SEVEN FMRFamide-RELATED AND TWO SCP-RELATED CARDIOACTIVE PEPTIDES FROM HELIX

By 1960, the neurotransmitters in molluscan ganglia had been sufficiently well-investigated that some of the cardioexcitatory activity present in ganglia could no longer be ascribed to the major, known cardioexcitatory transmitter 5-hydroxytryptamine (5-HT). Among the studies suggesting the occurrence of unknown factors in ganglia, two of the earliest (Meng, 1960; Kerkut and Laverack, 1960) were conducted on species of Helix. Jaeger (1966) found that the novel activity in a related land pulmonate was inactivated by pronase and concluded that the active principle must be a peptide. Frontali et al. (1967) also attributed to peptides several peaks of cardioexcitatory activity separated chromatographically from clam and whelk ganglia..

Several of these putative molluscan cardioactive peptides have been identified and sequenced in the past 13 years. They fall into two families, the peptides related to FMRFamide and those related to the SCPs, and their distributions among the Mollusca and other phyla are under study (Price and Greenberg, 1989; Price et al. 1989). In this paper we look at the complement of cardioactive peptides in a single species, Helix aspersa, the relative potencies of these peptides as cardioactive agents and, ultimately, their roles (individually and in the aggregate) as regulators of the circulation in this snail.

Price and Greenberg (1977) identified the tetrapeptide amide FMRFamide from clam ganglia. Shortly afterwards, in collaboration with Cottrell, they began to search for FMRFamide in Helix aspersa using the radula protractor muscle from a whelk as a bioassay. No authentic FMRFamide was found, but only related peptides that seemed to have a more potent effect on the Helix heart than FMRFamide itself (Cottrell et al. 1981). Abandoning the bioassay, they began to use radioimmunoassay (RIA) to follow the purification of these FMRFamide-like peptides. Eventually, one of them, pQDPFLRFamide, was sequenced, and FMRFamide itself was identified in Helix, but several peaks remained uncharacterised (Price et al. 1985).

Three natural heptapeptide analogues of pQDPFLRFamide were soon discovered in other pulmonate snails; their general sequence was: X-DPFLRFamide, where X could be serine, glycine or asparagine (Price et al. 1931a; Ebberink et al. 1987). Preliminary analyses of two immunoreactive peaks in Helix led to the speculation that one of these heptapeptides (NDPFLRFamide), as well as a novel analogue (NDPYLRFamide), was present (Price et al. 1987b). We will show that this speculation was only partially correct; each peak actually contains two peptides: SDPFLRFamide together with NDPFLRFamide and SEPYLRFamide together with NDPYLRFamide.

Even as the FMRFamide-like peptides of Helix were being investigated on the basis of their immunoreactivity, Lloyd (1978, 1980) was using the Helix heart to assay the cardioexcitatory activity of ganglion extracts from this snail. Though he found a peak of activity on Sephadex G-15 at an elution time similar to that of FMRFamide, FMRFamide is such a poor excitor of the Helix heart that it could not account for the activity. Therefore, Lloyd (1978) designated the activity peaks ‘SCP’, an abbreviation for small cardioactive peptide. Later, he and his associates found a similar peak of activity in Aplysia, and showed that it is composed of two related peptides, a nonapeptide sequenced from A. brasiliana, SCP_B (Morris et al. 1982), and an undecapeptide sequenced from A. californica, SCP_A (Mahon et al. 1985; Lloyd et al. 1987).

Since both Aplysia SCPs are potent excitors of the Helix heart (Lloyd et al. 1987), homologous peptides should be present in Helix itself. Indeed, immunoreactive SCP-like activity was previously found in Helix, and one peptide had an amino acid composition that suggested it was homologous but not identical to Aplysia SCP_A (Price, 1987).

We have developed RIAs that are highly specific for each peptide family (Price, 1987) and have used them to follow the purification of those members present in Helix. Here we report the complete identification and synthesis of the FMRF-amide- and SCP-related peptides so far found in the snail, as well as a description and evaluation of their activity on isolated ventricles. These data provide a foundation for further studies on the physiological roles of peptides in cardiovascular regulation.

