- Email: [email protected]
Characterization of the murine Inpp4b gene and identification of a novel isoform
Gene 376 (2006) 152 – 161 www.elsevier.com/locate/gene
Characterization of the murine Inpp4b gene and identification of a novel isoform☆ Mathieu Ferron
a,b
, Jean Vacher
a,b,c,⁎
a
b
Institut de recherches cliniques de Montreal, 110 av. des Pins O., Montreal, Qc, Canada H2W 1R7 Programme de biologie moléculaire, Faculté des études supérieures de l'Université de Montréal, Canada c Faculté de Médecine de l'Université de Montréal, Canada Received 20 December 2005; received in revised form 14 February 2006; accepted 21 February 2006 Available online 17 March 2006 Received by M. D'Urso
Abstract Inositol polyphosphate phosphatases and phosphoinositides second messengers have been associated with major cellular functions as growth, differentiation, apoptosis, protein trafficking and motility. To characterize the role of inositol phosphatases in cell physiology, we have isolated the mouse Inositol polyphosphate 4-phosphatase type II (Inpp4b) cDNA. The murine Inpp4b locus was mapped on chromosome 8 in a synthenic region of the human 4q27-31 interval between Il-15 and Usp38. The mouse Inpp4b proteins, α and β isoforms, encoded by this locus contained 927 and 941 amino acids respectively with a consensus phosphatase catalytic site and a conserved C2 domain that are highly similar with the human and rat homologues. Interestingly, we characterized a novel shorter isoform of Inpp4bα resulting from an alternative translation initiation site and exon 5 skipping. Inpp4b C2 domain interacted with preferential affinity to phosphatidic acid and phosphatidylinositol 3,4,5-triphosphate (PI(3,4,5)P3) lipids. While analysis of Inpp4b transcript and protein expression demonstrated a broad tissue distribution for the α isoform, as for the paralogue Inpp4aα and β isoforms, it also displayed a limited hematopoietic lineage distribution whereas the Inpp4bβ isoform had a highly restricted pattern. Importantly, the Inpp4bβ localized to the Golgi apparatus whereas Inpp4bα was mainly cytosolic, suggesting a different cellular function for this isoform. Together our characterization of the murine Inpp4b gene expression pattern, cellular sublocalization and interacting lipids support highly specific function for individual Inpp4 phosphatase proteins. © 2006 Elsevier B.V. All rights reserved. Keywords: Inositol phosphatase; Lipids binding; Golgi; C2 domain
1. Introduction Phosphatidylinositol lipids or phosphoinositides are implicated in major cellular mechanisms including cell growth, differentiation, apoptosis, protein trafficking and motility (CantAbbreviations: bp, base-pair; cDNA, complementary DNA; EGFP, enhanced green fluorescence protein; EST, expressed sequence tag; GST, glutathione S-transferase; kb, kilobase; kDa, kiloDalton; Mb, megabase; PCR, polymerase chain reaction; RT, reverse transcription; TBS-T, Tris buffer saline-tween. ☆ GenBank accession numbers: Mouse Inpp4b alpha cDNA, DQ296611; mouse Inpp4b alpha short cDNA, DQ296612; mouse Inpp4b beta cDNA, DQ296613. ⁎ Corresponding author. Institut de recherches cliniques de Montreal, 110 av. des Pins O., Montreal, Qc, Canada H2W 1R7. Tel.: +1 514 987 5734; fax: +1 514 987 5585. E-mail address: [email protected] (J. Vacher). 0378-1119/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2006.02.022
ley, 2002). Acting as second messengers, their intracellular concentration levels are tightly regulated by the opposite activities of phosphoinositide kinases and phosphatases (De Matteis and Godi, 2004). Phosphoinositides can be phosphorylated at three distinct positions, D-3, D-4 and D-5 of the inositol head group, generating 7 unique derivatives. One of the critical and most studied phosphoinositide kinases, the phosphatidylinositol 3-kinase (PI3-Kinase), phosphorylates specifically position D-3. This PI3-Kinase upon activation by recruitment at the membrane is responsible for the production of phosphatidylinositol 3,4,5-triphosphate (PI(3,4,5)P3) from the substrate phosphatidylinositol 4,5-biphosphate (PI(4,5)P2) (Cantley, 2002). Reversibly, newly synthesized PI(3,4,5)P3 can be rapidly dephosphorylated by specific inositol phosphatases (Majerus et al., 1999). PI(3,4,5)P3 can be dephosphorylated at the D-3 position by PTEN to
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
153
generate PI(4,5)P2 (Leslie and Downes, 2002) or at the D-5 position by SHIP1 or SHIP2 inositol polyphosphate 5-phosphatases that generate the PI(3,4)P2 second messenger (Ware et al., 1996). Dephosphorylation at the D-4 position on the inositol ring to generate PI(3)P occurs via activation of the inositol polyphosphate 4-phosphatase isoenzymes type I (Inpp4a) and type II (Inpp4b) (Norris et al., 1995, 1997). Two human and rat 4-phosphatase isoenzymes (type I or INPP4A, and type II or INPP4B) have been cloned, sharing 37% amino acid identity. Both isoenzymes contain a catalytic “CX5R” motif also found in protein tyrosine phosphatases. These phosphatases have dual specificity for protein and lipids. In addition, both these INPP4A and INPP4B isoenzymes can exist in two splice variants, α and β (Norris et al., 1995, 1997). While the INPP4Aα and INPP4Bα isoforms have a hydrophilic C-terminus region, the INPP4Aβ and INPP4Bβ isoforms have a hydrophobic C-terminus which contains a potential transmembrane domain. The murine Inpp4a and Inpp4b genes are not well characterized, but few studies on Inpp4a in mice and cell culture have been reported. In vivo, the mouse mutant weeble that corresponds to a null allele of Inpp4a, demonstrated a crucial role for Inpp4a phosphatase in neuronal survival (Nystuen et al., 2001). Inpp4a was also suggested to play a role in the megakaryocytic lineage as a negative regulator of cell growth acting downstream of the GATA-1 transcription factor (Vyas et al., 2000) and by forming a complex with PI3-Kinase (Munday et al., 1999). While Inpp4a cellular role in vitro was often associated with inhibition of a signaling pathway by dephosphorylation of the second messenger PI(3,4)P2 at the plasma membrane, recent studies suggested that an increase in production of PI(3)P pool by Inpp4a on endosomal membranes has a stimulatory effect on signaling (Ivetac et al., 2005; Shin et al., 2005). By contrast, no in vivo precise functional role has yet been assigned to Inpp4b. Interestingly however, Inpp4b expression was recently shown to be transcriptionally regulated by erythropoietin in erythroid preleukemic cells (Barnache et al., 2006). To gain insight into the cellular role of Inpp4b, Inpp4b cDNA was isolated, the locus mapped and the gene organization defined. Upon comparison to the human and rat homologues, the mouse Inpp4bα and β spliced isoforms demonstrated high conservation. Interestingly, we also identified and characterized a novel Inpp4bα isoform called Inpp4b alpha short (Inpp4bαs). The Inpp4b isoforms, Inpp4bα,αs,β showed distinct expression pattern, and subcellular localization, indicating specific functions for these isoforms.
and the EST BG145663 for the reverse primer 5′-GTGATGACATCATAGAGAGC-3′. This assay, when tested on mouse spleen cDNA, amplified a 669 bp cDNA specific fragment, and 6 different cDNA clones were sequenced. One of them, sP4D12 was found to contain the complete open reading frame of Inpp4b except for 250 pb in the 5′ region of the cDNA. Consequently, the 5′ half of the coding sequence was amplified by RT-PCR amplification on spleen cDNA with the forward primer 5′AACCGGAGCTCATTATTTGGTACCATTTG-3′ and the reverse primer 5′-CTCTGTGCTGACATTAGGAG-3′. The amplified product was digested by KpnI and SacI restriction enzymes to obtain a 1389 bp fragment (5′ half) and cloned with the 1507 bp SacI/BglI fragment (3′ half) obtained from sP4D12 clone in pBlueScript in order to reconstitute a 2890 bp fragment containing the complete 2781 bp open reading frame of Inpp4b.
2. Materials and methods
Total RNA extracted from tissues and hematopoietic cells of C57BL/6 mice were subjected to RT-PCR using primers specific for Inpp4aα and Inpp4aβ: forward 5′-AGCCTGTCCTCTTCAACGTG-3′, alpha reverse 5′-TGCATATTTGCGACTTCCAAC-3′ and beta reverse 5′-ACGCTATGCTCAGAAAGAGC-3′, and for Inpp4bα and Inpp4bβ: forward 5′-GATTAAAGAAGGCCAGTTGC-3′, alpha reverse 5′ACTTGGGGAAAGCCCATCAGC-3′ and beta reverse 5′AGAAGTGTCGAGAAGTTGGC-3′. Total protein extracts were prepared from tissues of C57BL/6 mice and analyzed by Western blotting as described (Rajapurohitam et al., 2001).
