Identification of the regions of second- Transfer of Fermentative Microbes Between
ary structure-and the choice of portions
of the protein to be included in the Generations in a Herbivorous Lizard drawing and of the sites to be labeled- are under control of the user through Abstract. Iguana iguana is herbivorous throughout life and utilizes a microbial input data. fermentation system in the elaborated hindgut to break down plant cell walls. The program is written entirely in Iguanas reared in captivity grew more slowly than wild hatchlings and failed to standard FORTRAN IV. The version develop the same complex populations offermentative microbes. Captive hatchlings that produced the line drawings in Figs. 1 fed fresh fecal material from an adult iguana acquired the complex microflora and and 2 has run on the IBM 370/165 and grew as rapidly as wild hatchlings. In the field, hatchlings actively associated with 168 and on the DEC VAX 11/780. These adults during the first weeks of life and obtained the complex microflora during this systems are available to many potential time. Acquisition of the fermentative microflora by neonatal iguanas apparently users. requires direct contact with older conspecifics. We constructed the program by ap- plication of well-known techniques [for Most herbivorous mammals satisfy hatchlings in low vegetation (1 to 4 m). I example, see (6)]. Orientations of a-heli- part of their energy requirements investigated how iguanas-reptiles that ces are found by superposition of the through the fermentative activity of spe- lack parental care-obtain the fermenta- backbone coordinates onto a standard cialized bacteria and protozoans in some tive microflora, and how important the helix oriented along the z-axis. For rib- portion of the digestive tract. These common microbe populations are for bon diagrams and strands of p-sheet, the complex populations of microbes break normal growth and utilization of plant orientation of each peptide plane is de- down plant cell wall materials, which are food. termined. For curved sheets, the midline otherwise indigestible to the herbivore, During 1979 and 1980, I hatched and of the arrow is computed by a spline fit and release wastes, including certain vol- reared iguanas in captivity (9) and found to the centers of gravity of successive atile fatty acids, which the herbivore that they always grew more slowly than peptides, and the orientation of the nor- absorbs and utilizes for energy via the hatchlings in the wild. In addition, they mal to the arrow is calculated by spline Krebs cycle (1). Young herbivorous failed to develop the same hindgut mi- fits to the direction cosines of the nor- mammals acquire their fermentative mi- croflora (10). Although it is generally mals to the successive peptide groups. croflora through contact with the moth- impossible to identify species of bacteria Whenever the program draws a (poten- er, usually by consuming fecal material. or to infer their biochemical function on tially) opaque structure, it saves the co- Protozoans and some bacteria can be the basis of appearance, many protozo- ordinates of the boundaries of one or transmitted to neonatal ruminants only ans and some bacterial forms are mor- more polygonal "windows" for use in a by very close contact, such as by the phologically unique and can be identified subsequent hidden-line removal step. consumption of feces or rumen eructate by simple microscopic examination. I ARTHUR M. LESK* (2). Neonates artificially reared in isola- found two such morphologically distinct Fairleigh Dickinson University, tion from older conspecifics fail to ac- microbes only in the hindguts of wild Teaneck, New Jersey 07666 quire all the normal species of microbes hatchlings. The first, a bacterial complex KARL D. HARDMANt and may exhibit impaired growth and or sarcina (Lampropedia merismope- Thomas J. Watson Research Center, other abnormalities (2), presumably be- dioides), also occurs in the hindguts of a IBM Corporation, P.O. Box 218, cause their fermentative capacity and variety of herbivorous reptiles, as well as Yorktown Heights, New York 10598 therefore nutrient intake are impaired. in the rumen of cattle and sheep (2, 11). Common iguanas (Iguana