Identification of a molluscan homologue of the neuroendocrine polypeptide 7B2.

In vertebrates, interaction of prohormone convertase 2 (PC2) with the highly conserved polypeptide 7B2 is essential for transport and maturation of proPC2 in the regulated secretory pathway. In vitro, 7B2 displays a strong inhibitory activity toward PC2. Here, we characterize a cDNA encoding the first invertebrate 7B2-related protein (L7B2) from the brain of the mollusc Lymnaea stagnalis The overall amino acid sequence identity between L7B2 and its vertebrate counterparts is surprisingly low (29%) and is restricted to a few small stretches of amino acid residues. Of particular interest are a conserved proline-rich region in the middle portion of the L7B2 sequence and a repeated conserved region in the carboxyl-terminal domain. Synthetic peptides corresponding to the carboxyl-terminal regions inhibit Lymnaea PC2 enzyme activity in extracts of insulin-producing neurons, in which both L7B2 and Lymnaea PC2 are abundantly expressed. Moreover, the peptides inhibit mouse PC2 enzyme activity. Our cloning of invertebrate 7B2 helps to delineate residues that are important for 7B2-PC2 interaction.

In vertebrates, interaction of prohormone convertase 2 (PC2) with the highly conserved polypeptide 7B2 is essential for transport and maturation of proPC2 in the regulated secretory pathway. In vitro, 7B2 displays a strong inhibitory activity toward PC2. Here, we characterize a cDNA encoding the first invertebrate 7B2-related protein (L7B2) from the brain of the mollusc Lymnaea stagnalis. The overall amino acid sequence identity between L7B2 and its vertebrate counterparts is surprisingly low (29%) and is restricted to a few small stretches of amino acid residues. Of particular interest are a conserved proline-rich region in the middle portion of the L7B2 sequence and a repeated conserved region in the carboxyl-terminal domain. Synthetic peptides corresponding to the carboxyl-terminal regions inhibit Lymnaea PC2 enzyme activity in extracts of insulin-producing neurons, in which both L7B2 and Lymnaea PC2 are abundantly expressed. Moreover, the peptides inhibit mouse PC2 enzyme activity. Our cloning of invertebrate 7B2 helps to delineate residues that are important for 7B2-PC2 interaction.
7B2 is strongly conserved among vertebrates. This applies particularly to the NT, which shows 84 -86% amino acid sequence identity between Xenopus and various mammalian species (1, 10 -12). This strong conservation suggests that large stretches of amino acid residues are crucial for the function of 7B2. Thus, a 7B2-related protein from a distantly related invertebrate would help to identify conserved residues that might be crucial for binding of 7B2 to PC2 or for inhibition of PC2 activity. We have now cloned a cDNA 2 encoding an invertebrate 7B2 protein from the mollusc Lymnaea stagnalis. We find that Lymnaea 7B2 (L7B2) displays a remarkably low (29%) degree of conservation between vertebrate and invertebrate 7B2, which is predominantly restricted to only a few small stretches of residues in both the NT and the CT of L7B2. Two regions in the CT of L7B2, designated LCT1 and LCT2, appears to inhibit LPC2 (13) enzyme activity in extracts of the Lymnaea neuroendocrine insulin-producing cells and recombinant purified mouse PC2 activity (14).

MATERIALS AND METHODS
PCR Analysis-Degenerate oligonucleotides were synthesized based on amino acid sequences conserved among vertebrates 7B2 proteins (10 -12). Oligos OL1 (5Ј-CGGAATTC(AG)A(TC)CCICCNAA(TC)CCNT-G(TC)CC-3Ј; based on P(D/N)PPNPCP, residues 88 -95; Ref. 11) and OL2 (5Ј-CAAGCTTGGNACN(GC)(AT)(CT)TT(CT)TTNGCNACNAC-3Ј; based on VVAKKSVP, residues 168 -175; Ref. 11), contained at the 5Ј end a recognition site for the endonucleases EcoRI and HindIII, respectively. PCR was performed on one animal equivalent brain hexanucleotide primed cDNA in a 100-l reaction volume with 1.0 unit of Super Taq DNA polymerase (Boehringer Mannheim) in two consecutive amplification rounds in a DNA thermal cycler (Perkin-Elmer) using 45 cycles of 94°C for 20 s, 53°C for 30 s, and 72°C for 1 min. Amplified cDNA was digested with EcoRI and HindIII and separated on an agarose gel. Fragments of the expected size were cloned and sequenced.
