Egg-laying hormone peptides in the aplysiidae family

Egg-laying behavior in the marine mollusc Aplysia californica is controlled by two clusters of peptidergic neurons. Each bag cell cluster contains several hundred homogeneous neurons located at the rostral end of the abdominal ganglion. These bag cell neurons synthesize an egg-laying hormone (ELH) precursor protein which, upon post-translational processing, results in multiple peptides, including ELH, α-, β- and ε-bag cell peptides (BCPs) and acidic peptide (AP), many of which have distinct physiological and behavioral roles (Conn and Kaczmarek, 1989). Because egg-laying behavior plays a crucial role in the reproductive capacity of gastropod molluscs, several studies have been carried out to characterize ELH genes or peptides encoded by these genes in different species. One of the first characterized peptides, the ELH of A. californica, has been sequenced and its gene was characterized soon afterwards (Scheller et al., 1983). Nagle et al. (1988a) demonstrated that the ELH of A. brasiliana is identical to that of A. californica. Furthermore, molecular genetic studies of the ELH gene family in A. parvula predict the synthesis of a 36-residue ELH-related peptide which is 78 % identical to A. californica ELH (Nambu and Scheller, 1986). Lastly, injection of abdominal ganglion extracts from A. californica induces egg laying and associated behaviors in A. vaccaria, A. brasiliana and A. dactylomela, which further suggests a high degree of homology of ELH peptide families among different species (Chiu and Strumwasser, 1981).

The current study focuses on the biochemical characterization of ELH peptides of two species from the Aplysiidae family, namely Aplysia vaccaria and Phyllaplysia taylori. As opposed to conventional peptide sequencing approaches, which require concentration and purification of a particular peptide from large amounts of starting material, we describe several mass spectrometric (MS) strategies that can be used to sequence ELH peptides fully using less starting material. We have previously demonstrated the use of matrix-assisted laser desorption/ionization (MALDI) MS to detect peptides in single neurons and connective tissues (Garden et al., 1996, 1998; Li et al., 1998). Single-cell MALDI MS of bag cells from A. vaccaria and P. taylori reveals similar peptide profiles to those from A. californica. Molecular mass measurement and subsequent sequence analysis using several mass spectrometric approaches have been performed. As many of these techniques may not be familiar, the Appendix contains a glossary of the mass spectrometric terms of importance. Post-source decay (PSD) techniques with MALDI MS indicated that the α-BCPs are identical among all three species examined. In addition, electrospray ionization (ESI) Fourier transformed ion cyclotron resonance (FTICR) MSⁿ analysis was used to obtain primary structure of several larger peptides from A. vaccaria. ε-BCP and ELH are found to have a high degree of homology to those of A. californica, each with only a few amino acid substitutions.

The Aplysiidae species selected for this study are native to the Pacific northwest USA. Aplysia californica (10–1000 g) were obtained from Aplysia Research Facility (Miami, FL, USA), Pacific Biomarine (Venice, CA, USA) and Marinus Inc. (Long Beach, CA, USA). Some of the larger animals (>1000 g) were collected off the Monterey Peninsula between January and July 1998. Aplysia vaccaria were obtained from Pacific Biomarine. Phyllaplysia taylori (body length 2–5 cm) were collected off the common eelgrass (Zostera) from San Juan Island, WA, USA, under the auspices of Friday Harbor Marine Station. Animals were maintained in tanks of aerated artificial sea water (Instant Ocean, Aquarium Systems, Mentor, OH, USA) at 14 °C.

Ganglia with intact connectives and commissures were removed after an injection of 390 mmol l⁻¹ MgCl₂ equal to half of each animal’s mass into the body cavity of the animal. In some cases, a moderate protease treatment (e.g. 1 % Protease Type IX for 30–60 min at 34 °C) was used to soften the connective tissues prior to cell dissection. Extracellular salts hinder MALDI MS analysis and were removed using a previously described approach (Garden et al., 1996). Briefly, the abdominal ganglion was isolated and pinned down in a dissection dish, and the physiological saline was replaced with an aqueous MALDI matrix solution, 10 mg ml⁻¹ 2,5-dihydroxybenzoic acid (ICN Pharmaceuticals, Costa Mesa, CA, USA). Tungsten needles were used to isolate and transfer the bag cell(s) onto a MALDI sample plate containing 0.5 μl of matrix solution. After drying at ambient temperature (20–25 °C), samples were either frozen at −20 °C for future analysis or analyzed immediately.

