HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 28, 21385-21395, July 14, 2000

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 275/28/21385    most recent C000159200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stabach, P. R. Articles by Morrow, J. S. Search for Related Content PubMed PubMed Citation Articles by Stabach, P. R. Articles by Morrow, J. S. Social Bookmarking                      What's this?

Identification and Characterization of V Spectrin, a Mammalian Ortholog of Drosophila _H Spectrin^*,

Paul R. Stabach and Jon S. Morrow

From the Department of Pathology and the Department of Molecular, Cellular, and Developmental Biology, Yale University, New Haven, Connecticut 06510

Received for publication, March 13, 2000, and in revised form, March 31, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Four mammalian -spectrin genes are currently recognized, all encode proteins of 240-280,000 M_r and display 17 triple helical homologous 106-residue repeat units. In Drosophila and Caenorhabditis elegans, a variant  spectrin with unusual properties has been recognized. Termed  heavy (_H), this spectrin contains 30 spectrin repeats, has a molecular weight in excess of 400,000, and associates with the apical domain of polarized epithelia. We have cloned and characterized from a human retina cDNA library a mammalian ortholog of Drosophila _H spectrin, and in accord with standard spectrin naming conventions we term this new mammalian spectrin  5 (V). The gene for human V spectrin (HUBSPECV) is on chromosome 15q21. The 11,722-nucleotide cDNA of V spectrin is generated from 68 exons and is predicted to encode a protein with a molecular weight of 416,960. Like its fly counterpart, the derived amino acid sequence of this unusual mammalian spectrin displays 30 spectrin repeats, a modestly conserved actin-binding domain, a conserved membrane association domain 1, a conserved self-association domain, and a pleckstrin homology domain near its COOH terminus. Its putative ankyrin-binding domain is poorly conserved and may be inactive. These structural features suggest that V spectrin is likely to form heterodimers and oligomers with  spectrin and to interact directly with cellular membranes. Unlike its Drosophila ortholog, V spectrin does not contain an SH3 domain but displays in repeat 5 a 45-residue insertion that displays 42% identity to amino acids 85-115 of the E4 protein of type 75 human papilloma virus. Human V spectrin is expressed at low levels in many tissues. By indirect immunofluorescence, it is detected prominently in the outer segments of photoreceptor rods and cones and in the basolateral membrane and cytosol of gastric epithelial cells. Unlike its Drosophila ortholog, a distinct apical distribution of V spectrin is inapparent in the epithelial cell populations examined, although it is confined to the outer segments of photoreceptor cells. The complete cDNA sequence of human V spectrin is available from GenBank^TM as accession number AF233523.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

First identified in association with the erythrocyte plasma membrane, spectrin is now appreciated as the central component in a ubiquitous and complex system linking membrane proteins, membrane lipids, and cytosolic factors with the major cytoskeletal filament systems of the cell. The functional unit of spectrin is typically a heterodimer composed of - and -type subunits joined noncovalently. Two mammalian -spectrin genes are recognized, encoding I and II spectrin; four mammalian  spectrin genes have been described, encoding I-IV spectrin. Combinations of different - and -spectrins, together with diversity generated by alternative mRNA splicing, correlate with the localization of different spectrins to different cells and tissues, as well their segregation to specific intracellular compartments, including localized plasma membrane domains, the Golgi, and other organelles (for reviews and relevant references see Refs. 1-3).

Despite this diversity, an enigma has been the apparent lack in mammals of an unusual spectrin first identified in Drosophila (4) and later in Caenorhabditis elegans (5). This spectrin, termed  heavy (_H), is unique in that it contains 30 spectrin repeat units rather than the 17 repeats characteristic of other  spectrins, may lack ankyrin binding capacity, and contains a SH3 domain. Like other  spectrins, it associates with  spectrin, binds actin, displays an 106-residue repeat unit size, and contains a pleckstrin homology domain near its COOH terminus. These features mark it as a member of the spectrin family rather than as a more distant relative of spectrin such as dystrophin or -actinin. In the Drosophila egg, maternally loaded _H spectrin is found in a uniform distribution along the plasma membrane (6). With the onset of cellularization and then gastrulation, _H spectrin redistributes to the apical-lateral furrows. In the adult fly, it persists as an apical protein and together with DE-cadherin contributes to the maintenance of adherens junctions and the apical terminal web in epithelial tissues (7). It is essential for normal development, with mutations in _H spectrin (karst locus) displaying pleiotropic larval phenotypes and frequent lethality. The few individuals that survive lack photoreceptor R7 and display bent wings, tracheal permeability defects, and other aberrations (7). These findings have suggested a role for _H spectrin in maintaining apical polarity, cell-cell contact, and cell signaling, although it does not per se appear to be required for the formation of simple apical versus basolateral polarity (7).

