- Email: [email protected]
Serotonin depletion inhibits feeding in carnivorous leeches (Haemopis)
BEHAVIORAL AND NEURAL BIOLOGY 61, 47--53 (1994)
Serotonin Depletion Inhibits Feeding in Carnivorous Leeches (Haemopis) VALERIE GOLDBURT, B E H I R A . SABBAN, AND A N N A L . K L E I N H A U S 1
Department of Cell Biology and Anatomy, New York Medical College, Valhalla, New York 10595
lossal Retzius cells contain the largest concentrations of serotonin. In addition, there are 5-HT-containing cells found only in the brain, such as the large lateral (LL) cells (Lent, Ono, Keyzer, & K a r t e n 1979). The importance of 5-HT in controlling feeding behavior in jawed, sanguivorous leeches is well documented (Lent, 1985). Injection of 5,7-DHT (a cytotoxic analog of 5-HT) directly into the blood sinus surrounding the nerve cord significantly lowered the 5-HT content of the nervous system and other serotonin-containing tissues (Lent, 1984). The depletion of 5-HT was accompanied by a decrease in biting frequency of the animals (Lent & Dickinson, 1984a). More recently, O'Gara, Chae, Latham, and Friesen (1991) confirmed these results by depleting the amines from the nervous system of H. medicinalis with reserpine. Reserpine is an alkaloid t h a t inhibits the storage of amines by binding specifically to the amine translocator of storage organelles (Shore, 1962). Haemopis marmorata is a generalized carnivore who may eat several types of meat on a daily basis. Its nervous system closely resembles t h a t of the bloodsucking Hirudo (Sawyer, 1986a, b, c). and contains the same number of identified serotonergic neurons (Lent, 1981). The role of serotonin in the regulation of Haemopis" feeding behavior has not been established. We therefore undertook this study to determine what effect reserpine t r e a t m e n t would have on the serotonin content of the nervous system and feeding behavior of Haemopis.
The effect of reserpine on the serotonin content of the central nervous system and the feeding behavior of the carnivorous leech Haemopis marmorata was investigated. Leeches were hand-fed to satiation by presentation of pieces of chicken liver held in forceps for three meals at 4-day intervals prior to and 1 week following three consecutive injections of 100 ~g reserpine in the crop. A group matched by weight and preinjection food intake was injected with the vehicle in a blind experimental design. Histochemical and chemical determinations showed that reserpine effectively depleted the serotonin content of the animal's central nervous system. Furthermore, the food intake of reserpine-treated animals was significantly less than that of the sham-injected group. However, the depleted animals continued to ingest approximately 40% of the amount consumed during the preinjection period. These results show that reserpine successfully depleted the serotonin in the carnivorous leech H. marmorata and decreased the food intake of this leech species. They suggest that feeding behavior in Haemopis is partially but not exclusively dependent on serotonin and that, in contrast to the sanguivorous leeches, additional factors may be necessary for its regulation. ©1994AcademicPress,Inc. INTRODUCTION Serotonin (5-HT), an endogenous neurotransmitter, plays a vital role in the regulation of swimming (Willard, 1981) and feeding behaviors of the leech Hirudo medicinalis (Lent & Dickinson, 1984b). There are 7 5-HT-containing cells in each of the 21 segmental ganglia of the central nervous system of the leech (Muller, Nicholls, & Stent, 1981). The co1 This work was supported by a g r a n t from the Whitehall Foundation and NIH G r a n t 2RO1NS1805048A2 to A.L.K. We t h a n k Alexis Mentor and William Braswell for their participation in some of the experiments, Anne Marie Snow for help with the photography, and Dr. E. Sabban for her invaluable assistance during the HPLC determinations. Address correspondence and reprint requests to A n n a L. Kleinhaus. Fax: (914) 993-4653. E-mail: [email protected].
MATERIALS AND METHODS Leeches of the species H. marmorata were obtained from commercial suppliers (St. Croix Biological, Minnesota) and placed immediately after their arrival in individual containers with approx47 0163-1047/94 $5.00 Copyright © 19~4 by Academic Press, Inc. All rights of reproduction in any form reserved.
48
GOLDBURT, SABBAN, AND KLEINHAUS
imately 100 ml of leech water (0.5 g Instant Ocean, Aquarium Systems)/liter distilled water. They were maintained at room temperature on a 12-h light12-h dark cycle and allowed to acclimate for a week before the start of the experiments. The experiments reported here were performed during J u l y and August of 1992. Similar experiments with essentially the same outcome were done in J u l y and August of 1991.
Behavioral Experiments Prior to experimentation, leeches of approximately the same size were selected and fed to satiation once every other day in order to determine the average amount of food ingested by each animal during a meal. The method for feeding was slightly modified from that described by Karrer and Sahley (1988). The animals were fed by placing pieces of chicken liver (cubes of approximately 0.6 g each) held in forceps directly in the vicinity of their head. We found that when we separated the meals by intervals of at least 4 days, the leeches' food intake per meal varied less than if the meals were given at closer intervals. We then selected a number of animals who had consumed similar amounts of food over four consecutive meals and segregated them into a control group (n = 9) and an experimental group (n = 13) to be treated with reserpine. At the onset the experimental group was larger than the control one to allow for attrition. We then fed the two groups of animals three additional meals each at 4-day intervals, injected them with reserpine and the vehicle, respectively, and let them recover for a week. After the time allotted for recovery of the ill effects of injection we fed them again for three meals separated by 4 days each. We tabulated the amount of food ingested by each leech in the preand postinjection periods. Five animals from the experimental group were eliminated during the experiment due to death or obvious deterioration and were not included in the final tally. Injections and feeding experiments were done by investigators who were unaware of the nature of the injection material or the state of the animals. This was accomplished by labeling each leech box with a different symbol the meaning of which was known only to an individual who was not otherwise involved in these experiments. Nonetheless, reserpine altered the demeanor of the treated specimens so that despite all precautions their identity revealed itself in the course of the feeding experiments. To avoid bias resulting from the visible effect of reserpine on the animal's appearance and behavior, determination
of serotonin content by HPLC or histochemical methods was done blindly; e.g., dissection and assays were done by different investigators.
