Journal of Chemical Ecology, Vol. 10, No. 6, 1984
CHEMICAL INDUCTION OF FEEDING IN CALIFORNIA SPINY LOBSTER, Panulirus interruptus (RANDALL): Responses to Molecular Weight Fractions of Abalone
RICHARD
K. Z I M M E R - F A U S T , W I L L I A M C. M I C H E L , E. T Y R E , a n d J A M E S F. C A S E
JEFFREY
Department of Biological Sciences and Marine Sciences Institute University of California, Santa Barbara, California 93106 (Received April 5, 1983; revised September 23, I983) Abstract--Molecular weight fractions of abalone muscle were tested for the ability to induce appetitive feeding and locomotor behavior in the spiny lobster, Panulirus interruptus. Fractions of 10,000 daltons were isolated by ultrafiltrations and gel chromatography from a seawater extract of abalone muscle. The two lower-molecularweight fractions ( 1000dalton fraction was also highly stimulatory, meaning that large and not small molecules were essential in initiating feeding. Finally, a 75% ethanolinsoluble component of the 1000 daltons) was actually tested in only one of these studies (Carr and Gurin, 1975). We have tested molecular weight fractions (10,000 daltons), prepared from abalone (Haliotis spp.) extract, for ability tO cause feeding in the California spiny lobster Panulirus interruptus (Randall). Abalone is available to Panulirus as carrion, and it is eaten alive by Panulirus in laboratory experiments (Carlberg, 1975). When available, at is the preferred bait used by southern California commercial lobster fisherman. Field experiments have already demonstrated that abalone muscle is attractive to lobsters, when presented in dialysis membranes releasing substances ~I0,000 daltons, but not when releasing substances 1000 daltons released from abalone muscle are attractive to lobsters (Zimmer-Faust and Case, 1982b). For the research described here, fractions were prepared from the standard FDAME. The 10,000-dalton molecular weight fractions along with combinations of these fractions were tested for stimulatory capacity. Fractions were prepared using a pressure ultrafiltration vessel (Amicon model 402) and membranes having molecular weight cutoffs of 1000 daltons (Amicon UM-2) and 10,000 daltons (Amicon UM-10). Ultrafiltrations were performed at 50 psi (N2) at 4~ Ultrafiltrates (10,000) were concentrated and rinsed with FSW to reduce the possibility of contamination by residual low-molecular-weight substances before dilution for tests, also at original concentrations. Recombinations (1000; 10,000) of ultrafiltrates together with their complementary ultraretentates were used to control for potential losses of attractivity caused by analytical procedures, Behavioral responses produced by recombined fractions were expected to equal those induced by whole extract (FDAME). A 1000- to 10,000-dalton fraction was prepared by first ultrafiltering FDAME using a UM-10 membrane, then collecting and ultrafittering the UM-10 ultrafiltrate with a UM-2 membrane. Ultraretentate from this second ultrafiltration was concentrated, rinsed with FSW, then diluted to its original concentration for assay. One liter of the ~10,000 dalton fraction was diluted to 4 liters with absolute ethanol, then held at 4~ for 24 hr. The ethanol-insoluble material was collected by centrifugation (16,300 g at 4 ~C for 30 rain), rinsed 2 times with 75% ethanol in FSW to remove residual soluble material, then resuspended to the original volume. The ethanol-soluble material was concentrated to dryness by rotary evaporation and resuspended to the
961
CHEMICAL INDUCTION OF FEEDING
original volume. The subfractions were stored frozen (-20~ in randomly assigned, numbered test tubes until testing. A schematic diagram of fractionation procedures is given in Figure 1. Chemical extracts and fractions were considered stimulatory when the proportions of animals responding differed significantly from the proportion responding to seawater controls (chi-square test: P ~ 0.05).
