Venom variation and chemoreception of the viperid Agkistrodon contortrix: evidence for adaptation?

P1: GCRJournal of Chemical Ecology [joec] pp900-joec-467947 July 10, 2003 15:15 Style file version June 28th, 2002

Journal of Chemical Ecology, Vol. 29, No. 8, August 2003 (!C 2003)

VENOM VARIATION AND CHEMORECEPTION OFTHE VIPERID Agkistrodon contortrix: EVIDENCE

FOR ADAPTATION?

ELI GREENBAUM,1,! NADEZHDA GALEVA,2 andMICHAEL JORGENSEN3

1Division of HerpetologyNatural History Museum & Biodiversity Research Centerand Department of Ecology and Evolutionary Biology

1345 Jayhawk BoulevardThe University of Kansas

Lawrence, Kansas 66045-7561, USA2Biochemical Research Service Laboratory

1251 Wescoe Hall DriveThe University of Kansas

Lawrence, Kansas 66045, USA31000 Emery Road, Lawrence, Kansas 66044, USA

(Received June 25, 2002; accepted April 13, 2003)

Abstract—Previous studies of chemoreceptive behavior in vipers suggest thatsnakes focus on the scent of envenomated tissue to track their prey follow-ing envenomation. Other studies have indicated a correlation between qualita-tive differences in venom biochemistry and geographic variation in diet. TheNorth American copperhead (Agkistrodon contortrix) varies geographically indiet and venom biochemistry; snakes were collected from three populations(Kansas, Texas, and Louisiana) that are known to have different prey prefer-ences. Behavioral experiments were conducted to assess whether copperheadspreferred envenomated preymore than nonenvenomated prey, as do other speciesof vipers studied thus far. Additional experiments tested the ability of copper-heads to distinguish between envenomated prey from different geographic pop-ulations, and between geographic populations of copperheads and two otherspecies of viper. Results indicated that copperheads prefer envenomated prey tononenvenomated prey. In envenomated-prey discrimination experiments, cop-perheads distinguished between envenomated prey from different geographicpopulations, and some snakes distinguished envenomated prey of A. contortrixfrom those of A. piscivorus and Sistrurus catenatus. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)was used to study the variation

! To whom correspondence should be addressed. E-mail: [email protected]

1741

0098-0331/03/0800-1741/0 C" 2003 Plenum Publishing Corporation


1742 GREENBAUM, GALEVA, AND JORGENSEN

of venom biochemistry in this species and two other taxa (A. piscivorus andS. catenatus), and confirmed intraspecific and interspecific variation of venomproteins. Relative potency of the venom from different populations as indicatedby time to immobilization experiments was in the order: Louisiana > Texas >

Kansas. The relative potency of the venom from each population matched theorder of preference in the chemoreception experiments. These results suggestthat chemoreception is sensitive to subtle differences in venom biochemistryand may reflect adaptation to improve efficiency of finding envenomated prey.

KeyWords—Agkistrodon contortrix, Sistrurus, chemoreception, venom, snake,SDS-PAGE, behavior, adaptation.

INTRODUCTION

Vipers are sit-and-wait, venomous predators that typically strike, envenomate, andrelease their prey (but see Kardong, 1986 for exceptions), and then use tongue-flicking (chemoreception) to trail and locate prey (Chiszar et al., 1992). Tongue-flicking has been used in numerous studies of squamate behavior as an index ofchemosensory investigation (Font andDesfilis, 2002, and citationswithin). Studiesof chemoreception in venomous snakes have demonstrated that individuals tongue-flick mammalian prey envenomated by a conspecific significantly more often thanprey that is not envenomated—this phenomenon of envenomated prey preferenceis common in viperids (e.g., Baumann, 1927, 1928, 1929: Vipera aspis and V.berus; Chiszar et al., 1992: Agkistrodon piscivorus andGloydius blomhoffi; Duvallet al., 1980; Robinson and Kardong, 1991; Chiszar et al., 1992; Lavin-Murcio andKardong, 1995; Chiszar et al., 1999: Crotalus atrox, C. durissus, C. enyo, C.horridus, and C. viridis). These studies suggest that after striking prey, vipersuse chemical cues from envenomated tissue to a larger extent than chemical cuesproduced by the prey (e.g., prey scent, urine, or alarm pheromones) to locate theirprey (Chiszar et al., 1992). Only conspecifics were used in these experiments; thus,it was not determined whether snakes could differentiate among venoms producedby different species.