Reserpine Treatment Reserpine (Sigma) was dissolved in 50 t~l of glacial acetic acid and diluted to the final concentration (100 t~g/0.2 ml) in leech Ringer (120 m M NaC1, 4 m M KC1, 2 m M CaC12, 1.8 m M MgC12, 10 m M T r i s HC1, 5 m M glucose; pH 7.4) on each experimental day. We essentially followed the injection procedure of (O'Gara et al., 1991), e.g., injection of 0.2 ml of the reserpine-containing solution into the crop of the animals. The second group of leeches was shaminjected with a solution of the same composition but without reserpine. While a single dose of the drug was sufficient to essentially deplete the ganglia of H. medicinalis of their serotonin (O'Gara et al., 1991), three injections were necessary for significant depletion of serotonin from the nervous system of Haemopis. Therefore, the results on food intake are from animals who received one daily injection of reserpine on three consecutive days.
Fluorescent Visualization of Serotonin Each midbody ganglion of the central nervous system of the leech contains seven serotonergic neurons. A number of supplemental neurons can be seen in the subesophageal ganglion ("brain") which consists of four fused ganglia (Lent et al., 1979). Neurons containing 5-HT can be visualized by the glyoxylic acid condensation method for amines as described by Lent (1982). For this procedure leech ganglia were isolated from the animal by a slightly modified method from that described earlier (Kleinhaus & Prichard, 1975). Briefly, ganglia were removed from animals which had been placed in icecold Ringer's or in carbonated water (Cooper, Mahaffey, & Applebee, 1986) for 15-30 min before dissection. Anesthetized leeches were then pinned out in wax-bottom dishes, and the nerve cord was exposed by a dorsal incision. The brain and midbody ganglia were surgically removed for analysis. In an attempt to minimize the loss of transmitter during handling of the ganglia, we raised the [MgC12]o in the leech Ringer bathing the nervous system to 20 m M (Lent, 1982). Isolated ganglia were placed on a glass slide, dried in a cool jet of air for approximately 10 min, and then immersed in a cold solution of SPG (0.5 g glyoxylic acid monohydrate, 3.4 g sucrose, and 1.6 g potassium phosphate dissolved in 31 ml distilled H20). Ganglia thus treated were dried again for approximately 30 min. They
49
SEROTONIN DEPLETION IN LEECHES
were then covered with glycerol and a coverslip and heated at 92°C for 3 min. Ganglia from sham- and reserpine-injected animals were processed simultaneously. Preparations were examined with an inverted fluorescence microscope (Nikon Diaphot) equipped with a 35-mm camera for photography.
~
0 n -0
~
High Pressure Liquid Chromatography (HPLC) HPLC was used to quantitate the amount of serotonin in sham-injected animals and to verify the efficacy of reserpine depletion. The brain and ganglia were placed in microcentrifuge tubes each containing 150 t~l ice-cold 0.1 N perchloric acid, 0.1 m M EDTA, and 7.6 m M sodium bisulfite and sonicated on ice. After one round of freezing and thawing, the solution was centrifuged for 5 min in an Eppendorf microfuge (17,000g) and the supernatant injected onto an ADS Phase II, 5-tLm reverse-phase HPLC column (Bioanalytical Systems) in 20-t~l aliquots. The HPLC consisted of a Waters Model 510 pump that pumped the mobile phase (0.1 M potassium phosphate, 0.4 m M octyl sulfate, 0.1 m M EDTA, 5% methanol, pH 3.0) at 1 ml/min. The serotonin was detected with a BAS LC-4B amperometric detector with the electrode set at + 0.75 V above the reference electrode. Peak areas were determined by the Schimatzu C-R3A chromoatopac recorder/integrator which calculated the amount of transmitter present in the samples by comparing them to standard curves. RESULTS
Feeding Behavior Despite our attempt to select animals who ate similar amounts of food during the preexperimental period, we found that variations in food intake were inevitable: an individual leech consumed from one to eight pieces of the cut liver pieces per meal. Furthermore, we found that the animals in both groups were not always alert and ready to feed. To alleviate skewing of the results, we attempted to rouse lethargic animals by repeatedly tipping the box until they responded to the presence of food. Once this had been achieved we offered as many liver pieces as the animal would ingest for a period of 10 min. The animals were left without food for 1 week after completion of the injections to minimize potential harmful effect of the injections themselves. There was no significant difference between the preinjection food intake of the control and experimental animals, t(15) = - 1 . 6 , p > 0.1. Sham injections did not alter the behavior of the animals nor the
1.0
1
0.8
o.6
tO
r-
P
"U 0
o
kl_
0.2
Sham
Reserpine
FIG. 1. Reserpine injection decreases food intake in Hae-
mopis. Data represent the ratio of post-/preinjection food intake from sham-injected (cross-hatched) (n = 9) and reserpine-injected (stipled) (n = 13) animals. Error bars are SEM.
