- Email: [email protected]
Identification and distribution of mouse carboxypeptidase A-6
Molecular Brain Research 137 (2005) 132 – 142 www.elsevier.com/locate/molbrainres
Research report
Identification and distribution of mouse carboxypeptidase A-6 Jose D. Fontenele-Neto1,2, Elena Kalinina1, Yun Feng1, Lloyd D. FrickerT,1 Department of Molecular Pharmacology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA Accepted 21 February 2005 Available online 1 April 2005
Abstract Carboxypeptidase A-6 (CPA6) was recently discovered in the human genome. To gain information regarding the potential function of this novel protein, the mouse homolog of CPA6 was identified using a combination of bioinformatics and reverse transcriptase-polymerase chain reaction (RT-PCR). In addition, homologs in rat, chicken, and frog were identified using a bioinformatics approach. The distribution of CPA6 mRNA in mouse tissues was examined using RT-PCR and in situ hybridization. A strong RT-PCR signal is detectable in olfactory bulb, and much lower levels are present in other regions such as the cerebral cortex, hippocampus, hypothalamus, striatum, and medulla. In peripheral tissues, a moderate RT-PCR signal is present in epididymis, and low levels are detectable in colon and spleen. The high level of CPA6 in adult mouse brain olfactory bulb was confirmed by in situ hybridization. Lower levels of CPA6 mRNA were found to be present in the cingulate cortex, lateral septum, pontine nucleus, and inferior olivary nucleus of the hindbrain. Within the olfactory bulb, CPA6 mRNA is enriched in the mitral and granular layer. A lower level of CPA6 mRNA is present in the internal and external plexiform layers, and no signal is detectable in the olfactory nerve layer. The distribution was also examined in whole embryos at embryonic day 14.5 and CPA6 mRNA was found to be enriched in eye, ear, osteoblasts, stomach, skin, dorsal root ganglia, and throughout the CNS. The presence of CPA6 mRNA in the rectus muscle layer of the eye at embryonic day 14.5 is consistent with the observation that the CPA6 gene is disrupted in a patient with Duane syndrome, a congenital eye defect. Taken together, the distribution of CPA6 suggests a specific role in a limited number of tissues, and it is possible that this role involves an aspect of cell migration. D 2005 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Gene structure and function Keywords: Peptide processing; Metallocarboxypeptidase; Neuropeptide; Olfactory bulb; Duane syndrome
1. Introduction Carboxypeptidases (CPs) perform many important functions within a range of organisms, including protein/ peptide digestion as well as more selective biosynthetic roles [6,27]. CPs that contain an active site zinc, termed metallocarboxypeptidases, are further divided into two
T Corresponding author. Fax: +718 430 8954. E-mail address: [email protected] (L.D. Fricker). 1 All authors contributed equally. 2 Present address: Department of Veterinary Medicine, Escola Superior de Agricultura de Mossoro´, BR 110 KM47 Mossoro´, RN 59625-900, Brazil. 0169-328X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2005.02.026
subfamilies based on their amino acid sequences [22]. One subfamily, termed the N/E subfamily, contains mammalian CPE, CPN, CPM, CPD, CPZ, CPX1, CPX2, and a protein termed aortic carboxypeptidase-like protein (ACLP) and AEBP1. Five of the members of this subfamily (CPE, N, M, D, and Z) cleave C-terminal basic amino acids (Lys or Arg) from the C-termini of peptides and are thought to function in the selective biosynthesis or processing of neuroendocrine peptides, growth factors, and other molecules involved in cell–cell signaling [14– 16,30,31]. The other three members of this family (CPX1, X2, and ACLP/AEBP1) are not active towards standard substrates, consistent with the absence of key catalytic residues [27]. It has been proposed that these
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
inactive members function as binding proteins, rather than active enzymes, although it is possible that they exhibit activity towards unknown molecules. The other subfamily of metalloCPs, termed the A/B family, includes mammalian CPB1, CPU (also termed CPB2, plasma CPB, and other names), CPA1, CPA2, CPA3 (also termed mast-cell CPA), CPA4 (previously named CPA3), CPA5, CPA6, and CPO [3–5,19,20,33,38]. Members of this A/B subfamily are produced as inactive zymogens that require activation, in contrast to the members of the N/E family which do not require an activation step. Once activated, CPB1 and CPU cleave C-terminal basic residues from peptides [5,19]. The substrate specificities of CPA1, 2, 3, and 5 have been examined; all are able to remove aliphatic or hydrophobic residues from the Ctermini of peptides [3,4,33,38]. The substrate specificities of CPA4, CPA6, and CPO have not been experimentally determined, and modeling suggests that CPA4 and 6 cleave aromatic/aliphatic residues while CPO cleaves acidic Cterminal residues [38]. The functions of several members of the A/B subfamily of CPs have been well studied. CPA1, CPA2, and CPB1 are secreted from the exocrine pancreas and function in the digestion of food in the intestine, following the action of chymotrypsin, trypsin, and pepsin [3–5]. CPA3 is secreted from mast cells and helps degrade a number of substances, following the action of chymase [33]. CPU is secreted from the liver and circulates in plasma as an inactive precursor form which is activated by plasmin or thrombin. Once active, CPU removes C-terminal Lys residues generated by the action of plasmin on fibrin, thus decreasing the binding of plasminogen to fibrin and inhibiting fibrinolysis [19]. Because the CPAs that have been characterized have overlapping enzyme substrate specificities, the key issue relating to the function of each enzyme is their unique distribution. Thus, one purpose of the present study was to determine the distribution of CPA6 in order to gain insights towards the function of this recently described member of the CP family. Previously, CPA6 mRNA expression was detected by RT-PCR in human fetal umbilical cord blood and in mouse brain [38]. In the present study, RT-PCR was used to determine which tissues express detectable levels of CPA6 mRNA. Then, in situ hybridization was used to precisely map the location of CPA6 mRNA in adult mouse brain and embryonic day 14.5 mouse tissues. A second goal of the current studies was to resolve discrepancies between the previously identified sequence of human CPA6 and the apparent mouse homologs of CPA6 in GenBank. All of these mouse homologs appear to lack exon 3, and as a result would not express functional protein. RTPCR was used to determine that the GenBank sequences do not represent the most abundant full-length form of mouse brain CPA6, which is very similar to human CPA6 and has an exon 3. In addition, a bioinformatics approach was used to identify putative homologs of CPA6 in rat, chicken, and
133
frog (Xenopus laevis). The observation that CPA6 appears to be highly conserved among these species suggests an important function for this protein.
2. Experimental procedures 2.1. Bioinformatics GenBank NR databases were searched with human CPA6 nucleotide and amino acid sequences using the blastn and tblastn programs (http://www.ncbi.nlm.nih.gov/BLAST). CPA6 sequences were also used to search genomic databases for mouse, rat, chicken (Gallus gallus), Drosophila melanogaster, Caenorhabditis elegans, and Saccharomyces cerevisiae. In those cases where the non-redundant GenBank database contained sequences that appeared to be missing an exon, based on the sequence of human CPA6, the genomic sequence was analyzed using the GT-AG rule for intron splice sites to find additional exons that would provide the missing sequence. 2.2. Animals Mice (C57BKS) were initially obtained from the Jackson Laboratory (Bar Harbor, ME) and bred within the barrier facility at Albert Einstein College of Medicine. Food (Purina rodent chow 5058) and water were provided ad libitum. The experimental protocol was approved by the committee for animal experimentation of the Albert Einstein College of Medicine. 2.3. RNA extraction and reverse transcriptase (RT)-PCR Total RNA from different brain regions (olfactory bulb, cerebral cortex, cerebellum, striatum, hypothalamus, hippocampus, and medulla) and several organs (testis, epididymis, ovary, lungs, spleen, liver, colon, and adrenal) from both male and female mice were obtained using an RNA easy kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. The RT-PCR was performed using the Titan one tube RT-PCR system (Roche Diagnostics GmbH, Mannheim, Germany) amplifying a sequence of 582 base pairs at the N-terminal region of CPA6 mRNA (exons 1 to 4) using TGCCTAAAGTGTCACATTTTGCTCA (nucleotides 62–86) and CTAAGGAGTGATACACC TCGTAGTTG (nucleotides 618–643) as forward and reverse primers, respectively. Briefly, 300 ng of total RNA was mixed with primers, dNTPs, and an enzyme mix containing avian myeloblastosis virus reverse transcriptase and a proof reading DNA polymerase. The reverse transcription was carried out for 30 min at 50 8C followed by a denaturation step at 94 8C for 2 min and by 35 PCR cycles (94 8C for 10 s, 59 8C for 45 s, 68 8C for 45 s). The resulting cDNA was separated in a 1% agarose gel containing ethidium bromide.
