimpose two radioactive spots over them (Fig. 8B). When microsomes from dog pancreas were added to the in vitro translation system, the two spots were not modified (data not shown). The results suggest two main conclusions: first, that the two spots belong to polypeptides encoded by different genes or one gene with differential splicing; and second, that the gene products are transferred to the embryo by the mother. Analysis of SSP5111 and SSP6107 polypeptides by MALDI-MS and Q-Tof-MS To obtain more details about the possible relationship between SSP5111 and SSP6107 polypetides, we analysed the MALDI-MS and Q-Tof-MS of the fragments resulting from the complete tryptic digestion of the polypeptides. In the SSP5111 MALDI-MS

Fig. 9. Presence of SSP5111 and SSP6107 polypeptides in different species of Drosophila. [35S]Methionine + [35S]cysteine-labelled polypeptides from wing imaginal disc of late third instar larvae of D. melanogaster and D. virilis resolved in 2D gel electrophoresis were transferred to nitrocellulose filters. The filters were assayed with the antiserum against SSP5111 polypeptide as described and exposed for autoradiography. (A) and (C) are the autoradiographs of (B) and (D) respectively. Arrows indicate polypeptides SSP5111 (left) and SSP6107 (right).

q FEBS 2000


spectra it was possible to identify seven peptides that covered 88 out of 194 amino acids of the protein (45.3%). Five of them were also present in the MALDI spectra corresponding to SSP6107 polypeptide (data not shown). The Q-Tof-MS analysis of both polypeptides was performed over the different peaks obtained in the HPLC chromatogram. In this way it was possible to assign specific sequences to 12 peaks in the chromatogram covering 85% of the total sequence of SSP5111 polypetide. All of the peaks were also assigned in the chromatogram obtained with SSP6107 polypeptide, indicating that both polypeptides share identity in at least this percentage of their amino-acids sequences.

IEF39 h). Moreover, it shows a level of expression `complementary' to that showed by SSP5111, in the sense that SSP6107 shows lower levels of expression in haltere, leg1, leg2, and leg3 than in wing or eye-antenna imaginal discs, but the level of both adds up to a constant value in all discs analysed. What is the relationship between these two polypeptides? The possibility of a post-translational modification was discarded by the ìn vitro' translation experiments, because both polypeptides were translated. Compatible with the fact that in the cytological mapping we obtain a single signal in the 11E region of the X chromosome, it is a differential splicing of the same gene or a gene duplication to a neighbouring position. MS analysis indicate that SSP5111 and SSP6107 share at least 85% of their amino-acid sequences. The isoelectric points of the polypeptides in D. melanogaster are considerably different from those in D. virilis suggesting they are separated by at least a mutational event and not a genetic posttranslational modification. The fact that both spots move to a different charge position while maintaining similar molecular mass strongly suggests that the two spots arise from differential splicing of a single gene separated in the two species by five million years of evolution. The fact that the expression profile of both polypeptides is complementary in different discs further support the notion of a similar transcriptional control but a different splicing pattern. Thus, the results suggest the existence of at least a third member in the Tpx family of Drosophila. The existence of multiple peroxidases could reflect multiple tissue distributions or specific combinations among them. Data of others [30] show that different Prxs serve restricted functions in a subcellular and tissue-specific manner. We have identified more than 30 proteins that exhibit homology, extended over the entire sequence, with the SSP5111 polypeptide. The sequence homology is particularly marked in the regions surrounding the two conserved cysteine residues that correspond to Cys47 and Cys168 of the SSP5111 polypeptide. The peroxidase reaction of Tpx proteins requires both of the conserved cysteine residues because the oxidized enzyme intermediate generated during the catalytic cycle is a dimer in which the subunits are linked by one or two intermolecular disulfide bonds. Concerning the function of these proteins, a primary role of this family of enzymes must be to act as peroxidases. Although H2O2 is generally considered to be a toxic by-product of respiration, increasing evidence suggests that the production of reactive oxygen species, such as H2O2 or superoxide anions, might be an integral component of membrane receptor signalling. In mammalian cells, a variety of extracelular stimuli, including tumour necrosis factor-a [15], platelet-defined growth factor [13], and epidermal growth factor [31], induce a transient increase in the intracellular concentration of H2O2. Many of these proteins that exhibit homology with SSP5111 polypeptide were discovered, as in our case, in connection with a variety of seemingly unrelated cellular processes such as proliferation, differentiation, natural killer cell activity and response to oxidative stress. For example, significant homology was found between polypeptide SSP5111 and MER5, a polypeptide which is preferentially expressed in erythroleukaemia cells during the early period of cell differentiation induced by SOMe2 [26]. Another case of homology with polypeptide SSP5111 involves one of the major proteins of red blood cells, the acidic peroxidoxin (TSA) or protector protein (PRP). This protein is induced at early stages of erythroid differentiation prior to haemoglobin accumulation, which suggests that it may play an important role in the differentiation programme of the erythroid cells, directed towards the detoxification of the reactive oxygen species [32]. The homology also includes a