amount of food ingested; following the injection regimen these animals consumed 91 +_ 10% of the amount they had ingested before. In contrast, the reserpine-treated animals behaved very differently. Leeches in this group became lethargic and were less responsive to stimuli than animals in the control group. Generally they showed little or no interest in food offered in the usual manner even after repeated attempts. Once the animals were successfully aroused by tipping the container, food was presented in close proximity to their mouth in the usual manner in order to minimize the effect of their overall decrease in responsiveness on food intake. All animals in this group ate less following the three reserpine injections than in the preinjection period 36 -+ 11% and significantly less than the shaminjected leeches t(15) = - 3 . 7 , p < .01. The graph in Fig. 1 compares the mean food intake of shaminjected animals with that of reserpine-treated leeches for the three consecutive meals following the injections relative to the amount of food ingested during three consecutive meals before the injection treatment. In contrast to the lack of difference in food intake between the two groups before treatment (see above), the difference in food consumption following treatment between the sham-injected and reserpine-treated animals was highly significant t(15) = 6.7, p < .001 (unpaired t test). A paired t test indicated that the amount of food ingested by the sham-injected leeches following treatment was not significantly different from the amount of food they ingested before the injections t(8) = 0.65, p > .5. On the other hand, the same comparison for the reserpine-treated animals indicated that
50
GOLDBURT, SABBAN, AND KLEINHAUS
their food intake was significantly reduced following treatment as compared to that in the preinjection period t(7) = 3.93, p < .01.
Effect of Reserpine on Serotonin-Content of Leeches Histochemical procedures indicated that the serotonin content of the reserpine-treated animals was lower than that of the controls. These results, however, were only qualitative--no quantitative determinant of serotonin content was attempted. Figure 2 shows an example of the fluorescence exhibited by the serotonergic neurons in cerebral ganglia from a sham-injected (A) and reserpine-treated (B) animals, respectively, after condensation with glyoxylic acid. In this picture, the outline of Retzius cells and the other serotonergic neurons in the ganglion from a reserpine-treated animal are barely visible which is in contrast to the bright fluorescence exhibited by the same neurons in a ganglion isolated from a sham-injected animal (Fig. 2A) suggesting that the reserpine effectively decreased the serotonin content in the neurons of this leech. The post-injection food intake of the leech whose cerebral ganglion is depicted in Fig. 2A increased to approximately 120% of its value prior to the injections (Fig. 3A) In contrast, the food intake of the leech whose brain is illustrated in Fig. 2B was decreased to about 43% of control.
Levels of Serotonin Determined by HPLC In pilot experiments we found that the levels of serotonin found in a single ganglion were below the detection level of our system. In order to determine whether there was a good correlation between the histochemical and chemical determinations of serotonin, we assayed the serotonin in the brains and the pooled first five segmental ganglia of one shaminjected and one naive leech and of two animals injected with reserpine. The brains isolated from the sham-injected and control specimens contained 48 and 53 pmol serotonin, respectively; the corresponding values for the pooled ganglia from the same leeches were 57 and 45 pmol. The samples from the reserpinized animals contained no detectable serotonin indicating that reserpine treatment depleted the nervous system of Haemopis as effectively as it did that of Hirudo. The brains in
control specimens contained an additional peak tentatively attributed to dopamine. This peak was no longer present following reserpine t r e a t m e n t but no quantitation of other amines was attempted in these assays. DISCUSSION The results of this study show that depletion of serotonin with reserpine treatment significantly reduced the food intake of the carnivorous leech H. marmorata. These results resemble those previously obtained in the bloodsucking leech Hirudo (Lent, 1984; O'Gara et al., 1991) and thus support our hypothesis that serotonin is crucial for the expression of normal feeding behavior in this leech species. This is interesting in light of the difference in eating behavior characteristic for Hirudo (a bloodsucker) and Haemopis (a carnivore), respectively. Hirudo feeds only approximately once a year in nature whereas Haemopis can eat daily. Furthermore, the sensory stimuli initiating feeding are different in both species. Whereas Hirudo is sensitive to vibration, temperature, and chemical stimuli (Phillips & Friesen, 1982; Lent, 1985; Elliott, 1986), Haemopis appears to respond mainly to chemical stimuli (unpublished results). While our study reveals a role for serotonin in the control of feeding in the carnivorous leech, several differences between our data on Haemopis and those on H. medicinalis are worth mentioning. A single injection of 100 ~g reserpine into the crop of Hirudo was sufficient to lower the 5-HT content of its nervous system by 99% (O'Gara et al., 1991). In the same study, the effect of reserpine on biting behavior reached a m a x i m u m 3 days after the injection and lasted for several weeks. In contrast, in Haemopis three injections of reserpine were required for complete serotonin depletion. Several possibilities might explain the need for repeated injections. First, it is possible that since Haemopis feeds and excretes on a daily basis insufficient concentrations of reserpine were absorbed to adequately deplete the nervous system. Second, the animal's digestive enzymes might denature the injected reserpine into an inactive metabolite. Amine depletion profoundly modified the demeanor of the treated leeches: they appeared more rigid and were sluggish compared to the controls. It is therefore possible that part of the decreased responsiveness
FIG. 2. Reserpine injections deplete serotenin from the suboesophageal ganglion of Haemopis. In this micrograph, the Retzius and other serotonergic cells in the cerebral ganglion from the sham-injected animal (A) are brightly fluorescent. In contrast, in (B) the neurons in the suboesophageal ganglion from a reserpine-injected animal are totally devoid of fluorescence. The calibration bar represents 200 t~M.