Peptides were extracted from two bag cell clusters from one A. vaccaria, whereas eight bag cell groups were pooled from four P. taylori. Sample preparation and microbore liquid chromatography (LC) separation were performed in a manner similar to that described by Floyd et al. (1999). The bag cell clusters were collected on dry ice and subsequently stored at −80 °C. Peptide extracts were made in 500 μl of acidified acetone (40:6:1 acetone:water:HCl) followed by 200 μl of 4 mol l⁻¹ urea. Samples were homogenized in a micro-homogenizer (Jencons Scientific Ltd, UK), sonicated for 5 min (Branson 2200, Danbury, CT, USA), and centrifuged for 10 min at 13 000 revs min⁻¹ in a microcentrifuge (Baxter, McGaw Park, IL, USA). The supernatant was removed, freeze-dried (Labconco, Fisher Scientific, Itasca, IL, USA) and resuspended in 50 μl of 2 % acetonitrile in 0.1 % aqueous trifluoroacetic acid (TFA). A sample (20 μl) of the extract was injected in a reversed-phase microbore LC instrument (Magic 2002, Michrom BioResources, Auburn, CA, USA) consisting of a Reliasil C-18 column (0.5 mm×150 mm) with 300 å packing. The flow rate was 25 μl min⁻¹ at ambient temperature. The column was equilibrated with solvent A (98 % H₂O, 0.1 % TFA, 1.9 % acetonitrile), and a gradient was developed from 0 % to 80 % of solvent B (90 % acetonitrile, 9.9 % H₂O, 0.1 % TFA) in 30 min and then from 80 % to 98 % of solvent B in 10 min. Samples were collected by a fraction collector (Gilson FC 203B, Middletown, WI, USA), and each fraction was screened by MALDI MS; 0.5 μl of each LC fraction was deposited onto a MALDI sample plate followed by the same volume of 2,5-dihydroxybenzoic acid matrix solution. Thus, more than 95 % of each fraction was available for further assays.

MALDI mass spectra were obtained using a Voyager Elite and a Voyager DE STR (PE Biosystems, Framingham, MA, USA) time-of-flight mass spectrometer equipped with delayed ion extraction. A pulsed nitrogen laser (337 nm) was used as the desorption/ionization source, and positive-ion mass spectra were acquired using both linear and reflectron mode. Each representative mass spectrum shown is the average of 128–256 laser pulses. Mass calibration was performed externally using either bovine insulin (Sigma, St Louis, MO, USA) and synthetic Aplysia α-bag cell peptide (BCP) (American Peptide Co., Sunnyvale, CA, USA) or a previously calibrated spectrum obtained from A. californica bag cells. Mass spectral peaks were assigned on the basis of comparison of observed and calculated masses. In the case of A. vaccaria and P. taylori, where peptide sequences are unknown, the initial spectral assignments were based on the similarity of the peptide profiles to those in A. californica. The identities and sequences of several peptides were subsequently determined using the MS-based sequencing methods described below.

Equal volumes of microbore LC fractions and matrix solution were either premixed in the vial or mixed on the MALDI sample plate. For MALDI PSD analysis, the matrix was α-cyano-4-hydroxycinnamic acid (10 mg ml⁻¹ in 6:3:1 acetonitrile:water:3 % TFA) (Aldrich, Milwaukee, WI, USA). The total acceleration voltage was 20 kV, and a delay time of 75 ns was used. Spectra were obtained by accumulating data from 100–256 laser shots. To obtain complete PSD spectra, a series of reflectron mass spectral segments was acquired, each optimized to focus fragment ions within different ranges of mass-to-charge ratio (m/z, where m is mass and z is charge number of an ion) (Spengler, 1997). Each segment was stitched together to generate a composite PSD spectrum.

In-source decay (ISD) mass spectra (Brown and Lennon, 1995) were acquired in the linear mode, the acceleration voltage was 20 kV, and laser power was applied well above the threshold value to facilitate fragmentation. The delay time used was 300–400 ns. Mass spectra were averaged over 100 laser shots and smoothed (19-point smoothing, Savitsky-Golay). Aqueous 2,5-dihydroxybenzoic acid matrix solution was used to prepare the samples for ISD analysis.