134
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
To determine if full-length mouse CPA6 mRNA is present in brain as predicted from the bioinformatics, oligonucleotides spanning the coding region were used for RT-PCR as described above, except that the elongation step was typically 3 min per cycle. Two forward oligonucleotides (TGCCTAAAGTGTCACATTTTGCTCA and GTTGTAAGCCACCTCCAGCCTC) and two reverse oligonucleotides (TCTCAAGGACATTTCTTTAGCAGGT and GTTTGCCAAGGTGCCTTTTGATCAG) were used and the resulting PCR product was subcloned and sequenced using a variety of internal primers. These oligonucleotides correspond to mouse CPA6 nucleotide positions 62–86, 143–164, 1511–1535, and 1576–1600, respectively. A minimum of two clones from independent RT-PCR reactions were sequenced to reduce the possibility of PCR-induced errors. 2.4. In situ hybridization Adult mice (2–3 months old) were sacrificed in a CO2 chamber. Brain, testis, epididymis, and colon were dissected out and frozen in n-hexane (Sigma, St Louis, MO) at 50 8C and stored at 80 8C. Brain coronal and sagittal sections (20 Am thick) were serially cut and transverse sections of the other tissues (15 Am thick) were also obtained. For the analysis of embryonic tissue, female mice were housed overnight with a male and the following morning examined for a vaginal plug. Those mice exhibiting vaginal plugs were sacrificed with CO2 vapor 14 days later. Embryos and their placenta were dissected out and frozen in n-hexane at 50 8C and stored at 80 8C. Transverse sections of the embryo and placenta were made. All sections were thaw mounted onto superfrostR/Plus slides (Fisher Scientific, Pittsburgh, PA), and stored at 80 8C until use. To generate the CPA6 probe, RT-PCR was used to generate a cDNA fragment of 860 base pairs corresponding to the N-terminal coding region of mouse CPA6 (exons 1–7). This cDNA fragment was subcloned into the EcoRI and BamHI sites of the pGEM-7Zf ( ) vector (Promega, Madison, WI). The resulting plasmid was used as a template to generate 35S-labeled riboprobes. Radiolabeled antisensestrand RNA probe was transcribed from an XbaI-linearized template with SP6 RNA polymerase (Invitrogen), along with 35S-uridine triphosphate (1250 Ci/mmol; PerkinElmer Life Sciences, Boston, MA). Labeled sense-strand probe was transcribed similarly from a SacI-linearized template with T7 RNA polymerase (Invitrogen). Unincorporated nucleotides were removed by chromatography with Sephadex G-50 (Amersham Biosciences), and probes were stored in 20 mM dithiothreitol (DTT) at 80 8C. Brain sections were warmed to room temperature, incubated for 5 min in 0.1 M sodium phosphate buffer (pH 7.4), fixed for 15 min with 4% paraformaldehyde in 0.1 M sodium phosphate buffer at 4 8C, and rinsed for 10 min in 0.1 M sodium phosphate buffer containing 0.15 M sodium chloride. Following two washes of 3 min each in 2
standard saline citrate (SSC; 300 mM sodium chloride, 30 mM sodium citrate, pH 7.0), sections were incubated in 0.1 M triethanolamine (pH 8.0) for 5 min and treated for 10 min with 0.25% acetic anhydride in 0.1 M triethanolamine. Finally, they were rinsed twice for 5 min in 2 SSC each and dehydrated in a graded ethanol series. Air-dried sections were stored at room temperature until hybridization. The probe was denaturated at 90 8C for 5 min and diluted to a final concentration of 107 cpm/ml in hybridization buffer consisting of 50% formamide, 10% dextran sulfate, 300 mM sodium chloride, 10 mM Tris–HCl (pH 7.5), 1 mM ethylenediaminetetraacetic acid (EDTA), 100 Ag/ml salmon sperm DNA, Denhardt’s solution (0.02% bovine serum albumin, 0.02% Ficoll, 0.02% polyvinylpyrolidone), and 20 mM DTT. Sections were covered with hybridization mix (100 Al/slide), coverslipped, sealed with rubber cement, and incubated for 18–20 h in a humidified chamber at 55 8C. Subsequently, the coverslips were carefully removed and the slides were rinsed twice at room temperature for 5 min each in 2 SSC. The slides were then incubated at 37 8C for 30 min with RNase A (Qiagen) diluted to 20 Ag/ml in a buffer containing 500 mM sodium chloride, 10 mM Tris–HCl (pH 7.5), 1 mM EDTA, and 1 mM DTT. They were washed twice for 15 min in 2 SSC at 37 8C, incubated for 30 min at 60 8C in 0.5 SSC, washed for 30 min at 65 8C in 0.1 SSC, and rinsed for 5 min in 0.1 SSC at room temperature. They were dehydrated in a graded ethanol series and allowed to air dry. A first set of autoradiograms was generated by apposition of labeled sections to X-Omatk Blue X-1 film (Eastman Kodak Co, Rochester, NY) for 4–10 days at room temperature. The slides were then dipped in Hypercoat LM-1 nuclear emulsion (Amersham Biosciences, Piscataway, NJ). Sections were stored in lightproof boxes at 4 8C for 2–4 weeks. Autoradiograms were developed for 4 min in D19 (Eastman Kodak) at 18 8C, rinsed in water, and fixed for 10 min in Kodak Polymax Fixer (Eastman Kodak). Sections were then counterstained lightly with Mayer’s hematoxylin, dehydrated in a graded ethanol series, defatted in xylene, mounted in cytosealk (Stephens Scientific, Riverdale, NJ), coverslipped, and analyzed under a Zeiss Axiophot microscope.
3. Results Two cDNA sequences with substantial sequence homology to human CPA6 are present in the mouse non-redundant GenBank database (NM177834 and AK078883). However, both of these sequences lack exon 3, which shifts the open reading frame and leads to a truncated protein that lacks the carboxypeptidase domain. An alternative methionine in the upstream region was predicted in the GenBank entries, but this would not produce a protein with a signal peptide and the resulting protein would not be expected to be expressed in the secretory pathway. To determine whether these forms
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
represent the major species of mouse CPA6, or if there were additional forms containing exon 3, RT-PCR was used with oligonucleotides in exons 1 and 4. The predicted product if no exon 3 was present was 457 nucleotides; with exon 3 the product would be 582 nucleotides. In olfactory bulb, a major band of approximately 582 nucleotides and a faint band at approximately 457 nucleotides were observed (Fig. 1). Sequence analysis of the PCR product confirmed the presence of the full-length insert. In addition, RT-PCR was performed with oligonucleotides located in the 5V and 3V untranslated regions of mouse CPA6 mRNA and the resulting cDNA product was sequenced using a variety of internal primers. The resulting sequence contained all the appropriate exons that were previously identified in human CPA6 (Fig. 2). The sequence missing from the previous GenBank entries for mouse CPA6 is indicated by the solid line and is clearly present in mouse CPA6. The N-terminal sequence is predicted to encode a signal peptide, with cleavage occurring between positions 30 and 31 (cleavage between Cys and His). Following this signal peptide is a prodomain of approximately 100 amino acids. The putative cleavage site of the prodomain (RNRR-S, with cleavage predicted to occur between the R and S) matches the general consensus site for proprotein convertases [29]. After the prodomain, mouse CPA6 has a 309 residue carboxypeptidase-like domain. All of the active site residues necessary for catalytic activity of other members of the metallocarboxypeptidase family are present in mouse CPA6, including the Zn2+-binding residues (H197, E200, and
Fig. 1. Analysis of the distribution of CPA6 mRNA in mouse tissues using RT-PCR. The oligonucleotides were located 582 base pair apart, as described in Experimental procedures. The sizes of DNA standards (in base pairs) are shown. Panel A: Analysis of RNA from different brain regions. Abbreviations: OB, olfactory bulb; Cx, cerebral cortex; Hip, hippocampus; Hyp, hypothalamus; Str, striatum; Cb, cerebellum; Med, medulla. Panel B: Analysis of RNA obtained from peripheral tissues. Abbreviations: Col, colon; Adr, adrenal; Spl, spleen; Liv, liver; Lng, lung; Ovr, ovary; Epid, epidymus; Test, testis. Results are shown for a representative experiment; each tissue/brain region was analyzed in three separate experiments with comparable results.
135
H325), the glutamic acid involved in catalytic activity (E399), and several residues involved in substrate binding (Fig. 2). One of these residues is methionine in position 384, which corresponds to the residue that sits in the bottom of the binding pocket for the side chain of the substrate and either conveys specificity for aromatic/aliphatic residues (I255 or L255 in CPA1, CPA2, and CPA3, using the numbering system of the mature form of bovine CPA1) or for basic residues (D255 in CPB and CPU, using the same numbering system). The deduced amino acid sequence of mouse CPA6 has 84% amino acid identity to human CPA6 and differs in length by a single amino acid (Fig. 3). Using a bioinformatics approach, a rat homolog of CPA6 was pieced together from genomic sequences (for exons 1–2) and from two GenBank non-redundant clones, one which contained only exons 3–6 (XM345496) and another which contained only exons 7–11 (XM232613). The predicted rat CPA6 homolog is the same length as mouse CPA6 and has 83% amino acid sequence identity to human CPA6. A chicken CPA6 homolog was also found, but the hypothetical protein encoded by the predicted sequence (AJ720827) does not contain a signal peptide. However, this GenBank sequence does not correspond to an actual cDNA sequence but was merely predicted from analysis of potential intron splice sites within the genome. If one considers an alternative intron splice site within the chicken CPA6 gene between exons 2 and 3 (maintaining the GT-AG rule for intron splicing) then the resulting protein would have a long open reading frame that has a signal peptide and substantial sequence identity (75%) with human CPA6 (Fig. 3). This hypothetical chicken protein would be 8 amino acids shorter than the mouse and rat homologs due to a shorter signal peptide. In addition to these GenBank sequences, a Xenopus sequence showed a strong match to CPA6; this Xenopus sequence appeared correct in the GenBank database and did not need any refinements to match with the other sequences. The Xenopus sequence has 65% amino acid sequence identity with human CPA6 and is 3 residues shorter (Fig. 3). All of the key active site residues in human CPA6 are conserved in the various homologs including H69, E72, R127, R145, H196, Y248, and E270 (by convention, the numbering system of mature bovine CPA1 is used). Importantly, the methionine in the position corresponding to the binding pocket of the substrate side chain (residue 255 in CPA/B) is present in CPA6 from all species (Fig. 3). Searches of Drosophila, C. elegans, and S. cerevisiae databases failed to detect any clear homologs of CPA6 in these species. All three of these organisms contain sequences with sequence similarity to general A/B subfamily carboxypeptidases, but none had greater overall similarity to CPA6 than to other members of the mammalian CPA/B subfamily. In considering the methionine in the position of the substrate binding pocket, three Drosophila sequences were found which contain an identical residue in this position. None of the C. elegans CPA-like sequences, or
136
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
Fig. 2. Nucleotide and deduced amino acid sequence of mouse CPA6. The nucleotide sequence was initially predicted from a bioinformatics approach. Then, RT-PCR was used to generate CPA6 cDNA corresponding to the full-length coding region, and this was sequenced to confirm that the prediction was correct. Arrows indicate the position of introns, and the numbers adjacent to these arrows indicate the exon. The nucleotide sequence that is underlined corresponds to exon 3, which was missing from previous mouse cDNA sequences deposited in GenBank. Active site or substrate binding residues that are mentioned in the text are boxed. The nucleotide sequence of full-length mouse CPA6 has been submitted to GenBank (AY773477).
the single S. cerevisiae CPA/B-like sequence, contained a methionine in the corresponding position. The gene for human CPA6 was previously found to be much larger than the genes for all other members of the A/B subfamily of carboxypeptidases [38]. To examine if this feature was conserved, the exon/intron junctions and the intron sizes were compared for mouse, human, and chicken CPA6 homologs. The overall length of the mouse CPA6 gene (c393 kb) is larger than that for the human gene (c315 kb), and the two longest introns are the same for both human and mouse CPA6 (Fig. 4). The chicken gene is considerably shorter (c86 kb) than the human and mouse
CPA6 genes, although still larger than a typical CPA/B gene [38]. Interestingly, the first intron is the longest in chicken CPA6 (41.8 kb), as is the case for human and mouse CPA6. The rat CPA6 gene also has a very large first intron (132.6 kb; data not shown). Thus, this feature of a large first intron for CPA6 is conserved in all species investigated. To gain an understanding as to possible functions of CPA6, the distribution was examined in mouse. Using RTPCR, CPA6 mRNA was detectable in the brain with the highest level of expression in the olfactory bulb (Fig. 1A). Several other brain regions (cerebral cortex, hippocampus, hypothalamus, striatum, and medulla) showed low levels of
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
137
Fig. 3. Alignment of human and mouse CPA6 with the predicted sequences of rat, chicken (G. gallus), and frog (X. laevis) CPA6. The sequence of frog CPA6 was obtained directly from GenBank. The rat and chicken sequences were pieced together from genomic sequences and partial GenBank sequences, as described in the Results section. Asterisks (bottom line) indicate those residues that are conserved (Cons.) in all species. The predicted cleavage site of the prodomain (RXRR) and the various active site or substrate binding residues are indicated. By convention, the numbering of these residues is based on the mature form of bovine CPA1. The amino acid sequence identity between CPA6 from each species and the human protein is indicated (bottom right).
