Prohibitins, Stomatins, and Plant Disease Response Genes Compose a Protein Superfamily That Controls Cell Proliferation, Ion Channel Regulation, and Death
2000; Elsevier BV; Volume: 275; Issue: 38 Linguagem: Inglês
10.1074/jbc.m002339200
ISSN1083-351X
AutoresRamgopal Nadimpalli, Nasser Yalpani, Gurmukh S. Johal, Carl R. Simmons,
Tópico(s)Plant Virus Research Studies
ResumoProhibitins, stomatins, and a group of plant defense response genes are demonstrated to belong to a novel protein superfamily. This superfamily is bound by similar primary and secondary predicted protein structures and hydropathy profiles. A PROSITE-formatted regular expression was generated that is highly predictive for identifying members of this superfamily using PHI-BLAST. The superfamily is named PID (proliferation,ion, and death) because prohibitins are involved in proliferation and cell cycle control, stomatins are involved in ion channel regulation, and the plant defense-related genes are involved in cell death. The plant defense gene family is named HIR (hypersensitive induced reaction) because its members are associated with hypersensitive reactions involving cell death and pathogen resistance. For this study, eight novel maize genes were introduced: four closely related to prohibitins (Zm-phb1, Zm-phb2, Zm-phb3, and Zm-phb4), one to stomatins (Zm-stm1), and three to a gene implicated in plant disease responses (Zm-hir1, Zm-hir2, and Zm-hir3). The maize Zm-hir3 gene transcript is up-regulated in a disease lesion mimic mutation (Les9), supporting a role in maize defense responses. Members of this gene superfamily are involved in diverse functions, but their structural similarity suggests a conserved molecular mechanism, which we postulate to be ion channel regulation. Prohibitins, stomatins, and a group of plant defense response genes are demonstrated to belong to a novel protein superfamily. This superfamily is bound by similar primary and secondary predicted protein structures and hydropathy profiles. A PROSITE-formatted regular expression was generated that is highly predictive for identifying members of this superfamily using PHI-BLAST. The superfamily is named PID (proliferation,ion, and death) because prohibitins are involved in proliferation and cell cycle control, stomatins are involved in ion channel regulation, and the plant defense-related genes are involved in cell death. The plant defense gene family is named HIR (hypersensitive induced reaction) because its members are associated with hypersensitive reactions involving cell death and pathogen resistance. For this study, eight novel maize genes were introduced: four closely related to prohibitins (Zm-phb1, Zm-phb2, Zm-phb3, and Zm-phb4), one to stomatins (Zm-stm1), and three to a gene implicated in plant disease responses (Zm-hir1, Zm-hir2, and Zm-hir3). The maize Zm-hir3 gene transcript is up-regulated in a disease lesion mimic mutation (Les9), supporting a role in maize defense responses. Members of this gene superfamily are involved in diverse functions, but their structural similarity suggests a conserved molecular mechanism, which we postulate to be ion channel regulation. hypersensitive reaction expressed sequence tag 4-morpholinepropanesulfonic acid Plants frequently respond to pathogen attack with the "hypersensitive reaction" (HR),1 a rapid localized necrosis at the site of infection that cordons off the pathogen and limits its spread (1Agrios G.N. Plant Pathology. Academic Press Ltd., London1988Google Scholar, 2Dangl J.L. Dietrich R.A. Richberg M.H. Plant Cell. 1996; 8: 1793-1807Crossref PubMed Scopus (723) Google Scholar, 3Greenberg J.T. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 12094-12097Crossref PubMed Scopus (499) Google Scholar). The HR cell death phenomenon bears similarities to programmed cell death or apoptosis observed in animals (3Greenberg J.T. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 12094-12097Crossref PubMed Scopus (499) Google Scholar). A group of tobacco genes were identified that caused the formation of HR-like lesions on tobacco leaves when ectopically expressed from a tobacco mosaic virus expression vector (4Karrer E.E. Beachy R.N. Holt C.A. Plant Mol. Biol. 1998; 36: 681-690Crossref PubMed Scopus (91) Google Scholar). One of these genes, represented by a cDNA called NG1, caused both HR-like lesion formation and induced expression of β-glucanase, a pathogenesis-related protein marker for defense activation. NG1 was thus interpreted to be a novel activator of the plant HR defense system (4Karrer E.E. Beachy R.N. Holt C.A. Plant Mol. Biol. 1998; 36: 681-690Crossref PubMed Scopus (91) Google Scholar). The predicted peptide for the cDNA NG1 was presented as 64 amino acids with no significant homology to any known proteins (4Karrer E.E. Beachy R.N. Holt C.A. Plant Mol. Biol. 1998; 36: 681-690Crossref PubMed Scopus (91) Google Scholar). Independently, an Arabidopsis NG1-like cDNA (gene 106) was reported to represent an mRNA induced by the plant defense activator isonicotinic acid (5Ryals, J. A., Alexander, D. C., Beck, J. J., Duesing, J. H., Goodman, R. M., Friedrich, L. B., Harms, C., Meins, F., Jr., Montoya, A., Moyer, M. B., Neuhaus, J.-M., Payne, G. B., Sperisen, C., Stinson, J. R., Uknes, S. J., Ward, E. R., and Williams, S. C. (March 25, 1997) U. S. Patent 5,614,395.Google Scholar). This induction was associated with systemic acquired resistance (reviewed in Ref. 6Neuenschwander U. Lawton K. Ryals J. Stacey G. Keen N.T. Plant-Microbe Interactions. Chapman and Hall, Inc., New York1996: 81-106Google Scholar) and occurred independent of de novo protein synthesis (5Ryals, J. A., Alexander, D. C., Beck, J. J., Duesing, J. H., Goodman, R. M., Friedrich, L. B., Harms, C., Meins, F., Jr., Montoya, A., Moyer, M. B., Neuhaus, J.-M., Payne, G. B., Sperisen, C., Stinson, J. R., Uknes, S. J., Ward, E. R., and Williams, S. C. (March 25, 1997) U. S. Patent 5,614,395.Google Scholar). Prohibitins are a group of highly conserved proteins that are thought to control the cell cycle, senescence, and tumor suppression (reviewed in Ref. 7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar). Prohibitins negatively control the cell cycle in the early G1 phase and specifically inhibit initiation of DNA synthesis (8Nuell M.J. Stewart D.A. Walker L. Friedman V. Wood C.M. Owens G.A. Smith J.R. Schneider E.L. Dell'Orco R. Lumpkin C.K. Danner D.B. McClung J.K. Mol. Cell. Biol. 1991; 11: 1372-1381Crossref PubMed Scopus (229) Google Scholar, 9Roskams A.J.I. Friedman V. Wood C.M. Walker L. Owens G.A. Stewart D.A. Altus M. Danner D.B. Liu X.-T. McClung J.K. J. Cell. Physiol. 1993; 157: 289-295Crossref PubMed Scopus (77) Google Scholar). Prohibitin genes appear to be expressed in many tissues and organisms, but with some modulation of expression consistent with a role in the cell cycle (7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar, 8Nuell M.J. Stewart D.A. Walker L. Friedman V. Wood C.M. Owens G.A. Smith J.R. Schneider E.L. Dell'Orco R. Lumpkin C.K. Danner D.B. McClung J.K. Mol. Cell. Biol. 1991; 11: 1372-1381Crossref PubMed Scopus (229) Google Scholar, 10Choongkittaworn N.M. Kim K.H. Danner D.B. Griswold M.D. Biol. Reprod. 1993; 49: 300-310Crossref PubMed Scopus (44) Google Scholar). Mutations or deletions of prohibitin are linked to some human breast and ovarian cancers, supporting the idea that prohibitin suppresses tumors as part of its antiproliferative function involving cell cycle control (11Legget B. Young J. Buttenshaw R. Thomas L. Young B. Chenevix-Trench G. Searle J. Ward M. Br. J. Cancer. 1995; 71: 1070-1073Crossref PubMed Scopus (33) Google Scholar, 12Sato T. Saito H. Swensen J. Olifant A. Wood C. Danner D. Sakamoto T. Takita K. Kasumi F. Miki Y. Skolnick M. Nakamura Y. Cancer Res. 1992; 52: 1643-1646PubMed Google Scholar, 13Foulkes W.D. Black D.M. Stamp G.W.H. Soloman E. Trowsdale J. Int. J. Cancer. 1993; 54: 220-225Crossref PubMed Scopus (95) Google Scholar). Prohibitins are also implicated in controlling senescence and aging, with which there may be a functional link to their antiproliferative function and cell cycle control (7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar). Prohibitins are localized largely in the mitochondria, especially in the inner mitochondrial membrane (7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar) near the periphery (14Ikonen E. Fiedler K. Parton R.G. Simons K. FEBS Lett. 1995; 358: 273-277Crossref PubMed Scopus (169) Google Scholar). Rat and human prohibitins possess a short transmembrane helix near their N termini, which may be integrated into mitochondrial membranes (7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar). As mitochondrial inner membrane proteins often control ion transport and ATP production, it has been speculated that prohibitins may be involved in these processes, in particular in mitochondrial calcium efflux, which regulates ATP formation (7McClung J.K. Jupe E.R. Dell'Orco R.T. Exp. Gerontol. 