2.1. cDNA cloning of Inpp4b To isolate the mouse Inpp4b cDNA, a C57BL/6 spleen cDNA mouse library LambdaZapII (Stratagene) was screened by PCR (Takumi and Lodish, 1994). Alignment of the rat Inpp4b sequence with mouse EST and TRACE databases sequences (www. ncbi.nlm.nih.gov) served to design specific primers for mouse Inpp4b cDNA. The sequence of the TRACE 8258396 was used for the forward primer 5′-AAGCTTCCTGGGCTGTGCCAG-3′
2.2. Chromosomal mapping The mouse chromosomal location of Inpp4b was determined by genotyping the BSS backcross mapping panel generated at The Jackson Laboratory (Rowe et al., 1994) by PCR with the primers designed from a (GATA)n repeat we found in the mouse Inpp4b intron 2 sequence. Each 94 genomic DNA sample (25 ng) was amplified using flanking primer (forward: 5′-TGGATACAGCCCAATGAGTTG-3′, reverse: 5′-TCCCTTCTGTCATAGGTAGG-3′). Genotype of each animal for the GATA repeat was determined based on the length of the PCR product: ∼260 pb for Mus musculus (C57BL/6J) and ∼250 pb Mus spretus. The results were analyzed with the Map Manager program at The Jackson Laboratory Backcross DNA Map Service. 2.3. Transcription start site and promoter analysis Reverse transcription (RT) reactions were carried out on total RNA isolated from muscle, spleen, heart, lung, thymus and brain using the [32P] radiolabeled primer: 5′-GACTTTCCTGCGTTCTCCAG-3′ as described (Benachenhou et al., 2002). The RT reactions were then run on a 6% acrylamide sequencing gel adjacent to a promoter–exon 1 region sequence reaction made with the same primer. To identify conserved putative transcription factor binding sites, the mouse and the human promoter region sequence were analyzed using the Consite website (http:// mordor.cgb.ki.se/cgi-bin/CONSITE/consite/). 2.4. Expression analysis by RT-PCR and by Western blot
154
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
Blots were probed with a Inpp4b rabbit polyclonal purified antiserum developed in our laboratory, which recognizes Inpp4b amino acids 2–235. 2.5. Identification of the Inpp4bαs isoform One Inpp4b cDNA clone isolated in our screen was lacking exon 5. A RT-PCR assay using primers in exon 3 and in exon 8 was designed to confirm this result on total RNA isolated from mouse tissues. The primers were: exon 3-forward 5′-AACCGGAGCTCATTATTTGGTACCATTTG-3′ and exon 8-reverse 5′-GGCACAGCCCAGGAAGCTTC-3′. Two specific amplicons (514 and 395 pb) were detected in multiple tissues, isolated and sequenced. The sequence demonstrated that the 436 bp amplicon lacks Inpp4b exon 5 (119 bp) like the original cDNA clone. 2.6. Generation of Inpp4b–EGFP fusion proteins Full length cDNA encoding mouse Inpp4bα without the stop codon was produced by PCR using tailed oligonucle-
otides to generate KpnI and XmaI restriction sites: forward 5′AATTGGTACCATTTGAAAAAGACATCGTGG-3′ and reverse 5′-AATTCCCGGGTGTCAGCTTTTCCATAAG-3′. The PCR product was digested and cloned into the expression pEGFP-N1 vector (Clontech) to generate pInpp4bα-EGFP. The cDNA encoding the alpha-short isoform, lacking the exon 5, was amplified and cloned with the same primers to generate pInpp4bαs-EGFP. For the pInpp4bβ-EGFP construct, pInpp4bα-EGFP was digested by BtrI and AgeI to remove the 3′ alpha coding region that was replaced by the Inpp4bβ 3′ coding region amplified with the following primer and digested by BtrI and AgeI: forward 5′-CATGGCAGTCGGCATTTCAG-3′ and reverse 5′-GATCACCGGTTGGTATTTGGCTAAGAGTAATG-3′. 2.7. Cell culture, transfection and immunofluorescence COS-7 (1 × 105) cells were plated on glass cover slips and grown in DMEM media supplemented with 10% FBS. After 24 h, cells were transfected with individual plasmid DNA (0.5 μg) using Lipofectamine (Invitrogen). After 48 h, cells
Fig. 1. Inpp4b genetic and physical maps. A. Haplotype analysis of the 94 (Mus spretus × C57BL/6) F1 × Mus spretus backcross progeny (BSS). Each column represents a chromosomal haplotype identified in the backcross and inherited from the F1 parents. The number of individuals is listed beneath each column. Open squares represent the C57BL/6 allele, and filled squares represent the Mus spretus allele as determined by PCR-VNTR for Inpp4b and RFLP and SSLP for other markers indicated at the left. B. Genetic localization of Inpp4b on mouse chromosome 8 on the BSS genetic map. The MGD integrated genetic map is also shown. Corresponding synteny homology regions of human chromosomes are represented at the right. C. Physical maps including Inpp4b on mouse chromosome 8 and INPP4B on human chromosome 4q27-31. Surrounding genes with transcripts orientation are represented.
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
155
Fig. 2. Identification of Inpp4b transcription start sites. A. Total RNA isolated from mouse tissues and yeast tRNA (tR) as negative control were subjected to 5′ extension reverse transcription reaction. The three major transcription start sites are indicated on the right by arrows. The star indicates the heart specific start site. B. Alignment of the mouse and human Inpp4b gene proximal promoter (− 347 to +254). Arrows indicate the positions of the three transcriptional start sites. The position of the oligonucleotide used for the primer extension reaction is boxed. Positions of potential conserved transcription factor binding sites are underlined (M, skeletal muscle; S, spleen; H, heart; L, liver; T, thymus; B, brain).
were fixed 15 min in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 in PBS for 15 min, washed in PBS, and incubated 30 min with 10% fetal bovine serum in PBS. Cells were then labeled with a 1:100 dilution of anti-Golgin 97 mouse monoclonal antibody (Molecular Probes) containing 10% fetal bovine serum for 60 min. After washing with PBS containing 10% fetal bovine serum, cells were stained with an Alexa Fluor® 546-conjugated goat anti-mouse secondary antibody. Nuclei were stained with TO-PRO® 3 (Molecular Probes) and cells were visualized by confocal microscopy (Zeiss) at 488 nm (EGFP), 543 nm (Alexa Fluor® 546) and 633 nm (TOPRO® 3). 2.8. Phospholipids binding assay To characterize Inpp4b lipid binding specificity, the mouse Inpp4b 5′ region (aa 2–235) including the C2-domain was amplified by PCR with the primers forward 5′CCGCCTCGAGTGAAATCAAAGAGGAAGG-3′ and reverse 5′-CAGACTCGAGACAGGATTCCTTAAGACTG-3′ including a restriction XhoI site. The amplification product was digested and cloned into the XhoI site of the pHGST.2T bacterial
expression vector (Coulombe and Meloche, 2002). Following induction and expression in BL21 cells, the HisX6-Inpp4b(2235)-GST recombinant fusion protein was purified using GST purification module (Amersham). The Inpp4b-C2 recombinant protein was released from GST following thrombin treatment. A panel of lipids spotted onto nitrocellulose membrane (PIPStrips™, Echelon Bioscience Inc.) was preincubated 1 h in TBS-T (Tris 10 mM, 150 mM NaCl, pH 8, 0.1% Tween-20) containing 0.1% ovalbumin, followed by incubation with Inpp4b-C2 recombinant protein (0.05 μg/ml) for 4 h. Membranes were washed 3 times in the same buffer and incubated 1 h with anti-Inpp4b rabbit polyclonal purified antiserum (1:5000), followed by incubation with anti-rabbit-HRP (1:10000). Control incubation was conducted in the same conditions without the recombinant protein. 2.9. Mouse and human Inpp4b gene structures To determine the Inpp4b exon–intron boundaries, mouse and human cDNA were aligned with the mouse and the human genomic sequence, respectively, using NCBI Blast program (www.ncbi.nlm.nih.gov).
156
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
3. Results 3.1. The murine Inpp4b gene: cDNA isolation, mapping and gene organization To isolate and characterize the murine Inpp4b gene, sequence alignment of the rat Inpp4b cDNA with mouse EST et TRACE databases sequences (www.ncbi.nlm.nih.gov) served to design murine specific primers. These oligonucleotides were used to screen by PCR a C57BL/6 spleen cDNA mouse library and one clone containing the full length Inpp4b cDNA was obtained (Supplemental Fig. S1). To genetically map Inpp4b in mouse, we first identified a microsatellite (GATA)n repeat polymorphism in the mouse Inpp4b locus and secondly, designed a specific PCR assay for this polymorphism. This PCR assay then served to localize Inpp4b by genotyping the C57BL/6 × Spretus (BSS) interspecific backcross panel consisting of 94 progenies developed at the Jackson's Laboratory. Genotyping screen defined two recombination events and localized Inpp4b on mouse Chromosome 8 in a 2.2 cM genetic interval, between the microsatellite marker D8Mit74 and the man2b1 gene (Fig. 1A). This mouse genetic interval shared a synteny homology with the human 4q27-31 region closed to the Gypa locus, near a break of synteny with human chromosome 19 p (Fig. 1B). Comparison of these two genomic physical intervals revealed a perfect conservation in genes order and transcription orientation in
mouse and human. In addition, the more detailed human physical map allowed to localize more precisely the murine Inpp4b locus in a smaller 1 megabase (Mb) interval between the Il15 and Usp38 genes (Fig. 1C). Sequence of the full length murine Inpp4b cDNA was aligned and compared to the genomic locus. The exon–intron boundaries were identified by consensus donor and acceptor signals in accordance with GT/AG rule (Supplemental Table S1). Inpp4b gene consisted of 25 exons with two 5′ untranslated exons. The 23 coding exons ranged in size from 45 to 1340 nucleotides long and intron length varied from 1.1 kb to more than 280 kb. Consequently, the entire murine Inpp4b locus is large and covered approximately 600 kb of genomic DNA. Based on the mouse gene organization, we then deduced the structure of the human INPP4B locus (Supplemental Table S2) and demonstrated a very high conservation with the mouse gene including the alternative exon 25 splicing, which produced the two INPP4B protein α and β isoforms. Remarkably, all internal coding exons have identical lengths in mouse and human, with the exception of exon 17 that is shorter by 9 bp in human in a non-conserved region (Supplemental Figure S2, and Table S1 and S2). 3.2. Transcriptional regulation of the murine Inpp4b gene To characterize the regulation of Inpp4b locus, identification of the transcription start site of Inpp4b gene was undertaken by primer extension. The template for this assay consisted of total RNA isolated from different adult mouse tissues (Fig. 2A). Interestingly, we mapped three different tissue-specific transcription start sites for Inpp4b in skeletal muscle (M), heart (H) and spleen–lung–thymus (S,L,T). The skeletal muscle transcription start site was the most proximal (+ 1) while the heartspecific was most distal (∼ − 300). The third initiation site common to spleen, lung and thymus was mapped at − 64 (Fig. 2A). Upstream sequences of the 3 start sites suggested that these were part of a TATA-less promoter. Consistently, two SP1 consensus binding sites were detected between − 140 and +104 bp. Sequence analysis of the promoter region defined several consensus sites for potential binding of known transcription factors common to mouse and human (Fig. 2B). 3.3. Inpp4 expression pattern
Fig. 3. mRNA expression patterns of the Inpp4b isoforms. A. Adult expression of the different Inpp4 isoforms in mouse tissues was determined by RT-PCR. B. Expression of Inpp4 isoforms in enriched mouse hematopoietic cells (T: T lymphocyte, B: B lymphocyte, NK: natural killer lymphocyte, MC: mast cell, MAC: peritoneal macrophage). Br: brain and C — control.