Similar conclusions have been noted in C. elegans, where the _H spectrin homolog, sma-1, resides on chromosome V. In sma-1 mutants, the extent of embryonic elongation is decreased, apparently because of impaired contraction of the actin skeleton in the epidermal cells enveloping the early embryo. Subsequently, sma-1 is expressed in the developing pharynx, intestine, and excretory cells, where it is postulated to participate in the formation of their apical domain (5). It also appears likely that a _H-like spectrin, albeit of lower molecular weight, might exist in avian enterocytes, because their brush border contains a variant spectrin (TW260/240) with an unusually large  subunit (8). Although this protein has never been cloned or sequenced, its apical location, size, and lack of ankyrin binding capacity (9) suggest that it differs from other  spectrins, features suggestive of an avian homolog of Drosophila _H. Paradoxically and unlike the situation in flies, nematodes, and avians, the terminal web of mammalian enterocytes is rich in the more generally expressed IIII spectrin (10).

We now report the identification and characterization of a full-length human ortholog of Drosophila _H spectrin. Cloned from a human retinal cDNA library and representing the largest mammalian spectrin gene yet identified, the human protein shares considerable overall structural and functional similarity to Drosophila _H. The mammalian protein also contains a unique sequence insertion homologous to a portion of the E4 protein of human papilloma virus type 75. Expressed most abundantly in cerebellum but also in other tissues and especially the retina and gastric epithelium, we anticipate that this unusual spectrin may play a role in organizing the actin skeleton associated with photoreceptor discs in the outer segment of the retina and the lateral and internal membranes of certain epithelial cells. In accord with standard spectrin naming conventions (1) and recognizing the recent identification of mammalian IV spectrin,¹ we term this new mammalian protein V spectrin.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cloning and Sequence Analysis of V Spectrin cDNA-- Unless otherwise stated, all molecular biological procedures followed standard methods (12). Candidate sequences were derived from GenBank^TM by BLAST searching using conserved sequences from human II spectrin. Oligonucleotides were designed to amplify and sequence the entire V spectrin cDNA from a CLONTECH Human Retina Marathon Ready cDNA library using platinum Taq high fidelity polymerase from Life Technologies, Inc., following the manufacturer's protocols. The oligonucleotides used are listed in Fig. 1 and Table I. All were designed with high predicted melting temperatures to suppress false priming. Primary PCR² reactions started with 2.5 µl of template DNA and contained 60 mM Tris-SO₄, pH 8.9, 18 mM ammonium sulfate, 5 mM MgSO₄, 1 mM each dNTP, 1 µM of each primer, and 1.0 unit of platinum Taq high fidelity polymerase in a final volume of 25 µl. The reaction profile utilized 27 cycles of denaturation at 94 °C for 1 min followed by a 68 °C annealing and extension step for 1 min/kb of expected amplimer. 5 µl of primary reaction was then used as template in a 50-µl nested PCR reaction under the same conditions. After amplification, 30 µl of the PCR reaction was loaded onto a 1% NuSieve GTG Low Melt Agarose gel, and amplimers of the appropriate size were excised from the gel using a clean razor blade. Extracted DNA was gel purified using GFX PCR gel band purification kit (Amersham Pharmacia Biotech) and cloned into the TA-TOPO vector (Invitrogen). Automated DNA sequence analysis was performed by the Keck Laboratory, Yale University. Adapter primers provided with the Marathon Ready cDNA kit were used in conjunction with V gene-specific primers to perform 5' and 3' rapid amplification of cDNA ends according to the manufacturer's instructions and treated as above. To eliminate errors associated with amplification, multiple overlapping PCR reactions representing the entire length of V spectrin were cloned, sequenced, and compared with each other and the sequence of Chromosome 15 deposited from the Whitehead Institute/MIT Center for Genome Research. Most sequence analysis was performed using the software Gene Construction Kit (Textco, Inc.) and a portfolio of analysis tools from DNAStar Inc. The intron/exon boundaries were established by comparing the entire cDNA sequence against the sequence of chromosome 15, followed by heuristic inspection of the resulting pairwise comparison for canonical splice junctions.

Northern Blot-- Northern blot analysis of multiple human tissues were performed according to the instruction of the manufacturer, using their multiple human tissue Northern blot (catalog number D1809-08, lot number 8904019; Invitrogen), as well as a human multiple tissue expression array (CLONTECH, Palo Alto, CA). V Spectrin cDNA corresponding to nucleotides 5424-6012 was gel purified and random labeled with [³²P]dATP using EZ-Strip probe labeling kit from Ambion. The loading of mRNA was verified by probing -actin mRNA with the probe provided with the multiple human tissue Northern blot kit.