CPA6 expression. In cerebellum, CPA6 mRNA levels were below the detection limit (Fig. 1A). Both male and female mice showed comparable levels of expression within the brain (data not shown). Of the peripheral tissues examined, CPA6 mRNA was detected at moderate levels in the epididymis and at low levels in colon, and spleen (Fig. 1B). No signal or extremely weak signals for CPA6 mRNA were observed in the adrenal, liver, lung, ovaries, and testis.
In situ hybridization carried out in adult brain sections revealed a similar pattern as observed using the RT-PCR technique, with strong labeling in the olfactory bulb (Fig. 5A). No signal was observed with control sense cRNA probes (not shown, and Fig. 5H). CPA6 mRNA was abundant in both main and accessory olfactory nuclei. The hybridization signal was observed mainly over the mitral and granular cell layers, with a lower hybridization signal
Fig. 4. Intron/exon junctions and intron sizes for mouse, human, and chicken CPA6. The human CPA6 gene organization was previously reported [38]. Intron lengths (in kb) are indicated.
138
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
Fig. 5. Analysis of the distribution of CPA6 in mouse brain using in situ hybridization. Panel A: Film autoradiogram of a mouse brain sagittal section showing the specific hybridization signal with the antisense cRNA probe for CPA6. No signal was detected when sense probes were used (not shown). Panels B–H: Dark-field (B–F and H) and bright-field (G) emulsion dipped autoradiograms of coronal sections of olfactory bulb hybridized for CPA6 mRNA. Note that all layers of the olfactory bulb are labeled and that both main and accessory nuclei show a hybridization signal. Intense labeling is observed over the mitral and granule layer of the main olfactory bulb. The control antisense probe shows no signal (H). Panels I–L: Dark-field emulsion dipped autoradiograms of mouse brain coronal sections hybridized with antisense CPA6 cRNA probe. Note that panel I shows a section through the basal forebrain, including the septum as well as the cingulate cortex, panels J and K show sections through the hippocampus, and panel L shows the dorsal nucleus of the inferior olivary complex and cerebellum. Abbreviations: AOB, accessory olfactory bulb; Cb, cerebellum; cc, corpus callosum; Cg, cingulate cortex; DG, Dentate gyrus; EPlA, external plexiform layer of the accessory olfactory bulb; EPl, external plexiform layer of the olfactory bulb; fi, fimbria of hippocampus; Hip, hippocampus; Hyp, hypothalamus; GrA, granular cell layer of the accessory olfactory bulb; GrO, granular cell layer of the olfactory bulb; IOD, inferior olivary complex; LS, lateral septum; LV, lateral ventricle; Med, medulla; Mi, mitral cell layer of the olfactory bulb; MiA, mitral cell layer of the accessory olfactory bulb; OB, olfactory bulb; Pn, pontine nucleus; Th, thalamus; vn, vomeronasal nerve. Scale bars: A = 2.0 mm; B and C = 0.3 mm; D = 0.6 mm; E–H = 0.4 mm; I = 1.3 mm; J = 0.1 mm; K = 0.05 mm; and L = 0.3 mm.
observed over the other layers of the olfactory bulb (Figs. 5B–G). It is interesting to note that periventricular areas surrounding the olfactory ventricle showed no labeling. The plexiform layers (both internal and external) showed a low hybridization signal as compared with the mitral and granular cell layers (Figs. 5E and F). No signal was observed over the olfactory nerve layer (Figs. 5B and C). Although the hybridization signal for CPA6 mRNA was strongest in the olfactory bulb, moderate CPA6 mRNA levels were detected in other areas such as the septal nucleus in the basal forebrain and in the cingulate cortex (Figs. 5A and I). In the hindbrain, hybridization signal was found over
the pontine nucleus and also over the inferior olivary nucleus (Figs. 5A and L). A very low signal was observed over the hippocampus (Figs. 5A, J, and K). CPA6 is expressed in several tissues during mouse development (Fig. 6). In brain, the distribution of CPA6 is broader in the embryonic brain than adult brain, with expression detected throughout the embryonic forebrain and cerebellum. The cortical mantle also showed a relative high CPA6 mRNA signal (Fig. 6A). CPA6 mRNA can be visualized in the pigmented epithelium of the retina and in the extrinsic muscle layer of the eye (Figs. 6A–C). CPA6 mRNA was also abundant in the ear and dorsal root ganglia
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
139
Fig. 6. Analysis of the distribution of CPA6 mRNA in E14.5 mouse embryo. (A) Dark-field image of a sagittal section. (B and C) Dark-field and bright-field image of the eye. (D–F) Bright-field and dark-field images of the rib and vertebrae. Abbreviations: bc, basioccipital cartilage; cb, cerebellum; drg, dorsal root ganglia; Sto, stomach. Scale bars: A = 1 mm; B and C = 0.3 mm, D–F = 0.1 mm.
(Fig. 6A). The skin of the embryo showed labeling (Fig. 6A), which was concentrated in the dermis. In the developing vertebrae, ribs and forelimbs CPA6 expression was abundant in the osteoblasts and was also present at a lower level in the hematopoietic stem cells within the newly formed bone marrow (Figs. 6A and D–F). Other organs that showed labeling included the muscle cell layer surrounding the stomach (Fig. 6A). Control hybridizations with sense cRNA probes showed no specific signal over any of the tissues (data not shown).
4. Discussion A major finding of the present study is that CPA6 has been conserved between human and mouse, and also appears to be highly conserved in rat, chicken, and frog. All of the predicted active site amino acids are conserved throughout the vertebrate species examined in the present study, including the unique methionine that is predicted to sit in the binding pocket for the substrate [38]. This high degree of conservation of CPA6 among vertebrates implies an important function. Determination of the function of
a peptidase requires knowledge of its distribution, as described in the present study (and discussed further below), and the enzymatic properties. Previously, modeling was used to predict that CPA6 cleaves C-terminal branched-chain aliphatic residues such as valine [38]. Attempts to express the active form of CPA6 in the baculovirus system were unsuccessful (unpublished observation), possibly due to the absence of the endopeptidase needed to process proCPA6 into the active form. The predicted cleavage site of the prodomain of CPA6 is RNRR-S in human and mouse and RKRR-S in frog [38]. These sequences fit with the consensus cleavage site for furin and other furin-like enzymes that are expressed in the trans Golgi network of many cell types [23,29,34]. In addition to furin, which has a broad tissue distribution, the furin-like enzymes PACE4E (an isoform of PACE4) and proprotein convertase 5/6A (PC5/6A) are both expressed in relatively high levels in the olfactory bulb [2,37]. Specifically, these two furin-like enzymes are enriched in the mitral cell layer and also show some positive cells in the plexiform and/or glomerular layers [2,37]. Because PACE4E and PC5/6A are likely to be expressed in the same cells as CPA6 and have the correct substrate specificity to cleave the prosegment of CPA6, it is
140
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
likely that one or both of these endopeptidases are physiologically involved in the activation of CPA6. Another important finding is that CPA6 mRNA is expressed in adult mouse brain, with the highest level of expression in olfactory bulb, and is also expressed in some peripheral adult tissues such as epididymis and in a variety of embryonic day 14.5 tissues. Taken together, the relatively high levels of expression of CPA6 mRNA in olfactory bulb and several embryonic day 14.5 tissues is consistent with a potential role for CPA6 in axon pathfinding signaling and/or other aspects of cell migration. CPA6 is observed in post mitotic neurons rather then in stem cells. For example, CPA6 mRNA is not observed over the rostral migratory stream that conveys the neuronal stem cells from the subventricular zone in the core of the olfactory bulb [17,21]. In addition, no CPA6 mRNA signal was observed in the periventricular area of the olfactory bulb or the subventricular area of the hippocampus, both of which contain relatively high levels of stem cells. Instead, CPA6 mRNA is expressed in post-mitotic neurons that migrate and populate the different zones within the olfactory bulb [9]. If CPA6 is involved in stem cell migration, there are several possible mechanisms; this CP could be involved in degradation of extracellular matrix or in the processing of one or more factors that regulate migration and/or axonal pathfinding (discussed below). The restricted distribution of CPA6 found in the present study is consistent with the general trend that members of the A/B subfamily of metalloCPs are expressed in relatively few cell types. CPA1, CPA2, and CPB1 are expressed mainly in the pancreatic acinar cells [3–5]. Normant et al. have reported that CPA1 and CPA2 mRNA are present in brain [24], although searches of GenBank sequences did not find any cDNAs for either gene product from brain cDNA libraries (search performed 9/04). CPU is produced only in the liver, from which it is secreted into blood where it functions after activation [19]. CPA3/mast cell CPA is only expressed in mast cells [33]. CPA4 (previously named CPA3) also shows a fairly restricted pattern of expression, although this has not been examined in detail [20]. CPA5 is predominantly expressed in germ cells of the testis [38]. In contrast to the members of the A/B subfamily, members of the N/E subfamily typically show much broader patterns of distribution. For example, CPD is present in most cell types [12,32,40], and CPE is expressed throughout the brain and other neuroendocrine tissues [22,28]. The restricted distribution of CPA6 mRNA observed in the present study is supported by bin silicoQ searches of GenBank databases. For example, searches for CPA6 cDNAs in mouse EST databases revealed hits only to sequences derived from olfactory bulb (BY719178, BB560354, and AV339318), neonatal brain (BB266208 and AV251788), and colon (BB072673 and BB625011) libraries. In the non-redundant database, a mouse CPA6 cDNA was isolated from a colon cDNA library (AK078883), a full-length human CPA6 cDNA was identified in a whole brain library (BC033684), and a