Preparation of Molecular Weight Fractions by Gel Chromatography. Molecular weight fractions were also prepared by gel chromatography to provide a comfirmatory assay of small molecules. Three hundred grams of Sephadex G-10 gel were hydrated in boiling nanopure water for 1 hr, cooled (20~ and packed in a glass column (60 X 5 cm) to a bed depth of 38 cm. Packing was protected by two Whatman GF/C filter papers and rinsed with 2 liters of nanopure water, prior to column use. A blue dextrin/fluorescein dye solution was used to visually calibrate elution volumes for high (>700 daltons) and low (~700 daltons) molecular weight fractions, respectively. When fractionating FDAME, 500 ml were carefully layered onto the column and washed into the resin by applying 25 ml of nanopure water. Nanopure water was then used to elute the fractions. High-molecular-weight substances (~700 daltons) emerged after 250 ml and were fully eluted at 750 ml. Lowmolecular-weight substances (~700 daltons) began to emerge after 800 ml and were fully eluted only after application of 4500 ml. Both fractions were evaporated to dryness and then resuspended to original concentrations before testing. A recombined solution (~700 + ~700) was prepared to control for analytical procedures. Amino Acid Analys&. A lithium citrate step-gradient system (Pierce) for examination of physiological fluids was used with a microbore cation exchange column (0.4 X 25 cm) to quantify the amino acids found in the FDAME and subfractions. Postcolumn derivatization with Fluropa (Dionex) yielded fluorescent products. Physiological fluid standards (Pierce) were used to identify the individual amino acids, and norleucine was used as an internal standard for quantification of the component amino acids. RESULTS
Description of Feeding and Locomotor Behavior and Selection of Behavioral Acts for Bioassay. To define aspects of behavior important in appetitive feeding and locomotion, we observed animals kept both in large tanks (1.3 X 0.74 X 0.20 m) and in smaller test aquaria as they responded to chemical extracts prepared from abalone muscle and mussels (Mytilus californianus). Responses to extracts were nearly identical to those occurring when whole food was placed near lobsters. No differences in behavior were observed associated with the two different types of tanks, although response latencies were often longer in the large tanks, as might be expected.
!
i:or,=0l
--7
JrecombiJne ultrafiltration 75O~xethano! ,,..,1,c (UreM-te!2ta,temembrane)l
1
ultrafimembrane) (UM-I Oltration F ' I I '( .r-,
>
~r
b~
b,9
CHEMICAL INDUCTION OF FEEDING
963
Inactive animals rested in shelters or in corners of aquaria with tails flexed and antennae contacting surfaces. Pereiopod movements were commonly associated with slight changes in body posture, while third maxillepeds frequently contacted other mouthparts. Occasionally, maxillepeds were extended, then flexed in a fanning motion, possibly associated with ventilation. Antennule flicking was relatively slow (~_I flick/sec) and occurred in bursts so that mean flick rates were not always indicative of shortterm variability. Following introduction of natural food or an effective chemical stimulus, antennule flicking increased but remained arrhythmic. Side-to-side sweeps of one or both antennules became frequent, and often lobsters pointed antennules in the direction of the prevailing current or towards a distant chemical source. Substrate probing movements with pereiopods were associated with increased antennular activity. Response sequences were frequently arrested at this point, particularly when only low concentrations (10-5-10 -7 g/liter FDAME) or small bits of food had been introduced. At higher stimulus intensities (e.g., > l0 -3 g/liter FDAME), lobsters continued to respond by rubbing opposing third maxillepeds together and by deflecting antennules downward into contact with third maxillepeds, leading to grasping and wiping of antennular flagellae. Wiping of one antennule typically alternated with wiping of the opposing antennule, in a series of successive acts. Induction of locomotion was commonly delayed for -->60 sec after introduction of an effective stimulus, and was initiated only when early pereiopod probings failed to contact food. For the present study, antennule flicking and wiping, leg probing, and locomotion were selected for assay. Both electrophysiological and behavioral investigations have shown that flicking is important to the detection of chemical stimuli (Snow, 1973; Pearson and Ola, 1977; Price and Ache, 1977; Pearson et al., 1979; Schmitt and Ache, 1979), while antennule wiping is believed to function either in cleaning and resetting antennule chemoreceptors (Snow, 1973) or in transferring stimulus molecules from antennules to mouthparts (Fuzessery and Childress, 1975). Probing and locomotion have obvious utility in food search. Feeding is defined here as the mutual occurrence of flicking, wiping, and probing, each within a 3-rain test period. These acts are important either in detecting or in searching for immediately nearby food. Because locomotion is also important in distant food search, it is treated independently of feeding. Definitions of these behavioral elements appear in Table 1. Appetitive Feeding Responses to Molecular Weight Fractions of FDAME. The stimulatory capacity of single fractions increased with molecular weight, but no single fraction evoked whole-feeding responses in a significant proportion of animals (Table 2). Feeding behavior approached
ZIMMER-FAusT ET AL.
