ABSTRACT Similar to other members of its genus, Conus striatus has a highly evolved venom apparatus that serves to inject paralyzing venom into its prey through a hollow, tubular radular tooth. Although much research is currently being conducted on the properties of the venom itself, little investigation has been done to elucidate the biomechanics behind the snail’s ability to inject its venom. Studies were carried out with individual radular teeth that defined the pressure-flow characteristics in order to estimate the pressure used by a snail to eject venom during prey capture. Pressure and flow were directly proportional over the range of investigated pressures (up to 4 cmHg) and flow rates (up to 2 ul s*). Behavioral observations of prey capture were carried out with C. catus, a closely related species that employs similar toxins. Video data revealed the approximate duration of venom ejection to be about 100 ms. The duration of venom ejection by C. striatus is probably similar, and the volume of ejected venom in this species can be 20 ul or more based on ’milking’ trials. Data suggest that the pressure necessary to carry out such an ejection must range from 2-4 atm. INTRODUCTION Predatory marine snails of the genus Conus have long been of interest because of their unique hunting strategies that employ peptide toxins to paralyze prey. In cone snails, the chitinous radular teeth resemble barbed hypodermic needles that are used to inject venom into the prey. Each species has a unique tooth structure, with varying size, shape, and number of barbs. These differences are often used as a basis for phylogenetic classifications (Kohn et al., 1999). When resting, the snail holds its proboscis retracted inside the rostrum. When prey is sensed by water-borne chemical signals (Spengler and Kohn, 1995), the proboscis is extended with a tooth held in place near the tip (Greene and Kohn, 1989). Searching behavior begins, and when contact with the prey is achieved, the snail impales the prey, paralyzing it, and pulls it into the expanded rhynchodeum by contraction of the proboscis (Kohn, 1956). Delivery of Conus venom involves generation and storage of radular teeth, transfer of a tooth to the proboscis, and final ejection of the venom. These processes are carried out within a complex venom apparatus that is relatively invariable between Conus species (Halstead, 1988). Marshall, et al. (in press) have detailed the anatomical correlates of the venom apparatus of the cone snail C. californicus (Fig. 1). It is highly likely that similar structures are found in piscivorous feeders such as C. striatus and C. catus. In Conus, the radular sac consists of two arms. Teeth are manufactured in the long arm and stored in the short arm (Marsh, 1977). Using scanning electron microscopy, Kohn et al. (1972) verified that each of the radular teeth is basically a sheet, presumably of chitin, rolled into a hollow tube. This tube has a large central lumen through which venom from the venom duct passes during the injection into the prey. In C. striatus, the venom enters the tooth via an adapical opening and leaves via an apical opening (Freeman, 1974). In preparation for prey capture, a single tooth is transported into the pharynx where it is surrounded by the relaxed proboscis and grasped by the sphincter near the tip of the proboscis (Greene and Kohn, 1989; Hermitte, 1946). Once at the tip of the proboscis, the lumen of the tooth is thought to be empty (Marsh, 1977), but findings from Marshall, et al. (in press) suggest that active peptides similar to those found in the anterior venom duct are present in the tooth lumen as well. Finally, the proboscis extends the tooth held at the tip, impales the prey, and injects the venom. How the force for final venom ejection out of the tooth or for tooth penetration is produced remains unknown. Although a muscular bulb posterior to the venom duct has often been hypothesized to provide the needed force, there is no direct evidence for this idea (Olivera, 1997). The muscular bulb does not appear to be capable of providing sufficient pressure to inject the venom (Songdahl, 1973). It is more likely that the proboscis itself provides the force for venom ejection (Kohn, et al. 1999). This report provides a description of the processes involved in venom ejection in C. striatus and C. catus, species that prey