1995; 30: 99-124Crossref PubMed Scopus (173) Google Scholar). Prohibitin and the prohibitin-like protein BAP37 (also called prohibitone) have also been localized to the plasma membrane of mouse lymphocytes, where they together interact with the IgM antigen receptor and may function in signaling apoptotic programmed cell death (15Terashima M. Kim K.-M. Adachi T. Nielsen P. Reth M. Kohler G. Lamers M.C. EMBO J. 1994; 13: 3782-3792Crossref PubMed Scopus (209) Google Scholar). Prohibitin and BAP37 interact directly with each other in animal mitochondria; and in yeast, they control replicative life span, possibly through control of mitochondrial membrane ionic potential (16Coates P.J. Jameison D.J. Smart K. Prescott A.R. Hall PA. Curr. Biol. 1997; 7: 607-610Abstract Full Text Full Text PDF PubMed Google Scholar). Stomatin is an integral membrane protein found in red blood cells. In genetic disorders in which this protein is missing, a hemolytic anemia called stomatocytosis results. In stomatocytosis, the red blood cells experience high passive diffusion of univalent cations and are often overhydrated due to an abnormally high amount of intracellular sodium and low amounts of potassium. These red blood cells assume a mouth-like shape, thus stomatocytosis, from "stoma," which is Greek for mouth (17Stewart G.W. Argent A.C. Dash B.C.J. Biochim. Biophys. Acta. 1993; 1225: 15-25Crossref PubMed Scopus (73) Google Scholar). Stomatin is thought to function as a negative regulator of univalent cation permeability. Stomatin has a single membrane-spanning region near its N terminus, with the rest of the protein thought to be cytoplasmic (18Stewart G.W. Hepworth-Jones B.E. Keen J.N. Dash B.C.J. Argent A.C. Casimir C.M. Blood. 1992; 79: 1593-1601Crossref PubMed Google Scholar). The molecular mechanism for stomatin function is unknown, but its cytoplasmic portion has been suggested to act as a ball and chain tether that can directly plug ion channels and may also interact with the cytoskeleton (17Stewart G.W. Argent A.C. Dash B.C.J. Biochim. Biophys. Acta. 1993; 1225: 15-25Crossref PubMed Scopus (73) Google Scholar). Northern blots detect stomatin mRNA expression in many human tissues besides red blood cells (18Stewart G.W. Hepworth-Jones B.E. Keen J.N. Dash B.C.J. Argent A.C. Casimir C.M. Blood. 1992; 79: 1593-1601Crossref PubMed Google Scholar). In this work, we present eight novel full-length cDNA sequences from maize, four of which are closely related to prohibitins, three to the hypersensitive response-inducing protein NG1, and one to stomatins. We demonstrate that these eight novel plant genes, along with many animal, bacterial, plant, and fungal sequences representing prohibitins, NG1-like proteins, stomatins, and other membrane proteins, compose a novel protein superfamily. Although these genes are involved in diverse physiological processes, their structural similarity suggests that they possess a related biochemical function. The eight full-length maize cDNAs presented in this study were identified in the EST collection at Pioneer Hi-Bred International, Inc. mRNA sources were from various tissues and treatments. The cDNA libraries were created at Pioneer Hi-Bred, and the ESTs were generated at Human Genome Sciences. The NG1 (HIR), prohibitin, and stomatin homologs were identified with the aid of the IRIS software package from Human Genome Sciences, which includes the BLAST algorithm, through which homology was indicated to tobacco NG1 (GenBankTM/EBI Data Bank accession number U66271), to prohibitins from various species, and to human stomatin (accession number U33925). Full-length insert sequences were produced at Pioneer Hi-Bred by primer walking using an ABI 377 sequencing machine. Sequences were assembled using SequencherTM Version 3.0 (Gene Codes, Ann Arbor, MI) and/or AssemblyLIGNTM (Eastman Kodak Co.) software. Initial public database searches were carried out using the BLASTP program (19Altschul S.F. Gish W. Miller W. Myers E.W. Lipman D.J. J. Mol. Biol. 1990; 215: 403-410Crossref PubMed Scopus (70731) Google Scholar) with a chickpea HIR-like gene (NCBI Protein Database accession number gi 3928150) as a probe, followed by PSI-BLAST (20Altschul S.F. Madden T.L. Schäffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59923) Google Scholar) with default parameters (Blosum 