partial clone (N-terminal through exon 8) was found in a library from human hematopoietic stem/progenitor cells (AF221594). These human and mouse tissues correspond to those identified in the present study as containing CPA6 mRNA. The large number of partial cDNA clones for CPA6 in the various data bases that are missing one or more exons, or with incorrectly spliced exons, may be due to the extremely large introns present in the CPA6 gene (Fig. 4). In the case of human [38] and mouse CPA6 (the present study), RT-PCR was used to verify that the partial clones reported in the data base did not represent the most abundant forms of the CPA6 mRNA. Although this has not yet been done for rat and chicken CPA6, the amino acid sequences predicted in the present study (Fig. 3) are likely to represent real forms of these proteins based on homology to the human and mouse proteins. Recently, CPA6 was identified as a carboxypeptidaselike gene on chromosome 8q that was disrupted by a balanced translocation in a Duane syndrome patient [26]. Duane syndrome is a congenital eye-movement disorder that is characterized by a failure of cranial nerve VI to develop, leading to aberrant innervation of the lateral rectus muscle by cranial nerve III. This causes oblique movements of the eye and impairment of adduction/abduction, retraction movements, and contractions of the palpebral fissure [13,25]. Although Duane syndrome is largely a sporadic disorder that affects 0.1% of the general population, there is also a genetic cause for some cases. Specifically, deletions of a locus on 8q31 have been associated with this syndrome, and the breakpoint was narrowed to a 40-kb region [7]. Subsequently, this 40-kb region was found to contain a portion of the coding region of the CPA6 gene [26]. In the present study, the finding that CPA6 mRNA is expressed in embryonic lateral rectus muscle cells raises the possibility that CPA6 protein is involved in the proper development of the neuromuscular junction with the cranial nerves. Taken together, the results of the present study and those of previous studies both on CPA6 and other related proteins enable some predictions as to the function of CPA6. All members of the metallocarboxypeptidase family function either within the secretory pathway of the cell or outside the cell after secretion. The gene for CPA6 encodes a precursor protein with an N-terminal signal peptide, supporting the idea that CPA6 will enter the secretory pathway. Based on homology with other members of the A/B subfamily of carboxypeptidases and alignment of the propeptide cleavage sites in these proteins, it is likely that proCPA6 is activated by an endopeptidase located in the trans Golgi network and could therefore function either in the late secretory pathway or outside the cell. Because most other carboxypeptidases function by cleaving C-terminal amino acids following the prior action of an endopeptidase, it is possible that CPA6 also follows an endopeptidase in an enzymatic pathway. Potential candidates for this endopeptidase include any enzyme that cleaves to the C-terminal side of aromatic/ aliphatic residues and which is expressed in olfactory bulb
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
and other tissues that express CPA6. Such candidates include caldecrin [35], enkephalinase/neutral endopeptides 24.11 [39], endopeptidase 24–16 [10], and numerous matrix metalloproteinases [36]. Although all of these enzymes are generally expressed in the same cells that contain CPA6 mRNA in the olfactory bulb, they are also present in other areas that do not contain detectable levels of CPA6 mRNA. Thus, it is possible that CPA6 primarily functions in a pathway with another endopeptidase that has not yet been reported. Some matrix metalloproteinases, such as matrix metalloproteinases 1, 10, and 12, cleave substrates to the C-terminal side of aliphatic or aromatic residues, and therefore a CPA-like enzyme could be involved in further processing of the matrix following the endopeptidase action. Because the extracellular matrix plays an important role in cell migration, this potential function would fit with the location of CPA6 in olfactory bulb and the link between the CPA6 gene and Duane syndrome. An alternative possibility is that CPA6 participates directly in the processing of a signal molecule that is important for axon pathfinding, such as semaphorin. The semaphorin family includes several members that are secreted [11], and some of these contain C-terminal hydrophobic residues (such as Semaphorin 3A, 3B). Interestingly, Semaphorin 3A is abundant in the mitral cell layer of the olfactory bulb and is also involved in the development of the oculomotor axonal projections in the chick embryo [8,18]. Adams and colleagues showed that the chemorepulsive activity of the secreted semaphorins is affected by a furin-like proteolytic activity [1]. In addition, they reported that the C-terminal 11 amino acids of Semaphorin 3A contributes to the chemorepulsive activity [1]. It is therefore conceivable that CPA6 removes the valine from the C-terminus of Semaphorin 3A and that this affects the chemorepulsive activity. Further studies are needed to test this intriguing possibility and identify the precise mechanism by which the deletion of the CPA6 gene causes the abnormal innervation of the eye that results in Duane syndrome. Acknowledgments This work was supported by National Institutes of Health grants DK-51271 and DA-04494 (to L.D.F.). The DNA sequencing facility of the Albert Einstein College of Medicine is supported in part by Cancer Center grant CA13330. Microscopy was performed in the Analytical Imaging Facility of the Albert Einstein College of Medicine. References [1] R.H. Adams, M. Lohrum, A. Klostermann, H. Betz, A.W. Puschel, The chemorepulsive activity of secreted semaphorins is regulated by furin-dependent proteolytic processing, EMBO J. 16 (1997) 6077 – 6086.
141
[2] T. Akamatsu, S. Daikoku, H. Nagamune, S. Yoshida, K. Mori, A. Tsuji, Y. Matsuda, Developmental expression of a novel Kexin family protease, PACE4E, in the rat olfactory system, Histochem. Cell Biol. 108 (1997) 95 – 103. [3] D.S. Auld, Carboxypeptidase A, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 812 – 821. [4] D.S. Auld, Carboxypeptidase A2, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 821 – 825. [5] F.X. Aviles, J. Vendrell, Carboxypeptidase B, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 831 – 833. [6] A.J. Barrett, N.D. Rawlings, J.F. Woessner, Introduction: clan MC containing metallocarboxypeptidases, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 1998, pp. 1318 – 1320. [7] G. Calabrese, L. Telvi, F. Capodiferro, E. Morizio, A. Pizzuti, L. Stuppia, R. Bordoni, A. Ion, D. Fantasia, R. Mingarelli, G. Palka, Narrowing the Duane syndrome critical region at chromosome 8q13 down to 40 kb, Eur. J. Hum. Genet. 8 (2000) 319 – 324. [8] J.K. Chilton, S. Guthrie, Development of oculomotor axon projections in the chick embryo, J. Comp. Neurol. 472 (2004) 308 – 317. [9] V. Coskun, M.B. Luskin, Intrinsic and extrinsic regulation of the proliferation and differentiation of cells in the rodent rostral migratory stream, J. Neurosci. Res. 69 (2002) 795 – 802. [10] P. Dauch, Y. Masuo, J.P. Vincent, F. Checler, Endopeptidase 24-16 in murines: tissue distribution, cerebral regionalization ontogeny, J. Neurochem. 59 (1992) 1862 – 1867. [11] F. de Castro, L. Hu, H. Drabkin, C. Sotelo, A. Chedotal, Chemoattraction and chemorepulsion of olfactory bulb axons by different secreted semaphorins, J. Neurosci. 19 (1999) 4428 – 4436. [12] W. Dong, L.D. Fricker, R. Day, Carboxypeptidase D is a potential candidate to carry out redundant processing functions of carboxypeptidase E based on comparative distribution studies in the rat central nervous system, Neuroscience 89 (1999) 1301 – 1317. [13] A. Duane, Congenital deficiency of abduction associated with impairment of adduction, retraction movements, contractions of the palpebral fissure and oblique movements of the eye, Arch. Ophthalmol. 34 (1905) 133 – 159. [14] L.D. Fricker, Carboxypeptidase E, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 840 – 844. [15] L.D. Fricker, Carboxypeptidase Z, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 844 – 846. [16] L.D. Fricker, Metallocarboxypeptidase D, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 848 – 851. [17] R. Galli, A. Gritti, L. Bonfanti, A.L. Vescovi, Neural stem cells: an overview, Circ. Res. 92 (2003) 598 – 608. [18] R.J. Giger, R.J. Pasterkamp, S. Heijnen, A.J. Holtmaat, J. Verhaagen, Anatomical distribution of the chemorepellent semaphorin III/collapsin-1 in the adult rat and human brain: predominant expression in structures of the olfactory-hippocampal pathway and the motor system, J. Neurosci. Res. 52 (1998) 27 – 42. [19] D.F. Hendriks, Carboxypeptidase U, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 825 – 828. [20] H. Huang, C.P. Reed, J.S. Zhang, V. Shridhar, L. Wang, D.I. Smith, Carboxypeptidase A3 (CPA3): a novel gene highly induced by histone deacetylase inhibitors during differentiation of prostate epithelial cancer cells, Cancer Res. 59 (1999) 2981 – 2988. [21] Z. Liu, L.J. Martin, Olfactory bulb core is a rich source of neural progenitor and stem cells in adult rodent and human, J. Comp. Neurol. 459 (2003) 368 – 391.
142
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
[22] D.R. Lynch, K.M. Braas, J.C. Hutton, S.H. Snyder, Carboxypeptidase E (CPE): immunocytochemical localization in the rat central nervous system and pituitary gland, J. Neurosci. 10 (1990) 1592 – 1599. [23] K. Nakayama, Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins, Biochem. J. 327 (1997) 625 – 635. [24] E. Normant, C. Gros, J.C. Schwartz, Carboxypeptidase A isoforms produced by distinct genes or alternative splicing in brain and other extrapancreatic tissues, J. Biol. Chem. 270 (1995) 20543 – 20549. [25] C.F. Parsa, E. Grant, W.P. Dillon Jr., L.S. du, W.F. Hoyt, Absence of the abducens nerve in Duane syndrome verified by magnetic resonance imaging, Am. J. Ophthalmol. 125 (1998) 399 – 401. [26] A. Pizzuti, G. Calabrese, M. Bozzali, L. Telvi, E. Morizio, V. Guida, V. Gatta, L. Stuppia, A. Ion, G. Palka, B. Dallapiccola, A peptidase gene in chromosome 8q is disrupted by a balanced translocation in a Duane syndrome patient, Invest. Ophthalmol. Visual Sci. 43 (2002) 3609 – 3612. [27] S.E. Reznik, L.D. Fricker, Carboxypeptidases from A to Z: implications in embryonic development and Wnt binding, Cell. Mol. Life Sci. 58 (2001) 1790 – 1804. [28] C.A. Ross, L.D. Fricker, S.H. Snyder, Carboxypeptidase E expression in rat brain, studied with in situ hybridization, Soc. Neurosci. 14 (1988) 302. [29] N.G. Seidah, A. Prat, Precursor convertases in the secretory pathway, cytosol and extracellular milieu, Essays Biochem. 38 (2002) 79 – 94. [30] R.A. Skidgel, Carboxypeptidase M, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 851 – 854.
[31] R.A. Skidgel, E.G. Erdos, Lysine carboxypeptidase, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 837 – 840. [32] L. Song, L.D. Fricker, Tissue distribution and characterization of soluble and membrane-bound forms of metallocarboxypeptidase D, J. Biol. Chem. 271 (1996) 28884 – 28889. [33] E.B. Springman, Mast cell carboxypeptidase, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 828 – 830. [34] G. Thomas, Furin at the cutting edge: from protein traffic to embryogenesis and disease, Nat. Rev., Mol. Cell Biol. 3 (2002) 753 – 766. [35] A. Tomomura, H. Yamada, K. Itagaki, K. Fujimoto, S. Katoh, Rat brain expresses serum calcium-decreasing factor (caldecrin), Neurosci. Lett. 317 (2002) 17 – 20. [36] T. Tsukatani, H.L. Fillmore, H.R. Hamilton, E.H. Holbrook, R.M. Costanzo, Matrix metalloproteinase expression in the olfactory epithelium, NeuroReport 14 (2003) 1135 – 1140. [37] P. Villeneuve, N.G. Seidah, A. Beaudet, Immunohistochemical distribution of the prohormone convertase PC5-A in rat brain, Neuroscience 92 (1999) 641 – 654. [38] S. Wei, S. Segura, J. Vendrell, F.X. Aviles, E. Lanoue, R. Day, Y. Feng, L.D. Fricker, Identification and characterization of three members of the human metallocarboxypeptidase gene family, J. Biol. Chem. 277 (2002) 14954 – 14964. [39] J.N. Wilcox, H. Pollard, J. Moreau, J.C. Schwartz, B. Malfroy, Localization of enkephalinase mRNA in rat brain by in situ hybridization: comparison with immunohistochemical localization of the protein, Neuropeptides 14 (1989) 77 – 83. [40] X. Xin, O. Varlamov, R. Day, W. Dong, M.M. Bridgett, E.H. Leiter, L.D. Fricker, Cloning and sequence analysis of cDNA encoding rat carboxypeptidase D, DNA Cell Biol. 16 (1997) 897 – 909.