964
TABLE 1. DEFINITIONSOF BEHAVIORALELEMENTS IN APPETITIVEFEEDINGAND LOCOMOTIONBY Panulirus Act Feeding Antennule flicking Leg probing Antennule wiping Locomotion
Definition Vertical deflection of a lateral antennular flagellum to a position nearly contacting the medial flagellum. A response was defined as >1.0 flick per second. Any nonlocomotor movement of a pereipod: either raking a dactyl across the substratum or elevating a dactyl to a position no longer in contact with the substratum. A downward and verticat deflection of an antennute, resulting in simultaneous contact of both antennular flagellae with the third maxiUepeds. A laterally or anteriorly directed movement of the body to a distance >89carapace length.
significance ( P = 0.06) in the presence of the >10,000 fraction, but activity associated with the < 1 0 0 0 fraction was nearly identical to that of seawater controls. This last result is i m p o r t a n t because it shows that low-molecularweight substances alone are ineffective feeding stimulants. It could be argued that the inability of single fractions to stimulate behavior resulted f r o m losses of attractivity associated with the fractionation procedure. This did not appear to be the case, since recombined (control) fractions wer e highly stimulatory (Table 2). It was also possible that the failure of the < 1 0 0 0 fraction to stimulate feeding was because tests were conducted exclusively at a single, high concentration. Low-molecular-weight substances, specifically amino acids, have been d e m o n s t r a t e d to be attractive when dilute, but nonattractive at high concentrations (Shelton and Mackie, t97 l; Carr, 1978). For this reason, we tested the < 1000 fraction at concentrations of 6.00 X 10-~ g/liter and 6.00 ;< i0 -7 g/liter. Only 22% ( N = 23) and i8% ( N = 23) o f tested animals exhibited feeding at these concentrations, respectively, indicating that the lack of attractivity was unrelated to stimulus intensity. We next tested c o m b i n a t i o n s o f two of the three single molecular weight fractions. Lobsters responded to these solutions equally as well as to the recombined (control) fractions, showing that c o m b i n e d attractants were essential to the p r o d u c t i o n of full behavioral activity (Table 2). Of particular interest was the ability of the > 1 0 0 0 fraction to evoke whole-feeding responses since this d e m o n s t r a t e d that molecules < 1 0 0 0 dattons are unnecessary for feeding to occur. Stimulants were f o u n d to v a r y widely in molecular weight, so that no single fraction (i.e., < 1 0 0 0 , t000-10,000, > t 0 , 0 0 0 ) was essential to induction of feeding.
CHEMICAL INDUCTIONOF FEEDING
965
TABLE 2. APPETITIVE FEEDING RESPONSESOF Panulirus TO F D A M E AND MOLECULAR WEIGHT FRACTIONSa
Test solution Whole extract (FDAME) Ultrafiltrations Single fractions < 1,000 1000-10,000 >10,000 Recombined fractions (control) 1,000 10,000 Combined fractions 1,000 10,000 Ethanolic extraction (< 10,000) Solute Precipitate Gel filtrations 700 700 Seawater (control)
Proportion feeding
Number tested
0.40 ***b
60
0.13 0.13 0.25
40 40 40
0.35** 0.33*
40 40
0.35** 0.38*** 0.40***
40 40 40
0.18 0.40"**
40 40
0.15 0.33* 0.37** 0.09
24 24 24 45
~ was defined as the occurrence of probing, wiping, and increased flicking, each within a 3-min trial period. hThe difference is significant between proportions of animals responding to test versus control (seawater) solutions (chi-square test: * P < 0.05, * * P < 0.01, * * * P < 0.001).