The instrumentation used in this study has been described in detail elsewhere (Winger et al., 1993). Briefly, the FTICR mass spectrometer is based on a 7 T superconducting magnet and is equipped with an external home-built ESI source. ESI was performed in the positive ion mode with a sample flow of 300 nl min⁻¹ using a Harvard Apparatus syringe pump (South Natick, MA, USA). The ESI source consists of a heated stainless steel ‘desolvation’ inlet capillary, a skimmer with an orifice diameter of 1 mm, and a short quadrupole segment, for collisional ion focusing, added to the set of two quadrupole ion guides used in the original configuration and operated in rf-only mode (approximately 750 kHz, approximately 500 V peak-to-peak, V_pp). Mass spectra were obtained utilizing standard experimental sequences employing in-trap ion accumulation (i.e. ion injection, pump-down and excitation/detection). A piezoelectric pulse valve (Laser techniques Inc., Albuquerque, NM, USA) was used to inject N₂ (to approximately 10⁻⁵ mmHg) for accumulated trapping of ions and for ion activation in collisionally induced dissociation (CID) experiments. Background pressure in the ICR trap was maintained at approximately 10⁻⁹ mmHg by a custom cryopumping assembly that provides pumping speeds in excess of 10⁵ l s⁻¹ and thus allows rapid transition between in-trap ion accumulation (i.e. 10⁻⁵ mmHg) and high performance ion excitation/detection (i.e. 10⁻⁹ mmHg) events. Typically, coherent ICR motion was excited by dipolar frequency sweep excitation (approximately 125 V_pp amplitude, excitation over a 200 kHz bandwidth with a 35 Hz μs⁻¹ sweep rate); Fourier transformation of the resulting discrete time domain signal (256×10³ data points at 270 kHz), followed by magnitude calculation and frequency-to-m/z conversion, yields a FTICR mass spectrum. Tuning and calibration of the mass spectrometer were achieved by using an acidic solution of methanol/water/acetic acid (50:49:1, v/v/v) containing peptide standards.

In MS² experiments, stored-waveform inverse Fourier transformed (SWIFT; Marshall et al., 1985) radial ejection was used to remove ions of all but a small selected m/z range (i.e. ion isolation). The m/z-selected parent ions are then translationally excited for dissociation by collisional activation employing low-amplitude (typically approximately 20 V_pp) sustained off-resonance irradiation (SORI; Gauthier et al., 1991) for 100 ms at a frequency approximately 1 kHz lower than the reduced ICR frequency of the parent ion. Optionally, after another SWIFT ejection event to isolate m/z-selected daughter ions, an additional SORI-CID event was performed in some cases to generate an MS³ spectrum. MSⁿ experiments (i.e. ion injection, isolation, dissociation and excitation/ detection) were repeated 25 times, and the resulting spectra were co-added before Fourier transformation.

An Odyssey data station (Finnigan, Madison, WI, USA) provided ICR trap control, data acquisition and storage. Data analysis was performed using ICR2LS, a software package developed at PNNL (Anderson and Bruce, 1995).

Within the Aplysiidae family, three different Aplysia species from the Pacific northwest USA including A. californica, A. vaccaria and P. taylori were selected for this study. Fig. 1 is a schematic diagram showing the position of bag cell neurons in these three species. The bag cell neurons in both A. californica and A. vaccaria form two bilaterally situated clusters, located at the junction of pleural–abdominal connectives and the abdominal ganglion, whereas in P. taylori approximately 100 bag cells are distributed at the top of the abdominal ganglion, forming less compact clusters. Furthermore, the central nervous system (CNS) of P. taylori is more asymmetric than that of A. californica, with the right pleural–abdominal connective significantly shorter than the left and the right cerebral–pleural connective much longer than the left in all specimens examined. Thus, the summed length of the two connectives is approximately the same for each side.

MALDI mass spectra of bag cells in A. californica, A. vaccaria and P. taylori show similar peptide profiles (Fig. 2). Essentially, all peaks in the spectra correspond to the protonated molecular masses of peptides. The signals in the mass spectra obtained for A. californica are assigned on the basis of predicted masses from the known ELH prohormone sequence (Garden et al., 1998). Both identified peptides and shortened processing forms are labeled in the spectrum. The signals in the mass spectra acquired from A. vaccaria and P. taylori are putatively assigned on the basis of their similarity to their counterparts in A. californica mass spectra. For example, peaks corresponding to α-BCP and its two shortened forms are detected in both A. vaccaria and P. taylori and have the same mass within 10 p.p.m. (±0.01 Da). β-BCP also appears to be identical in A. vaccaria and A. californica (they have the same mass within 16 p.p.m.). In A. vaccaria, the high abundance signals are observed at 4284.2 Da and 2880.1 Da (masses are given as average molecular mass unless stated otherwise), which resemble ELH (monoprotonated molecular mass 4385.1 Da) and AP (2961.3 Da) in A. californica, respectively. The mass differences may be due to amino acid substitutions in peptide sequences between different species. Because of the apparent similarity of the ELH peptides in A. vaccaria to those in A. californica, these peptides are named vε-BCP, vAP and vELH, respectively. The elucidation of the sequences of these peptides is discussed below.