Antibody Production-- V Spectrin cDNA corresponding to amino acids 2059-2270 was subcloned into both pGEX-4T3 (Amersham Pharmacia Biotech) and pTRC-HIS-C (Invitrogen) and overexpressed in DH5 strain of Escherichia coli with the addition of 0.1 mM isopropyl-1-thio--D-galactopyranoside. Soluble bacterial proteins were removed with a 1% Triton X-100 extraction in 50 mM NaCl and 50 Tris, pH 8.0, following sonication. The insoluble glutathione S-transferase fusion peptide was analyzed by SDS/polyacrylamide gel electrophoresis and stained lightly with Coomassie Blue, and the recombinant proteins were sliced from the gel, emulsified, and injected subcutaneously into New Zealand White rabbits as before (13). Sera were affinity purified against the recombinant His-tagged proteins using Millipore Immobilon-P membrane blots (14). Affinity purified antibodies were eluted from the membranes with 100 mM glycine, pH 2.5, and neutralized with volume of 1 M Tris, pH 7.5. Alternatively, antibodies were generated using His₆ fusion peptides (amino acids 1732-2088), and the immune serum was affinity-purified against glutathione S-transferase fusion proteins containing the same amino acids. Western blots demonstrating the specificity and activity of these antibodies are available as a supplementary figure in the on-line version of this manuscript.

Immunofluorescence-- Fresh rat tissues were quick frozen in OCT^TM embedding compound in isopentane immersed in liquid N₂. After frozen sectioning, 5-µm slices were fixed on glass slides for up to 2 h in ice-cold acetone. Fixed sections were rehydrated in cold PBS and blocked with 1% bovine serum albumin (w/v) in PBS for 1 h. Primary antibodies were applied overnight in a humidified chamber followed by PBS rinse. Cy3-labeled secondary goat anti-rabbit antibodies were then applied for 2 h. After washing, slides were mounted with glass coverslips and visualized by epifluorescence using an Olympus AX-70 microscope. Alternatively, human tissues were obtained at autopsy under protocols approved by the Yale Human Investigation committee (numbers 2422 and 3388) and fixed in buffered formalin. Fresh tissues from Harlan Sprague-Dawley rats were harvested into phosphate-buffered formalin. These human or rat tissues were fixed for 4-12 h and then embedded into paraffin and sectioned as before (15). Slides prepared in this way were suitable for immunostaining with the protocols used for frozen sectioning, provided that the deparaffinized sections were autoclaved for 5-7 min in 6.5 mM sodium citrate, pH 6.0 (15).

Other-- For Western blotting, cells or tissues were lysed in lysis buffer (2% SDS in PBS plus protease inhibitors) and separated by SDS/polyacrylamide gel electrophoresis. After transfer to polyvinylidene difluoride membrane, proteins of interest were detected with affinity purified antibodies. Protein determinations were carried out using the Pierce BCA method, as described in the Pierce Catalog and Handbook (16).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Identification of V Spectrin-- V spectrin was identified in silico by pairwise comparison of II spectrin amino acid sequences with the High Throughput Genomic Sequence data base using TBLASTN, a program that dynamically translates the nucleotide sequence into all six reading frames and compares them against a protein query sequence (17). A match was made with GenBank^TM accession number AC009877, a sequence deposited from the Whitehead Institute of the MIT Center for Genome Research. This sequence consisted of 19 unordered contigs of 166,613 base pairs located on chromosome 15. A BLAST search using these contigs identified 11 direct matches from the nonredundant data base of expressed sequence tags. Of these matches, two were derived from libraries of pooled human tissues (GenBank^TM accession numbers AI743728 and AI803545), one was from a Gessler Wilms tumor library (accession number AA599654), two were from human fetal heart (accession numbers W95544, and W95287), and the remaining six were from human retina libraries (accession numbers AA020814, R84914, H87737, AA021476, R85945, and W96166). Sequences at both the 5' and 3' ends were extended by rapid amplification of cDNA ends using a Marathon-Ready cDNA library from human retina (CLONTECH). Oligonucleotide primers flanking selected sites within the putative sequence were selected so as to cover the entire length of the expressed gene, and PCR was used to amplify the requisite sequences from a human retina cDNA library (CLONTECH). The oligonucleotides used to span the sequence of V spectrin are summarized in Fig. 1 and in Table I. All reaction products were sequenced in both directions from at least two independent amplification reactions. Where discrepancies were identified between reactions or with the available expressed sequence tag or genomic sequences (see below), additional reactions and sequencing were carried out to assure the fidelity of the reported sequence. The complete cDNA sequence of V spectrin derived from this analysis is available from GenBank^TM as accession number AF233523. The full-length cDNA comprises 11,722 nucleotides and includes 5'- and 3'-untranslated regions of 288 and 469 nucleotides, respectively. Not shown is the polyadenylation sequence at the 3' end, which is present on clones abutting this end of the gene. The open coding region consists of 11,025 nucleotides, predicting a 3,675-amino acid, 416,960-kDa protein (Fig. 2). A satisfactory Kozak initiation sequence is present immediately upstream of the first ATG, and an in-frame stop codon occurs at nucleotide 88. We are thus confident that the first ATG beginning at nucleotide 229 is the initiator methionine of the derived protein.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of the V spectrin cDNA and the locus of oligonucleotides used to amplify the complete sequence. The specific sequences corresponding to each of these oligonucleotides are given in Table I. All sequences used in the determination of the complete sequence were derived from a human retinal cDNA library. All sequences were confirmed by bidirectional sequencing from a minimum of two independent PCR amplifications. The cDNA sequence of human V spectrin is available from GenBankTM as accession number AF233523.