iguana) are The second, a large ciliate protozoan, is References and Notes unusual among lizards in that they are probably an undescribed species (12), 1. J. J. Holbrook, A. Liljas, S. J. Steindel, M. G. herbivorous throughout life (3). Iguanas but resembles the holotrichs of the ru- Rossmann, in The Enzymes, P. D. Boyer, Ed. (Academic Press, New York, ed. 3, 1970), vol. depend on a microbial fermentation sys- men (2). The occurrence of the Lampro- 11, p. 210. tem in the elaborated hindgut (4) for as pedia and the protozoan exclusively in 2. R. J. Feldmann, Atlas of Macromolecular Struc- ture on Microfiche (Tracor Jitco, Rockville, Md., 1976). much as 30 percent of their daily energy the hindguts of wild hatchlings suggested 3. See, for example, D. M. Blow, in Proteinase requirements (5). Unlike their mammali- that these animals were exposed to addi- Inhibitors, J. Fritz, H. Tschesche, L. J. Greene, an counterparts, neonatal iguanas have tional sources of fermentative microbes, E. Truscheit, Eds. (Springer-Verlag, Berlin, 1974), p. 473. no direct contact with their mothers and possibly by direct contact with other 4. The pictures of sperm whale myoglobin in the thus lack this obvious mechanism for iguanas. figures were computed from coordinates sup- plied by S. E. V. Phillips; pictures of other obtaining the specialized populations of In 1981, I conducted experimejts to proteins were computed from coordinates in the hindgut microbes. In the Gatun Lake locate sources of the morphologically Brookhaven National Laboratories protein data bank (5). 5. F. C. Bernstein et al., J. Mol. Biol. 112, 535 area of Panama, female iguanas travel to distinctive microbes and to determine, (1977). nesting sites in February, excavate under controlled conditions, whether 6. W. M. Newman and R. F. Sproull, Principles of hests, deposit their eggs, and then return possession of the more complex micro- Interactive Computer Graphics (McGraw-Hill, New York, ed. 2, 1979). - to their home ranges, which may be as flora was associated with improved 7. R. E. Dickerson, in The Proteins, H. Neurath, far as several kilometers from the nesting growth rate and digestive efficiency. I Ed. (Academic Press, New York, ed. 2., 1964), vol. 2, p. 634. site (6). The hatchlings emerge in May, used the ciliate protozoan as a marker, 8. We thank Douglas Richardson of University College, London, for advice in the early stages leave the nest area, and often settle in without assuming any useful function of of this project. This work was supported in part habitats quite distant from any adult the microbe itself. Because protozoans by grant PCM80-12007 from the National Sci- ence Foundation and by the IBM Thomas J. iguanas (7, 8). Even in those areas where have the most stringent requirements for Watson Research Center. both adults and young are found, such as transmission (2), conditions that allow * Current address: Medical Research Council, Laboratory of Molecular Biology, Hills Road, the outer edge of the forest around Gatun passage of the recognizable protozoan Cambridge CB2 2QH, England. Lake, iguanas of different ages typically between iguanas should also permit t Current address: Department of Chemistry, Harvard University, Cambridge, Mass. 02138. occupy different habitats; adults dwell in transfer of the important, but visually 27 July 1981; revised 28 December 1981 the forest canopy (15 to 30 m high) and unrecognizable, bacterial species. 540 0036-8075/82/0430-0540$01.00/0 Copyright 1982 AAAS SCIENCE, VOL. 216, 30 APRIL 1982 Newly hatched iguanas consume ap- Table 1. Acquisition of hindgut microbes and growth rates of hatchling iguanas under preciable amounts of soil within the nest experimental conditions. chamber and also lick the eggshells and Protozoa Mean growth Mn-hte each other (13). Hatchlings could obtain from fermentative microbes in this manner if Treatment N present after rate 0 to 5 weeks U test for growth rate* their mother contaminated the nest 3 weeks (mm/day) chamber soil, or the eggs, with her feces. Caged outdoors I examined over 20 iguana nests between Hatched in natural nests 6 - 0.16 - U = 15, P = .467 1979 and 1981 without detecting any fe- Hatched in captivity 8 - 0.18 U = 26, P = .525 cal