Single-cell MALDI of bag cell neurons from the three species examined reveal the presence of α-BCP, α-BCP_1–8 and α-BCP_1–7. To verify the sequences of the BCPs and to confirm that the same mass implies the same sequence, additional mass spectrometric techniques are needed. In MALDI MS, PSD provides information on the sequence of short peptides from inherent ion fragmentation (Spengler, 1997). Using established nomenclature, cleavage of an amide bond produces y-type ions retaining the C terminus and b-type ions retaining the N terminus (Siuzdak, 1996). In the case of PSD, immonium ions indicative of amino acid composition are also observed. PSD of microbore LC fractions containing α-BCP (1122.6 Da) from all three species were performed and compared (Fig. 3). The very similar peak patterns obtained for α-BCPs from all three species (Fig. 3) indicate that these peptides have the same sequence. Using the MS-Tag database (available at http://prospector.ucsf.edu) and the masses of only the major peaks, the sequence of α-BCP was obtained, confirming our initial assignments. Therefore, α-BCP is 100 % conserved across all the Aplysiidae species examined. The BCP data for multiple species are summarized in Table 1.

Because of the similar relative LC retention time, a peptide of m/z at 1788 is putatively identified to be ε-BCP in A. vaccaria (vε-BCP). If this is correct, the 74 Da mass difference compared with its counterpart in A. californica indicates the presence of amino acid substitutions. Both MALDI and ESI MS are soft ionization methods that provide mainly molecular mass information without producing fragment ions, a vital additional step in sequencing peptides. Larger peptides such as vε-BCP can be difficult to fragment using PSD, and complete sequence analysis cannot always be obtained. Thus, an alternative mass spectrometric approach has been used, specifically ESI-FTICR coupled with SORI-CID. Dissociation of a parent ion is particularly simple using an FTICR instrument, where sequential stages of an MSⁿ experiment are separated in time. In SORI-CID, ions are excited by irradiation with an off-resonance electric field and slowly activated by repeated collisions. Thus, only the lowest energy pathways for dissociation occur, resulting primarily in b-and y-type ions and/or the loss of neutrals (ammonia or water). However, MS² generally will not yield complete sequence information from larger peptides. FTICR offers the ability to perform multistage MSⁿ analysis to gain enhanced sequence information. Using MSⁿ techniques, along with the high-resolution and mass accuracy primary spectrum, we have completely and unambiguously assigned the sequence to vε-BCP.

Fig. 4A illustrates the main steps and results of multistage SORI-CID of vε-BCP. SORI-CID of the SWIFT isolated molecular ion (M+2H)²⁺ (m/z 895) yields a series of fragments. To obtain more complete sequence information, four high-abundance product ions, i.e. m/z 850 (2+), 1092

(1+), 1388 (1+) and 1475 (1+), were selected and isolated, for yet another SORI-CID event resulting in the MS³ product ion spectra shown in the bottom panels of Fig. 4A. After mass analyses of fragment ions from each MS stage and comparison of the results with dissociation patterns of A. californica ε-BCP (cε-BCP), we confirmed that ions b1–b11 of the two peptides are the same. This indicates that the first 11 amino acid residues of vε-BCP are identical to those of A. californica ε-BCP. The difference appears at position 12, where a 42 Da mass difference is noted, resulting from a Glu in ε-BCP being substituted by a Ser in vε-BCP. Similar fragment ion comparison allows two other amino acid substitutions to be located at position 16 and 19. A complete sequence of vε-BCP is shown in Table 2. The high resolution and mass measurement accuracy (approximately 1 p.p.m.) achievable with FTICR MS allowed confirmation of the sequence of vε-BCP (see Fig. 4B). The mass error in this case is 0.06 p.p.m. (±0.001 Da), which further confirms the derived sequence, since few other amino acid substitutions can result in identical molecular masses within this level of uncertainty.

As shown in Fig. 2, the mass of A. vaccaria ELH (vELH) is approximately 101 Da less than that of A. californica ELH (cELH). A common procedure for obtaining the sequence information on larger peptides involves a combination of cleavage enzymes and mass spectrometry. However, we have previously shown that, in A. californica, a number of shortened ELH peptides are observed, in essence providing us with the information we need without the addition of proteolytic enzymes. For example, we documented novel ELH processing in single A. californica bag cells with cleavage at Leu–Leu and Leu–Arg bonds (Garden et al., 1998). The MALDI spectrum of bag cells from A. vaccaria also indicates similar processing products. By comparing these shortened forms (Fig. 2A,B), we have noted that vELH_1–14 has an identical molecular mass (to within 50 p.p.m.) to that A. californica ELH_1–14, indicating that the sequence of the first 14 amino acids of both ELHs are probably the same. Thus, amino acid substitutions occur at the second half of the peptide, vELH_15–36, which is approximately 101 Da lighter than its counterpart in A. californica. In addition, a second cleavage at Leu29–Arg30 is observed in A. californica bag cells, yielding two shortened forms, ELH_1–29 and ELH_30–36. Similar processing forms are detected in A. vaccaria bag cells, with approximately 44 and 57 Da mass differences, respectively. This observation implies that at least one amino acid substitution occurs between the fifteenth and twenty-ninth residue, resulting in a mass difference of 44 Da, and at least one amino acid difference occurs between the thirtieth and thirty-sixth residue, yielding a mass difference of 57 Da.