Research report
Identification and distribution of mouse carboxypeptidase A-6 Jose D. Fontenele-Neto1,2, Elena Kalinina1, Yun Feng1, Lloyd D. FrickerT,1 Department of Molecular Pharmacology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA Accepted 21 February 2005 Available online 1 April 2005
Abstract Carboxypeptidase A-6 (CPA6) was recently discovered in the human genome. To gain information regarding the potential function of this novel protein, the mouse homolog of CPA6 was identified using a combination of bioinformatics and reverse transcriptase-polymerase chain reaction (RT-PCR). In addition, homologs in rat, chicken, and frog were identified using a bioinformatics approach. The distribution of CPA6 mRNA in mouse tissues was examined using RT-PCR and in situ hybridization. A strong RT-PCR signal is detectable in olfactory bulb, and much lower levels are present in other regions such as the cerebral cortex, hippocampus, hypothalamus, striatum, and medulla. In peripheral tissues, a moderate RT-PCR signal is present in epididymis, and low levels are detectable in colon and spleen. The high level of CPA6 in adult mouse brain olfactory bulb was confirmed by in situ hybridization. Lower levels of CPA6 mRNA were found to be present in the cingulate cortex, lateral septum, pontine nucleus, and inferior olivary nucleus of the hindbrain. Within the olfactory bulb, CPA6 mRNA is enriched in the mitral and granular layer. A lower level of CPA6 mRNA is present in the internal and external plexiform layers, and no signal is detectable in the olfactory nerve layer. The distribution was also examined in whole embryos at embryonic day 14.5 and CPA6 mRNA was found to be enriched in eye, ear, osteoblasts, stomach, skin, dorsal root ganglia, and throughout the CNS. The presence of CPA6 mRNA in the rectus muscle layer of the eye at embryonic day 14.5 is consistent with the observation that the CPA6 gene is disrupted in a patient with Duane syndrome, a congenital eye defect. Taken together, the distribution of CPA6 suggests a specific role in a limited number of tissues, and it is possible that this role involves an aspect of cell migration. D 2005 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Gene structure and function Keywords: Peptide processing; Metallocarboxypeptidase; Neuropeptide; Olfactory bulb; Duane syndrome
1. Introduction Carboxypeptidases (CPs) perform many important functions within a range of organisms, including protein/ peptide digestion as well as more selective biosynthetic roles [6,27]. CPs that contain an active site zinc, termed metallocarboxypeptidases, are further divided into two
T Corresponding author. Fax: +718 430 8954. E-mail address: [email protected] (L.D. Fricker). 1 All authors contributed equally. 2 Present address: Department of Veterinary Medicine, Escola Superior de Agricultura de Mossoro´, BR 110 KM47 Mossoro´, RN 59625-900, Brazil. 0169-328X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2005.02.026
subfamilies based on their amino acid sequences [22]. One subfamily, termed the N/E subfamily, contains mammalian CPE, CPN, CPM, CPD, CPZ, CPX1, CPX2, and a protein termed aortic carboxypeptidase-like protein (ACLP) and AEBP1. Five of the members of this subfamily (CPE, N, M, D, and Z) cleave C-terminal basic amino acids (Lys or Arg) from the C-termini of peptides and are thought to function in the selective biosynthesis or processing of neuroendocrine peptides, growth factors, and other molecules involved in cell–cell signaling [14– 16,30,31]. The other three members of this family (CPX1, X2, and ACLP/AEBP1) are not active towards standard substrates, consistent with the absence of key catalytic residues [27]. It has been proposed that these
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
inactive members function as binding proteins, rather than active enzymes, although it is possible that they exhibit activity towards unknown molecules. The other subfamily of metalloCPs, termed the A/B family, includes mammalian CPB1, CPU (also termed CPB2, plasma CPB, and other names), CPA1, CPA2, CPA3 (also termed mast-cell CPA), CPA4 (previously named CPA3), CPA5, CPA6, and CPO [3–5,19,20,33,38]. Members of this A/B subfamily are produced as inactive zymogens that require activation, in contrast to the members of the N/E family which do not require an activation step. Once activated, CPB1 and CPU cleave C-terminal basic residues from peptides [5,19]. The substrate specificities of CPA1, 2, 3, and 5 have been examined; all are able to remove aliphatic or hydrophobic residues from the Ctermini of peptides [3,4,33,38]. The substrate specificities of CPA4, CPA6, and CPO have not been experimentally determined, and modeling suggests that CPA4 and 6 cleave aromatic/aliphatic residues while CPO cleaves acidic Cterminal residues [38]. The functions of several members of the A/B subfamily of CPs have been well studied. CPA1, CPA2, and CPB1 are secreted from the exocrine pancreas and function in the digestion of food in the intestine, following the action of chymotrypsin, trypsin, and pepsin [3–5]. CPA3 is secreted from mast cells and helps degrade a number of substances, following the action of chymase [33]. CPU is secreted from the liver and circulates in plasma as an inactive precursor form which is activated by plasmin or thrombin. Once active, CPU removes C-terminal Lys residues generated by the action of plasmin on fibrin, thus decreasing the binding of plasminogen to fibrin and inhibiting fibrinolysis [19]. Because the CPAs that have been characterized have overlapping enzyme substrate specificities, the key issue relating to the function of each enzyme is their unique distribution. Thus, one purpose of the present study was to determine the distribution of CPA6 in order to gain insights towards the function of this recently described member of the CP family. Previously, CPA6 mRNA expression was detected by RT-PCR in human fetal umbilical cord blood and in mouse brain [38]. In the present study, RT-PCR was used to determine which tissues express detectable levels of CPA6 mRNA. Then, in situ hybridization was used to precisely map the location of CPA6 mRNA in adult mouse brain and embryonic day 14.5 mouse tissues. A second goal of the current studies was to resolve discrepancies between the previously identified sequence of human CPA6 and the apparent mouse homologs of CPA6 in GenBank. All of these mouse homologs appear to lack exon 3, and as a result would not express functional protein. RTPCR was used to determine that the GenBank sequences do not represent the most abundant full-length form of mouse brain CPA6, which is very similar to human CPA6 and has an exon 3. In addition, a bioinformatics approach was used to identify putative homologs of CPA6 in rat, chicken, and
133
frog (Xenopus laevis). The observation that CPA6 appears to be highly conserved among these species suggests an important function for this protein.
2. Experimental procedures 2.1. Bioinformatics GenBank NR databases were searched with human CPA6 nucleotide and amino acid sequences using the blastn and tblastn programs (http://www.ncbi.nlm.nih.gov/BLAST). CPA6 sequences were also used to search genomic databases for mouse, rat, chicken (Gallus gallus), Drosophila melanogaster, Caenorhabditis elegans, and Saccharomyces cerevisiae. In those cases where the non-redundant GenBank database contained sequences that appeared to be missing an exon, based on the sequence of human CPA6, the genomic sequence was analyzed using the GT-AG rule for intron splice sites to find additional exons that would provide the missing sequence. 2.2. Animals Mice (C57BKS) were initially obtained from the Jackson Laboratory (Bar Harbor, ME) and bred within the barrier facility at Albert Einstein College of Medicine. Food (Purina rodent chow 5058) and water were provided ad libitum. The experimental protocol was approved by the committee for animal experimentation of the Albert Einstein College of Medicine. 2.3. RNA extraction and reverse transcriptase (RT)-PCR Total RNA from different brain regions (olfactory bulb, cerebral cortex, cerebellum, striatum, hypothalamus, hippocampus, and medulla) and several organs (testis, epididymis, ovary, lungs, spleen, liver, colon, and adrenal) from both male and female mice were obtained using an RNA easy kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. The RT-PCR was performed using the Titan one tube RT-PCR system (Roche Diagnostics GmbH, Mannheim, Germany) amplifying a sequence of 582 base pairs at the N-terminal region of CPA6 mRNA (exons 1 to 4) using TGCCTAAAGTGTCACATTTTGCTCA (nucleotides 62–86) and CTAAGGAGTGATACACC TCGTAGTTG (nucleotides 618–643) as forward and reverse primers, respectively. Briefly, 300 ng of total RNA was mixed with primers, dNTPs, and an enzyme mix containing avian myeloblastosis virus reverse transcriptase and a proof reading DNA polymerase. The reverse transcription was carried out for 30 min at 50 8C followed by a denaturation step at 94 8C for 2 min and by 35 PCR cycles (94 8C for 10 s, 59 8C for 45 s, 68 8C for 45 s). The resulting cDNA was separated in a 1% agarose gel containing ethidium bromide.