Viper venom consists mainly of a complex mixture of peptides and proteins,which immobilize, kill, and digest prey (Thomas and Pough, 1979; Russell, 1980).Crotaline venom proteins have numerous deleterious effects on prey including, butnot limited to, local tissue and systemic vascular damage, hemolysis, and fibrinol-ysis (Johnson and Ownby, 1993; Gomez and Dart, 1995). Venom compositionand mode of action can vary at all taxonomic levels from populations to families(Jones, 1976; Broad et al., 1979; Mebs, 1999). Presumably, tissue envenomated bydifferent species should produce different chemical cues owing to variation in thechemical composition of the venoms. Because species with venoms of relativelyhigh toxicity can kill prey faster (Leon et al., 1999; Mebs, 1999), the distanceneeded to travel to envenomated prey is decreased. We hypothesize that as venomtoxicity increases, the ability of snakes to detect envenomated prey from a distance(i.e., salience), or ability to follow envenomated prey trails increases. Thus, the


CROTALINE VENOM AND CHEMORECEPTION 1743

ability to distinguish between envenomated prey from snakeswith disparate venomcomposition could reflect adaptations of local populations to improve chemore-ceptive ability. Chiszar et al. (1999) demonstrated the minimal perceptible venomdose required for effective rattlesnake chemoreception (#6.0 mg), but the effectsof qualitative venom variation on the chemoreceptive ability of vipers has neverbeen tested.

Agkistrodon contortrix is a medium-sized crotaline viper, which ranges fromextreme northeastern Mexico to Massachusetts. Currently, five subspecies are rec-ognized (Gloyd and Conant, 1990). Rodents are the main prey item throughoutits range, although local populations may concentrate on other types of prey (e.g.,Louisiana: frogs, Dundee and Rossman, 1989; Texas: cicadas and lepidopteranlarvae, Lagesse and Ford, 1996; Kansas: mammals, Fitch, 1960). Populationsof A. contortrix vary both in venom biochemistry and diet (Fitch, 1960; Jones,1976; Dundee and Rossman, 1989; Lagesse and Ford, 1996). Correlations be-tweenvenomanddiet have been suggested by studies of genes controlling venom inCrotalus and Trimeresus (Forsdyke, 1995, and citations within), and were demon-strated in the Asian viperid Calloselasma rhodostoma (Daltry et al., 1996). Thisgeneral information provides the basis for testing several hypotheses related tovariation in venom biochemistry and chemoreception in populations of the cop-perhead, A. contortrix.

To address whether variation in snake venom can affect chemoreceptivesearching, we tested the behavioral responses of copperheads (Agkistrodon con-tortrix) to: (1) conspecific envenomated vs. nonenvenomated mouse prey; (2) con-specific envenomated prey from snakes in different populations (Kansas, Texas,Louisiana)with documented variation in venombiochemistry anddiet; and (3) con-specific envenomated prey vs. envenomated prey from A. piscivorus and Sistruruscatenatus. We also tested how quickly a bite from snakes in the three copper-head populations could immobilize mammalian prey. We analyzed the venombiochemistry of three species of crotaline snakes (A. contortrix, A. piscivorus, andS. catenatus) to determine intraspecific and interspecific variation, and to identifyproteins that might be responsible for differences in chemoreceptive behavior.

METHODS AND MATERIALS

Animal Collection and Maintenance. Adult copperheads (Agkistrodon con-tortrix) from three populations (Kansas, Texas, Louisiana)with documented differ-ences in diet preferences (mammal, lepidopteran larvae, and amphibians, respec-tively) were collected from May 15 to June 30, 2000 (Greenbaum, 2003). Ninesnakes were collected from the vicinity of Lawrence, Kansas; 13 from the vicinityofWhitehouse, Texas; and eight from the vicinity ofMonroe, Louisiana. Individualsnakes were housed in 10-gal glass cages (26$ 31$ 51 cm) with opaque sides toprevent distractions to the animals. The bottomof each cagewas linedwith newspa-per, and a plastic water bowl allowed ad libitum drinking. All animals were housed



in a limited-access roomwith a relatively constant temperature (27.0± 1.0%C), hu-midity (50.0± 5%), and fluorescent lighting set for a daily 12 hr L:D cycle. Cagesand water bowls were cleaned regularly with disinfectant. Snakes were allowed toacclimate to captivity for 6 wk prior to behavioral experiments, which commencedin mid-August 2000. Two adult A. piscivorus from Bastrop, Louisiana, and oneadult Sistrurus catenatus from Kansas were housed in the same room and used asvenom donors for envenomated-prey experiments.

Behavioral Experiments. The following experimental protocols were modi-fied from those of Chiszar et al. (1999) andMain and Bull (1996). We began a trialby removing the cage lid of the test snake and placing a small, 60-W lamp overthe cage; this allowed us to observe the snake and made it difficult for the snaketo see us. Each snake’s cage served as a test arena for each experiment. After eachsnake was acclimated for 1–2min, it was allowed to strike a live mouse, which wasremoved immediately (Chiszar et al., 1999). Then a prey-presentation device wasplaced into the arena opposite the snake. The prey-presentation device consisted oftwo mouse carcasses (both lab-raised, 15–20 g), each in separate nylon bags thatwere suspended (10 cm apart and 2 cm from floor of cage) from a wooden beamacross the top of the aquarium. Location of each prey item in either the right orleft nylon bag was determined randomly. Nylon bags were washed with detergentand allowed to air dry between trials to remove prey scent and venom residue.