Screening of a Lymnaea cDNA Library of Cerebral Ganglia-Approximately 80,000 clones of an amplified ZAP II cDNA library (13) were plated at a density of 10 5 plaque-forming unit/400 cm 2 and absorbed to charged nylon membranes (Boehringer Mannheim). Clones were purified by screening at a lower plaque density. The L7B2 PCR product was used as a random primed probe, labeled with [␣-32 P]dATP (specific activity, Ͼ10 9 cpm/g). Membranes were hybridized in 6 ϫ SSC (1 ϫ SSC ϭ 0.15 M NaCl and 0.015 M sodium citrate), 0.2% SDS, 5 ϫ Denhardt's solution, and 10 g/ml herring sperm DNA at 65°C for 18 h. The filters were washed in 0.2 ϫ SSC, 0.2% SDS at 65°C for 30 min and autoradiographed.
Subcloning and Nucleotide Sequence Analysis-L7B2 PCR products and pBluescript II L7B2 cDNA generated by in vivo excision were sequenced in both orientations according to the dideoxy chain termination method (15), using T7 DNA polymerase (U. S. Biochemical Corp.). Following sequencing from universal primer sites present in the vectors, the sequence information was used to design new primers and sequencing was continued. Sequence alignments were made using the program Clustal V.
In Situ Hybridization and Immunocytochemistry-Specific ␣-35 S-UTP (DuPont)-labeled cRNA probes were made on 200 ng of linearized cDNA of LPC2 (nucleotides 1521-2632 (13)) and L7B2 (nucleotides 1-462). Separate in vitro transcription reactions were performed at 37°C, using either T3 or T7 RNA polymerase (Boehringer Mannheim) containing 1 mM G/A/CTP and 3 M [␣-35 S]UTP as described by Smit et * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
al. (16). Both antisense and sense (control) cRNA probes had specific activities between 1 ϫ 10 7 and 1 ϫ 10 8 cpm/g of RNA. Serial 7-m sections of 1% paraformaldehyde/1% acetic acid-fixed cerebral ganglia were used for in situ hybridization. After pretreatment and prehybridization of the slides (16), hybridization was carried out overnight, at 50°C, by applying 35 S-labeled probe at a final activity of 1.5 ϫ 10 6 cpm per slide. Slides were rinsed at 2 ϫ SSC/50% formamide at 50°C (16). Finally, radioactivity was visualized by dipping slides in melted, diluted (1:3) Ilford K5 emulsion (Ilford). Sections were exposed for 5-10 days in the dark at 4°C, and autoradiographs were developed in Kodak D19 developer and fixed in 24% sodium thiosulfate and Ilford rapid fixative. After dehydration in graded ethanols and clearance in xylene, coverslips were mounted with entallan. Alternate slides were used for immunolabeling with anti-insulin-related peptide antibodies. Immunolabeling grids were incubated in Tris-buffered saline-gelatin buffer (0.1 M Tris/HCl, 0.05% Tween 20, 0.25% gelatin, pH 7.4), and slides were immunolabeled using the procedure described by Van Minnen et al. (17). Enzymatic Assays-Light green cell (LGC) clusters and salivary glands were dissected separately and homogenized in native lysis buffer. Lysates were cleared by triplicate centrifugation at 16,000 ϫ g for 5 min. Cell extract was either used immediately or incubated with rabbit antiserum (control) or with LPC2 antiserum at 4°C for 1.5 h, and then protein A-agarose was added and incubated for 1 h. Agarose beads were removed by centrifugation and the supernatant was used. Triplicate (LPC2) and duplicate (mPC2 and mPC1/3) reactions were performed in 0.1 M NaAc, pH 5.0/5 mM CaCl 2 /0.1% Brij/5 g of bovine serum albumin, synthetic peptides