134
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
To determine if full-length mouse CPA6 mRNA is present in brain as predicted from the bioinformatics, oligonucleotides spanning the coding region were used for RT-PCR as described above, except that the elongation step was typically 3 min per cycle. Two forward oligonucleotides (TGCCTAAAGTGTCACATTTTGCTCA and GTTGTAAGCCACCTCCAGCCTC) and two reverse oligonucleotides (TCTCAAGGACATTTCTTTAGCAGGT and GTTTGCCAAGGTGCCTTTTGATCAG) were used and the resulting PCR product was subcloned and sequenced using a variety of internal primers. These oligonucleotides correspond to mouse CPA6 nucleotide positions 62–86, 143–164, 1511–1535, and 1576–1600, respectively. A minimum of two clones from independent RT-PCR reactions were sequenced to reduce the possibility of PCR-induced errors. 2.4. In situ hybridization Adult mice (2–3 months old) were sacrificed in a CO2 chamber. Brain, testis, epididymis, and colon were dissected out and frozen in n-hexane (Sigma, St Louis, MO) at 50 8C and stored at 80 8C. Brain coronal and sagittal sections (20 Am thick) were serially cut and transverse sections of the other tissues (15 Am thick) were also obtained. For the analysis of embryonic tissue, female mice were housed overnight with a male and the following morning examined for a vaginal plug. Those mice exhibiting vaginal plugs were sacrificed with CO2 vapor 14 days later. Embryos and their placenta were dissected out and frozen in n-hexane at 50 8C and stored at 80 8C. Transverse sections of the embryo and placenta were made. All sections were thaw mounted onto superfrostR/Plus slides (Fisher Scientific, Pittsburgh, PA), and stored at 80 8C until use. To generate the CPA6 probe, RT-PCR was used to generate a cDNA fragment of 860 base pairs corresponding to the N-terminal coding region of mouse CPA6 (exons 1–7). This cDNA fragment was subcloned into the EcoRI and BamHI sites of the pGEM-7Zf ( ) vector (Promega, Madison, WI). The resulting plasmid was used as a template to generate 35S-labeled riboprobes. Radiolabeled antisensestrand RNA probe was transcribed from an XbaI-linearized template with SP6 RNA polymerase (Invitrogen), along with 35S-uridine triphosphate (1250 Ci/mmol; PerkinElmer Life Sciences, Boston, MA). Labeled sense-strand probe was transcribed similarly from a SacI-linearized template with T7 RNA polymerase (Invitrogen). Unincorporated nucleotides were removed by chromatography with Sephadex G-50 (Amersham Biosciences), and probes were stored in 20 mM dithiothreitol (DTT) at 80 8C. Brain sections were warmed to room temperature, incubated for 5 min in 0.1 M sodium phosphate buffer (pH 7.4), fixed for 15 min with 4% paraformaldehyde in 0.1 M sodium phosphate buffer at 4 8C, and rinsed for 10 min in 0.1 M sodium phosphate buffer containing 0.15 M sodium chloride. Following two washes of 3 min each in 2
standard saline citrate (SSC; 300 mM sodium chloride, 30 mM sodium citrate, pH 7.0), sections were incubated in 0.1 M triethanolamine (pH 8.0) for 5 min and treated for 10 min with 0.25% acetic anhydride in 0.1 M triethanolamine. Finally, they were rinsed twice for 5 min in 2 SSC each and dehydrated in a graded ethanol series. Air-dried sections were stored at room temperature until hybridization. The probe was denaturated at 90 8C for 5 min and diluted to a final concentration of 107 cpm/ml in hybridization buffer consisting of 50% formamide, 10% dextran sulfate, 300 mM sodium chloride, 10 mM Tris–HCl (pH 7.5), 1 mM ethylenediaminetetraacetic acid (EDTA), 100 Ag/ml salmon sperm DNA, Denhardt’s solution (0.02% bovine serum albumin, 0.02% Ficoll, 0.02% polyvinylpyrolidone), and 20 mM DTT. Sections were covered with hybridization mix (100 Al/slide), coverslipped, sealed with rubber cement, and incubated for 18–20 h in a humidified chamber at 55 8C. Subsequently, the coverslips were carefully removed and the slides were rinsed twice at room temperature for 5 min each in 2 SSC. The slides were then incubated at 37 8C for 30 min with RNase A (Qiagen) diluted to 20 Ag/ml in a buffer containing 500 mM sodium chloride, 10 mM Tris–HCl (pH 7.5), 1 mM EDTA, and 1 mM DTT. They were washed twice for 15 min in 2 SSC at 37 8C, incubated for 30 min at 60 8C in 0.5 SSC, washed for 30 min at 65 8C in 0.1 SSC, and rinsed for 5 min in 0.1 SSC at room temperature. They were dehydrated in a graded ethanol series and allowed to air dry. A first set of autoradiograms was generated by apposition of labeled sections to X-Omatk Blue X-1 film (Eastman Kodak Co, Rochester, NY) for 4–10 days at room temperature. The slides were then dipped in Hypercoat LM-1 nuclear emulsion (Amersham Biosciences, Piscataway, NJ). Sections were stored in lightproof boxes at 4 8C for 2–4 weeks. Autoradiograms were developed for 4 min in D19 (Eastman Kodak) at 18 8C, rinsed in water, and fixed for 10 min in Kodak Polymax Fixer (Eastman Kodak). Sections were then counterstained lightly with Mayer’s hematoxylin, dehydrated in a graded ethanol series, defatted in xylene, mounted in cytosealk (Stephens Scientific, Riverdale, NJ), coverslipped, and analyzed under a Zeiss Axiophot microscope.
3. Results Two cDNA sequences with substantial sequence homology to human CPA6 are present in the mouse non-redundant GenBank database (NM177834 and AK078883). However, both of these sequences lack exon 3, which shifts the open reading frame and leads to a truncated protein that lacks the carboxypeptidase domain. An alternative methionine in the upstream region was predicted in the GenBank entries, but this would not produce a protein with a signal peptide and the resulting protein would not be expected to be expressed in the secretory pathway. To determine whether these forms
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
represent the major species of mouse CPA6, or if there were additional forms containing exon 3, RT-PCR was used with oligonucleotides in exons 1 and 4. The predicted product if no exon 3 was present was 457 nucleotides; with exon 3 the product would be 582 nucleotides. In olfactory bulb, a major band of approximately 582 nucleotides and a faint band at approximately 457 nucleotides were observed (Fig. 1). Sequence analysis of the PCR product confirmed the presence of the full-length insert. In addition, RT-PCR was performed with oligonucleotides located in the 5V and 3V untranslated regions of mouse CPA6 mRNA and the resulting cDNA product was sequenced using a variety of internal primers. The resulting sequence contained all the appropriate exons that were previously identified in human CPA6 (Fig. 2). The sequence missing from the previous GenBank entries for mouse CPA6 is indicated by the solid line and is clearly present in mouse CPA6. The N-terminal sequence is predicted to encode a signal peptide, with cleavage occurring between positions 30 and 31 (cleavage between Cys and His). Following this signal peptide is a prodomain of approximately 100 amino acids. The putative cleavage site of the prodomain (RNRR-S, with cleavage predicted to occur between the R and S) matches the general consensus site for proprotein convertases [29]. After the prodomain, mouse CPA6 has a 309 residue carboxypeptidase-like domain. All of the active site residues necessary for catalytic activity of other members of the metallocarboxypeptidase family are present in mouse CPA6, including the Zn2+-binding residues (H197, E200, and
Fig. 1. Analysis of the distribution of CPA6 mRNA in mouse tissues using RT-PCR. The oligonucleotides were located 582 base pair apart, as described in Experimental procedures. The sizes of DNA standards (in base pairs) are shown. Panel A: Analysis of RNA from different brain regions. Abbreviations: OB, olfactory bulb; Cx, cerebral cortex; Hip, hippocampus; Hyp, hypothalamus; Str, striatum; Cb, cerebellum; Med, medulla. Panel B: Analysis of RNA obtained from peripheral tissues. Abbreviations: Col, colon; Adr, adrenal; Spl, spleen; Liv, liver; Lng, lung; Ovr, ovary; Epid, epidymus; Test, testis. Results are shown for a representative experiment; each tissue/brain region was analyzed in three separate experiments with comparable results.
135
H325), the glutamic acid involved in catalytic activity (E399), and several residues involved in substrate binding (Fig. 2). One of these residues is methionine in position 384, which corresponds to the residue that sits in the bottom of the binding pocket for the side chain of the substrate and either conveys specificity for aromatic/aliphatic residues (I255 or L255 in CPA1, CPA2, and CPA3, using the numbering system of the mature form of bovine CPA1) or for basic residues (D255 in CPB and CPU, using the same numbering system). The deduced amino acid sequence of mouse CPA6 has 84% amino acid identity to human CPA6 and differs in length by a single amino acid (Fig. 3). Using a bioinformatics approach, a rat homolog of CPA6 was pieced together from genomic sequences (for exons 1–2) and from two GenBank non-redundant clones, one which contained only exons 3–6 (XM345496) and another which contained only exons 7–11 (XM232613). The predicted rat CPA6 homolog is the same length as mouse CPA6 and has 83% amino acid sequence identity to human CPA6. A chicken CPA6 homolog was also found, but the hypothetical protein encoded by the predicted sequence (AJ720827) does not contain a signal peptide. However, this GenBank sequence does not correspond to an actual cDNA sequence but was merely predicted from analysis of potential intron splice sites within the genome. If one considers an alternative intron splice site within the chicken CPA6 gene between exons 2 and 3 (maintaining the GT-AG rule for intron splicing) then the resulting protein would have a long open reading frame that has a signal peptide and substantial sequence identity (75%) with human CPA6 (Fig. 3). This hypothetical chicken protein would be 8 amino acids shorter than the mouse and rat homologs due to a shorter signal peptide. In addition to these GenBank sequences, a Xenopus sequence showed a strong match to CPA6; this Xenopus sequence appeared correct in the GenBank database and did not need any refinements to match with the other sequences. The Xenopus sequence has 65% amino acid sequence identity with human CPA6 and is 3 residues shorter (Fig. 3). All of the key active site residues in human CPA6 are conserved in the various homologs including H69, E72, R127, R145, H196, Y248, and E270 (by convention, the numbering system of mature bovine CPA1 is used). Importantly, the methionine in the position corresponding to the binding pocket of the substrate side chain (residue 255 in CPA/B) is present in CPA6 from all species (Fig. 3). Searches of Drosophila, C. elegans, and S. cerevisiae databases failed to detect any clear homologs of CPA6 in these species. All three of these organisms contain sequences with sequence similarity to general A/B subfamily carboxypeptidases, but none had greater overall similarity to CPA6 than to other members of the mammalian CPA/B subfamily. In considering the methionine in the position of the substrate binding pocket, three Drosophila sequences were found which contain an identical residue in this position. None of the C. elegans CPA-like sequences, or
136
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
Fig. 2. Nucleotide and deduced amino acid sequence of mouse CPA6. The nucleotide sequence was initially predicted from a bioinformatics approach. Then, RT-PCR was used to generate CPA6 cDNA corresponding to the full-length coding region, and this was sequenced to confirm that the prediction was correct. Arrows indicate the position of introns, and the numbers adjacent to these arrows indicate the exon. The nucleotide sequence that is underlined corresponds to exon 3, which was missing from previous mouse cDNA sequences deposited in GenBank. Active site or substrate binding residues that are mentioned in the text are boxed. The nucleotide sequence of full-length mouse CPA6 has been submitted to GenBank (AY773477).
the single S. cerevisiae CPA/B-like sequence, contained a methionine in the corresponding position. The gene for human CPA6 was previously found to be much larger than the genes for all other members of the A/B subfamily of carboxypeptidases [38]. To examine if this feature was conserved, the exon/intron junctions and the intron sizes were compared for mouse, human, and chicken CPA6 homologs. The overall length of the mouse CPA6 gene (c393 kb) is larger than that for the human gene (c315 kb), and the two longest introns are the same for both human and mouse CPA6 (Fig. 4). The chicken gene is considerably shorter (c86 kb) than the human and mouse
CPA6 genes, although still larger than a typical CPA/B gene [38]. Interestingly, the first intron is the longest in chicken CPA6 (41.8 kb), as is the case for human and mouse CPA6. The rat CPA6 gene also has a very large first intron (132.6 kb; data not shown). Thus, this feature of a large first intron for CPA6 is conserved in all species investigated. To gain an understanding as to possible functions of CPA6, the distribution was examined in mouse. Using RTPCR, CPA6 mRNA was detectable in the brain with the highest level of expression in the olfactory bulb (Fig. 1A). Several other brain regions (cerebral cortex, hippocampus, hypothalamus, striatum, and medulla) showed low levels of
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
137
Fig. 3. Alignment of human and mouse CPA6 with the predicted sequences of rat, chicken (G. gallus), and frog (X. laevis) CPA6. The sequence of frog CPA6 was obtained directly from GenBank. The rat and chicken sequences were pieced together from genomic sequences and partial GenBank sequences, as described in the Results section. Asterisks (bottom line) indicate those residues that are conserved (Cons.) in all species. The predicted cleavage site of the prodomain (RXRR) and the various active site or substrate binding residues are indicated. By convention, the numbering of these residues is based on the mature form of bovine CPA1. The amino acid sequence identity between CPA6 from each species and the human protein is indicated (bottom right).
CPA6 expression. In cerebellum, CPA6 mRNA levels were below the detection limit (Fig. 1A). Both male and female mice showed comparable levels of expression within the brain (data not shown). Of the peripheral tissues examined, CPA6 mRNA was detected at moderate levels in the epididymis and at low levels in colon, and spleen (Fig. 1B). No signal or extremely weak signals for CPA6 mRNA were observed in the adrenal, liver, lung, ovaries, and testis.