Prior to placement in nylon bags, envenomatedmicewere observed for 10minto ensure that venom had been injected, and all physical movements had ceased.The number of tongue flicks directed toward each prey item (within 2 cm of thebottom of the nylon bag) was recorded for 10 min. At the end of each trial, weremoved both prey from the bags and offered them to snakes; the purpose wasto evaluate whether the initial strike was a defensive or predatory response to theprey item. If the snake ate one of the items within 5 min, we classified its behavioras “predatory” (Chiszar et al., 1999). If a snake attempted to consume a prey itemthrough a bag during the final 2 min of a trial, we also interpreted the behavioras “predatory,” the trial was terminated, and the bitten prey item was recorded as“consumed” prey. If a snake crawled out of the cage, or remained stationary forthe first 5 min, the trial was terminated and not counted.

One experiment tested whether snakes could distinguish between enveno-mated vs. nonenvenomated prey. We hypothesized that Agkistrodon contortrixwould possess envenomated prey preference, as do other crotaline species (Chiszaret al., 1999). A nonenvenomated mouse (killed by CO2 overdose) was placed inone bag, and a mouse killed by envenomation from a snake in the same popu-lation as the test subject was placed in the remaining bag. The control experi-ment recorded the number of tongue flicks directed at two nonenvenomated mice.We also tested whether the snakes could distinguish between envenomated preyfrom conspecifics in their own population vs. conspecifics in one of two otherpopulations, or from individuals of A. piscivorus and Sistrurus catenatus. We



hypothesized that copperheads would direct more tongue flicks at prey enveno-mated by local conspecifics than at prey envenomated by conspecifics from otherpopulations. We also hypothesized that specimens of A. contortrix would directmore tongue flicks at prey envenomated by local conspecifics than at prey enven-omated by A. piscivorus or S. catenatus.

All studies followed a randomized, complete-blocks design to minimize dif-ferences in behavior caused by time of year or other nonexperimental factors (Sokaland Rohlf, 1995). Because some individual snakes rarely cooperated for experi-ments, there are small differences in sample sizes among experiments within eachpopulation; some individuals were tested more than once for a given experiment.A minimum of 10 wk elapsed between duplicate trials for individual snakes, thusminimizing the chance of one trial affecting the outcome of another. To test thisassumption, we subjected these data to runs tests to search for autocorrelationsamong snakes (Sokal and Rohlf, 1995). To reduce variability between individualsnakes and facilitate comparisons on the same scale, we converted behavioral data(number of tongue flicks directed toward each prey item) to percentages of totalnumber of tongue flicks for each trial (Duvall et al., 1980; Chiszar et al., 1999).We performed an arcsine transformation on these percentages to ensure normalityand homogeneity of variance (Duvall et al., 1980). We analyzed these data witha one-way ANOVA to test whether individual snakes differed in preference; ifthis test and runs tests did not indicate significant differences we combined trialsfrom each population. We analyzed these data with the Student’s 1-sample t test.Because of small sample sizes, it was not possible to make adjustments to thesignificance level (! = 0.05) to account for multiple t tests; thus, it is possible thata small number of Type I errors were made. Data on the type of prey consumed atthe end of each trial were analyzed with the sign test.

Prey Immobilization. During random trials of behavioral experiments foreach population of Agkistrodon contortrix, we timed how long it took an enveno-mated mouse (15–20 g) to become immobilized from a bite; immobilization oc-curred when a mouse could no longer walk. If immobilization did not occur within5 min, the trial was discounted. To minimize variation from amount of venom in-jected or location of envenomation, five trials were conducted for each snake andaveraged before the analysis. Data from these experiments were analyzed with aone-way analysis of variance with unequal sample sizes (ANOVA).

Venom Biochemistry Analysis. To examine the venom biochemistry of thesnakes, venom samples were collected from one Agkistrodon piscivorus and oneSistrurus catenatus, as well as three individuals from each copperhead popula-tion. Venom was collected by allowing snakes to bite plastic collection cups withparafilm covers three times. Samples were collected from snakes in all popula-tions at the same time; all individuals had eaten exactly 7 d prior to venom collec-tion. Samples were lyophilized and stored at 4.0%C; crude venom was dissolved in10mMTris-HCl (pH 8.6) (Johnson andOwnby, 1993). Total protein concentration



FIG. 1. Results of a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of crude snake venom using a Criterion 15% Tris-HCl gel. Lanes 1–3, Agkistrodoncontortrix from Louisiana; Lanes 4–6, A. contortrix from Texas; Lanes 7–9, A. contortrixfrom Kansas; Lane 10, A. piscivorus (PI); Lane 11, Sistrurus catenatus (CA). Note thevariation in presence and relative concentrations of venom proteins among the populationsof A. contortrix, and between the species. Asterisk denotes Phospholipase A2 HomologMyotoxin precursor (ca. 16.0 kDa); present in all A. contortrix. The last lane consists ofmolecular weight standards.

in the venom solutions was determined spectrophotometrically (595 nm detec-tion) with the Bio-Rad protein assay dye reagent. Sample aliquots containing20µg of protein were diluted with denaturing sample buffer (5%mercaptoethanol,2% sodium dodecyl sulfate, 0.06 M Tris-HCl [pH 6.8], 10% glycerol), heated at95%C for 5 min, and examined with sodium dodecyl sulfate-polyacrylamide gelelectrophoresis (SDS-PAGE) with a Criterion 15% Tris-HCl gel (Bio-Rad). Theelectrophoresis was performed at 200 V for 60 min, and gels were stained withCoomassie Brilliant Blue G-250 in 20% methanol to facilitate viewing of bands.The gel imagewas acquiredwith aMolecular Imager FX (Bio-Rad). Proteinmolec-ular weights were estimated by visual alignment to molecular weight standardson the SDS-PAGE gel and their relative concentrations were quantified by thedensitometry method (Quantity One Software, Bio-Rad).