ESI-FTICR MS was used to analyze an LC fraction containing intact vELH. As shown in Fig. 5A, vELH and cELH were mixed and analyzed together to obtain an accurate mass difference. The measured molecular mass (monoisotopic mass, M_i) of vELH is 4280.4934 Da, and the mass difference between two forms of ELH is 100.95 Da (Fig. 5A). To determine the complete sequence of vELH, multistage SORI-CID, as outlined for vε-BCP sequencing, was performed. The MS² spectrum showed a series of b and y ions (Fig. 5B, top trace), and numerous y ions and several b ions were detected in the MS³ spectrum of b₂₈³ daughter ion (Fig. 5B, bottom trace), significantly enriching the available sequence information. Because of the relatively large size of the peptide, SORI-CID did not yield fragment ions that give complete sequence information. However, on the basis of the high degree of homology of these two peptides and the fragments obtained for the vELH peptides, it is possible to deduce its sequence. The sequence of vELH is further confirmed using MALDI ISD analysis, in which a series of Cn-type ions were observed. Fig. 5C shows the MALDI spectrum of ISD analysis, with amino acid residue deduced by the mass difference between two consecutive ladder ions. Since the low mass gate was set at m/z 500, only the ladder ions with a mass greater than 500 were detected and used for sequence assignment. Thus, the observed series of Cn-type ions confirms the sequence from sixth to thirty-sixth residue. Compared with cELH, six amino acid substitutions occur at positions 15, 16, 17, 21, 24 and 31 in vELH. The high mass accuracy measurement of vELH (theoretical and experimentally observed masses agree within 1 p.p.m.) further supports the deduced sequence. Table 3 compares the sequence and amino acid substitutions for both ELHs from A. californica and A. vaccaria. The sequence of the A. vaccaria ELH and the predicted proteolytic processing agree with the single cell MALDI spectra presented earlier.

The ELH prohormone produces several additional peptides. One of the most notable peptides is acidic peptide (AP), found adjacent to ELH on the prohormone. Presumably, AP is localized to the same vesicles as ELH and ε-BCP, and it has been shown to be released during the electrical activity of the bag cell neurons, although no biological role has been determined (Conn and Kaczmarek, 1989). AP undergoes the same additional processing as ELH (i.e. cleavage at Leu–Leu bonds) with more than 10 shortened forms detected in A. californica. We have briefly compared vAP with A. californica AP and detected a significant number of amino acid substitutions. Using SORI-CID, a partial vAP sequence was obtained, with residues at the N and C termini being: Phe-Leu-Gly-Ala-Leu-Gly-Glu-Ser-Val-Asn-Ser-Thr-Ser-Gly-Lys-Leu-Leu-Glu. Because of the lack of known bioactivity, further characterization was not attempted.

Examination of the ELH prohormone peptide products provides information about the organization and evolution of the ELH genes, and possible information about the portion of a peptide important for receptor recognition through examination of homologous regions between species. The A. californica ELH precursor encodes α-BCP, β-BCP, γ-BCP, δ-BCP, ε-BCP, ELH and the AP peptides. In A. californica, α-BCP_1–8 and α-BCP_1–7 have higher activities than the full-length nine-residue peptide (Sigvardt et al., 1986). Our results indicate that α-BCP has 100 % homology among the three species examined. Interestingly, all three forms of α-BCPs (i.e. α_1–9, α_1–8 and α_1–7) were detected in bag cell neurons from A. vaccaria, P. taylori and A. californica, which supports our previous observation that α-BCP in A. californica bag cells undergoes intracellular processing (and not extracellular cleavages) to generate the more potent α_1–8 and α_1–7 forms. Previous studies showed that α-BCP in A. parvula is identical to its counterpart in A. californica (Nambu and Scheller, 1986), and α-CDCP in Lymnaea stagnalis have five consecutive amino acid residues in common with α-BCP, suggesting that α-BCPs and α-CDCP have similar receptors and may have similar functions. The α-BCP in A. californica has been reported to have autoexcitatory feedback effects on the bag cell neurons responsible for its release (Rothman et al., 1983; Rock et al., 1986; Brown and Mayeri, 1989). However, these findings contrast with the observation that α-BCP terminates the bag cell discharge and inhibits the elevation of cyclic AMP levels in bag cells, which suggests an inhibitory effect (Kauer et al., 1987; Berry, 1988). A more recent study suggests a temperature-dependent peptidergic feedback effect for α-BCP in seasonal egg laying in Aplysia californica (Redman and Berry, 1991), which implies that α-BCP is autoinhibitory at typical winter temperatures, but becomes autoexcitatory as ocean temperatures rise in the summer, and indicates this peptide may serve to regulate egg laying in response to seasonal temperature variations. The conservation of α-BCP among all Aplysiidae species examined confirms the important role of this peptide.