An additional test using products from the ethanolic extraction of the 700 daltons) and recombined (700) fractions were effective in stimulating whole-feeding responses. The
CHEMICAL INDUCTION OF FEEDING IN CALIFORNIA SPINY LOBSTER, Panulirus interruptus (RANDALL): Responses to Molecular Weight Fractions of Abalone
RICHARD
K. Z I M M E R - F A U S T , W I L L I A M C. M I C H E L , E. T Y R E , a n d J A M E S F. C A S E
JEFFREY
Department of Biological Sciences and Marine Sciences Institute University of California, Santa Barbara, California 93106 (Received April 5, 1983; revised September 23, I983) Abstract--Molecular weight fractions of abalone muscle were tested for the ability to induce appetitive feeding and locomotor behavior in the spiny lobster, Panulirus interruptus. Fractions of 10,000 daltons were isolated by ultrafiltrations and gel chromatography from a seawater extract of abalone muscle. The two lower-molecularweight fractions ( 1000dalton fraction was also highly stimulatory, meaning that large and not small molecules were essential in initiating feeding. Finally, a 75% ethanolinsoluble component of the 1000 daltons) was actually tested in only one of these studies (Carr and Gurin, 1975). We have tested molecular weight fractions (10,000 daltons), prepared from abalone (Haliotis spp.) extract, for ability tO cause feeding in the California spiny lobster Panulirus interruptus (Randall). Abalone is available to Panulirus as carrion, and it is eaten alive by Panulirus in laboratory experiments (Carlberg, 1975). When available, at is the preferred bait used by southern California commercial lobster fisherman. Field experiments have already demonstrated that abalone muscle is attractive to lobsters, when presented in dialysis membranes releasing substances ~I0,000 daltons, but not when releasing substances 1000 daltons released from abalone muscle are attractive to lobsters (Zimmer-Faust and Case, 1982b). For the research described here, fractions were prepared from the standard FDAME. The 10,000-dalton molecular weight fractions along with combinations of these fractions were tested for stimulatory capacity. Fractions were prepared using a pressure ultrafiltration vessel (Amicon model 402) and membranes having molecular weight cutoffs of 1000 daltons (Amicon UM-2) and 10,000 daltons (Amicon UM-10). Ultrafiltrations were performed at 50 psi (N2) at 4~ Ultrafiltrates (10,000) were concentrated and rinsed with FSW to reduce the possibility of contamination by residual low-molecular-weight substances before dilution for tests, also at original concentrations. Recombinations (1000; 10,000) of ultrafiltrates together with their complementary ultraretentates were used to control for potential losses of attractivity caused by analytical procedures, Behavioral responses produced by recombined fractions were expected to equal those induced by whole extract (FDAME). A 1000- to 10,000-dalton fraction was prepared by first ultrafiltering FDAME using a UM-10 membrane, then collecting and ultrafittering the UM-10 ultrafiltrate with a UM-2 membrane. Ultraretentate from this second ultrafiltration was concentrated, rinsed with FSW, then diluted to its original concentration for assay. One liter of the ~10,000 dalton fraction was diluted to 4 liters with absolute ethanol, then held at 4~ for 24 hr. The ethanol-insoluble material was collected by centrifugation (16,300 g at 4 ~C for 30 rain), rinsed 2 times with 75% ethanol in FSW to remove residual soluble material, then resuspended to the original volume. The ethanol-soluble material was concentrated to dryness by rotary evaporation and resuspended to the
961
CHEMICAL INDUCTION OF FEEDING
original volume. The subfractions were stored frozen (-20~ in randomly assigned, numbered test tubes until testing. A schematic diagram of fractionation procedures is given in Figure 1. Chemical extracts and fractions were considered stimulatory when the proportions of animals responding differed significantly from the proportion responding to seawater controls (chi-square test: P ~ 0.05).