View this table: [in this window] [in a new window]   Table I Oligonucleotides used to identify human V spectrin

View larger version (91K): [in this window] [in a new window]   Fig. 2.   The complete derived amino acid sequence of human V spectrin. The predicted protein product of the HUBSPECV gene encompasses 3,675 residues and has a predicted molecular weight of 416,960. Like other spectrins, it displays three distinct regions. Region 1 contains a modestly conserved canonical actin-binding sequence (bold type). Region 2 displays 29 full 106-residue spectrin repeat units and a partial 30th repeat. Clustal alignment (MegalignTM; DNAStar, Inc.) of each repeat reveals an overall strong conservation of the putative triple helical motif. The shaded areas represent the approximate extent of these helical regions, based on a comparison with the three-dimensional structure of Drosophila  spectrin as shown at the bottom (46). Based on the structure of a two-repeat unit, the true extent of the A helix may be somewhat less than that predicted from the Drosophila structure (43). Within this repeating domain, nonhomologous inserts occur in repeat unit 1 and in repeat unit 5 (see Fig. 4). Like many other -spectrins, region 3 contains a pleckstrin homology domain (bold type).

The derived protein sequence of V spectrin displays many unusual features that define it both as a -spectrin, as well as a mammalian ortholog of Drosophila _H spectrin. A characteristic feature of all spectrins is the presence of an 106-amino acid repetitive unit composed of three -helices; typically within this unit there is strong conservation of tryptophane followed by a hydrophobic residue at positions 17 and 18. Leucines are also typically found at positions 94 and 104. In vitro studies suggest that these residues are needed for maximal conformational stability (18), although they may be absent in regions of specialized function without disrupting the 106-residue phasing. Nonhomologous sequences that bestow special functionality may also appear within a repeat unit, such as the SH3 domain or sequences bestowing calmodulin binding activity within helix C of the 10th repeat of II spectrin (19, 20). Beginning at codon 305, the sequence of V spectrin displays 29 complete spectrin repeats and a partial 30th repeat unit (Fig. 2). Like other  spectrins, the first repeat is 120 amino acids in length and contains two short stretches of nonhomologous sequence to which have been attributed a role in membrane binding (21, 22) and heterodimer nucleation (23, 24). Repeats 2, 3, 7, 9, 11, 15, 17, 20, 21, and 28 lack the conserved tryptophane at position 17 but are otherwise conventional. All repeats align with good fidelity to the 106-residue repeat length, and all are anticipated to form triple helical structures (Fig. 2). Other predicted features of this protein along with its composite composition are shown in Fig. 3.

View larger version (47K): [in this window] [in a new window]   Fig. 3.   Composition and predicted biophysical features of V spectrin. The protein is predicted to have an isoelectric point of 6.5, and its composition is conventional. These analyses were carried out using the program ProteanTM.

In _H spectrin, and unlike other  spectrins, a SH3 domain sequence appears in repeat 5, at a position corresponding to the junction between helices B and C. V spectrin lacks a SH3 domain, but in this same region of repeat 5 displays a novel 45-residue insert (codons 816-860). This inserted sequence is unique among animal proteins in GenBank^TM. It is proline-rich and contains a single PXXP motif generally required for binding to SH3 domains (25). Surprisingly, it shares 42% identity and 50% similarity to residues 85-115 of the human papilloma virus type 75 E4 protein and shows a similar degree of homology to a putative equine herpesvirus protein (Fig. 4A).