The expression pattern of the Inpp4b transcripts encoding α and β isoforms was determined in mice from various tissues by semi-quantitative RT-PCR (Fig. 3A). The Inpp4bα isoform showed a more broad tissue expression with higher levels in brain, skeletal muscle, lung, spleen, testis and thymus. In contrast, the Inpp4b beta isoform displayed restricted expression with highest levels in brain and small intestine, and lower levels in skeletal muscle and heart. Since the Inpp4b alpha isoform was expressed in spleen and in thymus, we also investigated specifically the hematopoietic lineages (Fig. 3B). Of particular interest, the Inpp4bα isoform was expressed in a subset of lineages with high levels in NK and mast enriched cell populations, moderately in B-lymphocytes and undetectable in
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
157
Fig. 4. The Inpp4b protein orthologues and paralogues. A. Schematic representation of the Inpp4b mouse, rat, human, Drosophila and nematode orthologues. Inpp4a mouse paralogue is also shown. Overall amino acid identity with mouse Inpp4b protein was indicated as percentage. Amino acid identity for the 4-Ptase active site and the C2 domains were indicated below each domain. B. Phospholipid binding specificity of the mouse Inpp4b C2 domain. Detection was achieved by hybridization with the anti-Inpp4b antibody. C. Inpp4b protein expression in mouse tissues was determined on Western blot using a specific Inpp4b antibody raised against the Nterminus region of the protein (aa 2–235). As controls, COS cells were transfected or not with the Inpp4bα cDNA. β-actin was used as internal loading control.
T-lymphocyte and peritoneal macrophage cell populations (Fig. 3B). Because the Inpp4b locus has a paralogue Inpp4a, expression of both isoforms α and β of Inpp4a was analyzed for comparison. Both Inpp4a isoforms were widely expressed in most tissues but detected at different levels. Noticeably, expression in the brain was highest for both paralogues and isoforms (Fig. 3A). 3.4. The mouse Inpp4b proteins The mouse Inpp4bα and β cDNAs encode 927 and 941 amino acids protein respectively, with a predicted molecular weight of ∼ 105 kDa. The mouse protein shares 96% and 90% identity with the human and rat proteins respectively (Supplemental Figure S2A). In addition, the
mouse Inpp4b shared 45% amino acid identity with the mouse paralogue Inpp4a (Fig. 4A). All mammalian Inpp4 proteins have in common a consensus Cys(X)5Arg phosphatase catalytic site and also a conserved C2 domain. These Inpp4 proteins were phylogenetically related to Caenorhabditis elegans 4-ptase and Drosophila melanogaster CG1846 (Fig. 4A) but no homologue could be found in yeast. In the C. elegans protein, instead of the C2 domain, the presence of another lipid binding domain a pleckstrin homology (PH) domain suggested convergent evolution (Overduin et al., 2001). The D. melanogaster Inpp4 orthologue contained both a C2 and a PH domain but the PH domain was unrelated to that of the C. elegans. Interestingly, the duplication of Inpp4 genes (Inpp4a and Inpp4b), as well as the presence of an alternative splice site generating the beta isoforms, was only found in vertebrates.
158
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
Since it was demonstrated that several C2 domains can bind lipids and subsequently recruit proteins at particular intracellular membranes (Mehrotra et al., 2000; Chung et al., 1998), we investigated which lipid family could interact with the Inpp4bC2 domain. Using an in vitro lipid-binding assay, we demonstrated that Inpp4b-C2 domain preferentially binds to phosphatidic acid and PI(3,4,5)P3 (Fig. 4B). In addition, the Inpp4b-C2 domain also binds but with a lower affinity to other phosphorylated PI and phosphatidyl serine, supporting specific lipid interaction. To analyze Inpp4b protein expression pattern and cellular localization, we have raised rabbit polyclonal antibodies against the N-terminus region of Inpp4b (aa 2–235) that can recognize both Inpp4b α and β isoforms. Inpp4b proteins were detected by Western blotting as a predominant 105 kDa band, consistent with the predicted amino acid sequence and with the full length Inpp4bα cDNA expressed in COS cells as positive control (Fig. 4C). Strong Inpp4b expression was observed in brain, spleen, small intestine, lung and testis. As determined for transcript expression pattern (Fig. 3A), the Inpp4b proteins were also expressed at highest levels in brain tissue.
Fig. 5. Cellular localization of Inpp4bα-EGFP and Inpp4bβ-EGFP isoforms. COS cells were transfected with either EGFP (A–C), Inpp4bα-EGFP (D–F) or Inpp4bβ-EGFP (G–L). Cellular localization of the different EGFP fusion proteins was detected using confocal microscopy (A, D, G, J). Cells were counterstained with anti-Golgin 97 antibody (B, E, H, K). Nuclei were stained using TO-PRO-3 dye. Merge pictures demonstrated colocalization of Inpp4bβ with Golgin (C, F, I, L).
3.5. Subcellular localization of Inpp4bα and Inpp4bβ proteins To determine the cellular localization of the different Inpp4b isoforms, full length cDNAs of Inpp4bα and Inpp4bβ were inserted in frame into an EGFP expression vector, and the chimaeric proteins were expressed in COS cells. Compared to EGFP control expressing cells (Fig. 5A–C), Inpp4bα-EGFP was distributed uniformly in the cytoplasm and absent from the nucleus (Fig. 5D–F). In contrast, Inpp4bβ-EGFP localization in COS cells was reproducibly detected in close association with the Golgi apparatus in most cells by fluorescence (Fig. 5G–I). Subsequently, immunostaining with Golgin 97, a specific marker of the Golgi apparatus, showed consistent colocalization of Golgin 97 and EGFP in Inpp4bβ-EGFP expressing cells (Fig. 5G–L). 3.6. Characterization of a novel Inpp4b isoform: Inpp4bαs Among several Inpp4b cDNA clones isolated, one of them appeared to contain all Inpp4bα exons except for exon 5 (Fig. 6A). RT-PCR carried out on mouse spleen tissues supported the synthesis of this novel isoform in vivo. With a single pair of primers localized upstream of exon 3 and in exon 8 respectively, two specific amplified products were detected of 514 and 395 bp respectively. Upon sequencing these two specific products, the shortest amplicon demonstrated lack of exon 5 sequence (119 bp) and was suspected to result from alternative splicing. This shorter amplicon provided the first evidence for a shorter cDNA, Inpp4bαs (Inpp4bα short), compared to the Inpp4bα or β full length forms. The presence of the full length and the putative short form of Inpp4bα was then investigated by RT-PCR in a wide range of mouse tissues using primers that can distinguish the two isoforms. Two amplicons were observed in different ratios depending on the individual tissues or organs, supporting the existence in vivo of a novel isoform of Inpp4bα. Interestingly, the Inpp4bαs form was virtually undetectable in the brain, suggesting a tissue-specific regulation for this alternative splicing (Fig. 6B). Based on the exclusion of exon 5 and usage of the full length Inpp4b translation initiation codon, the Inpp4bαs transcript should undergo a frameshift that generates a premature stop codon in exon 6. Because a protein band of ∼ 7.5 kDa (65 aa) for Inpp4bαs was not detected by Western (data not shown) and that the Inpp4bαs cDNA contained all Inpp4bα sequences downstream of exon 6, we thus hypothesized that an alternative translation initiation site for the Inpp4bαs mRNA could be used. We found one other possibility for a translation codon at 50 nucleotides downstream from the original start codon (Fig. 6C,D). To test this hypothesis, we fused in frame the Inpp4bαs or Inpp4bα cDNA upstream of EGFP in expression vectors. These expression vectors were transfected in COS cells and expression of the resulting fusion protein species was characterized by Western blot. Anti-EGFP detected a ∼ 135 and ∼126 kDa band for the Inpp4bα-EGFP and Inpp4bαsEGFP fusion protein extracts, respectively (Fig. 6E). This Inpp4bαs cDNA could be transcribed in cells and was identical
M. Ferron, J. Vacher / Gene 376 (2006) 152–161 Fig. 6. Characterization of Inpp4bαs, a novel Inpp4b isoform. A. Schematic representation of Inpp4bα and Inpp4bαs 5′ mRNAs regions. Position of the deduced start codon (M) is indicated. Arrows indicate the position of the primers used in RT-PCR. B. Tissue expression pattern of Inpp4bαs was determined by RT-PCR. The specific shorter 395 pb amplification product is specific of Inpp4bαs. C and D. cDNA and protein sequences of Inpp4bα and Inpp4bαs respectively. The sequence corresponding to exon 5, excluded in Inpp4bαs, is underlined and in bold. The alternative N-terminus sequence of Inpp4bαs, absent in Inpp4bα, is bolded in D. Upstream in frame stop codons are underlined in both sequences. The start codon of Inpp4bα is bolded in the Inpp4bαs sequence. Arrows represent primers used in RT-PCR. E. The Inpp4bαs mRNA contains a functional open reading frame. cDNAs corresponding to Inpp4bα and Inpp4bαs were cloned upstream of EGFP in an expression vector and transfected in COS cells. Expression of the resulting fusion proteins was assessed by Western blotting using antibodies against Inpp4b or EGFP on protein extracts. In the right panel, equal amounts of protein extracts were analyzed (25 μg), and only a 1/100 (0.25 μg) of Inpp4bα-EGFP protein extract were analyzed in the middle and left panels. F. Subcellular localization of Inpp4bα-EGFP and of Inpp4bαs-EGFP in COS cells (upper panels). Nuclei were stained with Hoechst (middle panels) and merged images demonstrated a cytoplasmic localization for both isoforms (bottom panels). 159
160
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