In situ hybridization carried out in adult brain sections revealed a similar pattern as observed using the RT-PCR technique, with strong labeling in the olfactory bulb (Fig. 5A). No signal was observed with control sense cRNA probes (not shown, and Fig. 5H). CPA6 mRNA was abundant in both main and accessory olfactory nuclei. The hybridization signal was observed mainly over the mitral and granular cell layers, with a lower hybridization signal
Fig. 4. Intron/exon junctions and intron sizes for mouse, human, and chicken CPA6. The human CPA6 gene organization was previously reported [38]. Intron lengths (in kb) are indicated.
138
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
Fig. 5. Analysis of the distribution of CPA6 in mouse brain using in situ hybridization. Panel A: Film autoradiogram of a mouse brain sagittal section showing the specific hybridization signal with the antisense cRNA probe for CPA6. No signal was detected when sense probes were used (not shown). Panels B–H: Dark-field (B–F and H) and bright-field (G) emulsion dipped autoradiograms of coronal sections of olfactory bulb hybridized for CPA6 mRNA. Note that all layers of the olfactory bulb are labeled and that both main and accessory nuclei show a hybridization signal. Intense labeling is observed over the mitral and granule layer of the main olfactory bulb. The control antisense probe shows no signal (H). Panels I–L: Dark-field emulsion dipped autoradiograms of mouse brain coronal sections hybridized with antisense CPA6 cRNA probe. Note that panel I shows a section through the basal forebrain, including the septum as well as the cingulate cortex, panels J and K show sections through the hippocampus, and panel L shows the dorsal nucleus of the inferior olivary complex and cerebellum. Abbreviations: AOB, accessory olfactory bulb; Cb, cerebellum; cc, corpus callosum; Cg, cingulate cortex; DG, Dentate gyrus; EPlA, external plexiform layer of the accessory olfactory bulb; EPl, external plexiform layer of the olfactory bulb; fi, fimbria of hippocampus; Hip, hippocampus; Hyp, hypothalamus; GrA, granular cell layer of the accessory olfactory bulb; GrO, granular cell layer of the olfactory bulb; IOD, inferior olivary complex; LS, lateral septum; LV, lateral ventricle; Med, medulla; Mi, mitral cell layer of the olfactory bulb; MiA, mitral cell layer of the accessory olfactory bulb; OB, olfactory bulb; Pn, pontine nucleus; Th, thalamus; vn, vomeronasal nerve. Scale bars: A = 2.0 mm; B and C = 0.3 mm; D = 0.6 mm; E–H = 0.4 mm; I = 1.3 mm; J = 0.1 mm; K = 0.05 mm; and L = 0.3 mm.
observed over the other layers of the olfactory bulb (Figs. 5B–G). It is interesting to note that periventricular areas surrounding the olfactory ventricle showed no labeling. The plexiform layers (both internal and external) showed a low hybridization signal as compared with the mitral and granular cell layers (Figs. 5E and F). No signal was observed over the olfactory nerve layer (Figs. 5B and C). Although the hybridization signal for CPA6 mRNA was strongest in the olfactory bulb, moderate CPA6 mRNA levels were detected in other areas such as the septal nucleus in the basal forebrain and in the cingulate cortex (Figs. 5A and I). In the hindbrain, hybridization signal was found over
the pontine nucleus and also over the inferior olivary nucleus (Figs. 5A and L). A very low signal was observed over the hippocampus (Figs. 5A, J, and K). CPA6 is expressed in several tissues during mouse development (Fig. 6). In brain, the distribution of CPA6 is broader in the embryonic brain than adult brain, with expression detected throughout the embryonic forebrain and cerebellum. The cortical mantle also showed a relative high CPA6 mRNA signal (Fig. 6A). CPA6 mRNA can be visualized in the pigmented epithelium of the retina and in the extrinsic muscle layer of the eye (Figs. 6A–C). CPA6 mRNA was also abundant in the ear and dorsal root ganglia
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
139
Fig. 6. Analysis of the distribution of CPA6 mRNA in E14.5 mouse embryo. (A) Dark-field image of a sagittal section. (B and C) Dark-field and bright-field image of the eye. (D–F) Bright-field and dark-field images of the rib and vertebrae. Abbreviations: bc, basioccipital cartilage; cb, cerebellum; drg, dorsal root ganglia; Sto, stomach. Scale bars: A = 1 mm; B and C = 0.3 mm, D–F = 0.1 mm.
(Fig. 6A). The skin of the embryo showed labeling (Fig. 6A), which was concentrated in the dermis. In the developing vertebrae, ribs and forelimbs CPA6 expression was abundant in the osteoblasts and was also present at a lower level in the hematopoietic stem cells within the newly formed bone marrow (Figs. 6A and D–F). Other organs that showed labeling included the muscle cell layer surrounding the stomach (Fig. 6A). Control hybridizations with sense cRNA probes showed no specific signal over any of the tissues (data not shown).
4. Discussion A major finding of the present study is that CPA6 has been conserved between human and mouse, and also appears to be highly conserved in rat, chicken, and frog. All of the predicted active site amino acids are conserved throughout the vertebrate species examined in the present study, including the unique methionine that is predicted to sit in the binding pocket for the substrate [38]. This high degree of conservation of CPA6 among vertebrates implies an important function. Determination of the function of
a peptidase requires knowledge of its distribution, as described in the present study (and discussed further below), and the enzymatic properties. Previously, modeling was used to predict that CPA6 cleaves C-terminal branched-chain aliphatic residues such as valine [38]. Attempts to express the active form of CPA6 in the baculovirus system were unsuccessful (unpublished observation), possibly due to the absence of the endopeptidase needed to process proCPA6 into the active form. The predicted cleavage site of the prodomain of CPA6 is RNRR-S in human and mouse and RKRR-S in frog [38]. These sequences fit with the consensus cleavage site for furin and other furin-like enzymes that are expressed in the trans Golgi network of many cell types [23,29,34]. In addition to furin, which has a broad tissue distribution, the furin-like enzymes PACE4E (an isoform of PACE4) and proprotein convertase 5/6A (PC5/6A) are both expressed in relatively high levels in the olfactory bulb [2,37]. Specifically, these two furin-like enzymes are enriched in the mitral cell layer and also show some positive cells in the plexiform and/or glomerular layers [2,37]. Because PACE4E and PC5/6A are likely to be expressed in the same cells as CPA6 and have the correct substrate specificity to cleave the prosegment of CPA6, it is
140
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
likely that one or both of these endopeptidases are physiologically involved in the activation of CPA6. Another important finding is that CPA6 mRNA is expressed in adult mouse brain, with the highest level of expression in olfactory bulb, and is also expressed in some peripheral adult tissues such as epididymis and in a variety of embryonic day 14.5 tissues. Taken together, the relatively high levels of expression of CPA6 mRNA in olfactory bulb and several embryonic day 14.5 tissues is consistent with a potential role for CPA6 in axon pathfinding signaling and/or other aspects of cell migration. CPA6 is observed in post mitotic neurons rather then in stem cells. For example, CPA6 mRNA is not observed over the rostral migratory stream that conveys the neuronal stem cells from the subventricular zone in the core of the olfactory bulb [17,21]. In addition, no CPA6 mRNA signal was observed in the periventricular area of the olfactory bulb or the subventricular area of the hippocampus, both of which contain relatively high levels of stem cells. Instead, CPA6 mRNA is expressed in post-mitotic neurons that migrate and populate the different zones within the olfactory bulb [9]. If CPA6 is involved in stem cell migration, there are several possible mechanisms; this CP could be involved in degradation of extracellular matrix or in the processing of one or more factors that regulate migration and/or axonal pathfinding (discussed below). The restricted distribution of CPA6 found in the present study is consistent with the general trend that members of the A/B subfamily of metalloCPs are expressed in relatively few cell types. CPA1, CPA2, and CPB1 are expressed mainly in the pancreatic acinar cells [3–5]. Normant et al. have reported that CPA1 and CPA2 mRNA are present in brain [24], although searches of GenBank sequences did not find any cDNAs for either gene product from brain cDNA libraries (search performed 9/04). CPU is produced only in the liver, from which it is secreted into blood where it functions after activation [19]. CPA3/mast cell CPA is only expressed in mast cells [33]. CPA4 (previously named CPA3) also shows a fairly restricted pattern of expression, although this has not been examined in detail [20]. CPA5 is predominantly expressed in germ cells of the testis [38]. In contrast to the members of the A/B subfamily, members of the N/E subfamily typically show much broader patterns of distribution. For example, CPD is present in most cell types [12,32,40], and CPE is expressed throughout the brain and other neuroendocrine tissues [22,28]. The restricted distribution of CPA6 mRNA observed in the present study is supported by bin silicoQ searches of GenBank databases. For example, searches for CPA6 cDNAs in mouse EST databases revealed hits only to sequences derived from olfactory bulb (BY719178, BB560354, and AV339318), neonatal brain (BB266208 and AV251788), and colon (BB072673 and BB625011) libraries. In the non-redundant database, a mouse CPA6 cDNA was isolated from a colon cDNA library (AK078883), a full-length human CPA6 cDNA was identified in a whole brain library (BC033684), and a
partial clone (N-terminal through exon 8) was found in a library from human hematopoietic stem/progenitor cells (AF221594). These human and mouse tissues correspond to those identified in the present study as containing CPA6 mRNA. The large number of partial cDNA clones for CPA6 in the various data bases that are missing one or more exons, or with incorrectly spliced exons, may be due to the extremely large introns present in the CPA6 gene (Fig. 4). In the case of human [38] and mouse CPA6 (the present study), RT-PCR was used to verify that the partial clones reported in the data base did not represent the most abundant forms of the CPA6 mRNA. Although this has not yet been done for rat and chicken CPA6, the amino acid sequences predicted in the present study (Fig. 3) are likely to represent real forms of these proteins based on homology to the human and mouse proteins. Recently, CPA6 was identified as a carboxypeptidaselike gene on chromosome 8q that was disrupted by a balanced translocation in a Duane syndrome patient [26]. Duane syndrome is a congenital eye-movement disorder that is characterized by a failure of cranial nerve VI to develop, leading to aberrant innervation of the lateral rectus muscle by cranial nerve III. This causes oblique movements of the eye and impairment of adduction/abduction, retraction movements, and contractions of the palpebral fissure [13,25]. Although Duane syndrome is largely a sporadic disorder that affects 0.1% of the general population, there is also a genetic cause for some cases. Specifically, deletions of a locus on 8q31 have been associated with this syndrome, and the breakpoint was narrowed to a 40-kb region [7]. Subsequently, this 40-kb region was found to contain a portion of the coding region of the CPA6 gene [26]. In the present study, the finding that CPA6 mRNA is expressed in embryonic lateral rectus muscle cells raises the possibility that CPA6 protein is involved in the proper development of the neuromuscular junction with the cranial nerves. Taken together, the results of the present study and those of previous studies both on CPA6 and other related proteins enable some predictions as to the function of CPA6. All members of the metallocarboxypeptidase family function either within the secretory pathway of the cell or outside the cell after secretion. The gene for CPA6 encodes a precursor protein with an N-terminal signal peptide, supporting the idea that CPA6 will enter the secretory pathway. Based on homology with other members of the A/B subfamily of carboxypeptidases and alignment of the propeptide cleavage sites in these proteins, it is likely that proCPA6 is activated by an endopeptidase located in the trans Golgi network and could therefore function either in the late secretory pathway or outside the cell. Because most other carboxypeptidases function by cleaving C-terminal amino acids following the prior action of an endopeptidase, it is possible that CPA6 also follows an endopeptidase in an enzymatic pathway. Potential candidates for this endopeptidase include any enzyme that cleaves to the C-terminal side of aromatic/ aliphatic residues and which is expressed in olfactory bulb
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
and other tissues that express CPA6. Such candidates include caldecrin [35], enkephalinase/neutral endopeptides 24.11 [39], endopeptidase 24–16 [10], and numerous matrix metalloproteinases [36]. Although all of these enzymes are generally expressed in the same cells that contain CPA6 mRNA in the olfactory bulb, they are also present in other areas that do not contain detectable levels of CPA6 mRNA. Thus, it is possible that CPA6 primarily functions in a pathway with another endopeptidase that has not yet been reported. Some matrix metalloproteinases, such as matrix metalloproteinases 1, 10, and 12, cleave substrates to the C-terminal side of aliphatic or aromatic residues, and therefore a CPA-like enzyme could be involved in further processing of the matrix following the endopeptidase action. Because the extracellular matrix plays an important role in cell migration, this potential function would fit with the location of CPA6 in olfactory bulb and the link between the CPA6 gene and Duane syndrome. An alternative possibility is that CPA6 participates directly in the processing of a signal molecule that is important for axon pathfinding, such as semaphorin. The semaphorin family includes several members that are secreted [11], and some of these contain C-terminal hydrophobic residues (such as Semaphorin 3A, 3B). Interestingly, Semaphorin 3A is abundant in the mitral cell layer of the olfactory bulb and is also involved in the development of the oculomotor axonal projections in the chick embryo [8,18]. Adams and colleagues showed that the chemorepulsive activity of the secreted semaphorins is affected by a furin-like proteolytic activity [1]. In addition, they reported that the C-terminal 11 amino acids of Semaphorin 3A contributes to the chemorepulsive activity [1]. It is therefore conceivable that CPA6 removes the valine from the C-terminus of Semaphorin 3A and that this affects the chemorepulsive activity. Further studies are needed to test this intriguing possibility and identify the precise mechanism by which the deletion of the CPA6 gene causes the abnormal innervation of the eye that results in Duane syndrome. Acknowledgments This work was supported by National Institutes of Health grants DK-51271 and DA-04494 (to L.D.F.). The DNA sequencing facility of the Albert Einstein College of Medicine is supported in part by Cancer Center grant CA13330. Microscopy was performed in the Analytical Imaging Facility of the Albert Einstein College of Medicine. References [1] R.H. Adams, M. Lohrum, A. Klostermann, H. Betz, A.W. Puschel, The chemorepulsive activity of secreted semaphorins is regulated by furin-dependent proteolytic processing, EMBO J. 16 (1997) 6077 – 6086.