For the most intense bands on the gel (approximately 16, 25, 29, and 60 kDa;Figure 1), we utilized the peptide mass-fingerprinting approach to identify themain protein components (Henzel et al., 1993). This method involves a massspectrometric analysis of peptide fragments from in-gel digested proteins fol-lowed by a search with the generated mass data in protein databases. The protein



bands (16, 25, 29, and 60 kDa) were excised by hand and digested with trypsin(Promega) as previously described by Rosenfeld et al. (1992). Extracted pep-tide mixtures were analyzed with Matrix-Assisted Laser Desorption IonizationTime-of-Flight (MALDI-TOF) mass spectrometry (Voyager-DE STR, PerSeptiveBiosystems), with CHCA (10 mg/ml solution of alpha-cyano-4-hydroxycinnamicacid in 50% acetonitrile/0.3% trifluoroacetic acid [50% MeCN/0.3% TFA]) asa matrix. Internal mass calibration was performed using trypsin autolysis peaks(MH+ 842.5021 and 2211.0968). The peptide mass fingerprinting data were an-alyzed with the MS-Fit program (Protein Prospector, http://prospector.ucsf.edu)in the protein databases SWISS-PROT (http://www.expasy.ch/sprot) and NCBInr(http://www.ncbi.nlm.nih.gov). The following parameters were used: molecularweight range 10–100 kDa; peptide mass tolerance 100 ppm; maximum missedcleavages 3; acrylamide modified Cys.

RESULTS

Chemoreception Experiments. In all experiments, the one-way ANOVA testand runs tests did not indicate significant differences; thus, individial snakes did notdiffer in preference and trials fromeach respective populationwere combined.Eachpopulation of Agkistrodon contortrix directed more tongue flicks at envenomatedprey than at nonenvenomated prey (Kansas: t = 2.26; df = 12; P < 0.05; Texas:t = 2.24; df = 11; P < 0.05; Louisiana: t = 3.65; df = 10; P < 0.01; Figure 2);this result supported our hypothesis. In the control experiment, there were nosignificant differences in number of tongue flicks directed at the two nonenveno-mated prey items for each population (Kansas: t = 0.04; df = 9; P > 0.05; Texas:t = 0.96; df = 10; P > 0.05; Louisiana: t = 1.26; df = 5; P > 0.05). Despitethese results, snakes did not consume significantly more envenomated prey thannonenvenomated prey (all sign tests: P > 0.05); the first prey encountered mostoften was the one consumed.

Snakes from Kansas preferred Texas-envenomated prey to their own (t =4.91; df = 11; P < 0.01), andLouisiana-envenomatedprey to their own (t = 2.74;df = 9; P < 0.05; Figure 3). However, snakes from Kansas did not differ innumber of tongue flicks between prey envenomated by Kansas snakes and thoseof Agkistrodon piscivorus (t = 0.33; df = 8; P > 0.05) or Sistrurus catenatus(t = 0.55; df = 10; P > 0.05). Snakes from Texas did not prefer either preyitem in the Texas vs. Louisiana trials (t = 0.12; df = 28; P > 0.05), but Texassnakes preferred their own envenomated prey over those from Kansas (t = 3.80;df = 10; P < 0.01), A. piscivorus (t = 2.62; df = 8; P < 0.05), and S. catena-tus (t = 2.95; df = 10; P < 0.05). Snakes from Louisiana directed more tongueflicks at envenomated prey from their own population than envenomated prey fromTexas (t = 3.72; df = 15; P < 0.01), Kansas (t = 2.68; df = 10; P < 0.05), A.piscivorus (t = 2.92; df = 10; P < 0.05), or S. catenatus (t = 2.59; df = 10;



FIG. 2. Mean (+SE) of arcsine-transformed percentages of tongue flicks of three popu-lations of Agkistrodon contortrix directed at envenomated or nonenvenomated prey items.Asterisks indicate significant differences (P < 0.05).

P < 0.05). As with the experiments to test ability to discern envenomated vs.nonenvenomated prey, snakes often consumed the first prey item encountered (allsign tests: P > 0.05).

The amount of venom collected from cooperative snakes did not differ amongpopulations (F = 2.21; df = 2, 23; P > 0.05; Table 1). Because there was no sig-nificant difference in length (F = 2.34; df = 2, 27; P > 0.05), weight (F = 1.94;df = 2, 27; P > 0.05), or venom yields among the populations of snakes in thisstudy, it is assumed that the average amount of venom injected during behavioralexperiments did not differ among populations.During these trials,mice succumbeddifferently to venom from each population of Agkistrodon contortrix. The aver-age number of seconds required for a 15–20 g mouse to become immobilizedafter a bite from the snakes differed among populations (Table 1). A one-wayANOVA confirmed that these means were significantly different from each other(F = 5.34; df = 2, 27; P < 0.05; Table 1). Therefore, it is assumed that the venomof snakes from Louisiana is most potent and that of snakes from Kansas is leastpotent.

