HPLC ANALYSIS OF SQUID INK: A QUANTIFICATION OF L-DOPA AND DOPAMINE IN THE INK OF Loligo opalescens Heraldo M. Farrington Hopkins Marine Station Stanford University Pacific Grove, CA 93950 1992 Spring Class (175H) ABSTRACT Potential chemical cues in squid ink were identified and quantified using high-performance, reverse-phase, liquid chromatography (HPLC), Primary amines in squid ink were conjugated with the flourescent com¬ pound ortho-phthaldialdehyde (OPA). Upon separation by the column. I-dopa and dopamine were detected by an end-column flourometer. Crude ink was manually extracted from the dissected ink sacs of freshly-killed Loligo opalescens collected from Monterey Bay, California. Protocols for preventing the oxidation of the active compounds were developed, thereby allowing for the storage of ink for later analysis. Fresh ink was never analyzed. Standard curves were established for both I-dopa (r2 = ,967: CV of slope= 8.2%) and dopamine (r2 = .939; CV of slope = 9.0%). These standard curves were used to transform the integrated peak areas from the HPLC analysis into absolute quantities. Ten ink sacs, collected from 8 adult males and 2 adult females, were used to create four discrete sam¬ ples for analysis. For l-dopa we report a mean concentration of L39 (+- 0.43) nanomoles per milligram of extracted ink. For dopamine we re¬ port a mean concentration of C698 (+- 0-02) nanomoles per milligram of o. 0.70 extracted ink. INTRODUCTION Inking by cephalopods has long been recognized as an adaptive response to predation and physical threat. The ink clouds produced by these animals allow for retreat from a threatening situation and leave behind either a diffuse "smoke screen" or a dense, long-lasting, decoy that serves to confuse predators. Additionally, ink clouds may alert con¬ specitics to danger, especially within the dense schools of certain cepha¬ lopods such as squid. However, the visual nature of such communication would limit its usefulness under conditions of little or no illumination. Why then, do stressed squid eject ink at night and also at great depths? (Fox, 1974) One answer may be that cephalopod ink contains chemical compounds which provide defensive "counter-attack" capabilities against predators. MacGinitie and MacGinitie (1964) have suggested that octopus ink can paralyze the olfactory senses of a formidable predator, the moray eel. Additionally, squid ink has been shown to depress the appetites of certain reef fish (DOOLITTLE, 19??). These studies indicate that, in addition to melanin pigments, cephalopod ink probably contains chemicals which pos¬ sess neuroactive capabilities. However, none of these putative chemical signals has been identified. If such neuroactive compounds are present in ink, they need not ne¬ cessarily act upon predators only. Gilly and Lucero (1992) have shown that crude squid ink stimulates high-pressure jet escape responses in the squid Loligo opalescens, and that I-dopa, a likely component of the ink, can mimic this action. They suggest that "although the color of ink could serve as a visual alarm signal under conditions of sufficient illumination, chemical messengers (in the ink) ... would be more effective in the darkness of night or at great depths." Several studies have concentrated on the melanogenesis pathway within the cells of the ink sac (Giroud, 1882; Szabo et. al. 1962, 1963). However, surprisingly little is known regarding the presence of potential chemical messengers within cephalopod ink itself. Prota et al. (1981) found evidence for the presence of the enzyme tyrosinase in the ejected ink of various cephalopods, including Loligo opalescens. They speculated that this enzyme, which mediates the conversion of tyrosine to melanin. is responsible for the olfactory chemosensory properties of the ejected ink. Many such potential chemical messengers might be present within cephalopod ink. One plausible mechanism for their presence involves necritic melanocytes. These cells have been observed to be shed directly into the lumen of the ink sac (Ortonne et al., 1981). It is entirely possible that these old, dead cells then break apart, and thereby introduce a wide variety of amino acids (as well as the enzymes of cellular machinery) into the ink. In light of these findings, a chemical analysis devoted to identifying and quantifying potential messenger compounds becomes logically neces- sary. Such an analysis would be pertinent to a wide range of investiga¬ tions. These might include: 1) Behavioral studies on cephalopod communication, 2) Ecological studies on predator-prey interactions. 