Whole snails (typically 25–75 animals, depending on size, to give about 40–50 ml) were crushed, their volume measured, and four volumes of acetone added. Ganglia were dissected from snails and dropped into a fourfold excess of acetone. The tissue was shaken in the acetone and put in the freezer for 12–48 h. The liquid was decanted, clarified by centrifugation and filtration through nylon filters (0.22 or 0.45 μm pore), and reduced in volume under a water aspirator vacuum with heating to 65°C until all of the acetone was removed. An equal volume of aqueous 0.1 % trifluoroacetic acid (TFA) or water was added, and the liquid again centrifuged and/or filtered to remove any precipitate.

The clarified aqueous solution was pumped through a reverse-phase HPLC column (Waters NovaPak C18, 3.9 mm × 150 mm, or Brownlee RP-18, 4.6 mm × 250mm, at 2mlmin⁻¹; Brownlee RP-300, 2.1mm×220mm, at 0.5mlmin⁻¹). After all the extract had been pumped through, the column was washed with 0.1 % aqueous TFA until the absorbance fell about 90 % of the way back to the preinjection level. The flow was then switched to the starting acetonitrile (ACN) composition. Once this solvent reached the detector (as shown by a rapid increase in the absorbance) an increasing, linear gradient of acetonitrile (slope of 0.8 % ACN min⁻¹) was started; 30s fractions were collected, and a 2 μl sample was taken from each fraction for radioimmunoassay.

The one or two most immunoreactive fractions from each peak were pooled and injected back onto the same column with either n-butanol or isopropanol as the organic solvent, and again with 0.1 % TFA throughout. Other conditions were the same as those for the first HPLC run, except that the flow rate was reduced to 1.5 ml min⁻¹ with the RP-18 column because of excessive pressure at 2 ml min⁻¹ with these alcohols. The peak one or two most immunoreactive fractions from each run were re-run through the ACN/TFA system. The two HPLC systems were used alternately until the peak was judged to be pure by its shape and by the correspondence between the amount of absorbance and immunoreactivity.

The SCPs - particularly SCP_B, which has two methionyl residues - oxidise readily, producing many different oxidation products (see Price, 1987, for examples). Furthermore, the SCP antiserum we use recognises the oxidised SCPs very well. Therefore, HPLC fractionation of SCP activity can give patterns that are difficult to interpret. One solution is to oxidise all the SCPs in the crude extract, but this can cause other problems. For example, Price (1987) previously reported an amino acid composition for a Helix SCP (reproduced here in Table 1) that had a lower than expected level of tyrosine, probably due to halogenation of the tyrosine by chloride present during oxidation. We have now avoided such complications by oxidising the peptide only after it is substantially pure, i.e. between the second and third HPLC steps. Since acidic conditions during oxidation favour the conversion of methionine to methionine sulphoxide over other products (see Neumann, 1967), we oxidised by direct addition of 50 μ1 ml⁻¹ of hydrogen peroxide (30 % solution) to the fraction containing 0.1 % TFA. The oxidation was terminated after 15 min by injection of the reaction mixture onto the HPLC.

The basic method of the RIA for FMRFamide has been described (Price et al. 1987a), and we have used the same procedure and antiserum (S253) for much of the work reported here. More recently we raised another antiserum, Q2, by an initial immunisation with a conjugate of pQDPFLRFamide followed by boosting with a conjugate of DDPFLRFamide (see Price, 1982, for conjugation and immunisation methods). Iodinated pQYPFLRFamide served as the trace for both FMRFamide assays.

For the SCP assay we used a rabbit antiserum to SCP_B from H. R. Morris with iodinated SCP_B as the trace. The iodination method, buffer and assay were carried out exactly as for the FMRFamide assays.

The method for obtaining FAB spectra of the FaRPs has been described (Bulloch etal. 1988); the SCPs were treated similarly.