Venom Biochemistry. Analysis of the SDS-PAGE image of the venom sam-ples (Figure 1) revealed generally similar protein compositions for the three popu-lations of Agkistrodon contortrix. Some variation in relative protein concentrationwas noted betweenvenomsamples ofKansas snakes (16kDa>29kDa>25kDa>60 kDa) and those of Louisiana and Texas (16 kDa> 25 kDa> 60 kDa> 29 kDa).



FIG. 3. Mean (+SE) of arcsine-transformed percentages of tongue flicks of three popula-tions ofAgkistrodon contortrix directed at envenomated prey items. a: Kansas,A. contortrix;b: Texas, A. contortrix; c: Louisiana, A. contortrix. Asterisks indicate significant differences(P < 0.05).



TABLE 1. COMPARISON OF AVERAGE LENGTH, WEIGHT, AND VENOM YIELD OFSNAKE TEST SUBJECTS, AND TIME TO IMMOBILIZATION OFMICE ENVENOMATED BYCOPPERHEADS (Agkistrodon contortrix) FROM KANSAS (N = 9), TEXAS (N = 13),

AND LOUISIANA (N = 8)a

Kansas Texas Louisiana

Length (cm) 49.88± 11.01 45.37± 5.97 53.54± 9.08Weight (g) 126.67± 92.87 77.69± 39.98 87.37± 27.19Venom yield (g) 0.043± 0.019 0.025± 0.015 0.034± 0.02Mouse immobilization 26.25± 11.74 39.5± 31.07 62.83± 20.28time (sec)

a Data are Means ± 1 SD.

The 16 kDa protein with the highest concentration in all populations of Agk-istrodon contortrix has been identified as Phospholipase A2 Homolog Myotoxinprecursor (SWISS-PROTaccession number P49121,MW15775Da).Nine peptidematches covered 51% of the protein sequence. The highest concentration protein(14 kDa) from the A. piscivorus sample has been identified as Phospholipase A2precursor (SWISS-PROT accession number P51972, MW 13989 Da). Eight pep-tide matches covered 44% of the protein sequence. Phospholipase A2 is a majorcomponent of most snake venoms, and numerous deleterious effects on prey havebeen noted (Forsdyke, 1995; Meier and Stocker, 1995).

Althoughmass spectra of other in-gel digested protein bands identified 19–35molecular peaks, each with a mass range of 700–3000 Da, none of these proteinswere matched to database proteins as of May 30, 2002. This result is not sur-prising because no Agkistrodon species have completely sequenced genomes atthis time. Although our analysis was not limited to the genus Agkistrodon, nohomologous sequences from other species were identified in our database search.Two venom proteins of A. contortrix, including a hemorrhagic metalloprotease at29 kDa (Johnson and Ownby, 1993) and a hyaluronidase at 61 kDa (Kudo andTu, 2001) have been isolated and characterized, but their sequences have not beendetermined and deposited in these databases. These proteins could be possiblematches with our high band intensities at 29 kDa and 60 kDa, respectively.

DISCUSSION

The results indicate that Agkistrodon contortrix can distinguish between en-venomated and nonenvenomated prey, a finding consistent with those of similarstudies that examined other viperids (Baumann, 1927, 1928, 1929; Duvall et al.,1980; Robinson and Kardong, 1991; Chiszar et al., 1992, 1999; Lavin-Murcioand Kardong, 1995). Behavioral observations confirm that A. contortrix exhibits



typical viperid poststriking behavior, which includes an increased rate of tongueflicking (Radcliffe et al., 1986), strike-induced chemosensory searching (Chiszaret al., 1992; Stiles et al., 2002), poststrike head orientation toward the departingpath of struck prey (Keith et al., 1985; Lee et al., 1988), and a tendency to strikeand hold relatively smaller prey (Kardong, 1986).

Relative to rattlesnakes, which are dietary specialists (Klauber, 1956), thecopperhead (Agkistrodon contortrix) is a dietary generalist; however, severalpopulations seem to prefer certain types of prey (e.g., Louisiana: frogs, Dundeeand Rossman, 1989; Texas: cicadas and lepidopteran larvae, Lagesse and Ford,1996; Kansas: mammals, Fitch, 1960). We do not know whether this geographicand populational variation is an artifact of prey availability or some other factor.There is no likely significance to the observation that all test subjects consumedthe first prey item they encountered at the termination of each trial—if a snake dis-covered a suitable prey item on the floor of the cage, the prey item was consumed.Additional experiments are needed to test the significance of prey consumption(but see Duvall et al., 1978). For the purposes of the current study, consumptionindicated that the behavior was predatory.