Localization of SSP5111 and SSP6107 polypeptides in different species of Drosophila We also searched for the localization of these two related polypeptides, SSP5111 and SSP6107, in D. virilis, a species that is evolutionarily separated from D. melanogaster by millions of years. We labelled wing imaginal discs of D. virilis in the same experimental conditions as those used for D. melanogaster. The filters were assayed with the antibody against SSP5111 with the results shown in Fig. 9. The immunoblot of D. melanogaster (Fig. 9B) and its corresponding autoradigraph (Fig. 9A) was as expected. In the case of D. virilis we also obtained two positive spots in the immunoblot (Fig. 9D) that correspond to two labelled spots in the autoradigraph (Fig. 9C), but shifted to more basic positions than those obtained for D. melanogaster. In the same way, the distance between the two spots is greater in D. virilis (24 mm) than in D. melanogaster (16 mm).

DISCUSSION This study is part of an effort by our laboratories to monitor changes in polypeptide synthesis involved in cell proliferation and differentiation associated with morphogenesis in Drosophila. We are using the wing imaginal disc as both a model system and as a quantitative protein database that can be used for comparative purposes in genetic studies. We have focused on the purification of a polypeptide (cataloged as SSP5111 in acidic gels), that showed different levels of expression in different imaginal discs. Preparative two dimensional gels followed by microsequencing revealed that polypeptide SSP5111 shared a very strong homology with the human pag, a cytoplasmic Tpx constitutively expressed in most human cells, which is induced to higher levels upon serum stimulation in untransformed and ras-transformed HBL100 cells [9]. A probe of such a gene enabled us to screen a Drosophila cDNA library from which two clones (Jafrac1 and Jafrac2) were isolated and for which complete cDNA sequence was determined. Jafrac1 encoded a polypeptide of 194 amino acids, corresponding to the spot SSP5111 searched, and Jafrac2 encoded a polypeptide of 242 amino acids not yet identified in the database. The sequences show homology of 58% that reaches 75% if the first 44 amino acids of the product encoded by Jafrac2 are not considered. Homology searches clearly indicate that both polypeptides are members of the Tpx family, a widely distributed class of enzymes that act as peroxidases but require Trx or a thiol-containing intermediate to carry out their peroxidase function. An antibody against SSP5111 not only recognizes this polypeptide, but also another polypeptide, SSP6107; interestingly this is also included in the set of 17 polypeptides differentially expressed in imaginal discs (catalogued as