View larger version (51K): [in this window] [in a new window]   Fig. 4.   V spectrin contains novel sequences homologous to viral proteins. A, repeat 5 of V spectrin contains a nonhomologous 45-amino acid insertion that intersects the junction between the putative B and C helices (cf. Fig. 2). A BLASTp search of this sequence against GenBankTM reveals no homology to other animal proteins, but strong homology to residues 85-115 of the E4 protein of human papilloma virus type 75, and a similar homology to residues 171-199 of a presumed protein 23 from equine herpesvirus 2. The degree of identity of these sequences to this region of V spectrin is presented. The degree of similarity of both to spectrin is 50%. B, alignment of the ankyrin-binding domains of all recognized -spectrins. The residues whose deletion abrogates ankyrin binding activity in I spectrin are designated by the solid bar; residues that diverge from the -spectrin consensus and that are thought to play a role in ankyrin binding are designated by the dashed bar (26). The 14-15 repeat of V spectrin is the region most similar to the 14-15 repeats of all of the other spectrins, as detected by either clustal analysis or the method of Jotun Hein using the program Megalign (DNAStar, Inc.). Note the marked divergence of V spectrin from the consensus specified by the other spectrins.

Other  spectrins possess an ankyrin-binding domain in repeat 14-15 (26). This motif is characterized by a well conserved sequence leading into repeat 15, replacement of tryptophane at position 17, and an 43-residue region of poor homology (to other repeats) within the putative B helix. A comparison of all known  spectrins, three of which have documented ankyrin binding activity (I, II, and III), reveals strong preservation of these features in  spectrins I to IV (Fig. 4B). However, this strong degree of conservation does not extend to the 14-15 repeat or any other repeat, of V spectrin. Thus, although the 14-15 repeats of V spectrin are the repeats most similar to the 14-15 repeat units of the other spectrins, the level of divergence in V spectrin in this region is such to suggest that this spectrin may not bind to ankyrin.

Finally, the last repeat unit of V spectrin, like the last repeat of other  spectrins, is only a partial repeat. Such partial repeat units, consisting of helices A and B, are characteristic of the heterodimer self-association site that mediates the assembly of the spectrin heterodimer to form the tetrameric unit (27). The presence of this site in V spectrin strongly suggests a role for V spectrin in forming mixed heterotetramers and oligomers with an  spectrin. Beyond the central repeats domain (region 2), in region 1 of V spectrin there is a putative actin-binding domain (codons 134-157) that is 58% similar to the canonical actin-binding sequence found in other spectrins and in many actin-binding proteins (28). In region 3, a pleckstrin homology domain is found (codons 3533-3641) (Fig. 2).

V Spectrin Is Most Homologous to Drosophila _H Spectrin and Defines a New Subset of Mammalian Spectrins-- A comparison of the sequence of V spectrin with other human spectrins, as well as with Drosophila _H spectrin, reveals it to be overall more similar to _H than to any other mammalian spectrin (Fig. 3A). When each repeat of region 2 is individually compared with the other mammalian spectrins, the overall degree of homology is on the order of 30% (25-42%) with about equal representation between the  versus  spectrins. These similarities are greater than they are to other spectrin-related proteins, such as dystrophin or -actinin. Homologies over regions 1 and 3 are slightly better (52 and 30% respectively), probably reflecting the functional specialization of these regions. A dendrogram of the relationships between these spectrins and a similarity matrix is presented in Fig. 5.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   The relatedness of the different  spectrins. A, dendrogram of the relationships between members of the human -spectrin family, constructed using the CLUSTAL procedure (11), which allows estimation of relatedness and provides an approximation to an evolutionary tree. Because this program constructs the dendrogram on the basis of pairwise matches and the formation of consensus sequences, it does not technically provide an evolutionary tree. It results in a dendrogram that shows relatedness as a function of length of each of the branches, where the length is proportional to sequence distance. B, overall level of sequence homology and divergence within the  spectrin gene family. The similarities and divergences shown are over full-length sequences. C, dendrogram of the relationship of V spectrin to the Drosophila H and the sma-1 protein of C. elegans.