exclusively on fishes. Toxic components exist in the venom and details can be found in studies by Freeman (1974), Ramilo, et al. (1992), Cruz (1996), and Hahin, et al. (1991). Details of the mechanisms by which venom enters the tooth and is ejected into the prey are lacking. We have used isolated teeth to examine the biomechanical properties of ejection. First, we have found that maximum flow rate is directly proportional to peak pressure. Second, we have identified some previously unreported morphological characteristics of the C. striatus radular tooth through video data and scanning electron microscopy. Video data of venom ejection in C. catus revealed that the pressure that expels the venom also seems to insert the tooth into the victim. Data also indicated the approximate duration of venom ejection. Manual ’milking’ of C. striatus revealed the approximate volume of venom released per ejection. From all of these data, we were able to estimate the approximate pressure necessary for a living C. striatus to eject its venom. MATERIALS AND METHODS Specimens of C. striatus were obtained from American Samoa, and specimens of C. catus were obtained from Hawaii. Extraction of the teeth The radular sacs of C. striatus were removed and dissected from the animal and stored in -80 °C for several months prior to use. Teeth were dissected from the short arm of the radular sac, rinsed and stored in distilled water, and kept in 4 °C up to 3 days prior to use in ejection studies. Teeth were kept for approximately 15 days prior to use in electron microscopy. Ejection studies Teeth stored in distilled water were air-dried prior to use. Because we wanted to determine differences in fluid flow with and without the ligament attached, some teeth were left with the ligament intact, while the ligament and base of a number of teeth were cut using a microknife prior to use. Each tooth was placed inside a segment of clear plastic tubing. The tubing was positioned with the tip of the tooth exposed, and the tooth was then sealed in place with epoxy and allowed to dry for 5- 10 minutes. The tooth and tubing were then mounted and connected to a microinjection apparatus that served as the pressure source. A schematic diagram of the setup is shown in Fig. 2. As pressure was applied to the tooth, blue fluid consisting of distilled water and .03% Blue #1 dye flowed out of the microinjection apparatus and filled the mounted tooth. Blue fluid from the tooth was ejected into a separate capillary tube filled with clear distilled water. The set-up was connected to a pressure transducer, which was connected to an oscilloscope that recorded the voltage changes during each timed pressure pulse applied with an eppendorf microinjector 5242 apparatus. Calibration of pressure transducer is as follows: cm H20 = 0.7529 + 0.5704 * mV (Fig. 3) The experimental procedure was recorded using a Sony SSC-C374 color video camera. Teeth were viewed through a Nikon SMZ-2T microscope. Video images of the tooth were used to analyze how much fluid was being ejected with each pressure pulse (Fig. 4). The volume of fluid released into the capillary tube was measured by calibrating the length of fluid inside the capillary tube when a known volume of fluid was added. Scanning electron microscopy Teeth were cleaned briefly in concentrated sodium hypochlorite. A separate set of teeth was kept free of sodium hypochlorite. Teeth were immersed in 4% glutaraldehyde and were dehydrated through increasing concentrations of ethanol. Prior to scanning, teeth were air-dried, mounted on stubs with double-sided carbon tape, and coated with a thin layer of gold-palladium in a vacuum evaporator. They were examined with a Hitachi S-450 scanning electron microscope. No significant differences in scanning electron microscopy were observed in teeth cleaned with sodium hypochlorite and teeth kept free of sodium hypochlorite. Video data from C. catus Feeding behavior of C. catus was filmed with the same camera and microscope used for the ejection experiment. A fish (guppy) was held with a pair of forceps and presented to the C. catus for feeding. Milking' studies of C. striatus Part of the fin of a fish was attached to a 1.7 ml micro-centrifuge tube. The tube was then sealed with a piece of latex. This target was presented to C. striatus to induce venom ejection into the tube (Fig. 5). Volume of venom