Preparation of Molecular Weight Fractions by Gel Chromatography. Molecular weight fractions were also prepared by gel chromatography to provide a comfirmatory assay of small molecules. Three hundred grams of Sephadex G-10 gel were hydrated in boiling nanopure water for 1 hr, cooled (20~ and packed in a glass column (60 X 5 cm) to a bed depth of 38 cm. Packing was protected by two Whatman GF/C filter papers and rinsed with 2 liters of nanopure water, prior to column use. A blue dextrin/fluorescein dye solution was used to visually calibrate elution volumes for high (>700 daltons) and low (~700 daltons) molecular weight fractions, respectively. When fractionating FDAME, 500 ml were carefully layered onto the column and washed into the resin by applying 25 ml of nanopure water. Nanopure water was then used to elute the fractions. High-molecular-weight substances (~700 daltons) emerged after 250 ml and were fully eluted at 750 ml. Lowmolecular-weight substances (~700 daltons) began to emerge after 800 ml and were fully eluted only after application of 4500 ml. Both fractions were evaporated to dryness and then resuspended to original concentrations before testing. A recombined solution (~700 + ~700) was prepared to control for analytical procedures. Amino Acid Analys&. A lithium citrate step-gradient system (Pierce) for examination of physiological fluids was used with a microbore cation exchange column (0.4 X 25 cm) to quantify the amino acids found in the FDAME and subfractions. Postcolumn derivatization with Fluropa (Dionex) yielded fluorescent products. Physiological fluid standards (Pierce) were used to identify the individual amino acids, and norleucine was used as an internal standard for quantification of the component amino acids. RESULTS
Description of Feeding and Locomotor Behavior and Selection of Behavioral Acts for Bioassay. To define aspects of behavior important in appetitive feeding and locomotion, we observed animals kept both in large tanks (1.3 X 0.74 X 0.20 m) and in smaller test aquaria as they responded to chemical extracts prepared from abalone muscle and mussels (Mytilus californianus). Responses to extracts were nearly identical to those occurring when whole food was placed near lobsters. No differences in behavior were observed associated with the two different types of tanks, although response latencies were often longer in the large tanks, as might be expected.
!
i:or,=0l
--7
JrecombiJne ultrafiltration 75O~xethano! ,,..,1,c (UreM-te!2ta,temembrane)l
1
ultrafimembrane) (UM-I Oltration F ' I I '( .r-,
>
~r
b~
b,9
CHEMICAL INDUCTION OF FEEDING
963
Inactive animals rested in shelters or in corners of aquaria with tails flexed and antennae contacting surfaces. Pereiopod movements were commonly associated with slight changes in body posture, while third maxillepeds frequently contacted other mouthparts. Occasionally, maxillepeds were extended, then flexed in a fanning motion, possibly associated with ventilation. Antennule flicking was relatively slow (~_I flick/sec) and occurred in bursts so that mean flick rates were not always indicative of shortterm variability. Following introduction of natural food or an effective chemical stimulus, antennule flicking increased but remained arrhythmic. Side-to-side sweeps of one or both antennules became frequent, and often lobsters pointed antennules in the direction of the prevailing current or towards a distant chemical source. Substrate probing movements with pereiopods were associated with increased antennular activity. Response sequences were frequently arrested at this point, particularly when only low concentrations (10-5-10 -7 g/liter FDAME) or small bits of food had been introduced. At higher stimulus intensities (e.g., > l0 -3 g/liter FDAME), lobsters continued to respond by rubbing opposing third maxillepeds together and by deflecting antennules downward into contact with third maxillepeds, leading to grasping and wiping of antennular flagellae. Wiping of one antennule typically alternated with wiping of the opposing antennule, in a series of successive acts. Induction of locomotion was commonly delayed for -->60 sec after introduction of an effective stimulus, and was initiated only when early pereiopod probings failed to contact food. For the present study, antennule flicking and wiping, leg probing, and locomotion were selected for assay. Both electrophysiological and behavioral investigations have shown that flicking is important to the detection of chemical stimuli (Snow, 1973; Pearson and Ola, 1977; Price and Ache, 1977; Pearson et al., 1979; Schmitt and Ache, 1979), while antennule wiping is believed to function either in cleaning and resetting antennule chemoreceptors (Snow, 1973) or in transferring stimulus molecules from antennules to mouthparts (Fuzessery and Childress, 1975). Probing and locomotion have obvious utility in food search. Feeding is defined here as the mutual occurrence of flicking, wiping, and probing, each within a 3-rain test period. These acts are important either in detecting or in searching for immediately nearby food. Because locomotion is also important in distant food search, it is treated independently of feeding. Definitions of these behavioral elements appear in Table 1. Appetitive Feeding Responses to Molecular Weight Fractions of FDAME. The stimulatory capacity of single fractions increased with molecular weight, but no single fraction evoked whole-feeding responses in a significant proportion of animals (Table 2). Feeding behavior approached
ZIMMER-FAusT ET AL.