at the indicated concentrations, and either 1 l of LGC (ϳ10 g total protein) or 1 l of salivary gland extract (ϳ15 g of total protein), or 0.6 g of purified recombinant mouse PC2 (14). Reactions were preincubated with peptide for 30 min at room temperature. Then 0.2 M fluorogenic substrate (Pyr-Arg-Thr-Lys-Arg-AMC; Peptides International) was added and reactions were incubated at 20°C (with LGC extract) or at 37°C (with purified recombinant mouse PC2 (14)) for 5 h. Liberated AMC was estimated at 380 nm excitation, 460 nm emission in a Perkin-Elmer fluorometer or a Cambridge Biotechnology fluorometer, respectively. Purified recombinant PC1/3 (87 kDa; 0.46 units; Ref. 18) was tested with 1 mM to 10 M LCT2 in parallel reactions. Production of fluorescent AMC was linear between 1 and 7 h of incubation. The rate of AMC production in control reactions of LGC extracts not containing peptide was ϳ125 pmol/h, and it was ϳ70 pmol/h in the control reactions of LPC2 immunodepleted LGC extracts. Nonlinear regression, using the program SYSTAT, was used to determine the IC 50 of enzymatic inhibition.

RESULTS AND DISCUSSION
Cloning of Lymnaea 7B2 cDNA-Because a PC2 convertase is present in various types of peptidergic neurons in the Lymnaea brain (13), we hypothesized that the activation of LPC2, like that of vertebrate PC2, requires interaction with a protein related to 7B2. Because vertebrate 7B2 proteins show a high degree of sequence identity, in principle many possibilities for the design of primers toward regions of interphylum sequence conservation exist. Therefore, a total of 10 degenerate oligonucleotide primers corresponding to different conserved parts of the vertebrate 7B2 sequences were designed and tested with PCR amplifications in 20 different combinations of primer sets. Only by using primers OL1 and OL2 was a PCR product of the expected size found. This PCR product was cloned and sequenced, and appeared to encode a sequence similar to that of vertebrate 7B2 (data not shown).
We used the PCR fragment to screen 80,000 independent clones of a cDNA library of the cerebral ganglia of L. stagnalis. positions Ϫ13 and Ϫ20, indicating that the coding region is complete at the 5Ј end (Fig. 1). Translation of the mRNA is therefore likely to be initiated at methionine residue 1. Northern blot analysis showed a transcript of ϳ1.6 kilobase pairs (data not shown), indicating that the cDNA clone is indeed full-length.
Sequence Analysis of L7B2 Demarcates Only a Few Small Evolutionarily Conserved Regions-The predicted L7B2 protein is organized similarly to vertebrate 7B2 (Figs. 1 and 2), with a hydrophobic leader sequence, an NT domain, and a CT PC2 inhibitory domain (see below). Cleavage of the signal peptide most likely occurs after residue Ala-17 (19), providing a signal sequence that is shorter than that of vertebrate 7B2 (20, 21) (Fig. 2A). The L7B2 protein (calculated molecular mass, 28 kDa) is considerably larger than vertebrate 7B2 (calculated molecular mass, 20.8 kDa). The 7B2 sequences characterized in vertebrates show a high amino acid sequence identity, overall ranging from 71 to 99%. The sequence alignment of L7B2 with vertebrate 7B2 (Fig. 2) reveals a remarkably low degree of amino acid sequence identity (29%) and similarity (46 -48%). Nevertheless, based on the structural organization and the overall degree of sequence conservation, we conclude that cDNA clone L7B2 encodes Lymnaea 7B2.