ejected was estimated by comparison with known volumes in other tubes. RESULTS The radular tooth The radular tooth of C. striatus is approximately 8 mm long with a diameter of 0.235 mm. The shaft is cylindrical and armed apically with a barb and a barbed blade together with a long backward pointing process with a recurved tip. A ligament is attached to the base of the tooth. A schematic diagram is shown in Fig. 6. Scanning electron microscopy of adapical and apical opening are shown in Figs. 7 and 8, respectively. Video data from ejection studies revealed a second opening associated with the large recurved barb. Fluid would flow out of this location, although exact site of fluid exit was unclear (Fig. 9). Ejection studies Minimal differences in total volume ejected at various pressures applied for 100 ms pulses were found between teeth with ligament attached and those without (Fig. 10a). Almost no differences in the flow rate at a constant applied pressure of 400 hPa were found in teeth with and without ligament (Fig. 10b). Öther measurements showed that pressure sensed at the basal end of the tooth was directly proportional to applied pressure. Table I shows the different variables measured as pressure from 100 hPa to 1000 hPa was applied for 100 ms pulses. Fig. 1 la shows the voltage changes recorded on the oscilloscope for 3 different pressure applications. Fig. 1 1b plots the volume of fluid released following the 100 ms pressure pulses. Initial slopes over the first second were calculated to determine fluid flow rate. Rate of flow of fluid ejection is directly proportional to peak pressure recorded on the oscilloscope (Fig.12). Venom ejection by C. catus Video footage of C. catus venom ejection revealed that the radular tooth was not located at the tip of the proboscis prior to ejection as previously observed in other Conus species (Marsh, 1977). Rather, the tooth was located approximately one tooth length away from the tip of the proboscis prior to ejection (Fig. 13). In every case, the tooth remained in this position until the apparent moment of venom ejection. Within one video frame, the tooth was forced out to the tip of the proboscis, presumably by the same pressure that enables venom ejection. The basal end appeared to be grasped by the tip of the proboscis, and the volume behind the tooth then swelled for several frames. Although a donut shaped ring, presumably of muscle, is present at the base of the proboscis (Fig. 15), no constriction or other obvious changes in the muscular ring occurred prior to, during, or after venom ejection. Although it was difficult to ascertain the actual duration of venom ejection, we based our estimate on the time elapsed from when the tooth impales the prey to when the snail starts to pull back on the tooth in preparation for swallowing (Fig. 16). In every case examined, this time was 3-4 video frames (n=6). DISCUSSION An initial hypothesis regarding the role of the ligament was that it acts as a protective barrier that prevents the contents of the tooth from seeping out of the base. Given this potential role, we hypothesized that flow rates of teeth with the ligament attached would be slower than that of teeth without the ligament. However, ejection studies using the experimental set-up described revealed that no significant differences in volume released were observed in teeth with and without ligament. In both cases, volume, flow, and pressure are all directly proportional. Whether the ligament indeed acts as a protective barrier remains to be seen. Future studies of the structure of the ligament using techniques such as transmission electron microscopy could be of help in bringing greater understanding of its role and function in venom ejection in Conus species. Ejection studies also revealed that a second opening located near the recurved barb of the radular tooth is present. We were unable to locate this second opening through scanning electron microscopy, mostly because of difficulties in achieving the correct orientation of the tooth. Further investigation of the exact location, size, and function of this opening would be helpful in understanding the mechanics behind venom ejection in cone snails. Video data from venom ejection by living Conus catus revealed that tooth penetration into prey, paralysis of prey, and time prior to retraction of the proboscis spans