964
TABLE 1. DEFINITIONSOF BEHAVIORALELEMENTS IN APPETITIVEFEEDINGAND LOCOMOTIONBY Panulirus Act Feeding Antennule flicking Leg probing Antennule wiping Locomotion
Definition Vertical deflection of a lateral antennular flagellum to a position nearly contacting the medial flagellum. A response was defined as >1.0 flick per second. Any nonlocomotor movement of a pereipod: either raking a dactyl across the substratum or elevating a dactyl to a position no longer in contact with the substratum. A downward and verticat deflection of an antennute, resulting in simultaneous contact of both antennular flagellae with the third maxiUepeds. A laterally or anteriorly directed movement of the body to a distance >89carapace length.
significance ( P = 0.06) in the presence of the >10,000 fraction, but activity associated with the < 1 0 0 0 fraction was nearly identical to that of seawater controls. This last result is i m p o r t a n t because it shows that low-molecularweight substances alone are ineffective feeding stimulants. It could be argued that the inability of single fractions to stimulate behavior resulted f r o m losses of attractivity associated with the fractionation procedure. This did not appear to be the case, since recombined (control) fractions wer e highly stimulatory (Table 2). It was also possible that the failure of the < 1 0 0 0 fraction to stimulate feeding was because tests were conducted exclusively at a single, high concentration. Low-molecular-weight substances, specifically amino acids, have been d e m o n s t r a t e d to be attractive when dilute, but nonattractive at high concentrations (Shelton and Mackie, t97 l; Carr, 1978). For this reason, we tested the < 1000 fraction at concentrations of 6.00 X 10-~ g/liter and 6.00 ;< i0 -7 g/liter. Only 22% ( N = 23) and i8% ( N = 23) o f tested animals exhibited feeding at these concentrations, respectively, indicating that the lack of attractivity was unrelated to stimulus intensity. We next tested c o m b i n a t i o n s o f two of the three single molecular weight fractions. Lobsters responded to these solutions equally as well as to the recombined (control) fractions, showing that c o m b i n e d attractants were essential to the p r o d u c t i o n of full behavioral activity (Table 2). Of particular interest was the ability of the > 1 0 0 0 fraction to evoke whole-feeding responses since this d e m o n s t r a t e d that molecules < 1 0 0 0 dattons are unnecessary for feeding to occur. Stimulants were f o u n d to v a r y widely in molecular weight, so that no single fraction (i.e., < 1 0 0 0 , t000-10,000, > t 0 , 0 0 0 ) was essential to induction of feeding.
CHEMICAL INDUCTIONOF FEEDING
965
TABLE 2. APPETITIVE FEEDING RESPONSESOF Panulirus TO F D A M E AND MOLECULAR WEIGHT FRACTIONSa
Test solution Whole extract (FDAME) Ultrafiltrations Single fractions < 1,000 1000-10,000 >10,000 Recombined fractions (control) 1,000 10,000 Combined fractions 1,000 10,000 Ethanolic extraction (< 10,000) Solute Precipitate Gel filtrations 700 700 Seawater (control)
Proportion feeding
Number tested
0.40 ***b
60
0.13 0.13 0.25
40 40 40
0.35** 0.33*
40 40
0.35** 0.38*** 0.40***
40 40 40
0.18 0.40"**
40 40
0.15 0.33* 0.37** 0.09
24 24 24 45
~ was defined as the occurrence of probing, wiping, and increased flicking, each within a 3-min trial period. hThe difference is significant between proportions of animals responding to test versus control (seawater) solutions (chi-square test: * P < 0.05, * * P < 0.01, * * * P < 0.001).
An additional test using products from the ethanolic extraction of the 700 daltons) and recombined (700) fractions were effective in stimulating whole-feeding responses. The