mitochondrial protein induced during differentiation of Friend erythroleukemia cells [33], a protein induced by oxidative stress in the macrophage [34], and two proteins showing a natural killer cell-enhancing action in vitro and named NKEF-A and -B [35]. What can we say about the function of these proteins in Drosophila? Western blot analysis showed significant differences in the levels of expression of SSP5111 polypeptide throughout the different stages development. The highest level was found in young embryos, decreasing dramatically until the mature larvae, then increasing until adult suggesting that SSP5111 polypeptide is required in specific developmental processes. The data are reinforced by the fact that later on SSP5111 polypeptide is also differentially expressed in imaginal discs. In a cellular system, such as imaginal discs, which undergo extensive proliferation, oxidative metabolism must be accelerated. However no significant changes in the level of expression of SSP5111 nor SSP6107 polypeptides were found in the analysis of two Drosophila tumour suppressor mutants with abnormal wing disc development: fat ( f t) and two different alleles of lethal (2) giant discs (l(2)gd ) [7]. The ìn vitro' translation experiments showed that although SSP5111 is a major product in the embryo, this polypeptide was not translated when polyA+ mRNA from embryos was used, suggesting the possibility that SSP5111 polypeptide could be the result of a maternal-effect gene. This result could be connected with the data concerning the maternal-effect deadhead (dhd ) gene of Drosophila [36]. Females homozygous for dhd mutations lay eggs that appear to be morphologically normal but most fail to initiate development. Interestingly, about 10% of the embryos from dhd mothers initiate development; however, their development is not normal. In contrast with wild-type embryos, in which the initial nuclear divisions are synchronous, the nuclear divisions in the mutant embryos are asynchronous. The interesting point is that the dhd gene encodes a protein with extensive sequence similarity to Trx [37], a highly conserved disulfide reducing protein involved in the redox chain providing electrons to Tpx for reducing H2O2. The data suggest again that redox regulation plays a role in the post-translational control of some of the earliest events in development. The SSP5111 polypeptide must be one of the gene products required for these early events. Indeed, recent studies in a number of systems have begun to establish that the reversible formation of disulfide bonds via redox-control is important in post-translational control.

ACKNOWLEDGEMENTS The authors thank A. GarcõÂa-Bellido for help during the course of this work, G. Goubin for providing the pag probe, S. Campuzano for providing the polyA+ mRNA, M. Puype for expert assistance in the peptide separation by HPLC and C. Extavour for helpful comments on the manuscript. This work was supported by Grant PB87-0449 from Comision Asesora de InvestigacioÂn CientõÂfica y TeÂcnica and by an institutional grant from FundacioÂn RamoÂn Areces.

q FEBS 2000

5.

6.

7.

8.

9.

10. 11.

12.

13.

14. 15.

16.

17.

18.

19.

20.

21.

REFERENCES 1. O'Farrell, P.H. (1975) High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250, 4007±4021. 2. O'Farrell, P.Z., Goodman, H.M. & O'Farrell, P.H. (1977) High resolution two-dimensional electrophoresis of basic as well as acidic proteins. Cell 12, 1133±1142. 3. SantareÂn, J.F. (1990) Towards establishing a protein data base of Drosophila. Electrophoresis 11, 254±267. 4. SantareÂn, J.F. & GarcõÂa-Bellido, A. (1996) 2D gene expression

22.

23.