The V Spectrin Gene on Chromosome 15 q21 Spans 47,420 Nucleotides and Includes 68 Exons-- A Blast search of the human genome data base using V sequences identified the V spectrin gene, which we term HUBSPECV, on chromosome 15. Two sequence tagged markers fall within this gene, G42451 and G42463, which together with the known locus of the flanking gene for cytosolic phospholipase A2  (cPLA2 , accession number AF065216) (29) allow a refinement of its locus to 15q21. Comparison of the genomic sequence against the cDNA sequence reveals that V spectrin is generated from 68 exons. Nearly all of these exons are flanked by canonical donor/acceptor spice sequences; these and the genomic structure of HUSPECV are summarized in Fig. 6. Based on the identified expressed sequence tag sequences and the sequences derived directly by PCR amplification from retinal and cerebellum cDNA libraries, it appears that the first exon may be alternatively spliced, yielding at least one alternative 5' transcript of V spectrin. A second interesting feature to emerge from the structure of the V gene is its close proximity to the gene immediately downstream, cPLA2 . No nucleotides separate the 3' end of the transcribed sequences of V spectrin from the 3' end of the transcribed gene for cPLA2 . The immediate adjacency of these two genes is supported by both the absence of identified expressed sequence tags that bridge this region, as well as by direct amplification across this region of genomic DNA by PCR, which confirms the absence of intervening sequences (data not shown).

View larger version (99K): [in this window] [in a new window]   Fig. 6.   The genomic organization of HUSPECV. Top panel, graphical representation of the V genomic DNA. Thick bars represent the 68 exons of this gene. The positions of the translation initiation start and stop sites are as indicated in the 2nd and 68th exons, respectively (arrows). The arrowhead indicates a break of undetermined distance in the intronic sequence. Open boxes represent the 3' end of the cytosolic phospholipase A2  gene (cPLA2 ); an arrow also marks the stop codon of this gene, which directly abuts the 3' end of HUSPECV. Bottom panel, summary of the flanking sequences (lowercase letters) and the intron-exon boundaries for each of the exons.

V spectrin Is Expressed in a Subset of Tissues and Localizes to the Apical Outer Segment Domain of Photoreceptor Cells-- A dot-blot hybridization analysis against a variety of human tissues revealed that V spectrin is expressed at very low levels in many tissues (Fig. 7A and Table II). The strongest signals were identified in cerebellum, spinal cord, stomach, pituitary gland, liver, pancreas, salivary gland, kidney, bladder, and heart, with lesser amounts in most but not all other tissues. Northern blot analysis of brain and lung revealed single significant V transcripts at 11-12 kb in cerebellum (Fig. 7B). No smaller transcripts suggestive of alternative splicing were identified.

View larger version (45K): [in this window] [in a new window]   Fig. 7.   V spectrin is expressed in a subset of tissues by Northern analysis. A, Northern dot-blot of a variety of human tissues and cell lines, probed for V spectrin. These results are summarized in Table II. Although negligible or trace levels of expression were detected in many tissues, strong hybridization was observed in cerebellum, stomach, heart, nerve, and other tissues. B, Northern blot of electrophoretically separated human poly(A)+ mRNAs (2 mg) from various adult human brain regions probed for V spectrin. The same blot was hybridized for actin as a control for RNA loading. The positions of standards are as indicated. The V spectrin mRNA in the cerebellum migrates at 11.3 kb.

View this table: [in this window] [in a new window]   Table II Tissue expression of V spectrin

The intracellular distribution of V spectrin was examined by indirect immunofluorescence in a subset of tissues, selected on the basis of their levels of V spectrin expression. Surprisingly, sections of rat cerebellum and spinal cord failed to display a clearly discernible pattern of staining. However, in retina, the tissue from which it was cloned, V spectrin was heavily concentrated in the apical outer segments of both the rod and cone photoreceptor cells (Fig. 8). Additional V spectrin was also detected in a patchy distribution near the center of the outer plexiform layer. This is a region where photoreceptors synapse with the bipolar and horizontal cells of the inner nuclear layer. It is unclear whether the V spectrin in this region is associated with the presynaptic terminus of the photoreceptor cell or post-synaptic structures in the apical dendrites of cells from the outer plexiform layer.

View larger version (47K): [in this window] [in a new window]   Fig. 8.   Distribution of V spectrin in the human retina. Because V spectrin was cloned from a human retinal cDNA library, its distribution in this tissue was of particular interest. Paraffin embedded sections of human retina were immunostained with affinity purified anti-V spectrin antibody (red). Nuclei were stained with DAPI (blue). An adjacent section was stained with hematoxylin and eosin. The merged fluorescent image is shown at top, and an enlarged view of the V distribution is at the bottom. Note that V spectrin is confined to the apical outer segments of photoreceptors (both rods and cones) and to a punctate region in the middle of the outer plexiform layer. Preimmune sera yielded negligible staining (not shown).