3) Neurological studies on chemoreception and motor pathways. There are two principal justifications for our selection of primary amines as the class of compounds for which to search. The first is based on the cytology of the cephalopod ink sac. If one accepts the Raper-Mason pathway of melanogenesis (Figure 1), at least up to the formation of dopa chrome (Prota, 1988), one might expect that necritic melanocytes can introduce the primary amines tyrosine, I-dopa, and dopaquinone into the ink of cephalopods. Additionally, albeit through a completely separate pathway, I-dopa can be converted into dopamine and thence into norepi¬ nephrine. Thus, one might expect these primary amines to be present as well. The second justification is based upon behavioral studies which demonstrated that millimolar concentrations of I-dopa mimicked high- pressure, jet escape responses in living squid (Gilly and Lucero, 1992). We thus decided to perform a chemical analysis on squid ink from Loligo opalescens, using HPLC fractionation for primary amines, in an attempt to quantify I-dopa, dopamine, and whichever amino acids might consistently be present. Behavioral investigations could then be conduc¬ ted using these biologically-relevant concentrations of potential chemical messengers. MATERIALS AND METHODS HPLC APPARATUS AND MATERIALS Chromatography was performed with two Waters pumps connected to a Waters Model 720 System Controller programmed to generate a dual¬ solvent, stepped-elution profile. Peaks were detected by a Flouro-Tec digital filter flourometer with an excitation wavelength of 320 nm and an emission wavelength of 450 nm. Peak areas were integrated by a Waters Data Module. Separation of OPA-derivatized primary amines was perfor- med on an Utrasphere 3 um Octadecasilane (ÖDS) column of 75 mm length and 4.5 mm internal diameter. A combination of aqueous and non-aqueous solvents was used to create an elution profile with a total run time of 23 minutes. The aqueous solution (solvent A) was 150 mM sodium acetate in 90% water, 9.5% methanol, and 0.5% tetrahydrofuran (THF). Prior to the addition of methanol and THF, the pH was brought down to 7.20. The non-aqueous solution (solvent B) was 100% methanol. All solvent materials were either HPLC-grade or were passed through a 0.2 um Millipore filter prior to use. A modified stepped-elution profile was developed (Table 1) that allowed for the clear separation and signal-integration of I-dopa and dopamine at a flow rate of 1.8 ml/min. The time course for solvent B was as follows: constant 10% from 0 to 1 minute, linear increase to 20% from 1 to 4 minutes, constant 20% from 4 to 9 minutes, linear increase to 50% from 9 to 13.5 minutes, constant 50% from 13.5 to 19 minutes, linear increase to 90% from 19 to 21 minutes, linear decrease to 10% from 21 to 23 minutes. At 23 minutes, the initial conditions were re-attained and the column was ready for the next injection. In order to protect the column from particulate pollution, all samples were passed through a 0.2 um Millipore filter prior to injection. Derivi¬ tization was carried out by adding 25% OPA by volume to the injection sample (for instance: 250 ul OPA to 1000 ul of sample). We note in pas¬ sing that between uses, OPA must be kept sterile, refrigerated, and out of the light. The derivatized samples were then vortexed briefly and spun in a centrifuge for 1 minute. Volumes of supernatant, ranging from 10 ul to a maximum of 1000 ul, were injected for analysis within 3 minutes of derivitization. Criteria for the acceptance of signal integration included both peak shape and peak size, and followed those spelled out by the Walker Data Module Manual. STANDARD CURVES L-dopa and dopamine were dissolved in a solution of 150 mM sodium bicarbonate and 10 mM ascorbic acid. Prior to the addition of ascorbic acid, the solution was slowly brought to a pH of 10.4. Ascorbic acid is an anti-oxidant, and was absolutely essential for inhibiting the oxidation of both I-dopa and dopamine. Sodium bicarbonate buffered the solution so that the final pH was 10.0. At this pH both I-dopa and dopamine readily dissolved. Initial concentrations for both compounds were 1mM. These were each diluted once to produce final working concentrations of 50 uM for I-dopa and 100 uM for dopamine. These concentrations yielded sharp signal peaks which could be well-integrated by the Walker Data Module. In order to prevent oxidation during preparation, all solutions were bubbled with N2 gas (purity = .99995) before and after each dilution. Samples awaiting injection were also bubbled, then stored at 0° C. All samples for standard curves were injected within six hours of prepara¬ tion. Chemicals used were of the highest purity available (Sigma ACS Rea¬ gent Grade or better) except for sodium bicarbonate, which was of Cell Culture Grade and which contained an unidentified amino acid (retention time = 5.9) Our standard curves were established over the range of injections that produced the best-integrated peaks. The I-dopa curve (Figure 2) is based upon 7 samples from one 50 uM concentration (50, 60, 70, 80, 90. 