material. Soil lining the nest chamber Caged indoors did not contain higher concentrations of Controls 12 - 0.16 bacteria than nearby soil did (14). Also, Caged with wild hatchling 3 + iguanas that had hatched and eaten soil Fed feces 13 + 0.22t U = 7, P = .033 in a natural nest failed to develop Lam- Wild hatchlings (1974 to 1979) Free outdoors; hatched in natural nests 12 + 0.23t U = 22, P = .025 propedia or protozoa (10) after being *Values of growth rate were compared with growth rates of controls. caged outdoors for 3 to 6 weeks (15) and values. tSignificantly different from control grew no faster than iguanas hatched and reared in captivity (Table 1). These re- sults indicate that maternal contamina- Table 2. Position of hatchling iguanas as a function of time after emergence. tion of the nest is not a significant source of fermentative microbes. Number of sightings Another potential source of microflora Age Perch height Distance from nearest adult was the air or surfaces of the outdoor (weeks) environment. Fermentative microbes Forest canopy Low vegetation Less than More than might be obtained from contamination (15 to 30 m) (O to 5 m) expected expected by other iguanas or from naturally occur- Oto 3 33 4 29 6 ring populations on plants. I reared natu- 4to6 0 16 3 13 rally hatched iguanas and iguanas x2= 38.18 x= 19.07 P<.001 P<.0O hatched in captivity in outdoor cages located where adult iguanas were rela- tively abundant (15). After 3 to 5 weeks, these groups of hatchlings had not devel- controls did (16). The feces-fed group tioned significantly closer to adults than oped the complex microflora and grew grew significantly faster than the control they would have been if they were dis- no faster than controls reared in captivi- group and at a rate similar to the average tributed randomly through the vegeta- ty (Table 1). These results suggest that rate for wild hatchlings in Panama (Table tion (20). Four to 6 weeks after hatching general environmental contamination is 1) (13). In a digestion trial comparing 5- time, the hatchlings were distributed ran- not sufficient for acquisition of the nor- week-old iguanas from the two groups domly with respect to adults. mal hindgut microflora. (17), the feces-fed hatchlings assimilated These results demonstrate that newly A logical source of fermentative mi- their food significantly better than the hatched iguanas are able to contact adult crobes was direct contact with other controls did (18). iguanas and leave their customary habi- hatchlings. Iguanas in their first weeks of These experiments demonstrate that a tats to do so. Although I did not directly life are unusually gregarious for lizards, hatchling iguana must consume fresh fe- observe interactions between hatchling often moving and sleeping in groups of cal material from an adult in order to and adult iguanas high in the forest cano- two to six or more (7, 8). I caged three acquire the hindgut microbes that best py, I identified fecal material in the stom- 20-day-old hatchlings reared in captivity promote food utilization and growth. I achs or small intestines of three of nine with a 1-month-old wild hatchling for 2 also investigated whether newly hatched wild hatchlings examined during 1981. weeks. All three of the captive hatch- iguanas in nature actually have close Hatchlings of terrestrial herbivorous liz- lings developed populations of the ciliate enough contact with adults to obtain the ard species have been observed consum- protozoan, which strongly suggested microbes in this manner. In 1981, I ing feces from adults (21). that transmission of the complete micro- searched for hatchlings in the vegetation In conclusion, hatchling iguanas must flora had occurred (Table 1). This shows along the margin of Gatun Lake (19), form temporary associations with tiem- that hatchling iguanas can transfer fer- beginning before hatching had started bers of a previous generation, in loca- mentative microbes among themselves, and continuing for the next 2 months. tions removed from both their hatching although it does not explain how they The first hatchling iguanas I found were sites and their home habitats, in order to obtain them initially. perched high in the forest