SEROTONIN DEPLETION IN LEECHES
51
52
GOLDBURT, SABBAN, AND KLEINHAUS
Bloo
A
-0 m I1) f-
~ 5o 0 L
Pre-injection Post-injection
Pre-injection Post-injection
FIG. 3. Postinjection food intake of serotonin-depleted animal (same as in Fig. 2B) is decreased (B), whereas that of sham-injected one (same as in Fig. 2A) is unaffected (A).
to food may have been due to a general decline in arousal and decreased motor function despite our attempts to minimize these effects. Similar changes in appearance were observed in the reserpinetreated Hirudo (O'Gara et al., 1991) when tested for their biting responses. Nevertheless, despite the general decrease in activity in the reserpinized Haemopis, these animals continued to ingest a measurable amount of food. Therein lies perhaps the most interesting difference between our results and those in Hirudo. Indeed, as can be seen in Figs. 2 and 3 the animals that were successfully depleted by reserpine continued to ingest about 40% of the amount they had consumed before the injections, whereas Hirudo did not bite at all. One interesting possibility to consider for this discrepancy may be that the variable measured in our study was actual food intake, whereas the work on Hirudo used biting as a measure of feeding. Indeed, a recent study suggests that treatment of sated Hirudo leeches with 5-HT in an attempt to entice them to eat again was met with very little success, while the biting behavior was clearly increased by the exposure to 5HT (West, Nichter, & Halpern, 1991). In Hirudo, serotonin activates all the motor components of feeding behavior, e.g., salivation, biting, contraction of the pharynx, distension of the body wall, and mucus secretion (Lent, 1985). To our knowledge, the cellular action of serotonin on the components of the feeding circuit in Haemopis have not been studied in detail as yet. Thus, while the central nervous system of Haemopis and Hirudo are remarkably similar with respect to their morphology, it is possible that other transmitters resistant to reserpine might contribute to the regulation of feeding in Haemopis. Indeed, the total integration of the behavior may require a balance of many processes operating in parallel. These other processes might be able to
partially compensate for the lower levels of serotonin seen under our experimental conditions and explain the residual feeding observed in reserpinized animals. Much of the research in the control of appetite and weight in mammals centers around investigations of neurochemical and neuroendocrine systems. Indeed, a current hypothesis attributes many eating disorders in humans to an imbalance in the serotonergic system and other transmitters localized in specialized regions in the brain. For instance, one clinical study showed that serotonergic activity is decreased in bulimics, whereas anorexics whose weight has been returned to normal exhibit increased serotonergic activity (Leibowitz, 1992). While feeding behavior in humans involves many factors other than those required in the invertebrate, there appear to be some striking similarities at least with regard to some of the basic mechanisms involved. First, serotonin exerts a vital role in the control of the behavior in both and, second, feeding itself modulates levels of serotonin in both. The carnivorous leech whose feeding cycle is closer to that of mammals m a y prove to be an ideal preparation for the study of the basic mechanisms involved in the control of feeding by serotonin. Furthermore, such studies may increase our understanding of the evolution of behavioral neurochemistry. REFERENCES Cooper, J. E., Mahaffey, P., & K. Applebee (1986). Anaesthesia of the medicinal leech (Hirudo medicinalis). Veterinary Record, 118, 589-590. Elliott, E. J. (1986). Chemosensory stimuli in feeding behavior of the leech Hirudo medicinalis. Journal of Comparative Physiology A, 159, 391-401. Karrer, T., & Sahley, C. L. (1988). Discriminative conditioning
SEROTONIN DEPLETION IN LEECHES alters food preferences in the leech, Haemopis marmorata. Behavioral and Neural Biology, 50, 311-324. Kleinhaus, A. L., & Prichard, J. W. (1975). Calcium dependent action potentials produced in leech Retzius cells by tetraethylammonium chloride. Journal of Physiology, 246, 351361. Leibowitz, S. F. (1992). Neurochemical-neuroendocrine systems in the brain controlling macronutrients intake and metabolism. Trends in Neurosciences, 15, 491-497. Lent, C. M. (1981). Morphology of neurons containing monoamines within leech segmental ganglia. Journal of Experimental Zoology, 216, 311-316. Lent, C. M. (1982). Fluorescent properties ofmonoamine neurons following glyoxylic acid treatment of intact leech ganglia. Histochemistry, 75, 77-89. Lent, C. M. (1984). Quantitative effects of a neurotoxin upon serotonin levels within tissue compartments of the medicinal leech. Journal of Neurobiology, 15, 309-323. Lent, C. M. (1985). Serotonergic modulation of the feeding behavior of the medicinal leech. Brain Research Bulletin, 14, 643 -655. Lent, C. M., & Dickinson, M. H. (1984a). Retzius cells retain functional membrane properties following 'ablation' by the neurotexin 5,7-DHT. Brain Research, 300, 167-171. Lent, C., & Dickinson, M. (1984b). Serotonin integrates the feeding behavior of the medicinal leech. Journal of Comparative Physiology A, 154, 457-471.
53
Lent, C. M., Ono, J., Keyzer, K., & Karten, H. (1979). Identification of serotonin within vital-stained neurons from leech ganglia. Journal of Neurochemistry, 32, 1559-1563. Muller, K. J., Nicholls, J. G., & Stent, G. S. (Eds.) (1981). Neurobiology of the leech. Cold Spring Harbor Laboratory. Cold Spring Harbor, NY. O'Gara, B. A., Chae, H., Latham, L. B., & Friesen, W. O. (1991). Modification of leech behavior patterns by reserpine-induced amine depletion. Journal of Neuroscience, 11, 96-110. Phillips, C. E., & Friesen, W. O. (1982). Ultrastructure of the water-movement-sensitive sensilla in the medicinal leech. Journal of Neurobiology, 13, 473-486. Sawyer, R. T. (1986a). Leech biology and behavior. II. Feeding biology, ecology and systematics, 1st ed., Vol. 2. Oxford, UK: Clarendon Press. Sawyer, R. T. (1986b). Leech biology and behaviour. I. Anatomy, physiology, and behaviour, 1st ed., Vol. 1. Oxford, UK: Clarendon Press. Sawyer, R. T. (1986c). Leech biology and behaviour. III. Bibliography, 1st ed., Vol. 3. Oxford, UK: Clarendon Press. Shore, P. A. (1962). Release of serotonin and catecholamines by drugs. Pharmacological Reviews, 14, 531-550. West, B. R., Nichter, L. S., & Halpern, D. (1991) Leech therapy: when once is not enough. Blood Coagulation, Fibrinolysis and Kinin, 2, 197-200. Willard, A. L. (1981). Effects of serotonin on the generation of the motor program for swimming by the medicinal leech. Journal of Neuroscience, 1, 936-944.