100, and 110 ul injections). The dopamine curve (Figure 3) is based upon 10 samples: 5 samples from each of two separate 100 uM concentrations (60, 70, 80, 90, and 100 ul injections). All injections within each set of samples were analyzed consecutively. Each curve was estimated by linear regression. The standard curves established the sensitivity of detection and inte gration for each compound. They were used to transform the integrated peak areas, which resulted from each analysis of ink, into absolute nano- molar quantities. For each individual ink sample, various injection volumes yield various peak areas. These peak areas are then converted into molar amounts with the aid of the appropriate standard curve. Ideal¬ ly, the molar amounts should converge upon one value. This value repre¬ sents our measurement of the amount of compound per mass of ink in that particular sample. BIOLOGICAL MATERIALS Live specimens of Loligo opalescens were collected in Monterey Bay, California. Squid were decapitated, ink sacs were removed by dissection and whole ink was manually recovered. While it was important to maximize the amount of ink collected from each sac, it was equally important to minimize pollution of the ink by the cellular products from damaged melanocytes. A procedure for extracting the ink without homogenizing the ink sac had to be developed. Various methods involving hypodermic needles were attempted, but these were less than satisfactory due to the viscosity of the ink. The solution, we found, was to use sterile micro-capillary pipettes. INK COLLECTION TECHNIQUE A dissected ink sac was placed on a sheet of parafilm and a small cut was made through the wall of the sac near the ink duct. A glass rod (fa¬ bricated by melting over the small end of a Pasteur pipette) was then used to slowly force the ink out of the sac and onto the parafilm. One end of a micro-capillary tube was introduced into the pool of ink and held there until it was about half-full. More tubes were used until all the ink was collected. Nitrogen gas was then used to gently blow the ink out of the tubes and into a pre-weighed Eppendorf vial which contained 1 ml of 150 mM sodium acetate and 10 mM ascorbic acid. This procedure allowed us to quickly collect and weigh most of the ink within an ink sac, and simulta¬ neously dissolve it in a solution containing an anti-oxidant, thereby mini¬ mizing the risk of oxidation. It also avoided overt damage to the ink sac, and thus minimized the risk of pollution. STORAGE AND ANALYSIS OF INK Due to equipment difficulties and a scarcity of live squid, we were forced to store the collected ink for later analysis. In order to prevent oxidation, vials of collected ink were placed within a 50 ml centrifuge tube. The top was sealed with parafilm, and an hypodermic needle was used to inject nitrogen gas. A second hole in the parafilm allowed all of the air originally within the tube to escape. The lid was then screwed down over the parafilm and the entire collection was stored at -70° C. Prior to analysis, each vial was quickly thawed and vortexed. The vial was then centrifuged for 10 minutes (in a cold room) so as to produce a dense pellet of ink granules. Of the available 1.0 ml, 900 ul of clear supernatant were removed and centrifuged in 0.2 um filter tubes for about 45 seconds. The filtered ink extract was then prepared for injection. Typically, a 100 ul sample was conjugated with 25 ul OPA for a 100 ul injection. The shape and size of the resultant peaks guided the selec- tion of the next injection volume. Volumes were chosen so as to obtain well-integrated peaks. Table 2 displays these volume selections for all samples analyzed. Whenever possible, replicates were injected at one particular volume in order to establish a measure of our "injection error“. SAMPLING TECHNIQUE Due to early difficulties in extracting the ink from each sac, we were forced to pool individual ink sacs in order to obtain a measurable amount of ink. (Later, as we developed the micro-capillary technique, we were able to collect virtually all the ink within a sac without overtly damaging the sac itself.) Thus, each sample included ink collected from 2 or 3 individuals of the same sex. Four separate samples were analyzed. As can be seen from Table 2, the smallest mass of ink was 13.30 mg (collected from 2 males) while the largest mass of ink was 32.91 mg (collected from 3 males). In theory, this design would allow for a measurement of variability between sexes. However, our small sample size precludes any such measurement. In fact, we were unable to measure the natural variability between individuals, since our reported