62, gap existence cost 11, per residue gap cost 1, λ ratio = 0.85, expect threshold 10). About 24 sequences that appeared as significant hits, both in terms of statistical threshold and the type, were, along with the eight maize sequences presented herein, multiply aligned by the ClustalW program with default parameters (21Thompson J.D. Higgins D.G. Gibson T.J. Nucleic Acids Res. 1994; 22: 4673-4680Crossref PubMed Scopus (55757) Google Scholar). The residues were reduced to a consensus sequence according to an 80% consensus generated using the CONSENSUS program of Nigel Brown (NIMR, London). To look for conserved motifs in the 32 members included in the multiple alignment, we applied the MEME algorithm, which resulted in the detection of three highly conserved motifs (26Grundy W.N. Bailey T.L. Elkan C.P. Baker M.E. Biochem. Biophys. Res. Commun. 1997; 231: 760-766Crossref PubMed Scopus (33) Google Scholar). Highly conserved residues based in part on the MEME motifs were identified to generate a PROSITE-formatted regular expression profile to perform further data base searches by the PHI-BLAST program (22Zhang Z. Schäffer A. Miller W. Madden T.L. Lipman D.J. Koonin E.V. Altschul S.F. Nucleic Acids Res. 1998; 26: 3986-3990Crossref PubMed Scopus (255) Google Scholar). Phylogenetic analysis was carried out by using an option within ClustalW (23Higgins D.G. Thompson J.D. Gibson T.J. Methods Enzymol. 1996; 266: 383-402Crossref PubMed Scopus (1288) Google Scholar) to generate multiple alignments, followed by distance calculations and tree constructions with the PROTDIST and neighbor-joining program of the PHYLIP package (24Felsenstein J. PHYLIP Phylogeny Inference Package , Version 3.5c. Department of Genetics, University of Washington, Seattle1993Google Scholar). Secondary structure predictions were carried out by the DSC algorithm using multiple sequence inputs (25King R.D. Saqi M. Sayle R. Sternberg M.J.E. Comput. Appl. Biosci. 1997; 13: 473-474PubMed Google Scholar). Further structural analyses were carried out by hydropathy profiles using the Kyte-Doolittle method with a 19-residue sliding window. Plant material for mRNA expression analysis was produced from the following three maize families, each with the Les9 mutation (a disease lesion mutation of maize) segregating 1:1 among the progeny: family 1 (Mo95 18-15 × sibling wild-type +/Les9; background M14/Mo20W), family 2 (Mo94S 16-35 × sibling wild-type +/Les9; background M14W23/W23r), and family 3 (Mo95 24-3 × sibling wild-type +/Les9; background M14/W23). Of the three families, only family 1 with the Mo20W background suppresses the Les9 lesion mimic phenotype. For the Affymetrix GeneChip® analysis, wild-type andLes9 mutant plants from all three families were used. Plants were grown in soil in the greenhouse to the V8 stage, which is when the characteristic Les9 lesions normally begin to appear. The young, upper leaf of Les9 phenotype plants that did not yet express a lesion phenotype on that leaf and corresponding tissue from wild-type sibling plants were harvested. Using duplicate equal 2-g samples representing each of these six tissues, total RNA was isolated by the TriReagent® method according to the manufacturer's recommendations (Molecular Research Center, Inc., Cincinnati, OH). Pooled tissue from three different plants formed one sample, and the plants used for each sample were distinct. For GeneChip® expression analysis, 1 mg of total RNA from each sample was used for poly(A)+ mRNA isolation by the OligoTex resin binding method according to the manufacturer's recommendations (QIAGEN Inc., Chatsworth, CA). Protocols for preparing in vitro transcribed biotinylated cRNA probes from poly(A)+ mRNA for Affymetrix GeneChip® gene expression analysis were according to the manufacturer's recommendations (Affymetrix, Santa Clara, CA) and have been described (27Wodicka L. Dong H. Mittmann M. Ho M.-H. Lockhart D.J. Nature Biotechnol. 