to the full length Inpp4bα. With our anti-Inpp4b primary antibodies that should recognize the Inpp4bαs, we confirmed the existence of Inpp4bαs-EGFP fusion protein of similar molecular weight as observed with the EGFP antibody and provided definitive evidence of a different translation initiation site for the novel Inpp4bαs cDNA. This Inpp4bαs cDNA consisted of all the protein domains of Inpp4bα except for the 85 N-terminal amino acids substituted by 24 different amino acids and was translated into a protein of ∼ 98 kDa (866 aa) that contained less than half of the C2 domain. To determine the cellular localization of Inpp4bαs isoform in comparison to Inpp4bα, the fusion EGFP expression vectors were used in COS cells. Interestingly, Inpp4bαs EGFP fusion proteins demonstrated a similar cytosolic localization as detected for the full length Inpp4bα (Fig. 6F). Together, these results demonstrated that the Inpp4bαs represent a novel isoform of the Inpp4b protein in addition to the differentially spliced α and β isoforms. 4. Discussion Coordinate regulations of phosphatidyl inositol kinases and phosphatases play a key role in signaling pathways that control cell proliferation, differentiation and survival, but several phosphatases remain to be identified and characterized. This manuscript reports the isolation and characterization of the mouse locus for the Inpp4b inositol phosphatase. Our results showed that the murine Inpp4b gene structure and protein α and β isoforms were highly similar to human and rat homologues. Interestingly, we identified a novel alternative form of Inpp4bα. Furthermore, the three Inpp4b protein isoforms have a differential expression pattern as well as a specific cellular localization. Because phosphatases are essential to counterbalance kinase production of active PIPs, the second messengers, we have isolated, mapped and analyzed the murine Inpp4b gene. Inpp4b was localized on chromosome 8 between Il15 and Usp38 in a region of synteny homologous to 4q27-31 in human. The mouse Inpp4b gene and protein displayed high structure similarity to the rat and human genes and proteins with conserved consensus C2 domain and phosphatase active site. The conservation of a Inpp4-like gene in more primitive organisms such as insects and nematodes, and absence in yeast, suggested that this gene was specific of metazoans. In addition, Inpp4b gene in vertebrates has a paralogue Inpp4a with close genomic similarity that has not been investigated for regulation. Our characterization of the mouse and human Inpp4b gene structure showed two isoforms, alpha and beta, resulting of alternative splicing of exon 25 as defined for rat and human Inpp4a. The Inpp4b locus has three major transcriptional initiation sites that appeared to be tissue-specific. Our analysis of the Inpp4b upstream regulatory sequences also detected several consensus binding sites for different classes of transcription factors, and suggested complex transcriptional regulation for both Inpp4b isoforms in individual tissues. Moreover, we have identified a novel isoform of Inpp4bα, called Inpp4bαs, that resulted from a downstream translation initiation site and
exon 5 skipping of Inpp4bα. Characterization of this novel Inpp4bαs showed that the 85 N-terminal amino acids present in Inpp4bα were substituted by a different sequences of 24 amino acids. The expression patterns of both Inpp4b and Inpp4a genes were conjointly established for comparison analysis due to their close similarity. The murine Inpp4bα isoform exhibited relatively ubiquitous expression with a wide tissue distribution very similar to the murine Inpp4aα isoform. In apparent opposition to this broad expression pattern, expression of Inpp4bα was confined to specific cells of the hematopoietic lineages. Of interest, only the Inpp4bα isoform displayed a high level of expression in enriched mast, lymphocyte, and NK cell subpopulations relative to the undetectable expression of Inpp4bβ and of Inpp4aα and β isoforms, suggesting most likely a unique role in these cells. In comparison to Inpp4bα, the Inpp4bβ isoform showed a more restricted expression pattern to brain, heart, intestine, skeletal muscle and spleen. Noticeably, both α and β isoforms of Inpp4b and Inpp4a displayed highest expression in the brain, consistent with the neuronal defect in the Inpp4a weeble mutant (Nystuen et al., 2001). Particularly striking is the complete absence of expression of the Inpp4bαs isoform in this tissue in contrast to its presence in all other tissues analyzed. Together, these results point to both overlapping and complementary role for each of the five Inpp4 isoforms examined in this study. Our study of the Inpp4b proteins has revealed that in contrast to the α isoform, the β isoform had a hydrophobic C-terminal domain and consistently, these isoforms displayed specific subcellular localization. While the intracellular localization of the full length and short form of the Inpp4bα protein was mainly cytosolic, that of the Inpp4bβ isoform was associated with the Golgi apparatus. Interestingly, the differential subcellular localization of the Inpp4bα and β phosphatase isoforms was similar to that observed for the two phosphatidylinositol 4-kinase type II (PI4KII) isoforms. Indeed, the PI4KII α isoform was located at the trans-Golgi network whereas PI4KIIβ was localized exclusively in the cytosol (Wang et al., 2003; Wei et al., 2002). Such localization of the Inpp4b isoforms was complementary to that of the Inpp4a isoform associated with endosomes (Shin et al., 2005; Ivetac et al., 2005). The localization of individual Inpp4 protein isoform in non-redundant cellular compartments could implicate different molecular interactions or cellular function. In accordance with this precise subcellular localization, our results of Inpp4b-C2 protein–lipid interaction demonstrated preferential binding to phosphatidic acid and to PI(3,4,5)P3. Of note, this preferential binding of Inpp4b-C2 appeared quite different from that of Inpp4a-C2 domain shown to interact mainly with PI(3,4)P2 and PI(4)P (Ivetac et al., 2005). Hence, the binding factors to the C2 domain and the targeted subcellular localization of the multiple isoforms of Inpp4b are likely to impact on the protein/substrate interactions specificity and play a critical role in the cellular function of these proteins. Taken together, our characterization of the murine Inpp4b and Inpp4a demonstrated high gene and protein conservation. The differential Inpp4b expression pattern, cellular sublocalization
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
and interacting factors are likely to ascribe highly specific function for Inpp4 phosphatase proteins. Future studies on the Inpp4b gene should be directed to decipher the in vivo functional role of Inpp4b in mouse. Acknowledgments We thank M. Pata for excellent technical assistance. We are grateful to Drs S. Meloche and A. Veillette for vector and RNA samples, and to Dr A. Kania for critical reading of the manuscript. This work was supported by Canadian Institutes of Health Research (MOP44079) and National Sciences and Engineering Council of Canada (OPG0155277) grants to J.V. and a doctoral training award to M.F. from the Canadian Institutes of Health Research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.gene.2006.02.022. References Barnache, S., Le Scolan, E., Kosmider, O., Denis, N., Moreau-Gachelin, F., 2006. Phosphatidylinositol 4-phosphatase type II is an erythropoietinresponsive gene. Oncogene 25, 1420–1423. Benachenhou, N., Massy, I., Vacher, J., 2002. Characterization and expression analyses of the mouse Wiskott–Aldrich syndrome protein (WASP) family member Wave1/Scar. Gene 290, 131–140. Cantley, L.C., 2002. The phosphoinositide 3-kinase pathway. Science 296, 1655–1657. Chung, S.H., et al., 1998. The C2 domains of Rabphilin3A specifically bind phosphatidylinositol 4,5-bisphosphate containing vesicles in a Ca2+dependent manner. In vitro characteristics and possible significance. J. Biol. Chem. 273, 10240–10248. Coulombe, P., Meloche, S., 2002. Dual-tag prokaryotic vectors for enhanced expression of full-length recombinant proteins. Anal. Biochem. 310, 219–222. De Matteis, M.A., Godi, A., 2004. PI-loting membrane traffic. Nat. Cell Biol. 6, 487–492. Ivetac, I., et al., 2005. The type Ialpha inositol polyphosphate 4-phosphatase generates and terminates phosphoinositide 3-kinase signals on endosomes and the plasma membrane. Mol. Biol. Cell 16, 2218–2233. Leslie, N.R., Downes, C.P., 2002. PTEN: the down side of PI 3-kinase signalling. Cell. Signal. 14, 285–295.
161
Majerus, P.W., Kisseleva, M.V., Norris, F.A., 1999. The role of phosphatases in inositol signaling reactions. J. Biol. Chem. 274, 10669–10672. Mehrotra, B., Myszka, D.G., Prestwich, G.D., 2000. Binding kinetics and ligand specificity for the interactions of the C2B domain of synaptogmin II with inositol polyphosphates and phosphoinositides. Biochemistry 39, 9679–9686. Munday, A.D., Norris, F.A., Caldwell, K.K., Brown, S., Majerus, P.W., Mitchell, C.A., 1999. The inositol polyphosphate 4-phosphatase forms a complex with phosphatidylinositol 3-kinase in human platelet cytosol. Proc. Natl. Acad. Sci. U. S. A. 96, 3640–3645. Norris, F.A., Auethavekiat, V., Majerus, P.W., 1995. The isolation and characterization of cDNA encoding human and rat brain inositol polyphosphate 4-phosphatase. J. Biol. Chem. 270, 16128–16133. Norris, F.A., Atkins, R.C., Majerus, P.W., 1997. The cDNA cloning and characterization of inositol polyphosphate 4-phosphatase type II. Evidence for conserved alternative splicing in the 4-phosphatase family. J. Biol. Chem. 272, 23859–23864. Nystuen, A., Legare, M.E., Shultz, L.D., Frankel, W.N., 2001. A null mutation in inositol polyphosphate 4-phosphatase type I causes selective neuronal loss in weeble mutant mice. Neuron 32, 203–212. Overduin, M., Cheever, M.L., Kutateladze, T.G., 2001. Signaling with phosphoinositides: better than binary. Mol. Interv. 1, 150–159. Rajapurohitam, V., Chalhoub, N., Benachenhou, N., Neff, L., Baron, R., Vacher, J., 2001. The mouse osteopetrotic grey-lethal mutation induces a defect in osteoclast maturation/function. Bone 28, 513–523. Rowe, L.B., et al., 1994. Maps from two interspecific backcross DNA panels available as a community genetic mapping resource. Mamm. Genome 5, 253–274. Shin, H.W., et al., 2005. An enzymatic cascade of Rab5 effectors regulates phosphoinositide turnover in the endocytic pathway. J. Cell Biol. 170, 607–618. Takumi, T., Lodish, H.F., 1994. Rapid cDNA cloning by PCR screening. Biotechniques 17, 443–444. Vyas, P., Norris, F.A., Joseph, R., Majerus, P.W., Orkin, S.H., 2000. Inositol polyphosphate 4-phosphatase type I regulates cell growth downstream of transcription factor GATA-1. Proc. Natl. Acad. Sci. U. S. A. 97, 13696–13701. Wang, Y.J., et al., 2003. Phosphatidylinositol 4 phosphate regulates targeting of clathrin adaptor AP-1 complexes to the Golgi. Cell 114, 299–310. Ware, M.D., Rosten, P., Damen, J.E., Liu, L., Humphries, R.K., Krystal, G., 1996. Cloning and characterization of human SHIP, the 145-kD inositol 5phosphatase that associates with SHC after cytokine stimulation. Blood 88, 2833–2840. Wei, Y.J., et al., 2002. Type II phosphatidylinositol 4-kinase beta is a cytosolic and peripheral membrane protein that is recruited to the plasma membrane and activated by Rac-GTP. J. Biol. Chem. 277, 46586–46593.