approximately 100 ms. We take this to be a minimum period for venom ejection. It is highly likely that the duration of venom ejection in C. striatus is similar. Manual milking of C. striatus revealed that approximately 20ul of venom is released per ejection (Dr. Joseph Schulz, personal communication). Given these observations, we were able to deduce that the rate of volume ejection by C. striatus would be 200 uls Extrapolating from our pressure-flow data (Fig. 12), we were able to ascertain that approximately 2-4 atm of pressure is required to inject 20 ul of fluid out of the C. striatus radular tooth in 100 ms. However, our data were gathered using distilled water as a venom replacement. Because venom is probably much more viscous, even greater pressures would be required. The literature to date characterizes the tooth to be located at the tip of the proboscis prior to penetration into prey (Marsh, 1977). The finding that prior to penetration, the tooth was actually located farther back into the proboscis is obviously in contrast to that idea. The mechanism by which the tooth is propelled from its initial location 5-8 mm away from the tip of the proboscis into the prey has yet to be determined. It appears that the same force that propels the venom forward also propels the tooth, although this observation has yet to be supported by conclusive evidence. The anatomical structure that exerts this force remains to be elucidated. ACKNOWLEDGEMENTS 1 would like to thank Dr. William F. Gilly for his steadfast guidance and unwavering patience throughout my learning experience. This project would not have been possible without his constant support and caring mentoring. I would also like to extend my gratitude to Chris Patton for his help with SEM, Dr. Joseph Schulz for providing advice, assistance, Conus specimens, and video data, and the entire Gilly lab for opening their doors to a novice learner. LITERATURE CITED: Cruz, L.J. 1996. Neuroactive peptides of the marine snail Conus striatus. J. of Natural Toxins. 5(1):122. Freeman, S.E., R.J. Turner, S.R. Silva. (1974). The venom and venom apparatus of the marine gastropod Conus striatus. Toxicon. 12(6):587-592. Greene, J.L. and A.J. Kohn, 1989. Functional morphology of the Conus proboscis (Mollusca: Gastropoda) J. Zool. Lond. 219:487-493. Hahin, R., G.K. Wang, B.I. Shapiro, B. Strichartz. 1991. Alterations in sodium channel gating produced by venom of the marine mollusk Conus striatus. Toxicon. 29(2):245-260. Halstead, B.W. 1988. Poisonous and Venomous Marine Animals of the World. Darwin Press, Princeton, NJ. Pp. 243-263. Hermitte, L.C.D. 1946. Venomous marine mollusks of the genus Conus. Trans. Roy. Soc. Trop. Med. Hygine 39: 485-512. Kohn, A. J. 1956. Picivorous gastropods of the genus Conus. Proceedings of the National Academy of Science. 42:168-172. Kohn, A.J., J.W. Nybakken, and J.J. Van Mol. 1972. Radula tooth structure of the gastropod Conus imperialis elucidated by scanning electron microscopy. Science. 176:49-51. Kohn, A.J., M. Nishis, and B. Pernet. 1999. Snail spears and scimitars: A character analysis of Conus radular teeth. J. Moll. Stud. 65:461-481. Marsh, H. 1977. The radular apparatus of Conus. Journal of Molluscan Studies. 43:1- 11. Marshall, J., W.P. Kelly, S.S. Rubakhin, J. Bingham, J.V. Sweedler, and W.F. Gilly. 2002. Biological Bulletin. In press. Anatomical Correlates of Venom Production in Conus Californicus. Olivera, B.M. 1997. Conus venom peptides, receptor and ion channel targets, and drug design: 50 million years of neuropharmacology. Molecular Biology of the Cell 8: 2101-2109. 11 Ramilo, C.A., G.C. Zafaralla, et al. 1992. Novel alpha- and omega- conotoxins from Conus striatus venom. Biochem. 31(41): 9919-9926 Songdahl, J.H. 1973. The venom and venom apparatus of the Atlantic cone, Conus spurious atlanticus (Clench). Bulletin of Marine Science 23:600-612. Spengler, H.A. and A.J. Kohn. 1995. Comparative external morphology of the Conus osphradium. Journal of Zoology (London). 