The significance of the ability of Agkistrodon contortrix to distinguish be-tween envenomated prey from different populations may not be understood untilfuture workers can demonstrate a correlation between specific venom componentsand chemoreceptive behavior. It is possible that the venom components affect thetissue of envenomated prey and that the resulting tissue damage is sensed by snakesduring chemoreceptive searching. Moreover, it is likely that there is an adaptivevalue to higher-potency venom in relation to effectiveness of prey immobilizationand digestion (Thomas and Pough, 1979; Pough and Groves, 1983) and ability ofsnakes to find envenomated prey. If prey are envenomated with a relatively po-tent venom, they travel a shorter distance before immobilization; thus, the snakesfind prey more efficiently after envenomation. This may be especially importantin mesic habitats such as Louisiana, because water exacerbates chemoreceptivetrailing of prey, at least in rattlesnakes (Brock, 1980). Because the relative po-tency of the venom from each population matches the order of preference in thechemoreception experiments, we hypothesize that the specific proteins responsiblefor the increased onset of death are also responsible (perhaps indirectly) for theincreased preference of the envenomated prey to snakes during chemoreception.The relatively weak potency of the venom in snakes fromKansas may explain whythis population could not distinguish between their envenomated prey and those ofthe other species. However, caution is needed in interpreting these results becauseof the dearth of venom samples from A. piscivorus and Sistrurus catenatus.

One question arises.Why do not all populations of copperheads have themostpotent (Louisiana) venom if it is potentially adaptive? We propose four possibleexplanations: (1) Disparate environmental effects resulting from climate or habitatmay affect venom toxicity (Russell, 1980); (2) The more potent venom in snakes



from Louisiana arose from random genetic drift in an isolated population (Thorpeet al., 1995); (3) Intraspecific variation in venom is a result of antipredator defenseagainst predator resistance to venom (Kilmon, 1976; Ovadia and Kochva, 1977;Werner and Vick, 1977; Perez et al., 1979; Rodriguez-Acosta et al., 1995; Oliveiraand Santori, 1999); and (4) Potent venom is energetically costly and may be nec-essary only in populations that need to overpower prey resistance to the venom(Perez et al., 1978, 1979; De Wit, 1982; Daltry et al., 1996; Heatwole et al., 1999;Martinez et al., 1999). Preymay vary intraspecifically in immunological resistanceor physiological susceptibility to venom (Poran et al., 1987; Mebs, 1999); there-fore caution is needed in interpreting results based on laboratory mice that likelydo not have a natural resistance to venom.

Although our venom biochemistry analysis did not identify venom proteincomponents that explain the observed differences in venom potency and behavioramong the populations of copperheads in this study, it is hoped that our preliminarydata will serve as a basis for future research into this interesting link. If individualvenom components can be isolated and injected into prey items, it may be possibleto identify the specific components that are responsible for the behavior elucidatedherein.

Acknowledgments—This work was approved by the Institutional Animal Care and Use Commit-tee of the University of Kansas (Proposal No. 76-02); we are especially grateful for the efforts of NancySchwarting, George Pisani, and Eric Rundquist. Michail Alterman permitted use of the materials andequipment of the Biochemical Research Service Laboratory. For useful comments on the manuscriptand grant proposals, we thank David Chiszar, Juan Pablo Domınguez, John Kelly, Charlotte Ownby,Jorge Porras, Christopher Raxworthy, Eric Rundquist, Norm Slade, and our colleagues in the Divi-sion of Herpetology. We thank the state agencies that issued collection permits—Kansas Departmentof Wildlife and Parks (Permit No. SC-084-2000); Texas Parks and Wildlife Department (Permit No.SPR0100-074); and Louisiana Department of Wildlife and Fisheries (Permit No. LNHP-00-057). EGthanks Ron Gutberlet, Rod Wittenberg, Alan Byboth, Neil Ford, Chad “Leaf” Keith, John Carr, SteveJensen, Mike Monlezun, Mike Harrell, Barbara Banbury, Christopher Sheil, Hugo Alamillo, OmarTorres, and Brett Benz for field assistance. This work was supported by grants to the senior author fromthe Texas Herpetological Society (James R. DixonMilleniumAward), ChicagoHerpetological Society,Kansas City Herpetological Society, and Panorama Society (Natural History Museum, The Universityof Kansas). Field equipment was kindly donated by Midwest Custom Products Inc. (www.tongs.com).

REFERENCES

BAUMANN, F. 1927. Experimente uber den Geruchssinn der Viper. Rev. Suisse Zool. 34:173–184.BAUMANN, F. 1928. Uber die Bedeutung des Bisses und des Geruchssinnes fur den Nahrungserwerb

Viper. Rev. Suisse Zool. 35:233–239.BAUMANN, F. 1929. Experimente uber den Geruchssinn und der Beuterwerb der Viper (Vipera aspis

L.). Z. Vergl. Physiol. 10:36–119.BROAD, A. J., SUTHERLAND, S. K., and COULTER, A. R. 1979. The lethality in mice of dangerous

Australian and other snake venom. Toxicon 17:661–664.



BROCK, O. G. 1980. Predatory Behavior of Eastern Diamondback Rattlesnakes (Crotalus adamanteus):Field Enclosure and Laboratory Studies Emphasizing Prey Trailing Behavior. Unpublished PhDDissertation. Florida State University, Tallahassee, Florida.