A second tissue examined that expressed significant levels of V spectrin was stomach. A section of stomach taken from the rat gastroesophageal junction and fundus is presented in Fig. 9. Although the esophagus is negative, the lateral and basal margins of the gastric mucin and parietal cells stain prominently. There is also cytoplasmic staining in the parietal cells. Surprisingly, there is little apical staining of these cells, suggesting that unlike the role proposed for Drosophila H spectrin, V spectrin may not function to organize the apical domain of these epithelial cells.

View larger version (111K): [in this window] [in a new window]   Fig. 9.   Distribution of V spectrin in the rat esophagus and stomach. Another tissue with significant V spectrin expression is stomach. Paraffin-embedded sections of rat stomach were prepared and immunostained as in Fig. 8. A, the distribution of V spectrin. B, the distribution of nuclei in the section, stained with DAPI (blue). C, merged image, demonstrating that V spectrin is present only in the gastric epithelium but not in the rat esophagus. D, adjacent section stained with hematoxylin and eosin, revealing the histology of the area. The gastro-esophageal junction is marked with an asterisk. sto, gastric epithelium; eso, the esophagus. E, higher power view of gastric glands from the fundus, immunostained with anti-V spectrin antibody. Note the distribution of V spectrin along the lateral cell borders of the mucin cells and along the borders with some cytoplasmic staining of parietal cells. There is minimal accumulation in the apical domain of any cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The studies presented here identify a novel member of the mammalian spectrin family, an ortholog of an unusual spectrin first identified in Drosophila. The distinguishing features of this spectrin include: 1) its size (417 kDa); 2) the presence of 30 spectrin repeat units; 3) conserved actin-binding, self-association, and pleckstrin homology domain motifs; 4) the presence of a putative novel protein-protein interaction domain with homology to two viral proteins; 5) localization to the outer segments of rod and cone photoreceptor cells and along the lateral membranes of gastric mucous cells and the lateral membrane and cytosol of gastric parietal cells; and, 6) an overall level of sequence homology that places it approximately midway evolutionarly between the  and  spectrins.

Given its sequence divergence, it might be argued that V spectrin in fact represents a new member of the spectrin gene superfamily, dissimilar enough to deserve a new category. We do not favor this interpretation because it would obscure the important similarities between V spectrin and the other  spectrin family members. These include its likely association with a paired  subunit; modest conservation of its actin-binding domain and membrane association domain 1; conservation of its heterodimer association site (needed to form spectrin tetramers and larger oligomers 27); and the conservation of its pleckstrin homology domain.

The identification of V spectrin in photoreceptor cells complements earlier work that has collectively identified several other putative components of a spectrin-ankyrin-actin skeleton in such cells. These include an extensive filamentous actin network that extends from the outer segments to the external limiting membrane, a network in close proximity to protein 4.1 isoforms in retinal cones (30); ankyrin isoforms of 190 and 210 kDa (31); and based on immunologic criteria at least two other forms of spectrin, presumably IIII (240/235) and II2 (240/235E) (32). Significantly, all of the ankyrins and spectrins previously identified in photoreceptor cells are confined to the cell body and axons and are not present in photoreceptor outer segments. Thus, V spectrin represents the first and only spectrin identified within this specialized apical domain. The other site of V localization in the retina was in the outer plexiform layer, presumably in the synapses of this region. Retinal dystrophin has also been localized to this region, and by immunoelectron microscopy is confined to the post-synaptic density (33). Because these studies did not report immunostaining in the outer segment, it is unlikely that they were detecting V spectrin. It will thus be of interest in future work to determine whether dystrophin and V spectrin are co-localized in the retina or whether they are confined to opposite sides of the synaptic junction.

A search of the Online Medelian Inheritance in Man data base for possible associations with heritable disorders reveals several interesting possibilities but no compelling candidates. The approximate locus of V spectrin is 15q21. A variant of autosomal dominant spinocerebellar ataxia type III has been linked to marker D15S1039, which maps to a 7.6-centimorgan interval encompassing 15q14-q21.3 (34). This condition generally develops late in life and is characterized by a slowly progressive loss of coordination. The abundance of V in the cerebellum and the recognized ataxia that develops in mice lacking ankyrin G (35), another component of the prototypical spectrin membrane skeleton, are interesting correlations. However, vision disturbances or other abnormalities are only rarely found in this heterogeneous group of disorders (36), features that might be expected to more commonly occur in a V spectrin disorder, given its tissue distribution. A second interesting candidate is recessive familial amyotrophic lateral sclerosis type 1. This disorder presents at an early age of onset with a slowly progressive ataxia leading to paralysis. It has been mapped to 5q15.1-q21.1 (37). The abundance of V in the spinal cord and other neuronal tissues might offer a role for V in this disorder, but again the lack of apparent phenotype in other tissues where V is found reduces the likelihood of this linkage. Other possibilities include primary autosomal microcephaly (38) and autosomal recessive sensorineural hearing loss (39, 40). Observations related to these conditions include the shortening of body size in C. elegans lacking sma-1 (perhaps a form of microcephaly) and the putative role of spectrin in maintaining the stiffness and function of outer hair cells (41).