means were themselves derived from "averaged ink". RESULTS QUANTITATIVE ANALYSIS Table 3 lists the mean concentrations of I-dopa and dopamine present ).40 in squid ink. For I-dopa we report a mean concentration of 129 (+/- 0.43) nanomoles per milligram of extracted ink. For dopamine we report a mean O.1 concentration of 0.70 (+- 4 nanomoles per milligrams of extracted A e ink. We conclude that -dopa is present at a significantly higher concen¬ 0.05 tration than dopamine (2-05 » P » 0.92). In the last column of this same table, we report a coefficient of vari¬ ation for the slope of each standard curve: 8.2% for I-dopa and 9.0% for do¬ pamine. These values are relatively high for standard curves, and reflect the low r2 for each regression: .967 for I-dopa and .939 for dopamine, (Figures 2 and 3). The means were calculated from 4 separate samples. Table 2 displays these sample values. For dopamine, only one analysis per sample was done. (Excepting Vial C, in which both concentrations are virtually identi¬ cal.) For I-dopa, however, numerous analyses were performed on each sample. Thus, the mean concentration reported for I-dopa is actually a mean of means. QUALITATIVE ANALYSIS In the course of quantifying l-dopa and dopamine, we were able to identity a tew other primary amines which also are present in squid ink, However, we were unable to quantify these compounds. In Table 4 they are listed along with their approximate retention times. In addition to argi¬ nine, tyrosine, ammonia/ammonium salts, and phenylalanine, two signifi¬ cant peaks which remain unidentified were present often enough to merit fürther investigation. It is likely that these peaks represent aspartate and glutamate. DISCUSSION Numerous reports have speculated on the presence of neuroactive compounds within the ink of cephalopods. To our knowledge, this study marks the first attempt at identifying, let alone quantifying, any of these potential chemical messengers. While it is evident that HPLC analysis, as a technique, shows great promise for this type of investigation, the confi¬ dence with which we report our results is less than satisfactory. We feel that with some important modifications, this study should be continued and expanded. Ihree areas within our investigation require more work. The first of these is the generation of standard curves. The low r2 for the linear re¬ gressions are due, in part, to the peculiar peak shapes generated by dopa mine, and to a smaller extent, by I-dopa (Figure 5). The large, asymmetric "shoulders" tend to confuse the integrator since any reduction in slope is often interpreted as a return to baseline. Thus, the peak signal integra¬ tions, for dopamine especially, are highly variable, even for identical in¬ jection volumes. We are confident that this problem is resolvable. By programming the Walker Data Module to halt integration of the do¬ pamine peak at a particular point within the latter part of its "shoulder", the variability of integration from standard to standard should be reduced, This, in turn, will allow for a linear regression of better fit. While the slope of the line may not change, the standard deviation of the slope wil be reduced. Thus, our mean concentrations may not be altered directly. but the confidence with which we report them will be greatly increased, Referring to Table 3, lower coefficients of variation for the standard curves will allow us to make more meaningful conclusions regarding any variation within the mean concentrations themselves. The high variability of these means points to the second area of our investigation which requires more work: our sampling technique. Due to the rapid oxidation of both I-dopa and dopamine, it is imperative that we üse fresh, rather than frozen, ink. By so doing, we would eliminate most of the uncertainty associated with the standard deviations of these means. We would then be free to ascribe such variance to natural varia¬ tion among individuals, rather than to the effects of differential oxida¬ tion. In addition, by using the micro-capillary technique to collect enough ink from each squid, we could analyze individuals separately, and thus ex¬ tend our investigation to an analysis between individuals, The third area requiring more work is the quantification of other po¬ tential chemical messengers. This promises to be relatively easy, since these compounds are stable amino acids. They conjugate well with the flourescent label OPA and they do not break down into other compounds. Thus, the integration of their peaks is highly reproducible, and the genera¬ tion of standard curves with low errors of slope is virtually quaranteed, Finally, we feel that this investigation should expand slightly in order to make full use of HPLC analytic capabilities. A quantification