The deduced amino acid sequence reveals that in L7B2 various pairs of basic residues are present that are putative sites for endoproteolytic cleavage (7) (Lys 201 -Arg 202 , Lys 208 -Lys 209 , Lys 215 -Arg 216 , and Lys 238 -Lys 239 ). Vertebrate 7B2 proteins contain three pairs of basic residues, except for salmon 7B2, which contains only two pairs, and mouse and rat 7B2, each of which contains four pairs. Three of the four pairs in L7B2 align with sites at analogous positions in vertebrate 7B2 (Fig. 2), but none are consensus sites for furin enzyme activity, a situation different from mammalian and Xenopus 7B2 but similar to salmon 7B2 (12). Mammalian 7B2 proteins are cleaved at two sites during their transport through the secretory pathway; e.g. porcine 7B2 is cleaved after Arg 150 -Arg-Lys-Arg-Arg 154 and after Lys 171 -Lys 172 (numbering of 7B2 proteins refers to that used in Fig. 2A) (20,21). Processing of L7B2 to NT and CT domains might occur at Lys 201 -Arg 202 , a site corresponding to the 7B2 cleavage site in Xenopus (22).
Of special interest are two CT domains in L7B2, namely CT1, comprising Leu 203 to Arg 217 , and CT2, from Ser 218 to His 256 (Fig. 2, B and C). The region in vertebrate 7B2 corresponding to CT2 displays a substantial degree of sequence identity with these domains, whereas the region in vertebrate 7B2 corresponding to CT1 has diverged. The vertebrate 7B2 region analogous to CT2 is indeed a potent inhibitor of PC2 enzyme activity (4 -6, 9), whereas the NT domain containing the CT1 region fails to be active (5). In particular, the Lys 171 -Lys 172 pair present in the inhibitory CT2 domain is essential (5,9). The Lymnaea CT1 and CT2 domains may be internally cleaved because each contains a Lys-Lys pair (Lys 208 -Lys 209 and Lys 238 -Lys 239 for CT1 and CT2, respectively) ( Figs. 1 and 2). As shown below, both LCT1 and LCT2 inhibit LPC2 enzyme activity.
The sequence identity between the NT domain of L7B2 and the vertebrate counterparts is predominantly restricted to a 10-amino acid proline-rich region (residues Pro 90 -Thr 99 , rat 7B2; Pro 108 -Thr 117 , L7B2) and several conserved scattered residues (Fig. 2). It has been suggested that the region with sequence similarity with a portion of the 60 kDa subclass of molecular chaperones (residues 1-90) might represent the region involved in PC2-7B2 interaction (2). However, unlike the total NT domain and total 7B2, the vertebrate 7B2 1-90 NT domain has no influence on proPC2 maturation (3). Also, residues 90 -185 do not support the maturation of proPC2 (3), indicating that in addition to the strongly conserved prolinerich region (see "Addendum") (residues 90 -99), other residues, e.g. those conserved between L7B2 and vertebrate 7B2, might be of importance for 7B2-PC2 complex formation.
LPC2 Enzyme Activity Is Inhibited by LCT1 and LCT2-To test the inhibitory effect of L7B2 CT1 and CT2 on LPC2 enzyme activity, the soluble fraction of extracts of the insulinrelated peptide-producing neuroendocrine LGC was used as an enriched source of LPC2. In situ hybridization on alternate sections of the Lymnaea brain revealed the cellular colocalization of LPC2 and L7B2 in the LGC of the cerebral ganglia (Fig.  3). Therefore, enzyme activity present in the LGC extract was used to hydrolyze the fluorogenic substrate Pyr-Arg-Thr-Lys-Arg-AMC in the presence or absence of synthetic LCT1, LCT2, and a control peptide not structurally related to L7B2. In the LGC, the furin-like convertase Lfur1 (13) is expressed at low levels, whereas Lfur2 (23) expression is not detectable. 3 In contrast to the LGC, the salivary gland is devoid of LPC2 (12) but expresses Lfur2 (23), and therefore it served as a control for endoprotease activity not related to LPC2.