CHISZAR, D., LEE, R. K. K., SMITH, H. M., and RADCLIFFE, C. W. 1992. Searching behaviors byrattlesnakes following predatory strikes, pp. 369–382, in J. A. Campbell and E. D. Brodie Jr.(eds.). Biology of the Pit Vipers. Selva, Tyler, Texas.

CHISZAR, D., WALTERS, A., URBANIAK, J., SMITH, H. M., and MACKESSY, S. P. 1999. Discriminationbetween envenomated and nonenvenomated prey by western diamondback rattlesnakes (Crotalusatrox): Chemosensory consequences of venom. Copeia 1999:640–648.

DALTRY, J. C.,WUSTER,W., andTHORPE, R. S. 1996.Diet and snake venomevolution.Nature 379:537–540.

DE WIT, C. A. 1982. Resistance of the prairie vole (Microtus ochrogaster) and woodrat (Neotomafloridana), in Kansas, to venom of the Osage copperhead (Agkistrodon contortrix phaeogaster).Toxicon 20:709–714.

DUNDEE, H. A. and ROSSMAN, D. A. 1989. The Amphibians and Reptiles of Louisiana. Louisiana StateUniversity Press, Baton Rouge, Louisiana.

DUVALL, D., CHISZAR, D., TRUPIANO, J., and RADCLIFFE, C. W. 1978. Preference for envenomatedrodent prey by rattlesnakes. Bull. Psychon. Soc. 11:7–8.

DUVALL, D., SCUDDER, M., and CHISZAR, D. 1980. Rattlesnake predatory behaviour: Mediation ofprey discrimination and release of swallowing by cues arising from envenomated mice. Anim.Behav. 28:674–683.

FITCH, H. S. 1960. Autecology of the copperhead. Univ. Kansas Publ. Mus. Nat. Hist. 13:85–288.FONT, E. and DESFILIS, E. 2002. Chemosensory recognition of familiar and unfamiliar conspecifics by

juveniles of the Iberian wall lizard Podarcis hispanica. Ethology 108:319–330.FORSDYKE, D. R. 1995. Conservation of stem-loop potential in introns of snake venom phospholipase

A-2 genes: An application of FORS-D analysis. Mol. Biol. Evol. 12:1157–1165.GLOYD, H. K. and CONANT, R. 1990. Snakes of the Agkistrodon Complex, a Monographic Review.

Society for the Study of Amphibians and Reptiles, Oxford, Ohio.GOMEZ, H. F. and DART, R. C. 1995. Clinical toxicology of snakebite in North America, pp. 619–644,

in J. Meier and J. White (eds.). Clinical Toxicology of Animal Venoms and Poisons. CRC Press,Boca Raton, Florida.

GREENBAUM, E. 2003. A practical and safe method for the capture and restraint of venomous snakesin the field. Bull. Chicago Herpetol. Soc. 38:1–3.

HEATWOLE,H., PORAN,N., andKING, P. 1999. Ontogenetic changes in the resistance of bullfrogs (Ranacatesbeiana) to the venom of copperheads (Agkistrodon contortrix contortrix) and cottonmouths(Agkistrodon piscivorus piscivorus). Copeia 1999:808–814.

HENZEL, W. J., BILLECI, T. M., STULTS, J. T., WONG, S. C., GRIMLEY, C., and WATANABE, C. 1993.Identifying proteins from two-dimensional gels bymolecular mass searching of peptide fragmentsin protein sequence databases. Proc. Natl. Acad. Sci. USA 90:5011–5015.

JOHNSON, E. K. and OWNBY, C. L. 1993. Isolation of a myotoxin from the venom of Agkistrodoncontortrix laticinctus (broad-banded copperhead) and pathogenesis of myonecrosis induced by itin mice. Toxicon 31:243–255.

JONES, J. M. 1976. Variations of venom proteins in Agkistrodon snakes from North America. Copeia1976:558–562.

KARDONG, K. V. 1986. Predatory strike behavior of the rattlesnake,Crotalus viridis oreganus. J. Comp.Psychol. 100:304–314.

KEITH, J., LEE, R., and CHISZAR, D. A. 1985. Spatial orientation in cottonmouths (Agkistrodon pis-civorus) after detecting prey. Bull. Maryland Herpetol. Soc. 21:145–149.

KILMON, J. A., SR. 1976. High tolerance to snake venomby theVirginia opossum,Didelphis virginiana.Toxicon 14:337–340.



KLAUBER, L. M. 1956. Rattlesnakes—Their Habits, Life Histories, and Influence on Mankind. Uni-versity of California Press, Berkeley, California.

KUDO, K. and TU, A. T. 2001. Characterization of hyaluronidase isolated from Agkistrodon contortrixcontortrix (southern copperhead) venom. Archiv. Biochem. Biophys. 386:154–162.

LAGESSE, L. A. and FORD, N. B. 1996. Ontogenetic variation in the diet of the southern copperhead,Agkistrodon contortrix, in northeastern Texas. Texas J. Sci. 48:48–54.

LAVIN-MURCIO, P. A. and KARDONG, K. V. 1995. Scents related to venom and prey as cues inthe post strike trailing behavior of rattlesnakes, Crotalus viridis oreganus. Herpetologica 51:39–44.