Serotonin Depletion Inhibits Feeding in Carnivorous Leeches (Haemopis) VALERIE GOLDBURT, B E H I R A . SABBAN, AND A N N A L . K L E I N H A U S 1
Department of Cell Biology and Anatomy, New York Medical College, Valhalla, New York 10595
lossal Retzius cells contain the largest concentrations of serotonin. In addition, there are 5-HT-containing cells found only in the brain, such as the large lateral (LL) cells (Lent, Ono, Keyzer, & K a r t e n 1979). The importance of 5-HT in controlling feeding behavior in jawed, sanguivorous leeches is well documented (Lent, 1985). Injection of 5,7-DHT (a cytotoxic analog of 5-HT) directly into the blood sinus surrounding the nerve cord significantly lowered the 5-HT content of the nervous system and other serotonin-containing tissues (Lent, 1984). The depletion of 5-HT was accompanied by a decrease in biting frequency of the animals (Lent & Dickinson, 1984a). More recently, O'Gara, Chae, Latham, and Friesen (1991) confirmed these results by depleting the amines from the nervous system of H. medicinalis with reserpine. Reserpine is an alkaloid t h a t inhibits the storage of amines by binding specifically to the amine translocator of storage organelles (Shore, 1962). Haemopis marmorata is a generalized carnivore who may eat several types of meat on a daily basis. Its nervous system closely resembles t h a t of the bloodsucking Hirudo (Sawyer, 1986a, b, c). and contains the same number of identified serotonergic neurons (Lent, 1981). The role of serotonin in the regulation of Haemopis" feeding behavior has not been established. We therefore undertook this study to determine what effect reserpine t r e a t m e n t would have on the serotonin content of the nervous system and feeding behavior of Haemopis.
The effect of reserpine on the serotonin content of the central nervous system and the feeding behavior of the carnivorous leech Haemopis marmorata was investigated. Leeches were hand-fed to satiation by presentation of pieces of chicken liver held in forceps for three meals at 4-day intervals prior to and 1 week following three consecutive injections of 100 ~g reserpine in the crop. A group matched by weight and preinjection food intake was injected with the vehicle in a blind experimental design. Histochemical and chemical determinations showed that reserpine effectively depleted the serotonin content of the animal's central nervous system. Furthermore, the food intake of reserpine-treated animals was significantly less than that of the sham-injected group. However, the depleted animals continued to ingest approximately 40% of the amount consumed during the preinjection period. These results show that reserpine successfully depleted the serotonin in the carnivorous leech H. marmorata and decreased the food intake of this leech species. They suggest that feeding behavior in Haemopis is partially but not exclusively dependent on serotonin and that, in contrast to the sanguivorous leeches, additional factors may be necessary for its regulation. ©1994AcademicPress,Inc. INTRODUCTION Serotonin (5-HT), an endogenous neurotransmitter, plays a vital role in the regulation of swimming (Willard, 1981) and feeding behaviors of the leech Hirudo medicinalis (Lent & Dickinson, 1984b). There are 7 5-HT-containing cells in each of the 21 segmental ganglia of the central nervous system of the leech (Muller, Nicholls, & Stent, 1981). The co1 This work was supported by a g r a n t from the Whitehall Foundation and NIH G r a n t 2RO1NS1805048A2 to A.L.K. We t h a n k Alexis Mentor and William Braswell for their participation in some of the experiments, Anne Marie Snow for help with the photography, and Dr. E. Sabban for her invaluable assistance during the HPLC determinations. Address correspondence and reprint requests to A n n a L. Kleinhaus. Fax: (914) 993-4653. E-mail: [email protected].
MATERIALS AND METHODS Leeches of the species H. marmorata were obtained from commercial suppliers (St. Croix Biological, Minnesota) and placed immediately after their arrival in individual containers with approx47 0163-1047/94 $5.00 Copyright © 19~4 by Academic Press, Inc. All rights of reproduction in any form reserved.
48
GOLDBURT, SABBAN, AND KLEINHAUS
imately 100 ml of leech water (0.5 g Instant Ocean, Aquarium Systems)/liter distilled water. They were maintained at room temperature on a 12-h light12-h dark cycle and allowed to acclimate for a week before the start of the experiments. The experiments reported here were performed during J u l y and August of 1992. Similar experiments with essentially the same outcome were done in J u l y and August of 1991.
Behavioral Experiments Prior to experimentation, leeches of approximately the same size were selected and fed to satiation once every other day in order to determine the average amount of food ingested by each animal during a meal. The method for feeding was slightly modified from that described by Karrer and Sahley (1988). The animals were fed by placing pieces of chicken liver (cubes of approximately 0.6 g each) held in forceps directly in the vicinity of their head. We found that when we separated the meals by intervals of at least 4 days, the leeches' food intake per meal varied less than if the meals were given at closer intervals. We then selected a number of animals who had consumed similar amounts of food over four consecutive meals and segregated them into a control group (n = 9) and an experimental group (n = 13) to be treated with reserpine. At the onset the experimental group was larger than the control one to allow for attrition. We then fed the two groups of animals three additional meals each at 4-day intervals, injected them with reserpine and the vehicle, respectively, and let them recover for a week. After the time allotted for recovery of the ill effects of injection we fed them again for three meals separated by 4 days each. We tabulated the amount of food ingested by each leech in the preand postinjection periods. Five animals from the experimental group were eliminated during the experiment due to death or obvious deterioration and were not included in the final tally. Injections and feeding experiments were done by investigators who were unaware of the nature of the injection material or the state of the animals. This was accomplished by labeling each leech box with a different symbol the meaning of which was known only to an individual who was not otherwise involved in these experiments. Nonetheless, reserpine altered the demeanor of the treated specimens so that despite all precautions their identity revealed itself in the course of the feeding experiments. To avoid bias resulting from the visible effect of reserpine on the animal's appearance and behavior, determination
of serotonin content by HPLC or histochemical methods was done blindly; e.g., dissection and assays were done by different investigators.