In addition, several other bag cell peptides also show high degree of interspecies homologies. β-BCP and γ-BCP both share a sequence of four amino acid residues in common with α-BCP (Scheller et al., 1983). As shown in Table 1, β-BCP and γ-BCP in A. vaccaria also appear to be identical to those in A. californica. Early evidence suggest that β-BCP and γ-BCP may also have direct autoreceptor-mediated actions on the bag cell neurons (Brown and Mayeri, 1986). Interestingly, β₁-CDCP in Lymnaea stagnalis is completely homologous to the Aplysia β-BCPs and also exhibits a high degree of homology with β₂-CDCP and β₃-CDCP (Geraerts et al., 1988). As another example of BCPs, albeit with a different primary structure, the overall sequence identity of vε-BCP with ε-BCP is 84 %. Although no biological function is known for ε-BCP, the high degree of homology of this peptide across different species of Aplysia, together with our previous results showing that ε-BCP is released from electrically stimulated A. californica bag cell neurons in a Ca²⁺-dependent manner (Garden et al., 1998), suggests that this peptide warrants further study of potential physiological or behavioral roles.

In fact, the highly conserved nature of BCPs has been usefully exploited by using antisera to α-BCP or α-CDCP to locate putative ELH-producing cells in other species. Painter et al. (1989) used an antibody specific to α-BCP to localize two additional populations of central neurons that contain immunoreactivity. Identical labeling patterns were observed in brasiliana and A. dactylomela. Using antibodies raised against α-CDCP allowed the identification of unknown egg-laying regulating systems in various species of gastropod molluscs (Van Minnen et al., 1992). Furthermore, anti-α-BCP was used to localize putative egg-laying neuroendocrine cells (intercerebral white cells) in the nudibranch Archidoris montereynsis (Wiens and Brownell, 1994).

The neuropeptide hormone ELH is known to cause egg deposition in sexually mature Aplysia and to induce a series of stereotyped egg-laying behavior patterns (Conn and Kaczmarek, 1989). Thus, the comparison of ELH primary structure across different species is of special interest with regard to the evolutionary history of the reproductive behavior of the gastropod species. Table 3 provides a comparison of amino acid sequences of ELH-related peptides. As mentioned above, ELH prohormones have been previously identified in the CNS of opisthobranchs including A. californica (Scheller et al., 1982), A. parvula (Nambu and Scheller, 1986) and A. brasiliana (Nagle et al., 1988a). The present study contributes to biochemical characterization of ELH in A. vaccaria. In addition, multiple ELH-related genes are expressed in the atrial gland of A. californica (Nagle et al., 1986). The ELH-related atrial gland peptides that have been chemically characterized are A-ELH, [Ala²⁷]A-ELH and [Gln²³, Ala²⁷]A-ELH (Nagle et al., 1986, 1988b; Rothman et al., 1986). Each is a 36-residue basic peptide and is as potent as bag cell ELH in inducing egg laying when injected into animals (Nagle et al., 1989). Besides opisthobranch molluscs, the freshwater snail Lymnaea stagnalis contains neurosecretory caudo dorsal cells (CDCs) whose peptide product, CDC hormone-I (CDCH-I), initiates and coordinates egg deposition (Vreugdenhil et al., 1988). In addition, a recent report described the identification and characterization of genomic DNA from an abalone ELH (aELH) in Haliotis rubra (Wang and Hanna, 1998). A high degree of nucleotide homology (95.4 %) between aELH and CDCH was found, and the deduced amino acid sequence of aELH is 97.2 % homologous to that of CDCH and 47.2 % homologous to that of A. californica ELH. Obviously, the homology of ELH-related peptides exists not only in Aplysiidae species but also in several other distantly related molluscan species. For example, Van Minnen and coworkers (Croll et al., 1993) have shown that antibodies raised against α-CDCP and α-CDCH labeled numerous neurons in the bivalves Mytilus edulis, Mya arenaria and Placopecten magellanicus, suggesting that related peptides might be involved in the reproduction of both gastropods and bivalve molluscs. Besides the examples presented above, a recent report demonstrated the presence of ELH-like immunoreactivity in the nervous systems of three prosobranch species (Ram et al., 1998). Lastly, a leech egg-laying-like hormone (lELH) has recently been described and is included in Table 3 because it represents the first biochemical characterization of an egg-laying hormone in another phylum (Salzet et al., 1997).