Characterization of the murine Inpp4b gene and identification of a novel isoform☆ Mathieu Ferron
a,b
, Jean Vacher
a,b,c,⁎
a
b
Institut de recherches cliniques de Montreal, 110 av. des Pins O., Montreal, Qc, Canada H2W 1R7 Programme de biologie moléculaire, Faculté des études supérieures de l'Université de Montréal, Canada c Faculté de Médecine de l'Université de Montréal, Canada Received 20 December 2005; received in revised form 14 February 2006; accepted 21 February 2006 Available online 17 March 2006 Received by M. D'Urso
Abstract Inositol polyphosphate phosphatases and phosphoinositides second messengers have been associated with major cellular functions as growth, differentiation, apoptosis, protein trafficking and motility. To characterize the role of inositol phosphatases in cell physiology, we have isolated the mouse Inositol polyphosphate 4-phosphatase type II (Inpp4b) cDNA. The murine Inpp4b locus was mapped on chromosome 8 in a synthenic region of the human 4q27-31 interval between Il-15 and Usp38. The mouse Inpp4b proteins, α and β isoforms, encoded by this locus contained 927 and 941 amino acids respectively with a consensus phosphatase catalytic site and a conserved C2 domain that are highly similar with the human and rat homologues. Interestingly, we characterized a novel shorter isoform of Inpp4bα resulting from an alternative translation initiation site and exon 5 skipping. Inpp4b C2 domain interacted with preferential affinity to phosphatidic acid and phosphatidylinositol 3,4,5-triphosphate (PI(3,4,5)P3) lipids. While analysis of Inpp4b transcript and protein expression demonstrated a broad tissue distribution for the α isoform, as for the paralogue Inpp4aα and β isoforms, it also displayed a limited hematopoietic lineage distribution whereas the Inpp4bβ isoform had a highly restricted pattern. Importantly, the Inpp4bβ localized to the Golgi apparatus whereas Inpp4bα was mainly cytosolic, suggesting a different cellular function for this isoform. Together our characterization of the murine Inpp4b gene expression pattern, cellular sublocalization and interacting lipids support highly specific function for individual Inpp4 phosphatase proteins. © 2006 Elsevier B.V. All rights reserved. Keywords: Inositol phosphatase; Lipids binding; Golgi; C2 domain
1. Introduction Phosphatidylinositol lipids or phosphoinositides are implicated in major cellular mechanisms including cell growth, differentiation, apoptosis, protein trafficking and motility (CantAbbreviations: bp, base-pair; cDNA, complementary DNA; EGFP, enhanced green fluorescence protein; EST, expressed sequence tag; GST, glutathione S-transferase; kb, kilobase; kDa, kiloDalton; Mb, megabase; PCR, polymerase chain reaction; RT, reverse transcription; TBS-T, Tris buffer saline-tween. ☆ GenBank accession numbers: Mouse Inpp4b alpha cDNA, DQ296611; mouse Inpp4b alpha short cDNA, DQ296612; mouse Inpp4b beta cDNA, DQ296613. ⁎ Corresponding author. Institut de recherches cliniques de Montreal, 110 av. des Pins O., Montreal, Qc, Canada H2W 1R7. Tel.: +1 514 987 5734; fax: +1 514 987 5585. E-mail address: [email protected] (J. Vacher). 0378-1119/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2006.02.022
ley, 2002). Acting as second messengers, their intracellular concentration levels are tightly regulated by the opposite activities of phosphoinositide kinases and phosphatases (De Matteis and Godi, 2004). Phosphoinositides can be phosphorylated at three distinct positions, D-3, D-4 and D-5 of the inositol head group, generating 7 unique derivatives. One of the critical and most studied phosphoinositide kinases, the phosphatidylinositol 3-kinase (PI3-Kinase), phosphorylates specifically position D-3. This PI3-Kinase upon activation by recruitment at the membrane is responsible for the production of phosphatidylinositol 3,4,5-triphosphate (PI(3,4,5)P3) from the substrate phosphatidylinositol 4,5-biphosphate (PI(4,5)P2) (Cantley, 2002). Reversibly, newly synthesized PI(3,4,5)P3 can be rapidly dephosphorylated by specific inositol phosphatases (Majerus et al., 1999). PI(3,4,5)P3 can be dephosphorylated at the D-3 position by PTEN to
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
153
generate PI(4,5)P2 (Leslie and Downes, 2002) or at the D-5 position by SHIP1 or SHIP2 inositol polyphosphate 5-phosphatases that generate the PI(3,4)P2 second messenger (Ware et al., 1996). Dephosphorylation at the D-4 position on the inositol ring to generate PI(3)P occurs via activation of the inositol polyphosphate 4-phosphatase isoenzymes type I (Inpp4a) and type II (Inpp4b) (Norris et al., 1995, 1997). Two human and rat 4-phosphatase isoenzymes (type I or INPP4A, and type II or INPP4B) have been cloned, sharing 37% amino acid identity. Both isoenzymes contain a catalytic “CX5R” motif also found in protein tyrosine phosphatases. These phosphatases have dual specificity for protein and lipids. In addition, both these INPP4A and INPP4B isoenzymes can exist in two splice variants, α and β (Norris et al., 1995, 1997). While the INPP4Aα and INPP4Bα isoforms have a hydrophilic C-terminus region, the INPP4Aβ and INPP4Bβ isoforms have a hydrophobic C-terminus which contains a potential transmembrane domain. The murine Inpp4a and Inpp4b genes are not well characterized, but few studies on Inpp4a in mice and cell culture have been reported. In vivo, the mouse mutant weeble that corresponds to a null allele of Inpp4a, demonstrated a crucial role for Inpp4a phosphatase in neuronal survival (Nystuen et al., 2001). Inpp4a was also suggested to play a role in the megakaryocytic lineage as a negative regulator of cell growth acting downstream of the GATA-1 transcription factor (Vyas et al., 2000) and by forming a complex with PI3-Kinase (Munday et al., 1999). While Inpp4a cellular role in vitro was often associated with inhibition of a signaling pathway by dephosphorylation of the second messenger PI(3,4)P2 at the plasma membrane, recent studies suggested that an increase in production of PI(3)P pool by Inpp4a on endosomal membranes has a stimulatory effect on signaling (Ivetac et al., 2005; Shin et al., 2005). By contrast, no in vivo precise functional role has yet been assigned to Inpp4b. Interestingly however, Inpp4b expression was recently shown to be transcriptionally regulated by erythropoietin in erythroid preleukemic cells (Barnache et al., 2006). To gain insight into the cellular role of Inpp4b, Inpp4b cDNA was isolated, the locus mapped and the gene organization defined. Upon comparison to the human and rat homologues, the mouse Inpp4bα and β spliced isoforms demonstrated high conservation. Interestingly, we also identified and characterized a novel Inpp4bα isoform called Inpp4b alpha short (Inpp4bαs). The Inpp4b isoforms, Inpp4bα,αs,β showed distinct expression pattern, and subcellular localization, indicating specific functions for these isoforms.
and the EST BG145663 for the reverse primer 5′-GTGATGACATCATAGAGAGC-3′. This assay, when tested on mouse spleen cDNA, amplified a 669 bp cDNA specific fragment, and 6 different cDNA clones were sequenced. One of them, sP4D12 was found to contain the complete open reading frame of Inpp4b except for 250 pb in the 5′ region of the cDNA. Consequently, the 5′ half of the coding sequence was amplified by RT-PCR amplification on spleen cDNA with the forward primer 5′AACCGGAGCTCATTATTTGGTACCATTTG-3′ and the reverse primer 5′-CTCTGTGCTGACATTAGGAG-3′. The amplified product was digested by KpnI and SacI restriction enzymes to obtain a 1389 bp fragment (5′ half) and cloned with the 1507 bp SacI/BglI fragment (3′ half) obtained from sP4D12 clone in pBlueScript in order to reconstitute a 2890 bp fragment containing the complete 2781 bp open reading frame of Inpp4b.
2. Materials and methods
Total RNA extracted from tissues and hematopoietic cells of C57BL/6 mice were subjected to RT-PCR using primers specific for Inpp4aα and Inpp4aβ: forward 5′-AGCCTGTCCTCTTCAACGTG-3′, alpha reverse 5′-TGCATATTTGCGACTTCCAAC-3′ and beta reverse 5′-ACGCTATGCTCAGAAAGAGC-3′, and for Inpp4bα and Inpp4bβ: forward 5′-GATTAAAGAAGGCCAGTTGC-3′, alpha reverse 5′ACTTGGGGAAAGCCCATCAGC-3′ and beta reverse 5′AGAAGTGTCGAGAAGTTGGC-3′. Total protein extracts were prepared from tissues of C57BL/6 mice and analyzed by Western blotting as described (Rajapurohitam et al., 2001).