Reserpine Treatment Reserpine (Sigma) was dissolved in 50 t~l of glacial acetic acid and diluted to the final concentration (100 t~g/0.2 ml) in leech Ringer (120 m M NaC1, 4 m M KC1, 2 m M CaC12, 1.8 m M MgC12, 10 m M T r i s HC1, 5 m M glucose; pH 7.4) on each experimental day. We essentially followed the injection procedure of (O'Gara et al., 1991), e.g., injection of 0.2 ml of the reserpine-containing solution into the crop of the animals. The second group of leeches was shaminjected with a solution of the same composition but without reserpine. While a single dose of the drug was sufficient to essentially deplete the ganglia of H. medicinalis of their serotonin (O'Gara et al., 1991), three injections were necessary for significant depletion of serotonin from the nervous system of Haemopis. Therefore, the results on food intake are from animals who received one daily injection of reserpine on three consecutive days.
Fluorescent Visualization of Serotonin Each midbody ganglion of the central nervous system of the leech contains seven serotonergic neurons. A number of supplemental neurons can be seen in the subesophageal ganglion ("brain") which consists of four fused ganglia (Lent et al., 1979). Neurons containing 5-HT can be visualized by the glyoxylic acid condensation method for amines as described by Lent (1982). For this procedure leech ganglia were isolated from the animal by a slightly modified method from that described earlier (Kleinhaus & Prichard, 1975). Briefly, ganglia were removed from animals which had been placed in icecold Ringer's or in carbonated water (Cooper, Mahaffey, & Applebee, 1986) for 15-30 min before dissection. Anesthetized leeches were then pinned out in wax-bottom dishes, and the nerve cord was exposed by a dorsal incision. The brain and midbody ganglia were surgically removed for analysis. In an attempt to minimize the loss of transmitter during handling of the ganglia, we raised the [MgC12]o in the leech Ringer bathing the nervous system to 20 m M (Lent, 1982). Isolated ganglia were placed on a glass slide, dried in a cool jet of air for approximately 10 min, and then immersed in a cold solution of SPG (0.5 g glyoxylic acid monohydrate, 3.4 g sucrose, and 1.6 g potassium phosphate dissolved in 31 ml distilled H20). Ganglia thus treated were dried again for approximately 30 min. They
49
SEROTONIN DEPLETION IN LEECHES
were then covered with glycerol and a coverslip and heated at 92°C for 3 min. Ganglia from sham- and reserpine-injected animals were processed simultaneously. Preparations were examined with an inverted fluorescence microscope (Nikon Diaphot) equipped with a 35-mm camera for photography.
~
0 n -0
~
High Pressure Liquid Chromatography (HPLC) HPLC was used to quantitate the amount of serotonin in sham-injected animals and to verify the efficacy of reserpine depletion. The brain and ganglia were placed in microcentrifuge tubes each containing 150 t~l ice-cold 0.1 N perchloric acid, 0.1 m M EDTA, and 7.6 m M sodium bisulfite and sonicated on ice. After one round of freezing and thawing, the solution was centrifuged for 5 min in an Eppendorf microfuge (17,000g) and the supernatant injected onto an ADS Phase II, 5-tLm reverse-phase HPLC column (Bioanalytical Systems) in 20-t~l aliquots. The HPLC consisted of a Waters Model 510 pump that pumped the mobile phase (0.1 M potassium phosphate, 0.4 m M octyl sulfate, 0.1 m M EDTA, 5% methanol, pH 3.0) at 1 ml/min. The serotonin was detected with a BAS LC-4B amperometric detector with the electrode set at + 0.75 V above the reference electrode. Peak areas were determined by the Schimatzu C-R3A chromoatopac recorder/integrator which calculated the amount of transmitter present in the samples by comparing them to standard curves. RESULTS
Feeding Behavior Despite our attempt to select animals who ate similar amounts of food during the preexperimental period, we found that variations in food intake were inevitable: an individual leech consumed from one to eight pieces of the cut liver pieces per meal. Furthermore, we found that the animals in both groups were not always alert and ready to feed. To alleviate skewing of the results, we attempted to rouse lethargic animals by repeatedly tipping the box until they responded to the presence of food. Once this had been achieved we offered as many liver pieces as the animal would ingest for a period of 10 min. The animals were left without food for 1 week after completion of the injections to minimize potential harmful effect of the injections themselves. There was no significant difference between the preinjection food intake of the control and experimental animals, t(15) = - 1 . 6 , p > 0.1. Sham injections did not alter the behavior of the animals nor the
1.0
1
0.8
o.6
tO
r-
P
"U 0
o
kl_
0.2
Sham
Reserpine
FIG. 1. Reserpine injection decreases food intake in Hae-
mopis. Data represent the ratio of post-/preinjection food intake from sham-injected (cross-hatched) (n = 9) and reserpine-injected (stipled) (n = 13) animals. Error bars are SEM.
amount of food ingested; following the injection regimen these animals consumed 91 +_ 10% of the amount they had ingested before. In contrast, the reserpine-treated animals behaved very differently. Leeches in this group became lethargic and were less responsive to stimuli than animals in the control group. Generally they showed little or no interest in food offered in the usual manner even after repeated attempts. Once the animals were successfully aroused by tipping the container, food was presented in close proximity to their mouth in the usual manner in order to minimize the effect of their overall decrease in responsiveness on food intake. All animals in this group ate less following the three reserpine injections than in the preinjection period 36 -+ 11% and significantly less than the shaminjected leeches t(15) = - 3 . 7 , p < .01. The graph in Fig. 1 compares the mean food intake of shaminjected animals with that of reserpine-treated leeches for the three consecutive meals following the injections relative to the amount of food ingested during three consecutive meals before the injection treatment. In contrast to the lack of difference in food intake between the two groups before treatment (see above), the difference in food consumption following treatment between the sham-injected and reserpine-treated animals was highly significant t(15) = 6.7, p < .001 (unpaired t test). A paired t test indicated that the amount of food ingested by the sham-injected leeches following treatment was not significantly different from the amount of food they ingested before the injections t(8) = 0.65, p > .5. On the other hand, the same comparison for the reserpine-treated animals indicated that
50
GOLDBURT, SABBAN, AND KLEINHAUS
their food intake was significantly reduced following treatment as compared to that in the preinjection period t(7) = 3.93, p < .01.