parameters of wing imaginal disc of Drosophila for developmental analysis. Dev. Genes Evol. 206, 349±354. SantareÂn, J.F. (1996) Two-dimensional gel analysis of proteins in cell lines from the central nervous system of larval Drosophila in vitro. Cell. Dev. Biol. 32, 434±440. SantareÂn, J.F. & GarcõÂa-Bellido, A. (1990) High-resolution twodimensional gel analysis of proteins in wing imaginal discs: a data base of Drosophila. Exp. Cell Res. 189, 169±176. SantareÂn, J.F., MilaÂn, M. & GarcõÂa-Bellido, A. (1998) Two-dimensional gel analysis of proteinas in the Drosophila wing imaginal disc mutants fat and lethal (2) giant discs. Exp. Cell Res. 243, 199±206. SantareÂn, J.F., Assiego, R. & GarcõÂa-Bellido, A. (1993) Patterns of protein synthesis in imaginal discs of Drosophila melanogaster, a comparison between different discs and stages. Roux's Arch. Dev. Biol. 203, 131±139. ProspeÂri, M.-T., Ferbus, D., Karczinski, I. & Goubin, G. (1993) A human cDNA corresponding to a gene overexpressed during cell proliferation encodes a product sharing homology with amoebic and bacterial proteins. J. Biol. Chem. 268, 11050±11056. Chae, H.Z., Chung, J.S. & Rhee, S.G. (1994) Thioredoxin-dependent peroxide reductase from yeast. J. Biol. Chem. 269, 27670±27678. Chae, H.Z., Robison, K., Poole, L.B., Church, G., Storz, G. & Rhee, S.G. (1994) Cloning and sequencing of thiol-specific antioxidant from mammalian brain: alkyl hydroperoxide reductase and thiolspecific antioxidant define a large family of antioxidant enzymes. Proc. Natl Acad. Sci. USA 91, 7017±7021. Jacobson, F.S., Morgan, R.W., Christman, M.F. & Ames, B.N. (1989) An alkyl hydroperoxide reductase from Salmonella typhimurium involved in the defense of DNA against oxidative damage. J. Biol. Chem. 264, 1488±1496. Sundaresan, M., Yu, Z.-X., Ferrans, V.J., Irani, K. & Finkel, T. (1995) Requirement for generation of H2O2 for platelet-derived growth factor signal transduction. Science 279, 296±299. Holmgren, A. (1989) Thioredoxin and glutaredoxin systems. J. Biol. Chem. 264, 13963±13966. Kang, S.W., Chae, H.Z., Seo, M.S., Kim, K., Baines, I.C. & Rhee, S.G. (1998) Mammalian peroxiredoxin isoforms can reduce hydrogen peroxide generated in response to mgrowth factors and tumor necrosis factor-a. J. Biol. Chem. 273, 6297±6302. Zhang, P., Liu, B., Kang, S.W., Seo, M.S., Rhee, S.G. & Obeid, L.M. (1997) Thioredoxin peroxidase is a novel inhibitor of apoptosis with a mechanism distinct from that of Bcl-2. J. Biol. Chem. 272, 30615±30618. Currie, D.A., Milner, M.J. & Evans, C.V. (1988) The growth and differentiation in in vitro of leg and wing imaginal disc cells from Drosophila melanogaster. Development 102, 805±814. Laskey, R.A. & Mills, A.D. (1975) Quantitative film detection of 3H and 14C in polyacrylamide gels by fluorography. Eur. J. Biochem. 56, 335±341. Bravo, R. (1984) Two-dimensional gel electrophoresis: a guide for the beginner. In Two-Dimensional Electrophoresis of Proteins (Celis, J.E. & Bravo, R., eds), pp. 4±36. Academic Press, New York. Rasmussen, H.H., Van Damme, J., Bauw, G., Puype, M., Gesser, B., Celis, J.E. & Vandekerckhove, J. (1991) Protein-electroblotting and microsequencing in establishing integrated human protein databases. In Methods in Protein Sequence Analysis (JoÈrnvall, H., HoÈoÈg, J.O. & Gustavsson, A.M., eds), pp. 103, BirkhaÈuser-Verlag, Basel. SantareÂn, J.F., Van Damme, J., Puype, M., Vandekerckhove, J. & GarcõÂa-Bellido, A. (1993) Identification of Drosophila wing imaginal discs proteins by two-dimensional gel analysis and microsequencing. Exp. Cell Res. 206, 220±226. Thompson, J.D., Higgins, D.G. & Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673±4680. Gevaert, K. Demol, H. Sklyarova, T. Vandekerckhove, J. & Houthaeve, T. (1998) A peptide concentration and purification method for protein characterization in the subpicomole range using matrix assisted laser

q FEBS 2000

24.

25.

26.

27.

28.

29.