1997; 15: 1359-1367Crossref PubMed Scopus (857) Google Scholar). In brief, 2 μg of poly(A)+mRNA/sample, described above in mRNA isolations, was used for the first strand cDNA synthesis. This involved a T7-(dT)24 oligonucleotide primer and reverse transcriptase SuperScript II (Life Technologies, Inc.). The second strand synthesis involved Escherichia coli DNA polymerase I (Life Technologies, Inc.). The double-stranded cDNA was then cleaned up using phenol/chloroform extraction and phase-lock gels (5 Prime → 3 Prime, Inc., Boulder, CO), followed by ethanol precipitation. For thein vitro transcription to produce cRNA, biotin-11-CTP and biotin-16-UTP, in addition to all four NTPs, were used with T7 transcriptase (Ambion Inc., Austin, TX). The in vitrotranscript product was cleaned up using RNeasy affinity resin columns (QIAGEN Inc.). In vitro labeled transcript yields ranged from 60 to 80 μg/sample. They were stored at −80 °C until used. The in vitro transcript products were fragmented in acetate buffer (pH 8.1) at 94 °C for 35 min prior to chip hybridization. Equal amounts (12 μg/sample) of in vitrolabeled transcript were used to probe each chip overnight. The biotinylated RNA hybridizing to the chips was labeled with a streptavidin-phycoerythrin conjugate and scanned using a confocal fluorescence microscope. Expression intensity was determined as described (27Wodicka L. Dong H. Mittmann M. Ho M.-H. Lockhart D.J. Nature Biotechnol. 1997; 15: 1359-1367Crossref PubMed Scopus (857) Google Scholar). Comparisons of mRNA abundance (cRNA abundance) were made per repetition between the Les9 and wild-type samples. The average -fold change and S.E. for all the repetitions per family were calculated and are presented. The GeneChip® used in these experiments was constructed by Affymetrix using a set of 1501 maize cDNA ESTs, representing nearly as many genes. The genes used to produce this GeneChip®encompass many physiological processes; perhaps one-third could be defense-related based on their homology to known or suspected defense-related genes. Two of the 1501 ESTs representedZm-hir3 (ESTs CMSAR19R and CBPCC63R). Each cRNA GeneChip® probing was replicated two or three times (repetitions A, B, and C). In brief, the 1.28 × 1.28-cm GeneChip® contains a high density array of 20-mer oligonucleotides affixed to a silicon wafer. These oligonucleotides were synthesized in situ on the silicon wafer by a light-dependent combinatorial chemical synthesis (27Wodicka L. Dong H. Mittmann M. Ho M.-H. Lockhart D.J. Nature Biotechnol. 1997; 15: 1359-1367Crossref PubMed Scopus (857) Google Scholar). The oligonucleotide sequences are complementary to the sense strand of cDNA ESTs from Pioneer Hi-Bred. For each gene, there are up to 40 20-mer oligonucleotides synthesized. Twenty of these oligonucleotides are exact matches to different, although sometimes overlapping, regions of the EST. The other 20 oligonucleotides contain one base mismatch in the center, which changes hybridization efficiency. (For a minority of genes, there were <20 oligonucleotide probe pairs, but never 80% identity) and to HIR-like proteins from tobacco, chickpea, andArabidopsis (>80% identity). Pairwise amino acid similarities of plant HIR and HIR-like proteins with maize prohibitins were between 28 and 36%, and those with maize stomatinZm-stm1 were between 34 and 37%. This suggested that the maize HIR proteins were somewhat closer in amino acid sequence to stomatins than to prohibitins. The non-redundant protein data base at NCBI was searched using the PSI-BLAST program (20Altschul S.F. Madden T.L. Schäffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59923) Google Scholar) with a hypothetical protein from chickpea (accession number gi 3928150) as a probe, which has >90% amino acid similarity to the maize HIR proteins. This search identified many genes, including stomatins and integral membrane proteins (E < 10−16), prohibitins (E < 10−8), and HFLK/HFLC proteins (E < 10−6). Twenty-four of these public sequences, along with the eight maize sequences introduced above, were used to generate an unrooted dendogram (Fig. 1). This dendogram revealed a large superfamily with at least four constituent families. The stomatins and integral membrane proteins, including a mechanosensor protein from Caenorhabditis elegans (accession number gi 2493263), formed a large family containing sequences from diverse phyla. A second family was composed of HIR and HIR-like sequences from plants. The family consisting of stomatins and integral membrane proteins was most closely related to the HIR family. The third family was composed of prohibitins and related sequences from diverse phyla. The bacterial membrane proteins HFLK/HFLC formed a small fourth family. Amino acid sequences for 32 members of this superfamily were also multiply aligned to reveal shared and diverged features (Fig.2). The coding region lengths for the HIR proteins (242–286 amino acids) are