2.1. cDNA cloning of Inpp4b To isolate the mouse Inpp4b cDNA, a C57BL/6 spleen cDNA mouse library LambdaZapII (Stratagene) was screened by PCR (Takumi and Lodish, 1994). Alignment of the rat Inpp4b sequence with mouse EST and TRACE databases sequences (www. ncbi.nlm.nih.gov) served to design specific primers for mouse Inpp4b cDNA. The sequence of the TRACE 8258396 was used for the forward primer 5′-AAGCTTCCTGGGCTGTGCCAG-3′
2.2. Chromosomal mapping The mouse chromosomal location of Inpp4b was determined by genotyping the BSS backcross mapping panel generated at The Jackson Laboratory (Rowe et al., 1994) by PCR with the primers designed from a (GATA)n repeat we found in the mouse Inpp4b intron 2 sequence. Each 94 genomic DNA sample (25 ng) was amplified using flanking primer (forward: 5′-TGGATACAGCCCAATGAGTTG-3′, reverse: 5′-TCCCTTCTGTCATAGGTAGG-3′). Genotype of each animal for the GATA repeat was determined based on the length of the PCR product: ∼260 pb for Mus musculus (C57BL/6J) and ∼250 pb Mus spretus. The results were analyzed with the Map Manager program at The Jackson Laboratory Backcross DNA Map Service. 2.3. Transcription start site and promoter analysis Reverse transcription (RT) reactions were carried out on total RNA isolated from muscle, spleen, heart, lung, thymus and brain using the [32P] radiolabeled primer: 5′-GACTTTCCTGCGTTCTCCAG-3′ as described (Benachenhou et al., 2002). The RT reactions were then run on a 6% acrylamide sequencing gel adjacent to a promoter–exon 1 region sequence reaction made with the same primer. To identify conserved putative transcription factor binding sites, the mouse and the human promoter region sequence were analyzed using the Consite website (http:// mordor.cgb.ki.se/cgi-bin/CONSITE/consite/). 2.4. Expression analysis by RT-PCR and by Western blot
154
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
Blots were probed with a Inpp4b rabbit polyclonal purified antiserum developed in our laboratory, which recognizes Inpp4b amino acids 2–235. 2.5. Identification of the Inpp4bαs isoform One Inpp4b cDNA clone isolated in our screen was lacking exon 5. A RT-PCR assay using primers in exon 3 and in exon 8 was designed to confirm this result on total RNA isolated from mouse tissues. The primers were: exon 3-forward 5′-AACCGGAGCTCATTATTTGGTACCATTTG-3′ and exon 8-reverse 5′-GGCACAGCCCAGGAAGCTTC-3′. Two specific amplicons (514 and 395 pb) were detected in multiple tissues, isolated and sequenced. The sequence demonstrated that the 436 bp amplicon lacks Inpp4b exon 5 (119 bp) like the original cDNA clone. 2.6. Generation of Inpp4b–EGFP fusion proteins Full length cDNA encoding mouse Inpp4bα without the stop codon was produced by PCR using tailed oligonucle-
otides to generate KpnI and XmaI restriction sites: forward 5′AATTGGTACCATTTGAAAAAGACATCGTGG-3′ and reverse 5′-AATTCCCGGGTGTCAGCTTTTCCATAAG-3′. The PCR product was digested and cloned into the expression pEGFP-N1 vector (Clontech) to generate pInpp4bα-EGFP. The cDNA encoding the alpha-short isoform, lacking the exon 5, was amplified and cloned with the same primers to generate pInpp4bαs-EGFP. For the pInpp4bβ-EGFP construct, pInpp4bα-EGFP was digested by BtrI and AgeI to remove the 3′ alpha coding region that was replaced by the Inpp4bβ 3′ coding region amplified with the following primer and digested by BtrI and AgeI: forward 5′-CATGGCAGTCGGCATTTCAG-3′ and reverse 5′-GATCACCGGTTGGTATTTGGCTAAGAGTAATG-3′. 2.7. Cell culture, transfection and immunofluorescence COS-7 (1 × 105) cells were plated on glass cover slips and grown in DMEM media supplemented with 10% FBS. After 24 h, cells were transfected with individual plasmid DNA (0.5 μg) using Lipofectamine (Invitrogen). After 48 h, cells
Fig. 1. Inpp4b genetic and physical maps. A. Haplotype analysis of the 94 (Mus spretus × C57BL/6) F1 × Mus spretus backcross progeny (BSS). Each column represents a chromosomal haplotype identified in the backcross and inherited from the F1 parents. The number of individuals is listed beneath each column. Open squares represent the C57BL/6 allele, and filled squares represent the Mus spretus allele as determined by PCR-VNTR for Inpp4b and RFLP and SSLP for other markers indicated at the left. B. Genetic localization of Inpp4b on mouse chromosome 8 on the BSS genetic map. The MGD integrated genetic map is also shown. Corresponding synteny homology regions of human chromosomes are represented at the right. C. Physical maps including Inpp4b on mouse chromosome 8 and INPP4B on human chromosome 4q27-31. Surrounding genes with transcripts orientation are represented.
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
155
Fig. 2. Identification of Inpp4b transcription start sites. A. Total RNA isolated from mouse tissues and yeast tRNA (tR) as negative control were subjected to 5′ extension reverse transcription reaction. The three major transcription start sites are indicated on the right by arrows. The star indicates the heart specific start site. B. Alignment of the mouse and human Inpp4b gene proximal promoter (− 347 to +254). Arrows indicate the positions of the three transcriptional start sites. The position of the oligonucleotide used for the primer extension reaction is boxed. Positions of potential conserved transcription factor binding sites are underlined (M, skeletal muscle; S, spleen; H, heart; L, liver; T, thymus; B, brain).
were fixed 15 min in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 in PBS for 15 min, washed in PBS, and incubated 30 min with 10% fetal bovine serum in PBS. Cells were then labeled with a 1:100 dilution of anti-Golgin 97 mouse monoclonal antibody (Molecular Probes) containing 10% fetal bovine serum for 60 min. After washing with PBS containing 10% fetal bovine serum, cells were stained with an Alexa Fluor® 546-conjugated goat anti-mouse secondary antibody. Nuclei were stained with TO-PRO® 3 (Molecular Probes) and cells were visualized by confocal microscopy (Zeiss) at 488 nm (EGFP), 543 nm (Alexa Fluor® 546) and 633 nm (TOPRO® 3). 2.8. Phospholipids binding assay To characterize Inpp4b lipid binding specificity, the mouse Inpp4b 5′ region (aa 2–235) including the C2-domain was amplified by PCR with the primers forward 5′CCGCCTCGAGTGAAATCAAAGAGGAAGG-3′ and reverse 5′-CAGACTCGAGACAGGATTCCTTAAGACTG-3′ including a restriction XhoI site. The amplification product was digested and cloned into the XhoI site of the pHGST.2T bacterial
expression vector (Coulombe and Meloche, 2002). Following induction and expression in BL21 cells, the HisX6-Inpp4b(2235)-GST recombinant fusion protein was purified using GST purification module (Amersham). The Inpp4b-C2 recombinant protein was released from GST following thrombin treatment. A panel of lipids spotted onto nitrocellulose membrane (PIPStrips™, Echelon Bioscience Inc.) was preincubated 1 h in TBS-T (Tris 10 mM, 150 mM NaCl, pH 8, 0.1% Tween-20) containing 0.1% ovalbumin, followed by incubation with Inpp4b-C2 recombinant protein (0.05 μg/ml) for 4 h. Membranes were washed 3 times in the same buffer and incubated 1 h with anti-Inpp4b rabbit polyclonal purified antiserum (1:5000), followed by incubation with anti-rabbit-HRP (1:10000). Control incubation was conducted in the same conditions without the recombinant protein. 2.9. Mouse and human Inpp4b gene structures To determine the Inpp4b exon–intron boundaries, mouse and human cDNA were aligned with the mouse and the human genomic sequence, respectively, using NCBI Blast program (www.ncbi.nlm.nih.gov).
156
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
3. Results 3.1. The murine Inpp4b gene: cDNA isolation, mapping and gene organization To isolate and characterize the murine Inpp4b gene, sequence alignment of the rat Inpp4b cDNA with mouse EST et TRACE databases sequences (www.ncbi.nlm.nih.gov) served to design murine specific primers. These oligonucleotides were used to screen by PCR a C57BL/6 spleen cDNA mouse library and one clone containing the full length Inpp4b cDNA was obtained (Supplemental Fig. S1). To genetically map Inpp4b in mouse, we first identified a microsatellite (GATA)n repeat polymorphism in the mouse Inpp4b locus and secondly, designed a specific PCR assay for this polymorphism. This PCR assay then served to localize Inpp4b by genotyping the C57BL/6 × Spretus (BSS) interspecific backcross panel consisting of 94 progenies developed at the Jackson's Laboratory. Genotyping screen defined two recombination events and localized Inpp4b on mouse Chromosome 8 in a 2.2 cM genetic interval, between the microsatellite marker D8Mit74 and the man2b1 gene (Fig. 1A). This mouse genetic interval shared a synteny homology with the human 4q27-31 region closed to the Gypa locus, near a break of synteny with human chromosome 19 p (Fig. 1B). Comparison of these two genomic physical intervals revealed a perfect conservation in genes order and transcription orientation in
mouse and human. In addition, the more detailed human physical map allowed to localize more precisely the murine Inpp4b locus in a smaller 1 megabase (Mb) interval between the Il15 and Usp38 genes (Fig. 1C). Sequence of the full length murine Inpp4b cDNA was aligned and compared to the genomic locus. The exon–intron boundaries were identified by consensus donor and acceptor signals in accordance with GT/AG rule (Supplemental Table S1). Inpp4b gene consisted of 25 exons with two 5′ untranslated exons. The 23 coding exons ranged in size from 45 to 1340 nucleotides long and intron length varied from 1.1 kb to more than 280 kb. Consequently, the entire murine Inpp4b locus is large and covered approximately 600 kb of genomic DNA. Based on the mouse gene organization, we then deduced the structure of the human INPP4B locus (Supplemental Table S2) and demonstrated a very high conservation with the mouse gene including the alternative exon 25 splicing, which produced the two INPP4B protein α and β isoforms. Remarkably, all internal coding exons have identical lengths in mouse and human, with the exception of exon 17 that is shorter by 9 bp in human in a non-conserved region (Supplemental Figure S2, and Table S1 and S2). 3.2. Transcriptional regulation of the murine Inpp4b gene To characterize the regulation of Inpp4b locus, identification of the transcription start site of Inpp4b gene was undertaken by primer extension. The template for this assay consisted of total RNA isolated from different adult mouse tissues (Fig. 2A). Interestingly, we mapped three different tissue-specific transcription start sites for Inpp4b in skeletal muscle (M), heart (H) and spleen–lung–thymus (S,L,T). The skeletal muscle transcription start site was the most proximal (+ 1) while the heartspecific was most distal (∼ − 300). The third initiation site common to spleen, lung and thymus was mapped at − 64 (Fig. 2A). Upstream sequences of the 3 start sites suggested that these were part of a TATA-less promoter. Consistently, two SP1 consensus binding sites were detected between − 140 and +104 bp. Sequence analysis of the promoter region defined several consensus sites for potential binding of known transcription factors common to mouse and human (Fig. 2B). 3.3. Inpp4 expression pattern
Fig. 3. mRNA expression patterns of the Inpp4b isoforms. A. Adult expression of the different Inpp4 isoforms in mouse tissues was determined by RT-PCR. B. Expression of Inpp4 isoforms in enriched mouse hematopoietic cells (T: T lymphocyte, B: B lymphocyte, NK: natural killer lymphocyte, MC: mast cell, MAC: peritoneal macrophage). Br: brain and C — control.