Effect of Reserpine on Serotonin-Content of Leeches Histochemical procedures indicated that the serotonin content of the reserpine-treated animals was lower than that of the controls. These results, however, were only qualitative--no quantitative determinant of serotonin content was attempted. Figure 2 shows an example of the fluorescence exhibited by the serotonergic neurons in cerebral ganglia from a sham-injected (A) and reserpine-treated (B) animals, respectively, after condensation with glyoxylic acid. In this picture, the outline of Retzius cells and the other serotonergic neurons in the ganglion from a reserpine-treated animal are barely visible which is in contrast to the bright fluorescence exhibited by the same neurons in a ganglion isolated from a sham-injected animal (Fig. 2A) suggesting that the reserpine effectively decreased the serotonin content in the neurons of this leech. The post-injection food intake of the leech whose cerebral ganglion is depicted in Fig. 2A increased to approximately 120% of its value prior to the injections (Fig. 3A) In contrast, the food intake of the leech whose brain is illustrated in Fig. 2B was decreased to about 43% of control.
Levels of Serotonin Determined by HPLC In pilot experiments we found that the levels of serotonin found in a single ganglion were below the detection level of our system. In order to determine whether there was a good correlation between the histochemical and chemical determinations of serotonin, we assayed the serotonin in the brains and the pooled first five segmental ganglia of one shaminjected and one naive leech and of two animals injected with reserpine. The brains isolated from the sham-injected and control specimens contained 48 and 53 pmol serotonin, respectively; the corresponding values for the pooled ganglia from the same leeches were 57 and 45 pmol. The samples from the reserpinized animals contained no detectable serotonin indicating that reserpine treatment depleted the nervous system of Haemopis as effectively as it did that of Hirudo. The brains in
control specimens contained an additional peak tentatively attributed to dopamine. This peak was no longer present following reserpine t r e a t m e n t but no quantitation of other amines was attempted in these assays. DISCUSSION The results of this study show that depletion of serotonin with reserpine treatment significantly reduced the food intake of the carnivorous leech H. marmorata. These results resemble those previously obtained in the bloodsucking leech Hirudo (Lent, 1984; O'Gara et al., 1991) and thus support our hypothesis that serotonin is crucial for the expression of normal feeding behavior in this leech species. This is interesting in light of the difference in eating behavior characteristic for Hirudo (a bloodsucker) and Haemopis (a carnivore), respectively. Hirudo feeds only approximately once a year in nature whereas Haemopis can eat daily. Furthermore, the sensory stimuli initiating feeding are different in both species. Whereas Hirudo is sensitive to vibration, temperature, and chemical stimuli (Phillips & Friesen, 1982; Lent, 1985; Elliott, 1986), Haemopis appears to respond mainly to chemical stimuli (unpublished results). While our study reveals a role for serotonin in the control of feeding in the carnivorous leech, several differences between our data on Haemopis and those on H. medicinalis are worth mentioning. A single injection of 100 ~g reserpine into the crop of Hirudo was sufficient to lower the 5-HT content of its nervous system by 99% (O'Gara et al., 1991). In the same study, the effect of reserpine on biting behavior reached a m a x i m u m 3 days after the injection and lasted for several weeks. In contrast, in Haemopis three injections of reserpine were required for complete serotonin depletion. Several possibilities might explain the need for repeated injections. First, it is possible that since Haemopis feeds and excretes on a daily basis insufficient concentrations of reserpine were absorbed to adequately deplete the nervous system. Second, the animal's digestive enzymes might denature the injected reserpine into an inactive metabolite. Amine depletion profoundly modified the demeanor of the treated leeches: they appeared more rigid and were sluggish compared to the controls. It is therefore possible that part of the decreased responsiveness
FIG. 2. Reserpine injections deplete serotenin from the suboesophageal ganglion of Haemopis. In this micrograph, the Retzius and other serotonergic cells in the cerebral ganglion from the sham-injected animal (A) are brightly fluorescent. In contrast, in (B) the neurons in the suboesophageal ganglion from a reserpine-injected animal are totally devoid of fluorescence. The calibration bar represents 200 t~M.
SEROTONIN DEPLETION IN LEECHES
51
52
GOLDBURT, SABBAN, AND KLEINHAUS
Bloo
A
-0 m I1) f-
~ 5o 0 L
Pre-injection Post-injection
Pre-injection Post-injection
FIG. 3. Postinjection food intake of serotonin-depleted animal (same as in Fig. 2B) is decreased (B), whereas that of sham-injected one (same as in Fig. 2A) is unaffected (A).