141
[2] T. Akamatsu, S. Daikoku, H. Nagamune, S. Yoshida, K. Mori, A. Tsuji, Y. Matsuda, Developmental expression of a novel Kexin family protease, PACE4E, in the rat olfactory system, Histochem. Cell Biol. 108 (1997) 95 – 103. [3] D.S. Auld, Carboxypeptidase A, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 812 – 821. [4] D.S. Auld, Carboxypeptidase A2, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 821 – 825. [5] F.X. Aviles, J. Vendrell, Carboxypeptidase B, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 831 – 833. [6] A.J. Barrett, N.D. Rawlings, J.F. Woessner, Introduction: clan MC containing metallocarboxypeptidases, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 1998, pp. 1318 – 1320. [7] G. Calabrese, L. Telvi, F. Capodiferro, E. Morizio, A. Pizzuti, L. Stuppia, R. Bordoni, A. Ion, D. Fantasia, R. Mingarelli, G. Palka, Narrowing the Duane syndrome critical region at chromosome 8q13 down to 40 kb, Eur. J. Hum. Genet. 8 (2000) 319 – 324. [8] J.K. Chilton, S. Guthrie, Development of oculomotor axon projections in the chick embryo, J. Comp. Neurol. 472 (2004) 308 – 317. [9] V. Coskun, M.B. Luskin, Intrinsic and extrinsic regulation of the proliferation and differentiation of cells in the rodent rostral migratory stream, J. Neurosci. Res. 69 (2002) 795 – 802. [10] P. Dauch, Y. Masuo, J.P. Vincent, F. Checler, Endopeptidase 24-16 in murines: tissue distribution, cerebral regionalization ontogeny, J. Neurochem. 59 (1992) 1862 – 1867. [11] F. de Castro, L. Hu, H. Drabkin, C. Sotelo, A. Chedotal, Chemoattraction and chemorepulsion of olfactory bulb axons by different secreted semaphorins, J. Neurosci. 19 (1999) 4428 – 4436. [12] W. Dong, L.D. Fricker, R. Day, Carboxypeptidase D is a potential candidate to carry out redundant processing functions of carboxypeptidase E based on comparative distribution studies in the rat central nervous system, Neuroscience 89 (1999) 1301 – 1317. [13] A. Duane, Congenital deficiency of abduction associated with impairment of adduction, retraction movements, contractions of the palpebral fissure and oblique movements of the eye, Arch. Ophthalmol. 34 (1905) 133 – 159. [14] L.D. Fricker, Carboxypeptidase E, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 840 – 844. [15] L.D. Fricker, Carboxypeptidase Z, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 844 – 846. [16] L.D. Fricker, Metallocarboxypeptidase D, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 848 – 851. [17] R. Galli, A. Gritti, L. Bonfanti, A.L. Vescovi, Neural stem cells: an overview, Circ. Res. 92 (2003) 598 – 608. [18] R.J. Giger, R.J. Pasterkamp, S. Heijnen, A.J. Holtmaat, J. Verhaagen, Anatomical distribution of the chemorepellent semaphorin III/collapsin-1 in the adult rat and human brain: predominant expression in structures of the olfactory-hippocampal pathway and the motor system, J. Neurosci. Res. 52 (1998) 27 – 42. [19] D.F. Hendriks, Carboxypeptidase U, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 825 – 828. [20] H. Huang, C.P. Reed, J.S. Zhang, V. Shridhar, L. Wang, D.I. Smith, Carboxypeptidase A3 (CPA3): a novel gene highly induced by histone deacetylase inhibitors during differentiation of prostate epithelial cancer cells, Cancer Res. 59 (1999) 2981 – 2988. [21] Z. Liu, L.J. Martin, Olfactory bulb core is a rich source of neural progenitor and stem cells in adult rodent and human, J. Comp. Neurol. 459 (2003) 368 – 391.
142
J.D. Fontenele-Neto et al. / Molecular Brain Research 137 (2005) 132–142
[22] D.R. Lynch, K.M. Braas, J.C. Hutton, S.H. Snyder, Carboxypeptidase E (CPE): immunocytochemical localization in the rat central nervous system and pituitary gland, J. Neurosci. 10 (1990) 1592 – 1599. [23] K. Nakayama, Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins, Biochem. J. 327 (1997) 625 – 635. [24] E. Normant, C. Gros, J.C. Schwartz, Carboxypeptidase A isoforms produced by distinct genes or alternative splicing in brain and other extrapancreatic tissues, J. Biol. Chem. 270 (1995) 20543 – 20549. [25] C.F. Parsa, E. Grant, W.P. Dillon Jr., L.S. du, W.F. Hoyt, Absence of the abducens nerve in Duane syndrome verified by magnetic resonance imaging, Am. J. Ophthalmol. 125 (1998) 399 – 401. [26] A. Pizzuti, G. Calabrese, M. Bozzali, L. Telvi, E. Morizio, V. Guida, V. Gatta, L. Stuppia, A. Ion, G. Palka, B. Dallapiccola, A peptidase gene in chromosome 8q is disrupted by a balanced translocation in a Duane syndrome patient, Invest. Ophthalmol. Visual Sci. 43 (2002) 3609 – 3612. [27] S.E. Reznik, L.D. Fricker, Carboxypeptidases from A to Z: implications in embryonic development and Wnt binding, Cell. Mol. Life Sci. 58 (2001) 1790 – 1804. [28] C.A. Ross, L.D. Fricker, S.H. Snyder, Carboxypeptidase E expression in rat brain, studied with in situ hybridization, Soc. Neurosci. 14 (1988) 302. [29] N.G. Seidah, A. Prat, Precursor convertases in the secretory pathway, cytosol and extracellular milieu, Essays Biochem. 38 (2002) 79 – 94. [30] R.A. Skidgel, Carboxypeptidase M, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 851 – 854.
[31] R.A. Skidgel, E.G. Erdos, Lysine carboxypeptidase, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 837 – 840. [32] L. Song, L.D. Fricker, Tissue distribution and characterization of soluble and membrane-bound forms of metallocarboxypeptidase D, J. Biol. Chem. 271 (1996) 28884 – 28889. [33] E.B. Springman, Mast cell carboxypeptidase, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.), Handbook of Proteolytic Enzymes, Academic Press, San Diego, 2004, pp. 828 – 830. [34] G. Thomas, Furin at the cutting edge: from protein traffic to embryogenesis and disease, Nat. Rev., Mol. Cell Biol. 3 (2002) 753 – 766. [35] A. Tomomura, H. Yamada, K. Itagaki, K. Fujimoto, S. Katoh, Rat brain expresses serum calcium-decreasing factor (caldecrin), Neurosci. Lett. 317 (2002) 17 – 20. [36] T. Tsukatani, H.L. Fillmore, H.R. Hamilton, E.H. Holbrook, R.M. Costanzo, Matrix metalloproteinase expression in the olfactory epithelium, NeuroReport 14 (2003) 1135 – 1140. [37] P. Villeneuve, N.G. Seidah, A. Beaudet, Immunohistochemical distribution of the prohormone convertase PC5-A in rat brain, Neuroscience 92 (1999) 641 – 654. [38] S. Wei, S. Segura, J. Vendrell, F.X. Aviles, E. Lanoue, R. Day, Y. Feng, L.D. Fricker, Identification and characterization of three members of the human metallocarboxypeptidase gene family, J. Biol. Chem. 277 (2002) 14954 – 14964. [39] J.N. Wilcox, H. Pollard, J. Moreau, J.C. Schwartz, B. Malfroy, Localization of enkephalinase mRNA in rat brain by in situ hybridization: comparison with immunohistochemical localization of the protein, Neuropeptides 14 (1989) 77 – 83. [40] X. Xin, O. Varlamov, R. Day, W. Dong, M.M. Bridgett, E.H. Leiter, L.D. Fricker, Cloning and sequence analysis of cDNA encoding rat carboxypeptidase D, DNA Cell Biol. 16 (1997) 897 – 909.