FMRFamide was purchased from Sigma, CRB or Peninsula Laboratories. FLRFamide came from Sigma and pQDPFLRFamide and SCP_B from Peninsula. SDPFLRFamide was synthesised, deprotected and purified by John Riehm (Ebberink et al. 1987); and NDPFLRFamide, NDPYLRFamide, SEPYLRFamide and SGYLAFPRMamide were synthesised by the Protein Chemistry Laboratory of the University of Florida Interdisciplinary Center for Biotechnology Research. SGYLAFPRMamide and NDPFLRFamide were removed from the resin and deprotected using the standard HF method, while the NDPYLRFamide and SEPYLRFamide were deprotected and removed from the resin by trifluorometh-anesulphonic acid (TFMSA) according to the Applied Biosystems protocol. These four peptides were purified by HPLC and quantified by amino acid analysis. The amino acid compositions are shown in Table 2.

The method of Payza (1987) was used to bioassay peptides on ventricles isolated from Helix aspersa. After collection in the vicinity of Kingsbarns, Scotland, by Mr J. Brown, the snails were allowed to aestivate for several months. Ventricles were isolated and perfused for a minimum of 30 min with saline (composition in mmol l⁻¹: NaCl, 80; KC1, 5; MgC12, 5; CaCh, 7; Hepes, 20; pH 7.5) before any peptides were added. The perfusion rate was set at a minimum of 600 μlmin⁻¹ and was then maintained at a constant level for a particular ventricle, although the rate varied slightly from one preparation to the next.

Only two peptides at a time, or three in the case of the heptapeptides, were tested on a particular ventricle; each pair or triplet was tested on a minimum of five ventricles, recorded either isometrically or isotonically. Roughly equipotent doses of each peptide in turn were applied 10 min apart. The percentage increase in tension or amplitude was plotted against the logarithm of the dose, and the potency ratio for the pair or triplet of agonists was estimated from the linear portion of the plots.

Several peaks of FMRFamide-like immunoreactivity can be distinguished after HPLC fractionation of either whole snail or ganglionic extracts; the exact number depends on the antisera used for their characterisation. By using two antisera of differing specificity in parallel, we can distinguish at least five peaks in ganglionic extracts (Fig. 1A), of which only two, i.e. those containing FMRFamide and pQDPFLRFamide, had been completely identified in earlier work on Helix. We isolated them once again and, as expected, they contained the molecular ions for FMRFamide and pQDPFLRFamide (Table 3), but neither showed evidence of containing additional immunoreactive species.

Our analysis of three of the remaining immunoreactive peaks is presented below.

One peak, at the elution position of FLRFamide, does not appear as a distinct peak with either RIA alone, but the intersection of the two immunoreactive profiles shows it clearly (Fig. 1A). This peptide is a minor component of the total immunoreactivity, so it tends to appear as a shoulder on the FMRFamide peak when assayed with most antisera, e.g. our S253. In contrast, our Q2 antiserum, which was raised to an extended FLRFamide, separates the FLRFamide and FMRFamide peaks, but merges FLRFamide with the following peak (Fig. 1A). FABms analysis of the putative FLRFamide peak showed a clear molecular ion at 581 (Table 3). Thus, its identity was confirmed.

The second unknown FaRP corresponds to a peak that had previously been reported to elute near FLRFamide and to contain a tyrosine for phenylalanine substitution (Price et al. 1987b). This peak is not very immunoreactive with the S253 antiserum, but reacts well with the Q2 antiserum (Fig. 1A). Our first batch, purified from an extract of 10 suboesophageal ganglia, was used for FABms analysis and showed a prominent ion at 923. This value is just 16 Da (one oxygen atom - the difference between Phe and Tyr) more than NDPFLRFamide (907 molecular ion), as expected from the amino acid composition we had reported earlier. From a somewhat larger batch (15–20 central nervous systems), we ended up with a peak that was obviously a doublet (Fig. 2A) as judged by its ultraviolet absorbance. Since both components of the doublet seemed immunoreactive, and the ratio of 280 to 210 nm absorbance was constant through the peak (Fig. 2A), we decided to analyse each of the two most immunoreactive fractions (17 and 18) separately. So we divided each in half - one half for sequencing and the other for FABms. From the earlier-eluting fraction we obtained only the sequence NDPYLRF (Fig. 2B) and a 923 molecular ion, confirming our identification of this component as NDPYLRFamide. The later-eluting immunoreactive fraction sequenced as expected for a mixture of two peptides which share a common C-terminal pentapeptide sequence (Fig. 2C). Since one is NDPYLRFamide, the other must be SEPYLRFamide, and both expected molecular ions occur in this fraction (923 and 910; Table 3), a further confirmation of this sequence assignment.