of I-dopa within ink extracted manually from ink sacs does little more than to esta blish an upper limit to the potential concentrations of this compound in the ejected ink and in the water column. We propose to measure the rate of oxidative decay of both I-dopa and dopamine within various volumes of natural seawater and/or within vari¬ ous time periods. Such an analysis would yield highly relevant, biological concentrations with which one could more accurately measure the neuro¬ active potential of these compounds. In conclusion, the results of this study are encouraging. Both I-dopa and dopamine are present in frozen squid ink in measurable amounts, and techniques were developed which promise to assist greatly the more ac¬ curate quantification of these and other potential chemical messengers within the ink of Loligo opalescens. ACKNOWLEDGEMENTS The authors would like to thank Dr. Richard Zimmerman for both his generosity in allowing us the use of the Alberte Lab HPLC equipment, as well as for his invaluable assistance providing instruction in its use. We would also like to thank Taylor Liu for obtaining live squid, and we are ap preciative of Jimmy Lucero's efforts both in maintaining them and in di¬ recting their dissection. Finally, we are grateful to the fishing commu¬ nity of Monterey Bay for providing us access to freshly-caught squid. REFERENCES Doolittle. (19) Unknown. Fox D. L. (1974) Biochemical and Biophysical Perspectives in Marine Bio¬ logy (edited by Malins D. C. and Sargent J. R.). 1, 170-209. Academic Press, New York. Gilly W. F., and Lucero M. T. (1992) Behaviorial Responses to Chemical Sti¬ mulation of the Olfactory Organ in the Squid Loligo opalescens. Jour- nal of Experimental Biology. 162, 202-229. Giroud P. (1882) Recherches sur la poche du noir des Cephalopodes des cotes de France. Arch. Zool. Exp. Gen. 10, 1-100. MacGinitie G. E. and MacGinitie N. (1968) Natural History of Marine Ani¬ mals, 2nd ed., p. 523. McGraw-Hill, New York. Ortonne J. P Voulot C., Khatchadourian C., Palumbo A., and Prota G. (1981) Phenotypic Expression in Pigment Cells. Pigment Cell 1981 (edited by Makoto Seiji), pp. 49-57. University of Tokyo Press. Prota G., Ortonne J. P., Voulot C., Katchadourian C., Nardi G., and Palumbo A. (1981) Occurence and Properties of Tyrosinase in the Ejected Ink of Cephalopods. Comp. Bioch. Phys. 68B, 415-419. Prota G. (1988) Some New Aspects of Eumelanin Chemistry. Advances in Pigment Cell Research, pp. 101-124. Szabo G. and Sims T. (1962) Biol. Bull. Wood's Hole. 123, 153. Szabo G. and Arnold J. M. (1963) Biol. Bull. Wood's Hole. 125, 393-394 TYROSINASE - Leoon oon oon NO. 0. Nn. 1 s1om SNN. Sn. HO HO accelerates b DOPAQUINONE DOPA TYROSINE amall amounte ot sep tast co. Ho. o HO. -COOH -COOH - 10 1 HO No LEUCODOPACHROME DOPACHROME 8.8-DIHYDROXYINDOLE Amar208 and 478 um (REDI o MEL ANOCHROME ME LANINIS) Amax 200 and 840 nm slow To (PURPLEI A INDOLE-S.S-QUINONE FIGURE 1 Raper-Mason pathway for melanogenesis 30 28 26 24 3 22 20- 18 16 14 12 10. L-DOPA 2.5 3.5 5 5.5 6 AMOUNT (NANOMOLES) R-Squared = .967; Slope = 4067 +/- 401(CV = 8%). FIGURE 2 o 24: 22 20 18 16 14- 12: 10 8 6- DOPAMINE 10 AMOUNT (NANOMOLES) R-Squared = .939; Slope = 3407 +/- 307 (CV = 9%) FIGURE 3 11 L I —— S — 13TNI TIME (MIN FLOW (mIMIN)%A%B CURVE — INITIAL 1.8 90 80 06 1.0 1.8 11 4.0 80 1.8 50 9.0 1.8 06 50 13.5 50 1.8 19.0 06 1.8 85 21.0 1.8 90 10 06 TABLE 1: Elution Profile. Curve values refer to fixed response curves by which the Walker Pump Controller shapes each time step in the profile. A value of 06 indicates a linear response; a value of 11 indicates an instantaneous response. SAMPLE: VIALF VIALG VIAL D VIALC 2 F NUM. AND SEX: 3 M 2 M 3M 21.61 TOTAL INK (mg): 13.30 13.71 32.91 FRACTION ANALYZED L-DOPA (Nanomoles/mg Ink) 1.77 0.033 1901.S0 0.050 1.78 0.86 1.94 0.88 1.08 0.97 0.100 1.05 1.76 1.19 1.09 (164 (0.95) (MEANS) (1.11) (1.86) FRACTION ANALYZED DOPAMINE (Nanomoles/mg ink) 0.73 0.020 0.74 0.400 0.55 0.88 0.650 0.63 TABLE 2. Determinations of I-dopa and dopamine concentrations in squid ink by HPLC analysis. Ästerisks occur where analysit either resulted in a non-integratable peak or simply was not carried out. CROND MEANCONC. S.D. COEFF. VAR. COEFEVAR. IMEAN) ISL CUBVE) (4/-) .430 32.8% 8.2% L-DOPA 1.39 0.70 9.0% DOPAMINE (+/-) .020 3.4% TABLE 3: Mean concentrations of I-dopa and dopamine in squid ink (nanomoles per mg of ink). L-dopa is present at a significantly higher concentration than dopamine: tg = 3-2058' (95» P » 04). 30 105 n = 4. e ne or suw level, RETENTION TIME PRIMARY AMINE (Minute 1.50 Aspartate (? 2.50 Glutamate (? 6.08 Arginine 9.00 Tyrosine AmmonialAmmonium 10.20/11.30 Phenylalanine 11.85 Table 4: Qualitative analysis of other primary amines present in squid ink. None of these were quantified.