The expression pattern of the Inpp4b transcripts encoding α and β isoforms was determined in mice from various tissues by semi-quantitative RT-PCR (Fig. 3A). The Inpp4bα isoform showed a more broad tissue expression with higher levels in brain, skeletal muscle, lung, spleen, testis and thymus. In contrast, the Inpp4b beta isoform displayed restricted expression with highest levels in brain and small intestine, and lower levels in skeletal muscle and heart. Since the Inpp4b alpha isoform was expressed in spleen and in thymus, we also investigated specifically the hematopoietic lineages (Fig. 3B). Of particular interest, the Inpp4bα isoform was expressed in a subset of lineages with high levels in NK and mast enriched cell populations, moderately in B-lymphocytes and undetectable in
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
157
Fig. 4. The Inpp4b protein orthologues and paralogues. A. Schematic representation of the Inpp4b mouse, rat, human, Drosophila and nematode orthologues. Inpp4a mouse paralogue is also shown. Overall amino acid identity with mouse Inpp4b protein was indicated as percentage. Amino acid identity for the 4-Ptase active site and the C2 domains were indicated below each domain. B. Phospholipid binding specificity of the mouse Inpp4b C2 domain. Detection was achieved by hybridization with the anti-Inpp4b antibody. C. Inpp4b protein expression in mouse tissues was determined on Western blot using a specific Inpp4b antibody raised against the Nterminus region of the protein (aa 2–235). As controls, COS cells were transfected or not with the Inpp4bα cDNA. β-actin was used as internal loading control.
T-lymphocyte and peritoneal macrophage cell populations (Fig. 3B). Because the Inpp4b locus has a paralogue Inpp4a, expression of both isoforms α and β of Inpp4a was analyzed for comparison. Both Inpp4a isoforms were widely expressed in most tissues but detected at different levels. Noticeably, expression in the brain was highest for both paralogues and isoforms (Fig. 3A). 3.4. The mouse Inpp4b proteins The mouse Inpp4bα and β cDNAs encode 927 and 941 amino acids protein respectively, with a predicted molecular weight of ∼ 105 kDa. The mouse protein shares 96% and 90% identity with the human and rat proteins respectively (Supplemental Figure S2A). In addition, the
mouse Inpp4b shared 45% amino acid identity with the mouse paralogue Inpp4a (Fig. 4A). All mammalian Inpp4 proteins have in common a consensus Cys(X)5Arg phosphatase catalytic site and also a conserved C2 domain. These Inpp4 proteins were phylogenetically related to Caenorhabditis elegans 4-ptase and Drosophila melanogaster CG1846 (Fig. 4A) but no homologue could be found in yeast. In the C. elegans protein, instead of the C2 domain, the presence of another lipid binding domain a pleckstrin homology (PH) domain suggested convergent evolution (Overduin et al., 2001). The D. melanogaster Inpp4 orthologue contained both a C2 and a PH domain but the PH domain was unrelated to that of the C. elegans. Interestingly, the duplication of Inpp4 genes (Inpp4a and Inpp4b), as well as the presence of an alternative splice site generating the beta isoforms, was only found in vertebrates.
158
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
Since it was demonstrated that several C2 domains can bind lipids and subsequently recruit proteins at particular intracellular membranes (Mehrotra et al., 2000; Chung et al., 1998), we investigated which lipid family could interact with the Inpp4bC2 domain. Using an in vitro lipid-binding assay, we demonstrated that Inpp4b-C2 domain preferentially binds to phosphatidic acid and PI(3,4,5)P3 (Fig. 4B). In addition, the Inpp4b-C2 domain also binds but with a lower affinity to other phosphorylated PI and phosphatidyl serine, supporting specific lipid interaction. To analyze Inpp4b protein expression pattern and cellular localization, we have raised rabbit polyclonal antibodies against the N-terminus region of Inpp4b (aa 2–235) that can recognize both Inpp4b α and β isoforms. Inpp4b proteins were detected by Western blotting as a predominant 105 kDa band, consistent with the predicted amino acid sequence and with the full length Inpp4bα cDNA expressed in COS cells as positive control (Fig. 4C). Strong Inpp4b expression was observed in brain, spleen, small intestine, lung and testis. As determined for transcript expression pattern (Fig. 3A), the Inpp4b proteins were also expressed at highest levels in brain tissue.
Fig. 5. Cellular localization of Inpp4bα-EGFP and Inpp4bβ-EGFP isoforms. COS cells were transfected with either EGFP (A–C), Inpp4bα-EGFP (D–F) or Inpp4bβ-EGFP (G–L). Cellular localization of the different EGFP fusion proteins was detected using confocal microscopy (A, D, G, J). Cells were counterstained with anti-Golgin 97 antibody (B, E, H, K). Nuclei were stained using TO-PRO-3 dye. Merge pictures demonstrated colocalization of Inpp4bβ with Golgin (C, F, I, L).
3.5. Subcellular localization of Inpp4bα and Inpp4bβ proteins To determine the cellular localization of the different Inpp4b isoforms, full length cDNAs of Inpp4bα and Inpp4bβ were inserted in frame into an EGFP expression vector, and the chimaeric proteins were expressed in COS cells. Compared to EGFP control expressing cells (Fig. 5A–C), Inpp4bα-EGFP was distributed uniformly in the cytoplasm and absent from the nucleus (Fig. 5D–F). In contrast, Inpp4bβ-EGFP localization in COS cells was reproducibly detected in close association with the Golgi apparatus in most cells by fluorescence (Fig. 5G–I). Subsequently, immunostaining with Golgin 97, a specific marker of the Golgi apparatus, showed consistent colocalization of Golgin 97 and EGFP in Inpp4bβ-EGFP expressing cells (Fig. 5G–L). 3.6. Characterization of a novel Inpp4b isoform: Inpp4bαs Among several Inpp4b cDNA clones isolated, one of them appeared to contain all Inpp4bα exons except for exon 5 (Fig. 6A). RT-PCR carried out on mouse spleen tissues supported the synthesis of this novel isoform in vivo. With a single pair of primers localized upstream of exon 3 and in exon 8 respectively, two specific amplified products were detected of 514 and 395 bp respectively. Upon sequencing these two specific products, the shortest amplicon demonstrated lack of exon 5 sequence (119 bp) and was suspected to result from alternative splicing. This shorter amplicon provided the first evidence for a shorter cDNA, Inpp4bαs (Inpp4bα short), compared to the Inpp4bα or β full length forms. The presence of the full length and the putative short form of Inpp4bα was then investigated by RT-PCR in a wide range of mouse tissues using primers that can distinguish the two isoforms. Two amplicons were observed in different ratios depending on the individual tissues or organs, supporting the existence in vivo of a novel isoform of Inpp4bα. Interestingly, the Inpp4bαs form was virtually undetectable in the brain, suggesting a tissue-specific regulation for this alternative splicing (Fig. 6B). Based on the exclusion of exon 5 and usage of the full length Inpp4b translation initiation codon, the Inpp4bαs transcript should undergo a frameshift that generates a premature stop codon in exon 6. Because a protein band of ∼ 7.5 kDa (65 aa) for Inpp4bαs was not detected by Western (data not shown) and that the Inpp4bαs cDNA contained all Inpp4bα sequences downstream of exon 6, we thus hypothesized that an alternative translation initiation site for the Inpp4bαs mRNA could be used. We found one other possibility for a translation codon at 50 nucleotides downstream from the original start codon (Fig. 6C,D). To test this hypothesis, we fused in frame the Inpp4bαs or Inpp4bα cDNA upstream of EGFP in expression vectors. These expression vectors were transfected in COS cells and expression of the resulting fusion protein species was characterized by Western blot. Anti-EGFP detected a ∼ 135 and ∼126 kDa band for the Inpp4bα-EGFP and Inpp4bαsEGFP fusion protein extracts, respectively (Fig. 6E). This Inpp4bαs cDNA could be transcribed in cells and was identical
M. Ferron, J. Vacher / Gene 376 (2006) 152–161 Fig. 6. Characterization of Inpp4bαs, a novel Inpp4b isoform. A. Schematic representation of Inpp4bα and Inpp4bαs 5′ mRNAs regions. Position of the deduced start codon (M) is indicated. Arrows indicate the position of the primers used in RT-PCR. B. Tissue expression pattern of Inpp4bαs was determined by RT-PCR. The specific shorter 395 pb amplification product is specific of Inpp4bαs. C and D. cDNA and protein sequences of Inpp4bα and Inpp4bαs respectively. The sequence corresponding to exon 5, excluded in Inpp4bαs, is underlined and in bold. The alternative N-terminus sequence of Inpp4bαs, absent in Inpp4bα, is bolded in D. Upstream in frame stop codons are underlined in both sequences. The start codon of Inpp4bα is bolded in the Inpp4bαs sequence. Arrows represent primers used in RT-PCR. E. The Inpp4bαs mRNA contains a functional open reading frame. cDNAs corresponding to Inpp4bα and Inpp4bαs were cloned upstream of EGFP in an expression vector and transfected in COS cells. Expression of the resulting fusion proteins was assessed by Western blotting using antibodies against Inpp4b or EGFP on protein extracts. In the right panel, equal amounts of protein extracts were analyzed (25 μg), and only a 1/100 (0.25 μg) of Inpp4bα-EGFP protein extract were analyzed in the middle and left panels. F. Subcellular localization of Inpp4bα-EGFP and of Inpp4bαs-EGFP in COS cells (upper panels). Nuclei were stained with Hoechst (middle panels) and merged images demonstrated a cytoplasmic localization for both isoforms (bottom panels). 159
160
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