Comparing the amino acid sequence of all egg-laying hormones reported to date (Table 3), A. vaccaria ELH shares 83 % sequence identity with A. californica and A. brasiliana ELH. In the case of A. parvula ELH, the sequence identity is 72 %. If one extends the comparison to L. stagnalis CDCH, vELH matches at 14 positions, making the sequence identity and similarity approach 39 % and 55 %, respectively. Furthermore, vELH shares 13 amino acid residues with aELH. Similarly, lELH matches vELH at 10 positions, yielding a sequence homology of 28 %.

As shown in Table 3, all the ELH-like peptides contain 36 amino acid residues. This observation indicates that the peptide chain length may be important for egg-laying activity, which has been confirmed by previous structure/activity studies using synthetic ELH analogs to induce egg-laying (Strumwasser et al., 1987). In these studies, removal of the N-terminal amino acid (ELH_2–36) or extension of C-terminal by one residue (ELH-Gly37) results in a loss of egg-laying activity. However, both ELH_1–34 and ELH_1–35 are at least moderately active in an egg-laying bioassay (Strumwasser et al., 1987). This result suggests that the C-terminal amide is not required for bioactivity. Furthermore, the residue heterogeneity at positions 35 and 36 in ELH-related peptides (shown in Table 3) indicates that the identity of these residues is less important for activity.

Also interesting when comparing the primary structure of ELH-related peptides is the pattern of the highly conserved regions near both N and C termini, with a more variable sequence in the middle region. Clearly, all seven sequences are identical at positions 1–14, with the exception of A. parvula ELH, which has a different residue at position 11. Substitutions are much more common in the middle segment, where five out of the six amino acid substitutions occur for A. vaccaria ELH (see Table 3). The majority of substitutions also occur for A. parvula ELH and A. californica atrial gland peptides in this segment. This region is followed by another highly conserved region between residues 29 and 34 at the C terminus, where only one substitution of Ala for Gln occurs at position 31 for A. vaccaria ELH. These data agree with the early observation and speculation that ELH and related peptides may be U-shaped molecules and that the conservation of N- and C-terminal sequences may be responsible for egg-laying activity in Aplysia (Nagle et al., 1989). Among all ten known ELH-related peptides, there are six identical amino acid residues, which are given in bold type in Table 3. Five of them occur within conserved regions (residues 1–14 and 29–34), and four of them are in the C-terminal region (Leu29, Arg30, Arg32 and Leu34), which may act as the critical receptor recognition/binding site.

The homology of the ELH-related peptides across different species is striking, especially considering the fact that leeches and molluscs are not in the same phylum. Other reported examples of highly conserved neuropeptides between molluscs and leeches include lysine-conopressin (Salzet et al., 1993) and the peptides GDPFLRFamide and SDPFLRFamide (Salzet et al., 1994). The structurally conserved neuropeptide families have been investigated in many distant animal groups (Boer et al., 1992). For example, the enkephalins, originally extracted from mammals, have been shown to be present in the nervous tissues of the molluscs Mytilus edulis and L. stagnalis (Leung et al., 1990). In addition, structurally closely related oxytocins and vasopressins have been identified in vertebrates (Hadley, 1988), insects (Proux et al., 1987) and gastropods (Sawyer et al., 1984; Cruz et al., 1987).

The conservation of molecular structure may imply a conservation of receptors and, in some cases, even of function. Interspecies injection experiments (Ram, 1982) demonstrated that nervous system extracts caused egg laying in different subfamilies within the order Anaspidea, whereas inter-order injections of nervous system extracts from Pleurobranchaea (Notaspidea) into Stylocheilus (Anaspidea) usually failed to induce egg laying. Such experiments suggest the evolution of ELH receptors in related opisthobranchs. ELH and α-BCP have been established as bioactive peptides in A. californica, and the highly conserved nature of these peptides in A. vaccaria suggests that these peptides also have important physiological functions in many Aplysiidae species.