The last unidentified peak appeared at the common elution position of the three peptides S-, N- and G-DPFLRFamide which are not separated by either of the solvent systems we used. All three had been identified previously in other pulmonate species (Price et al. 1987a; Ebberink et al. 1987), but not in Helix. FAB mass spectrometry of this purified peak clearly shows molecular ions corresponding to SDPFLRFamide at 880, and to NDPFLRFamide at 907 (Fig. 3; Table 3), but not to GDPFLRFamide which would appear at 850.

When extracts of only a few whole Helix or of dissected nervous systems were chromatographed, we observed three peaks of SCP immunoreactivity (Fig. IB). Large batches of material (i.e. from 40 or more individuals) produced only two broader peaks. The peak eluting after the position expected for MNYLAFPRM-amide (Fig. IB) seems to be a form of MNYLAFPRMamide which reverts back to the normal form during further purification (data not shown). We think that it may be a complex of the SCP and a small molecule, but we do not have rigorous evidence for this.

The later-eluting of the two largest SCP peaks is MNYLAFPRMamide. It elutes at the position expected for authentic SCP_B, it has the amino acid composition of SCP_B (Table 1), and the oxidised peptide has a molecular ion corresponding to SCP_B with two additional oxygens (Table 3).

The earliest SCP-like peak corresponds to a peptide reported previously, but incompletely identified (Price, 1987). Analysis of this peak by FABms gives a molecular ion corresponding to that expected from its amino acid composition (Table 1) assuming the methionyl residue is the sulphoxide. Microsequencing of this peptide yielded X-Gly-Tyr-Leu-Ala-Phe-Pro-Arg-Met (Fig. 4), where the first position was ambiguous because of contamination with free amino acids. Serine is one of the predominant amino acids in the first sequencing cycle, and the only full sequence consistent with the observed molecular ion is Ser-Gly-Tyr-Leu-Ala-Phe-Pro-Arg-Met-amide (SGYLAFPRMamide). This peptide was synthesised and found to elute at the same time as the natural peptide.

All the peptides, both the FaRPs and the SCPs, affected the beat amplitude and tone of the ventricle more than the beat rate. Log dose-response curves were linear and approximately parallel in the middle of the concentration range, so potency comparisons are based on the peptide concentrations required to give equivalent responses in this range. Exemplary matches of equivalent responses of pairs (or triplets) of the peptides on the same ventricle are shown (Figs 5 and 6), as are our quantitative estimates of potency (Fig. 7).

Equipotent doses of the tetrapeptide and heptapeptide FaRPs produce cardioexcitations that are indistinguishable (Fig. 5). Of the tetrapeptides, FMRFamide is about 10 times less potent than FLRFamide (Fig. 5A). The five heptapeptides were approximately equipotent (Fig. 5D-F), about 10 times more potent than FLRFamide (Fig. 5B), and about 100 times more potent than FMRFamide (Fig. 5C).

The heptapeptides and nonapeptides all had inhibitory effects on isolated Helix ventricles, particularly towards the high end of their concentration ranges. But these effects were idiosyncratic, occurring more frequently with some peptides, and on some ventricles, than others. Even when inhibiting the beat, however, the heptapeptide FaRPs were approximately equipotent (Fig. 5E).

The two SCPs of Helix, MNYLAFPRMamide and SGYLAFPRMamide, also gave very similar responses, though SGYLAFPRMamide was consistently about three times more potent (Fig. 6A). Since the doses of these two peptides had been quantified by HPLC, this difference, though small, is probably reliable. The highest doses of the SCPs were inhibitory (Fig. 6D).