comparable to those of prohibitins (272–289 amino acids) and many of the stomatins and other membrane-associated proteins (249–481 amino acids). Relative to prohibitins and stomatins, the HIR proteins are typically shorter at the N terminus. Several regions of the protein superfamily are highly conserved and aligned well with fewer gaps. Two residues, Asp and Ala (corresponding to amino acids 64 and 167, respectively, in Zm-HIR1, being used here as a reference superfamily member), are completely conserved among all the proteins, suggesting a critical role for these residues in the biological function of these proteins. Other amino acids and structural groups of amino acids are also conserved in the PID superfamily, as shown in the consensus sequence (Fig. 2). Also depicted in Fig. 2 are the consensus DSC predicted secondary structures of each of the families within the PID superfamily. Each of the four families within the PID superfamily share secondary structural features in the same general relative positions, further indicating a relationship between these proteins.Figure 2Fig. 2. Multiple sequence alignment and consensus and DSC secondary structure predictions for the PID superfamily. A multiple alignment of 32 amino acid sequences for representative members of the PID superfamily was constructed by the ClustalW program and then manually refined: HIR proteins (green bar), stomatins and membrane proteins (red bar), HFLK proteins (brown bar), and prohibitins (blue bar). Highlighted are identical (red) or similar (blue) residues shared by at least three families within the PID superfamily. The alignment was used to generate a consensus sequence, which spans the region corresponding to amino acids 1–253 on Zm-HIR1. The consensus sequence at the bottom was based on conservation of a residue at any given position in >80% of sequences. Amino acids conserved 80% or more are shown in uppercase red letters, and two 100% conserved amino acids (Asp and Ala) are shown in uppercase andred-underlined letters. The abbreviations for amino acid structural groups are in lowercase letters as follows: o, alcohol (Ser, Thr); l, aliphatic (Ile, Leu, Val); a, aromatic (Phe, Trp, Tyr);c, charged (Asp, Glu, His, Lys, Arg); h, hydrophobic (Ala, Cys, Phe, Gly, His, Ile, Lys, Leu); −, negative (Asp, Glu); p, polar (Cys, Asp, Glu, His, Lys, Asn, Gln, Arg, Ser, Thr); +, positive (His, Lys, Arg); s, small (Ala, Ser, Thr, Val); u, tiny (Ala, Gly, Ser); t, turn-like (Ala, Cys, Asp, Glu, Gly, His, Lys, Asn, Gln, Arg, Ser, Thr); and dot, any residue or gap. The location of the PID superfamily regular expression is identified by arrows. The location of the stomatin PROSITE signature is similarly indicated. Shown below the consensus sequence is the secondary structure predictions that were carried out using the latest version of the DSC algorithm (44Ross K.D. Sternberg M.J.E. Protein Sci. 1996; 5: 2298-2310Crossref PubMed Scopus (399) Google Scholar), the prediction accuracy of which is >72%: DSC_ALL, DSC_HIR, DSC_PHB, DSC_STM, and DSC_HFLK represent consensus predictions for all the PID superfamily sequences and all the members of the respective families of HIR proteins, prohibitins, stomatins, and HFLK proteins, respectively. C, coil; E, β-strands;H, helix; dot, gaps.View Large Image Figure ViewerDownload Hi-res image Download (PPT) A systematic search for conserved motifs among the aligned sequences was performed using the MEME algorithm. The MEME motifs have been indicated as reliable indicators of family membership (26Grundy W.N. Bailey T.L. Elkan C.P. Baker M.E. Biochem. Biophys. Res. Commun. 1997; 231: 760-766Crossref PubMed Scopus (33) Google Scholar). The search resulted in the identification of three conserved motifs (Fig. 2). Using Zm-HIR1 as reference again, the amino acid positions of these motifs are as follows: motif 1, 108–167; motif 2, 56–80; and motif 3, 23–88. The relative spatial positions of these three motifs in all these genes appear to be spatially well conserved, indicating the possibility for a similar structural orientation in three-dimensional space. Motif 2 is a subset of motif 3 and is conserved in all members in the alignment. All three motifs are present in all members of the superfamily, except the HFLK/HFLC proteins, which contain only motif 2. HFLK and HFLC are bacterial membrane proteins with protease activity and are involved in lysogenization. They appea
Referência(s)