Dose-response analysis revealed that in the LGC extract, the inhibition of protease activity by LCT2 displays a biphasic character (Fig. 4A). To discriminate between the inhibition of LPC2 and of non-LPC2 activity, LPC2 was specifically removed from the cell extract by immunoprecipitation (Fig. 4B). The high affinity inhibition by LCT2 was lost after LPC2 immunodepletion, whereas the remaining activity was inhibited only at a high dose (IC 50 of 31 Ϯ 3 M). The inhibition profile on LPC2 activity was determined by subtraction of non-LPC2 (open circles) inhibition of LPC2 activity calculated by subtraction of LPC2-immunodepleted activity ( filled squares and filled circles) from total activity in the LGC extract. D, as in A, but using extracts of the salivary gland. About 10% of the total enzyme activity cannot be inhibited by 100 mM EDTA nor by LCT2 at millimolar concentrations in extracts of the LGC and salivary gland. Experiments were performed in triplicate. IC 50 s were calculated by nonlinear regression using the program SYSTAT. activity from the activity in the LGC extract (Fig. 4C). A high affinity inhibition of LCT2 is found at 1.3 Ϯ 0.3 nM, which is in the same range as the IC 50 of vertebrate CT on PC2 (5,9). Because LCT2 shows no inhibition toward Ca 2ϩ -dependent proteases present in the soluble fraction of the salivary gland (Fig. 4D) and because previous experiments revealed that the CT domain is not an inhibitor of PC1/3 activity (5), the nanomolar inhibition of LCT2 very likely involves LPC2 activity, whereas the inhibition at high concentrations concerns other, as yet unidentified enzymes. Addition of 100 mM EDTA to either the LGC extract or the salivary gland extract resulted in a conversion of 10%, displaying the residual activity of Ca 2ϩindependent proteases. Although LCT1 is much less potent than LCT2, it also shows a biphasic curve, and upon LPC2 immunodepletion from the LGC extracts, the residual activity is inhibited at 40 Ϯ 3 M (Fig. 4, A and B). The IC 50 of LCT1 (ϳ2.5 Ϯ 0.2 M) toward LPC2 activity was determined by subtraction of non-LPC2 activity from the activity in the LGC extract (Fig. 4C).
Interspecies Conservation of PC2-7B2 Interaction-To examine an interspecies functional conservation of the inhibitory LCT, both LCT1 and LCT2, as well as the control peptide, were tested on recombinant purified mouse PC2 and PC1/3 (14,18). LCT2 inhibits mouse PC2 with an IC 50 of 36 Ϯ 3 M, and the less active LCT1 inhibits mouse PC2 with an IC 50 of 154 Ϯ 4 M), whereas mouse CT inhibits PC2 activity at 45 Ϯ 3 nM. PC1/3 enzyme activity is not inhibited by either of these peptides (data not shown). Thus, the LCT1 and LCT2 peptides inhibit LPC2 at low micromolar and nanomolar concentrations, respectively, whereas they inhibit mouse PC2 only at higher concentrations. This is likely due to sequence divergence and reflects the evolutionary distance between vertebrates and invertebrates (ϳ600 million years). So the region that resembles the carboxyl terminus of vertebrate 7B2 most (i.e. LCT1) has the least potency with respect to the inhibition of (L)PC2 activity. This finding demonstrates that carboxyl-terminal regions of L7B2 with quite different amino acid sequences are able to inhibit the catalytic site of (L)PC2. Our results reveal that during evolution the inhibitory action of the CT domain on PC2 activity has been conserved from vertebrates to invertebrates.
Addendum-In a recent publication, Zhu et al. (24) suggest the importance of a proline-rich region in 7B2-PC2 interaction. Interestingly, all prolines in this region are conserved in L7B2.