canopy, usual- acquire their fermentative microflora. Hatchling iguanas, like herbivorous ly 1 to 5 m from adult iguanas. Hatch- Later, the composition of the hindgut mammals, could obtain their fermenta- lings and adults perched in a variety of microflora may be adjusted through con- tive microflora by consuming fresh feces tree species, and many trees of the same tact with other hatchlings. Acquisition of from an adult. I reared two groups of species had no iguanas in them. In the fermentative microbes may be as impor- hatchlings as before (9), but fed the first 3 weeks of the hatching season, a tant a determinant of the unusual gregari- hatchlings in one group fresh fecal mate- significant portion of the hatchlings ob- ousness of hatchling iguanas as detection rial from a caged adult iguana-about 10 served were in the forest canopy, where- of predators (7, 8). Burghardt (7) has mg daily from the third to the twentieth as 4 to 6 weeks after hatching, most likened the apparent social organization day of life. At 20 days of age, 12 of the 13 hatchlings were in low vegetation (Table of hatchling iguanas to the herding be- feces-fed hatchlings had ciliated protozo- 2). Furthermore, during the first 3 havior of dinosaurs. My work suggests ans in the hindgut, while none of the 12 weeks, hatchling iguanas were posi- an additional comparison between mod- 30 APRIL 1982 541 em and ancient herbivorous reptiles, es- 19. Iguanas were sighted on their sleeping perches Nature (London) 229, 172 (1971); J. H. Ostrum, at night by means of a 200,000-candlepower Palaeogeogr. Palaeoclimatol. Palaeoecol. 11, pecially if, as seems likely, many large hand-held light; they were observed from a boat 287 (1972); K. Troyer, in preparation. herbivorous dinosaurs also depended on passing along the lakeshore. Age, sex, perch 23. I thank A. S. Rand for technical assistance and height, and tree species were recorded, and discussions. I thank G. M. Burghardt, A. S. microbial fermentation systems. In the location was plotted on an outline map for each Rand, M. J. Ryan, and J. A. Stamps for reading iguana sighted. Distances between individuals and commenting on the manuscript and the absence of parental care, each new gen- were estimated for iguanas closer together than many temporary residents of Barro Colorado eration must actively contact a previous 10 m; greater distances were measured on the Island for discussions. Supported by a Regents maps. fellowship from the University of California, one for acquisition of fermentative mi- 20. E. C. Pielou, An Introduction to Mathematical Davis, by assistantships and other aid from the crobes. Attraction of young animals to Ecology (Wiley, New York, 1969). Smithsonian Tropical Research Institute, and by 21. D. Werner (Amblyrhynchus cristatus) and J. B. graduate research grants from the University of older conspecifics for this purpose could Iverson (Cyclura carinata), personal communi- California, Davis, and from Sigma Xi. represent the mechanism for origin of cation. 22. R. T. Bakker, Sci. Am. 232, 58 (April 1975); 30 November 1981 multigenerational dinosaur social groups, whose fossilized footprints have recently come to light (22). KATHERINE TROYER Department of Zoology, University Of California, Davis 95616, and Autoantibodies to Insulin Receptor Spontaneously Smithsonian Tropical Research Develop as Anti-Idiotypes in Mice Immunized with Insulin Institute, APO Miami, Florida 34002 Abstract. Mice immunized with insulin developed antibodies to both insulin and References and Notes the insulin receptor. The antibodies to insulin receptor displaced labeled insulin from 1. R. E. Hungate, Annu. Rev. Ecol. Syst. 6, 39 insulin receptors and mimicked the actions of insulin in stimulating the oxidation of (1975); R. H. McBee, ibid. 2, 165 (1971). glucose and its incorporation into lipids, and in inhibiting lipolysis. The antibodies to 2. R. E. Hungate, The Rumen and Its Microbes insulin receptor could be blocked by or bound to the antibodies to insulin, and (Academic Press, New York, 1966). 3. A. S. Rand, in The Ecology of Arboreal Foli- therefore were identified as anti-idiotypes. Thus, immunization against a hormone vores, G. G. Montgomery, Ed. (Smithsonian may activate spontaneously an idiotype-anti-idiotype network resulting in antibodies Institution, Washington, D. C., 1978). 