30.

desorption/ionization-postsource decay (MALDI-PSD) sequencing. Electrophoresis 19, 909±917. Pardue, M.L. (1994) Looking at Drosophila polytene chromosomes. In Drosophila melanogaster: Practical Uses in Cell and Molecular Biology (Goldstein, L.S.B & Fyrberg, E.A., eds), pp. 334±351. Academic Press, San Diego. Gatti, M., Bonaccorsi, S. & Pimpinelli, S. (1994) Looking at Drosophila mitotic chromosomes. In Drosophila melanogaster: Practical Uses in Cell and Molecular Biology (Goldstein, L.S.B & Fyrberg, E.A., eds), pp. 372±391. Academic Press, San Diego. Yamamoto, T., Matsui, Y., Natorui, S. & Obinata, M. (1989) Cloning of a housekeeping-type gene (MER5) preferentially expressed in murine erythroleukemia cells. Gene 80, 337±343. Lim, Y.-S., Cha, M.-K., Kim, H.-K. & Kim, I.I.-H. (1994) The thiolspecific antioxidant protein from human brain: gene cloning and analysis of conserved cysteine regions. Gene 140, 279±284. Torian, B.E., Flores, B.M., Stroeher, V.L., Hagen, F.S. & Stamm, W.E. (1990) cDNA sequence analysis of a 29-kDa cysteine rich surface antigen of pathogenic Entamoeba histolytica. Proc. Natl Acad. Sci. USA 87, 6358±6362. Kang, S.W., Baines, I.C. & Rhee, S.G. (1998) Characterization of a mammalian peroxiredoxin that contains one conserved cysteine. J. Biol. Chem. 273, 6303±6311. Rozell, B., Hansson, H.A., Luthman, M. & Holmgren, A. (1985) Immunohistochemical localization of thioredoxin and thioredoxin reductase in adult rats. Eur. J. Cell Biol. 38, 79±86.

Thioredoxin peroxidase of Drosophila (Eur. J. Biochem. 267) 497 31. Bae, Y.S., Kang, S.W., Seo, M.S., Baines, I.C., Tekle, E., Chock, P.B. & Rhee, S.G. (1997) Epidermal growth factor (EGF)-induced generation of hydrogen peroxide. J. Biol. Chem. 272, 217±221. 32. Rabilloud, T., Berthier, R., Vincon, M., Ferbus, D., Goubin, G. & Lawrence, J.-J. (1995) Early events in erythroid differentiation: accumulation of the acidic peroxidoxin (PRP/TSA/NKEF-B) Biochem. J. 312, 699±705. 33. Watabe, S., Hasegawa, H., Takimoto, K., Yamamoto, Y. & Takahashi. S.Y. (1995) Possible function of SP-22, a substrate of mitochondrial ATP-dependent protease, as a radical scavenger. Biochem. Biophys. Res. Comm. 213, 1010±1016. 34. Ishii, T., Yamada, M., Sato, H., Matsue, M., Taketani, S., Nakayama, K., Sugita, Y. & Bannai, S. (1993) Cloning and characterization of a 23-kDa stress-induced mouse peritoneal macrophage protein. J. Biol. Chem. 268, 18633±18636. 35. Shau, H. & Kim, A. (1994) Identification of natural killer enhancing factor as a major antioxidant in human red blood cells. Biochem. Biophys. Res. Commun. 199, 83±88. 36. Salz, H.K., Flickinger, T.W., Mittendorf, E., Pellicena-PalleÂ, A., Petschek, J.P. & Albrecht, E.B. (1994) The Drosophila maternal effect locus deadhead encodes a thioredoxin homolog required for female meiosis and early embryonic development. Genetics 136, 1075±1086. 37. Pellicena-PalleÂ, A., Stitzinger, S.M. & Salz, H.K. (1997) The function of the Drosophila thioredoxin homologue encoded by the deadhead gene is redox-dependent and blocks the initiation of development but not DNA synthesis. Mech. Dev. 62, 61±65.