to the full length Inpp4bα. With our anti-Inpp4b primary antibodies that should recognize the Inpp4bαs, we confirmed the existence of Inpp4bαs-EGFP fusion protein of similar molecular weight as observed with the EGFP antibody and provided definitive evidence of a different translation initiation site for the novel Inpp4bαs cDNA. This Inpp4bαs cDNA consisted of all the protein domains of Inpp4bα except for the 85 N-terminal amino acids substituted by 24 different amino acids and was translated into a protein of ∼ 98 kDa (866 aa) that contained less than half of the C2 domain. To determine the cellular localization of Inpp4bαs isoform in comparison to Inpp4bα, the fusion EGFP expression vectors were used in COS cells. Interestingly, Inpp4bαs EGFP fusion proteins demonstrated a similar cytosolic localization as detected for the full length Inpp4bα (Fig. 6F). Together, these results demonstrated that the Inpp4bαs represent a novel isoform of the Inpp4b protein in addition to the differentially spliced α and β isoforms. 4. Discussion Coordinate regulations of phosphatidyl inositol kinases and phosphatases play a key role in signaling pathways that control cell proliferation, differentiation and survival, but several phosphatases remain to be identified and characterized. This manuscript reports the isolation and characterization of the mouse locus for the Inpp4b inositol phosphatase. Our results showed that the murine Inpp4b gene structure and protein α and β isoforms were highly similar to human and rat homologues. Interestingly, we identified a novel alternative form of Inpp4bα. Furthermore, the three Inpp4b protein isoforms have a differential expression pattern as well as a specific cellular localization. Because phosphatases are essential to counterbalance kinase production of active PIPs, the second messengers, we have isolated, mapped and analyzed the murine Inpp4b gene. Inpp4b was localized on chromosome 8 between Il15 and Usp38 in a region of synteny homologous to 4q27-31 in human. The mouse Inpp4b gene and protein displayed high structure similarity to the rat and human genes and proteins with conserved consensus C2 domain and phosphatase active site. The conservation of a Inpp4-like gene in more primitive organisms such as insects and nematodes, and absence in yeast, suggested that this gene was specific of metazoans. In addition, Inpp4b gene in vertebrates has a paralogue Inpp4a with close genomic similarity that has not been investigated for regulation. Our characterization of the mouse and human Inpp4b gene structure showed two isoforms, alpha and beta, resulting of alternative splicing of exon 25 as defined for rat and human Inpp4a. The Inpp4b locus has three major transcriptional initiation sites that appeared to be tissue-specific. Our analysis of the Inpp4b upstream regulatory sequences also detected several consensus binding sites for different classes of transcription factors, and suggested complex transcriptional regulation for both Inpp4b isoforms in individual tissues. Moreover, we have identified a novel isoform of Inpp4bα, called Inpp4bαs, that resulted from a downstream translation initiation site and
exon 5 skipping of Inpp4bα. Characterization of this novel Inpp4bαs showed that the 85 N-terminal amino acids present in Inpp4bα were substituted by a different sequences of 24 amino acids. The expression patterns of both Inpp4b and Inpp4a genes were conjointly established for comparison analysis due to their close similarity. The murine Inpp4bα isoform exhibited relatively ubiquitous expression with a wide tissue distribution very similar to the murine Inpp4aα isoform. In apparent opposition to this broad expression pattern, expression of Inpp4bα was confined to specific cells of the hematopoietic lineages. Of interest, only the Inpp4bα isoform displayed a high level of expression in enriched mast, lymphocyte, and NK cell subpopulations relative to the undetectable expression of Inpp4bβ and of Inpp4aα and β isoforms, suggesting most likely a unique role in these cells. In comparison to Inpp4bα, the Inpp4bβ isoform showed a more restricted expression pattern to brain, heart, intestine, skeletal muscle and spleen. Noticeably, both α and β isoforms of Inpp4b and Inpp4a displayed highest expression in the brain, consistent with the neuronal defect in the Inpp4a weeble mutant (Nystuen et al., 2001). Particularly striking is the complete absence of expression of the Inpp4bαs isoform in this tissue in contrast to its presence in all other tissues analyzed. Together, these results point to both overlapping and complementary role for each of the five Inpp4 isoforms examined in this study. Our study of the Inpp4b proteins has revealed that in contrast to the α isoform, the β isoform had a hydrophobic C-terminal domain and consistently, these isoforms displayed specific subcellular localization. While the intracellular localization of the full length and short form of the Inpp4bα protein was mainly cytosolic, that of the Inpp4bβ isoform was associated with the Golgi apparatus. Interestingly, the differential subcellular localization of the Inpp4bα and β phosphatase isoforms was similar to that observed for the two phosphatidylinositol 4-kinase type II (PI4KII) isoforms. Indeed, the PI4KII α isoform was located at the trans-Golgi network whereas PI4KIIβ was localized exclusively in the cytosol (Wang et al., 2003; Wei et al., 2002). Such localization of the Inpp4b isoforms was complementary to that of the Inpp4a isoform associated with endosomes (Shin et al., 2005; Ivetac et al., 2005). The localization of individual Inpp4 protein isoform in non-redundant cellular compartments could implicate different molecular interactions or cellular function. In accordance with this precise subcellular localization, our results of Inpp4b-C2 protein–lipid interaction demonstrated preferential binding to phosphatidic acid and to PI(3,4,5)P3. Of note, this preferential binding of Inpp4b-C2 appeared quite different from that of Inpp4a-C2 domain shown to interact mainly with PI(3,4)P2 and PI(4)P (Ivetac et al., 2005). Hence, the binding factors to the C2 domain and the targeted subcellular localization of the multiple isoforms of Inpp4b are likely to impact on the protein/substrate interactions specificity and play a critical role in the cellular function of these proteins. Taken together, our characterization of the murine Inpp4b and Inpp4a demonstrated high gene and protein conservation. The differential Inpp4b expression pattern, cellular sublocalization
M. Ferron, J. Vacher / Gene 376 (2006) 152–161
and interacting factors are likely to ascribe highly specific function for Inpp4 phosphatase proteins. Future studies on the Inpp4b gene should be directed to decipher the in vivo functional role of Inpp4b in mouse. Acknowledgments We thank M. Pata for excellent technical assistance. We are grateful to Drs S. Meloche and A. Veillette for vector and RNA samples, and to Dr A. Kania for critical reading of the manuscript. This work was supported by Canadian Institutes of Health Research (MOP44079) and National Sciences and Engineering Council of Canada (OPG0155277) grants to J.V. and a doctoral training award to M.F. from the Canadian Institutes of Health Research. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.gene.2006.02.022. References Barnache, S., Le Scolan, E., Kosmider, O., Denis, N., Moreau-Gachelin, F., 2006. Phosphatidylinositol 4-phosphatase type II is an erythropoietinresponsive gene. Oncogene 25, 1420–1423. Benachenhou, N., Massy, I., Vacher, J., 2002. Characterization and expression analyses of the mouse Wiskott–Aldrich syndrome protein (WASP) family member Wave1/Scar. Gene 290, 131–140. Cantley, L.C., 2002. The phosphoinositide 3-kinase pathway. Science 296, 1655–1657. Chung, S.H., et al., 1998. The C2 domains of Rabphilin3A specifically bind phosphatidylinositol 4,5-bisphosphate containing vesicles in a Ca2+dependent manner. In vitro characteristics and possible significance. J. Biol. Chem. 273, 10240–10248. Coulombe, P., Meloche, S., 2002. Dual-tag prokaryotic vectors for enhanced expression of full-length recombinant proteins. Anal. Biochem. 310, 219–222. De Matteis, M.A., Godi, A., 2004. PI-loting membrane traffic. Nat. Cell Biol. 6, 487–492. Ivetac, I., et al., 2005. The type Ialpha inositol polyphosphate 4-phosphatase generates and terminates phosphoinositide 3-kinase signals on endosomes and the plasma membrane. Mol. Biol. Cell 16, 2218–2233. Leslie, N.R., Downes, C.P., 2002. PTEN: the down side of PI 3-kinase signalling. Cell. Signal. 14, 285–295.
161
Majerus, P.W., Kisseleva, M.V., Norris, F.A., 1999. The role of phosphatases in inositol signaling reactions. J. Biol. Chem. 274, 10669–10672. Mehrotra, B., Myszka, D.G., Prestwich, G.D., 2000. Binding kinetics and ligand specificity for the interactions of the C2B domain of synaptogmin II with inositol polyphosphates and phosphoinositides. Biochemistry 39, 9679–9686. Munday, A.D., Norris, F.A., Caldwell, K.K., Brown, S., Majerus, P.W., Mitchell, C.A., 1999. The inositol polyphosphate 4-phosphatase forms a complex with phosphatidylinositol 3-kinase in human platelet cytosol. Proc. Natl. Acad. Sci. U. S. A. 96, 3640–3645. Norris, F.A., Auethavekiat, V., Majerus, P.W., 1995. The isolation and characterization of cDNA encoding human and rat brain inositol polyphosphate 4-phosphatase. J. Biol. Chem. 270, 16128–16133. Norris, F.A., Atkins, R.C., Majerus, P.W., 1997. The cDNA cloning and characterization of inositol polyphosphate 4-phosphatase type II. Evidence for conserved alternative splicing in the 4-phosphatase family. J. Biol. Chem. 272, 23859–23864. Nystuen, A., Legare, M.E., Shultz, L.D., Frankel, W.N., 2001. A null mutation in inositol polyphosphate 4-phosphatase type I causes selective neuronal loss in weeble mutant mice. Neuron 32, 203–212. Overduin, M., Cheever, M.L., Kutateladze, T.G., 2001. Signaling with phosphoinositides: better than binary. Mol. Interv. 1, 150–159. Rajapurohitam, V., Chalhoub, N., Benachenhou, N., Neff, L., Baron, R., Vacher, J., 2001. The mouse osteopetrotic grey-lethal mutation induces a defect in osteoclast maturation/function. Bone 28, 513–523. Rowe, L.B., et al., 1994. Maps from two interspecific backcross DNA panels available as a community genetic mapping resource. Mamm. Genome 5, 253–274. Shin, H.W., et al., 2005. An enzymatic cascade of Rab5 effectors regulates phosphoinositide turnover in the endocytic pathway. J. Cell Biol. 170, 607–618. Takumi, T., Lodish, H.F., 1994. Rapid cDNA cloning by PCR screening. Biotechniques 17, 443–444. Vyas, P., Norris, F.A., Joseph, R., Majerus, P.W., Orkin, S.H., 2000. Inositol polyphosphate 4-phosphatase type I regulates cell growth downstream of transcription factor GATA-1. Proc. Natl. Acad. Sci. U. S. A. 97, 13696–13701. Wang, Y.J., et al., 2003. Phosphatidylinositol 4 phosphate regulates targeting of clathrin adaptor AP-1 complexes to the Golgi. Cell 114, 299–310. Ware, M.D., Rosten, P., Damen, J.E., Liu, L., Humphries, R.K., Krystal, G., 1996. Cloning and characterization of human SHIP, the 145-kD inositol 5phosphatase that associates with SHC after cytokine stimulation. Blood 88, 2833–2840. Wei, Y.J., et al., 2002. Type II phosphatidylinositol 4-kinase beta is a cytosolic and peripheral membrane protein that is recruited to the plasma membrane and activated by Rac-GTP. J. Biol. Chem. 277, 46586–46593.