There have been revolutionary advances in mass spectrometric technology over the last decade. Several of the newer mass spectrometric methods have been used here to determine the sequences of the ELH-related peptides. MALDI MS is used to provide accurate masses for the peptides in single cells without elaborate sample treatment, so that rapid peptide profiles can be obtained. This allows, for example, confirmation that the cells being examined in P. taylori are indeed the bag cells. To obtain sequence information, the peptide ions must be fragmented. We employed several mass spectrometric sequencing approaches to achieve this goal. Compared with conventional Edman sequencing, the main advantages of mass spectrometric peptide sequencing are its higher sensitivity, its compatibility with chemical modifications and the possibility of analyzing multiple component samples and poorly purified samples (especially in the case of MALDI-PSD). In the case of peptide sequencing for A. vaccaria, a single animal was used, demonstrating the higher sensitivity of MS-based approaches. For complete sequencing of the smaller peptides, PSD with MALDI MS is fairly straightforward. Multistage ESI-FTICR SORI-CID were applied to perform structural characterization of larger peptides such as vε-BCP and vELH.

One of the challenges for researchers interested in identifying new bioactive peptides is identifying the active peptides from the large number of peptides present at significant levels in most neuronal networks. Even if a putative neuropeptide gene is discovered and sequenced, it is difficult to determine which peptide cleaved from the prohormone is responsible for biological activity. As one strategy, one can use the fast-interganglionic transport of peptides to locate candidate peptides and their genes (Lloyd, 1989; Li et al., 1998). Specifically, MALDI MS can be used to determine directly (without sample treatment) the accurate masses of many peptides in the major connectives of the molluscs (Li et al., 1998). Next, by comparing peptide profiles between closely related species for conserved peptides (masses), a subset of candidate peptides for further characterization can be rapidly generated. We are currently using this approach in the Aplysiidae family.

A fragmentation technique used to obtain primary structural information. This approach utilizes collision with a target gas to induce fragmentation of a mass-selected ion.

A sample introduction/ionization method used to produce gas-phase ions from molecules in solution. This soft ionization technique, together with MALDI, revolutionized mass spectrometric analysis of large molecules, particularly biopolymers. The process generates ions by spraying a solution (aqueous or organic solvent) through a charged inlet. As the solvent evaporates, ions in the highly charged droplets are ejected. The ions are then electrostatically directed into the mass analyzer. This ion source commonly produces multiply charged ions.

A mass analyzer based on the principle of a charged particle orbiting in the magnetic field. FTICR mass analyzer offers unrivaled mass measurement accuracy and resolution, as well as multistage tandem mass spectrometric capabilities (MSⁿ).

A fragmentation technique used with MALDI-TOF MS to obtain structural information typically from peptides greater than 1500 Da. The decay of the precursor ions occurs in the MALDI ion source, yielding a series of product ions indicative of amino acid sequence.

Tandem mass spectrometry (where n refers to the number of generations of fragment ions being analyzed) allows one to induce fragmentation of a particular mass analyte and perform successive MS experiments on these fragment ions. For example, MS² refers to fragmentation of precursor ion and analysis of the fragment ions.

A sample introduction/ionization method that generates ions by desorbing them from a solid matrix material with a pulsed laser beam. This ionization method is tolerant of high levels of impurities.

A fragmentation technique used with MALDI-TOF MS to obtain structural information typically from peptides less than 2000 Da. The decay of a mass-selected precursor ion occurs after the ion leaves the ion source in the flight tube, with the fragment ions being separated in the reflectron.

A device used in a time-of-flight mass spectrometer to increase the mass resolution. This device retards and then reverses ion velocities to correct for the flight times of ions having different kinetic energies and increase the effective flight path.

A method for extended excitation of ions for collisional activation in an FTICR mass analyzer (i.e. ions excited slightly off-resonance will be alternately accelerated and decelerated with a period equal to the difference between the excitation frequency and the ion cyclotron frequency). With FTICR, multi-stage SORI-CID events can be utilized to obtain enhanced structural information.

One of the simplest mass analyzers that is used to separate ions on the basis of their mass-to-charge ratio (m/z). The time of an ion traveling through the flight tube is correlated to its m/z, where m is its mass and z is its charge number, with lighter ions arriving earlier and heavier ions arriving later.

This study is supported by National Institutes of Health (NIH) Grant NS 31609 and National Science Foundation Grant CHE 96-22663 (to J.V.S.). Aplysia californica were partially provided by the National Resource for Aplysia at the University of Miami under NIH National Center for Research Resources Grant RR 10294. Pacific Northwest National Laboratory is operated by Battelle Memorial Institute for the US Department of Energy, Office of Environmental and Biological Research, through Contract No. DE-AC06-76RLO 1830. J.V.S. acknowledges the superb facilities and support of Friday Harbor Marine Laboratories, University of Washington, San Juan, WA, USA.