A final consideration is the role of V spectrin. It is now well recognized that spectrins may contribute to the establishment and maintenance of membrane order in a variety of membrane compartments, including intracellular organelles and at the plasma membrane. One concept of how they might function is by linking mosaics of integral and cytosolic proteins, collected as a consequence of the polyfunctonality of spectrin, to actin and the other filament systems. By this "linked mosaic" model of spectrin function (42), membrane microdomains or membrane-bound transport containers might be organized, fixed to other cytoskeletal systems, or tethered to the motors of intracellular transport. An initial appraisal of the intracellular disposition of V spectrin suggests that it distributes between the cytosol and the plasma membrane. In gastric epithelial cells, it is concentrated along the lateral margins of the cell, although cytosolic concentrations are present. These features distinguish it from the apical distribution reported for Drosophila _H spectrin but are similar to the embryonic pattern of _H observed during cellularization. One attractive hypothesis is that V spectrin interacts with adherens junctions situated along the borders of epithelial cell-cell contacts, as does _H spectrin, and organizes and perhaps stiffens the basolateral domain of cells. Gastric parietal cells also contain an elaborate intracellular canalicular membrane network that effects gastric acid secretion; perhaps an extensive V spectrin network is needed to support and organize this internal membrane system. In this regard, gastric epithelial cells can be considered similar to the outer segments of photoreceptor cells, because outer segments also contain an elaborate system of internal membranes in the form of their photopigment-rich discs. Thus, one role for V spectrin might be to organize specialized internal membrane compartments and tether them to the plasma membrane and the cytoskeleton.

A feature of spectrin that might facilitate this task is the flexible nature of the -helical repeat unit (43), a property that would allow compliant and flexible coupling, albeit at a defined length, between membranes and actin or other cytoskeletal filaments. It is unclear why such a long spectrin molecule would be needed for this task, with exactly 30 repeat units. The multiple coiled coil repeats in both - and -spectrin, -actinin, and dystrophin have presumably arisen from a common ancestral gene (possibly -actinin) by duplication events that occurred prior to the separation of vertebrates and invertebrates 545 million years ago. (44, 45). The fact that _H spectrin and V spectrin each have independently maintained a 30-repeat length throughout evolution suggests a crucial role for either a longer actin-membrane cross-linker or the need for greater extensible flexibility than can be provided by the other smaller conventional spectrins.

	ACKNOWLEDGEMENTS

We thank Drs. James Barbeau, Carol Cianci, David Rimm, John Sinard, Soojung Je, Mike Stankewich, Tom D'Aquila, and Phillip Davis for advice and assistance with the procurement and preparation of tissues for immunostaining and analysis and for assistance with other aspects of this study. Dr. M. Solimena is thanked for providing unpublished information on IV spectrin.

	FOOTNOTES

^* This work was supported by grants from the National Institutes of Health (to J. S. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF233523.

The on-line version of this article (available at http://www.jbc.com) contains a supplementary figure.

To whom correspondence should be addressed: Dept. of Pathology, Yale University, 310 Cedar St., New Haven, CT 06510. Tel.: 203-785-3624; Fax: 203-785-7037; E-mail: jon.morrow@yale.edu.

Published, JBC Papers in Press, April 11, 2000, DOI 10.1074/jbc.C000159200

¹ S. Berghs, D. Aggujaro, R. Dirkx, Jr., E. Maksimova, P. R. Stabach, J.-M. Hermel, J.-P. Zhang, W. Philbrick, V. Slepnev, and M. Solimena, submitted for publication.

	ABBREVIATIONS

The abbreviations used are: PCR, polymerase chain reaction; PBS, phosphate-buffered saline; contig, group of overlapping clones.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Das, C. Base, S. Dhulipala, and R. R. Dubreuil Spectrin functions upstream of ankyrin in a spectrin cytoskeleton assembly pathway J. Cell Biol., October 23, 2006; 175(2): 325 - 335. [Abstract] [Full Text] [PDF]

				M. C. Stankewich, P. R. Stabach, and J. S. Morrow Human Sec31B: a family of new mammalian orthologues of yeast Sec31p that associate with the COPII coat. J. Cell Sci., March 1, 2006; 119(Pt 5): 958 - 969. [Abstract] [Full Text] [PDF]