to food may have been due to a general decline in arousal and decreased motor function despite our attempts to minimize these effects. Similar changes in appearance were observed in the reserpinetreated Hirudo (O'Gara et al., 1991) when tested for their biting responses. Nevertheless, despite the general decrease in activity in the reserpinized Haemopis, these animals continued to ingest a measurable amount of food. Therein lies perhaps the most interesting difference between our results and those in Hirudo. Indeed, as can be seen in Figs. 2 and 3 the animals that were successfully depleted by reserpine continued to ingest about 40% of the amount they had consumed before the injections, whereas Hirudo did not bite at all. One interesting possibility to consider for this discrepancy may be that the variable measured in our study was actual food intake, whereas the work on Hirudo used biting as a measure of feeding. Indeed, a recent study suggests that treatment of sated Hirudo leeches with 5-HT in an attempt to entice them to eat again was met with very little success, while the biting behavior was clearly increased by the exposure to 5HT (West, Nichter, & Halpern, 1991). In Hirudo, serotonin activates all the motor components of feeding behavior, e.g., salivation, biting, contraction of the pharynx, distension of the body wall, and mucus secretion (Lent, 1985). To our knowledge, the cellular action of serotonin on the components of the feeding circuit in Haemopis have not been studied in detail as yet. Thus, while the central nervous system of Haemopis and Hirudo are remarkably similar with respect to their morphology, it is possible that other transmitters resistant to reserpine might contribute to the regulation of feeding in Haemopis. Indeed, the total integration of the behavior may require a balance of many processes operating in parallel. These other processes might be able to
partially compensate for the lower levels of serotonin seen under our experimental conditions and explain the residual feeding observed in reserpinized animals. Much of the research in the control of appetite and weight in mammals centers around investigations of neurochemical and neuroendocrine systems. Indeed, a current hypothesis attributes many eating disorders in humans to an imbalance in the serotonergic system and other transmitters localized in specialized regions in the brain. For instance, one clinical study showed that serotonergic activity is decreased in bulimics, whereas anorexics whose weight has been returned to normal exhibit increased serotonergic activity (Leibowitz, 1992). While feeding behavior in humans involves many factors other than those required in the invertebrate, there appear to be some striking similarities at least with regard to some of the basic mechanisms involved. First, serotonin exerts a vital role in the control of the behavior in both and, second, feeding itself modulates levels of serotonin in both. The carnivorous leech whose feeding cycle is closer to that of mammals m a y prove to be an ideal preparation for the study of the basic mechanisms involved in the control of feeding by serotonin. Furthermore, such studies may increase our understanding of the evolution of behavioral neurochemistry. REFERENCES Cooper, J. E., Mahaffey, P., & K. Applebee (1986). Anaesthesia of the medicinal leech (Hirudo medicinalis). Veterinary Record, 118, 589-590. Elliott, E. J. (1986). Chemosensory stimuli in feeding behavior of the leech Hirudo medicinalis. Journal of Comparative Physiology A, 159, 391-401. Karrer, T., & Sahley, C. L. (1988). Discriminative conditioning
SEROTONIN DEPLETION IN LEECHES alters food preferences in the leech, Haemopis marmorata. Behavioral and Neural Biology, 50, 311-324. Kleinhaus, A. L., & Prichard, J. W. (1975). Calcium dependent action potentials produced in leech Retzius cells by tetraethylammonium chloride. Journal of Physiology, 246, 351361. Leibowitz, S. F. (1992). Neurochemical-neuroendocrine systems in the brain controlling macronutrients intake and metabolism. Trends in Neurosciences, 15, 491-497. Lent, C. M. (1981). Morphology of neurons containing monoamines within leech segmental ganglia. Journal of Experimental Zoology, 216, 311-316. Lent, C. M. (1982). Fluorescent properties ofmonoamine neurons following glyoxylic acid treatment of intact leech ganglia. Histochemistry, 75, 77-89. Lent, C. M. (1984). Quantitative effects of a neurotoxin upon serotonin levels within tissue compartments of the medicinal leech. Journal of Neurobiology, 15, 309-323. Lent, C. M. (1985). Serotonergic modulation of the feeding behavior of the medicinal leech. Brain Research Bulletin, 14, 643 -655. Lent, C. M., & Dickinson, M. H. (1984a). Retzius cells retain functional membrane properties following 'ablation' by the neurotexin 5,7-DHT. Brain Research, 300, 167-171. Lent, C., & Dickinson, M. (1984b). Serotonin integrates the feeding behavior of the medicinal leech. Journal of Comparative Physiology A, 154, 457-471.
53
Lent, C. M., Ono, J., Keyzer, K., & Karten, H. (1979). Identification of serotonin within vital-stained neurons from leech ganglia. Journal of Neurochemistry, 32, 1559-1563. Muller, K. J., Nicholls, J. G., & Stent, G. S. (Eds.) (1981). Neurobiology of the leech. Cold Spring Harbor Laboratory. Cold Spring Harbor, NY. O'Gara, B. A., Chae, H., Latham, L. B., & Friesen, W. O. (1991). Modification of leech behavior patterns by reserpine-induced amine depletion. Journal of Neuroscience, 11, 96-110. Phillips, C. E., & Friesen, W. O. (1982). Ultrastructure of the water-movement-sensitive sensilla in the medicinal leech. Journal of Neurobiology, 13, 473-486. Sawyer, R. T. (1986a). Leech biology and behavior. II. Feeding biology, ecology and systematics, 1st ed., Vol. 2. Oxford, UK: Clarendon Press. Sawyer, R. T. (1986b). Leech biology and behaviour. I. Anatomy, physiology, and behaviour, 1st ed., Vol. 1. Oxford, UK: Clarendon Press. Sawyer, R. T. (1986c). Leech biology and behaviour. III. Bibliography, 1st ed., Vol. 3. Oxford, UK: Clarendon Press. Shore, P. A. (1962). Release of serotonin and catecholamines by drugs. Pharmacological Reviews, 14, 531-550. West, B. R., Nichter, L. S., & Halpern, D. (1991) Leech therapy: when once is not enough. Blood Coagulation, Fibrinolysis and Kinin, 2, 197-200. Willard, A. L. (1981). Effects of serotonin on the generation of the motor program for swimming by the medicinal leech. Journal of Neuroscience, 1, 936-944.