The differences in the effects of the SCPs and FaRPs were examined by comparing the actions of SGYLAFPRMamide with those of FLRFamide or pQDPFLRFamide on several ventricles (Fig. 6B-D)’. A response to an SCP was usually longer lasting than that to an equipotent dose of a FaRP, giving a broader peak (Fig. 6B-D). Moreover, the positive inotropic actions of the SCPs were often accompanied by a decrease in diastolic tone (e.g. Fig. 6A), an effect never seen with the FaRPs. SGYLAFPRMamide was approximately eight times more potent than pQDPFLRFamide and the other heptapeptide FaRPs (Fig. 6B), and sixty times more effective than FLRFamide (Fig. 6C).

Fig. 7 summarises the differences in potency between pairs of peptides.

We have now identified nine endogenous peptides with excitatory effects on the Helix ventricle, and so some of the agents responsible for the cardioexcitatory activity of Helix ganglia are now more clearly delimited almost 30 years after their first detection. Although these nine peptides seem to account for the majority of the non-5-HT cardioexcitatory activity found in the ganglia, other minor stimulatory components remain to be discovered. The larger, and still unanswered, question is how the effects of these peptides are integrated to regulate cardiovascular function.

There is some evidence that the FMRFamide-related peptides can function as both cardioregulatory hormones and transmitters: immunoreactive FMRFamide is present in Helix blood (Price et al. 1985), and also in the heart itself (Lehman and Price, 1987). In contrast, no SCP bioactivity could be found in the blood or heart of either Helix (Lloyd, 1978) or Aplysia (Lloyd et al. 1985). Nevertheless, the SCPs are the most potent cardioexcitors of the nine, so they might still influence cardioregulation at levels not detectable by gel chromatography and bioassay.

The structures of the two SCPs we have isolated from Helix are compatible with an SCP precursor very similar to that found in Aplysia (Mahon et al. 1985). Since one SCP of Helix is identical to SCP_B of Aplysia, the other would seem to be the Helix counterpart of SCP_A. Indeed, the replacement of a single cytosine with a thymidine in the Aplysia SCP precursor gene would change the proline in the third position of SCP_A to a serine, and the resultant precursor would be processed to the novel SCP reported here.

Earlier work on the FMRFamide-related heptapeptides of the pulmonates indicated that each species has two different heptapeptides (Ebberink et al. 1987; Price et al. 1987a,b), and we expected Helix to have only two: pQDPFLRFamide and one other. In retrospect, our previous work on Helix had revealed a mixture of NDPFLRFamide and SDPFLRFamide (see amino acid composition in Price et al. VJKlb), but we could not interpret it as such, because serine, glycine and aspartic acid are common free amino acid contaminants. The use of FABms, as in this study, reduces the problem created by free amino acid contaminants.

Price et al. (1987b) hypothesised that two heptapeptides are encoded by a gene derived from the 5′end of the FMRFamide precursor. This speculation is still untested but it cannot be entirely accurate. Instead, either two heptapeptide genes occur in Helix, or a single heptapeptide precursor must contain at least five heptapeptides. However, our further speculation - that one or more heptapeptide precursors originated from the N-terminal region of the FMRFamide precursor - is strengthened by the discovery of two heptapeptides ending -YLRFamide; i.e. the N-terminal region of the Aplysia precursor does encode such a peptide.

The work reported here illustrates the utility of FABms in the identification of peptides. Though the three heptapeptides (N-, S- and G-DPFLRFamide) cannot be separated with acidic solvent systems on several C-18 columns, we can continue to use these otherwise desirable acidic systems if we identify the peptides that are present with FABms. In addition, the FABms identification of the molecular ion provides a more positive test of identity than does HPLC co-elution.

This work was supported by NIH grant HL28440 to MJG and DAP, the Committee of Principals and Vice Chancellors (Britain), and the University of St Andrews (Scotland) (WL), and by a grant from the Interdisciplinary Center for Biotechnology Research (ICBR) of the University of Florida. We thank the Protein Chemistry Core Facility of the University of Florida for peptide sequences and peptide syntheses, and we are grateful to M. L. Milstead for her assistance with the figures.