235: 439-453. TABLES TABLE I. MEASURED VARIABLES AS PRESSURE APPLIED VARIES (100 ms pulses) n=3 for 100-800 hPa pulses; n=1 for 1000 hPa pulse Applied Total vol Mean peak Mean Flow Standard Standard Pressure pressure Deviation Rate (uV/s) Deviation (ul) (cm H20) Pressure Flow Rate 100 hPa 0.368 3.71 0.5154 0.009292 0.073 1.127 0.04949 200 hPa 9.854 0.214 0.344 400 hPa 2.195 19.69 0.654 0.895 0.008021 800 hPa 4.05 40.37 0.035119 1.387 0.568 5.965 1000 hPa 30.24 none none 22 FIGURE LEGENDS: Fig. 1. Diagram of venom apparatus of a typical cone snail. Modified from Marshall, et al. (in press) Fig. 2. Diagram of experimental set-up. Tooth was inserted into a clear plastic tubing Tip of tubing was sealed with epoxy (yellow). Tooth was connected to a microinjector apparatus that pumps blue fluid into the tooth. Blue fluid flows out of the tooth and into a capillary tubing filled with clear distilled water. Set-up was connected to a pressure transducer that was also connected to an oscilloscope. Experiments were filmed using a color video camera (not shown). Fig. 3. Graph of pressure transducer calibration. Fig. 4. Diagram of video measurements. Measurements were made after every 3 frames (100 ms). Length of fluid released into the capillary tube filled with distilled water was measured. Calibration of fluid volume in the capillary tubing enabled us to determine how much volume was being released at each frame. See text for additional details. Fig. 5. Schematic diagram of milking procedure of C. striatus. Fig. 6. Schematic diagram of a radular tooth from C. striatus. Not to scale. Fig. 7. Scanning electron micrograph of adapical opening of the radular tooth from C. striatus with ligament attached. Fig. 8. Scanning electron micrograph of apical end of the radular tooth from C. striatus. Fig. 9. Scanning electron micrograph of barbed aperture of the radular tooth from C. striatus. Second opening is presumed to be at an unspecified location within the boxed region. Possible locations include the tip of the barbed aperture itself or an opening located at the shaft of the tooth beneath the barbed aperture. Fig. 10a. Graph of total volumes ejected as pressure applied with 100 ms pulses was varied for teeth with and without ligament. Fig. 1Ob. Graph of incremental volume ejected per 100 ms with an applied pressure of 400 hPa in teeth with and without ligament. Duration of pressure was 100 ms. Fig. 11a. Pressure transients in a tooth with ligament as pressure was applied at 200 hPa, 400 hPa, and 800 hPa. Duration of pulse was 100 ms. 14 Fig. 11b. Graph of incremental volume ejected over time corresponding to the pressure transients illustrated in Panel a. Fig. 12. Graph of flow rate vs. pressure. Data points are means (+ 5.0) for pressure pulses of 100 hPa, 200 hPa, 400 hPa, and 800 hPa (applied pressures, 100 ms pulses). n=3 for all, except n=1 for 1000 hPa. Fig. 13. Modified diagram of venom apparatus. Arrows indicate the observed location of the tooth as well as the muscular ring inside the proboscis. Fig. 14. Relevant frames depicting duration of venom ejection in C. catus. Arrow in Frame 1 depicts the radular tooth located a few millimeters away from the tip of the proboscis as observed prior to ejection. Frame II shows that the radular tooth has already penetrated the prey. Frame III shows the radular tooth still impaled to the prey. Arrow in Frame IV reveals that the cone snail has started pulling back on the radular tooth out of the prey in preparation for swallowing. Estimated duration of venom ejection is spread out over Frames II-IV. Time lapse is 1/3 seconds per frame. Fig. 15. Extrapolated pressure-volume curve given a time pulse of 100 ms and a volume of 20 ul. e FIGURES radular sac long arm muscular bulb short arm proboscis — — esophagus k- pharynx anterior Figure 1 ) venom duct - posterior Figure 2 H Pressure Transducer Pressul Source (microinjec apparatus Oscilloscope 0 100 - 80- 60 - 20- High-pressure transducer calibration cm H20 = 0.7529 + 0.5704'mV (n-3, bidirectional) 170 Pressure transducer out (mV) Figure 3 Figure 4 1915.03 E 19.15.06 19:15:09 Figure 5 fishfin latex Figure 6 100um Figure 7 Figure 8 50 um Figure 9 500 um wand wo lig PV — 100 200 300 400 500 600 700 800 900 PrSSu (NPA) Figure 10a 400HPA TV fines Figure 10b 12 14 16 4-0NPAI A Figure 1la 6.5 cm H20 Pressure N 55 Figure 11b 10- os- 8 00 2 3 4 5 ûme (s) Figure 11a-b 100 ms Pressure Pulses — — 319. 10- 805 oo 012345 time (s) o 012345 time (s) C e Figure 12 Flow Rate vs Pressure Mean Flow Rate ulls proboscis muscular ring Figure 13 radular sac muscular bulb oo venom duct sophagus — pharynx anterior posterior e .. 10000 1 atm 1000 Figure 15 2-4 atm pressure range : 9 0.01 100 1000 Flow (ulis) 20 ul injected in 100 ms