LEE, R. K. K., CHISZAR, D. A., and SMITH, H. M. 1988. Post-strike orientation of the prairie rattlesnakefacilitates location of envenomated prey. J. Ethol. 6:129–134.

LEON, G., STILES, B., ALAPE, A., ROJAS, G., and GUTIEREZ, J. M. 1999. Comparative study on theability of IgGandF(ab&)2 antivenoms toneutralize lethal andmyotoxic effects inducedbyMicrurusnigrocinctus (coral snake) venom. Am. J. Trop. Med. Hyg. 61:266–271.

MAIN, A. R. and BULL, C. M. 1996. Mother-offspring recognition in two Australian lizards, Tiliquarugosa and Egernia stokesii. Anim. Behav. 52:193–200.

MARTINEZ, R. R., PEREZ, J. C., SANCHEZ, E. E., and CAMPOS, R. 1999. The antihemorrhagic factor ofthe Mexican ground squirrel, (Spermophilus mexicanus). Toxicon 37:949–954.

MEBS, D. 1999. Snake venom composition and evolution of Viperidae. Kaupia 8:145–148.MEIER, J. andSTOCKER,K. F. 1995.Biology and distribution of venomous snakes ofmedical importance

and the composition of snake venoms, pp. 367–412, in J. Meier and J. White (eds.). ClinicalToxicology of Animal Venoms and Poisons. CRC Press, Boca Raton, Florida.

OLIVEIRA, M. E. and SANTORI, R. T. 1999. Predatory behavior of the opossum Didelphis albiventrison the pitviper Bothrops jararaca. Stud. Neotrop. Fauna Environ. 34:72–75.

OVADIA, M. and KOCHVA, E. 1977. Neutralization of Viperidae and Elapidae snake venoms by sera ofdifferent animals. Toxicon 15:541–547.

PEREZ, J. C., HAWS, W. C., and HATCH, C. H. 1978. Resistance of woodrats (Neotoma micropus) toCrotalus atrox venom. Toxicon 16:198–200.

PEREZ, J. C., PICHYANGKUL, S., and GARCIA, V. E. 1979. The resistance of three species of warm-blooded animals to western diamondback rattlesnake (Crotalus atrox) venom. Toxicon 17:601–607.

PORAN, N. S., COSS, R. G., and BENJAMINI, E. 1987. Resistance of California ground squirrels (Sper-mophilus beecheyi) to the venom of the northern Pacific rattlesnake (Crotalus viridis oreganus):A study of adaptive variation. Toxicon 25:767–777.

POUGH, F. H. and GROVES, J. D. 1983. Specializations of the body form and food habits of snakes. Am.Zool. 23:443–454.

RADCLIFFE, C. W., ESTEP, K., BOYER, T., and CHISZAR, D. 1986. Stimulus control of predatory be-haviour in red spitting cobras (Naja mossambica pallida) and prairie rattlesnakes (Crotalus v.viridis). Anim. Behav. 34:804–814.

ROBINSON, B. G. and KARDONG, K. V. 1991. Relocation of struck prey by venomoid (venom-less)rattlesnakes, Crotalus viridis oreganus. Bull. Maryland Herpetol. Soc. 27:23–30.

RODRIGUEZ-ACOSTA, A., AGUILAR, I., and GIRON,M. E. 1995. Antivenom activity of opossum (Didel-phis marsupialis) serum fraction. Toxicon 33:95–98.

ROSENFELD, J., CAPDEVIELLE, J., GUILLEMOT,G.C., and FERRARA, P. 1992. In-gel digestion of proteinsfor internal sequence analysis after one- or two-dimensional gel electrophoresis. Anal. Biochem.203:173–179.

RUSSELL, F. E. 1980. Snake Venom Poisoning. J. B. Lippincott, Philadelphia, Pennsylvania.SOKAL, R. R. and ROHLF, F. J. 1995. Biometry, 3rd edn. Freeman, New York, New York.STILES, K., STARK, P., CHISZAR, D., and SMITH, H. M. 2002. Strike-induced chemosensory searching

(SICS) and trail-following behavior in copperheads (Agkistrodon contortrix), pp. 413–418, in



G.W. Schuett, M. Hoggren, M. E. Douglas, and H.W. Greene (eds.). Biology of the Vipers. EagleMountain Publishing, LC, Eagle Mountain, Utah.

THOMAS, R. G. and POUGH, F. H. 1979. The effect of rattlesnake venom on digestion of prey. Toxicon17:221–228.

THORPE, R. S., MALHOTRA, A., BLACK, H., DALTRY, J. C., andWUSTER,W. 1995. Relating geographicpattern to phylogenetic process. Phil. Trans. R. Soc. Lond. B 349:61–68.

WERNER,R.M. andVICK, J.A. 1977.Resistance of the opossum (Didelphis virginiana) to envenomationby snakes of the family Crotalidae. Toxicon 15:29–33.

Venom variation and chemoreception of the viperid Agkistrodon contortrix: evidence for adaptation?

Documents

Transcript of Venom variation and chemoreception of the viperid Agkistrodon contortrix: evidence for adaptation?