4. T. S. Parsons and J. E. Cameron, iq Biology of to the hormone receptor. the Reptilia, C. Gans and T. S. Parsons, Eds. (Academic Press, New York, 1977), vol. 6, p. 189; J. B. Iverson, J. Morphol. 163, 79 (1980). Several diseases of man can be related which in turn induce anti-idiotypes that 5. R. H. McBee, in Iguanas of the World, G. M. Burghardt and A. S. Rand, Eds. (Noyes, Park to the action of autoantibodies binding to can feedback to shut off or modify the Ridge, N.J., in press). receptors on the surface of normal cells original idiotypic response. Antibodies 6. A. S. Rand, Copeia 1968, 552 (1968); G. G. Montgomery, A. S. Rand, M. E. Sunquist, ibid. of the individual. For example, patients to idiotypes can readily be prepared by 1973, 620 (1973). with Graves' disease have antibodies immunizing animals against purified idio- 7. G. M. Burghardt, Am. Zool. 17, 177 (1977). 8. , H. W. Greene, A. S. Rand, Science 195, that abnormally stimulate the thyroid types; anti-idiotypes so produced have 689 (1977); H. W. Greene, G. M. Burghardt, B. A. Dugan, A. S. Rand, J. Herpetol. 12, 169 gland by activating receptors for thyroid- been shown to influence immune reac- (1978). stimulating hormone (TSH) (1). These tions (7). However, the spontaneous 9. Iguana eggs were excavated from nest sites near Barro Colorado Island, Panama, in late April or antibodies to TSH receptor cause hyper- generation of anti-idiotypes in response early May and kept at ambient temperature in secretion of thyroid hormones and hy- to immunization against an antigen has the screened Kodak House animal room on Barro Colorado Island until they hatched. perthyroidism. Myasthenia gravis is seldom been detected (8), and therefore Hatchlings were weighed, measured, toe- caused by antibodies that bind to and it has been difficult to demonstrate a clipped, and marked with a fiber-tip pen. Marked iguanas were caged according to their treatment groups; they were given free access to block the receptor for acetylcholine at physiological function for the hypotheti- fresh Tetragonia expansa leaves; and hatchlings the neuromuscular junction (2). Another cal idiotype-anti-idiotype network. received a weekly vitamin D and calcium sup- instance of autoimmunity to receptor is Sege and Peterson (9) proposed that a plement. Cages were illuminated and heated by incandescent light bulbs on a cycle of 12 hours exemplified by those patients with anti- hormone could be used as an antigen to of light and 12 hours of darkness. bodies to the insulin receptor who suffer stimulate idiotypic antibodies that recog- 10. Weighed samples of the hindgut contents were preserved with I ml of 3 percent Formalin from severe insulin-re§istant diabetes nize the hormone. Such antibodies might buffered to pH 7 and examined at a magnifica- (3). The pathophys ology of autoimmuni- be used to induce anti-idiotypes that tion of x 150 for the presence of protozoa. Concentrations of microbes were determined from a 1:100 dilution of the same solution eNam- ty to hormone receptors is not well un- sterically fit the combining sites of the ined in a bacterial counting chamber at x600. derstood. How might an individual's im- idiotypes. Thus, the anti-idiotypes could 11. G. A. Schad, R. Knowles, E. Meerovitch, Can. mune system be triggered to produce have a three-dimensional configuration J. Microbiol. 10, 891 (1964). 12. R. E. Hungate, personal communication. antibodies that bind to seemingly normal similar to that of the hormone antigen, 13. A. S. Rand, personal communication. membrane components? and such anti-idiotypes might be able to 14. Samples were taken from the soil lining the nest chamber or from soil near but not contiguous to One explanation for receptor autoim- bind to the hormone receptor. Accord- the nest. Mean concentrations of bacteria in the munity is that it arises in the course of ingly, Sege and Peterson immunized rab- two sets of samples were: nest chamber soil (N = 6), 1.49 x 101 cells per gram; other soil (N = 3), 1.59 x 109 cells per gram; Mann-Whitney, anti-idiotypic regulation of the immune bits against rat antibodies to insulin and U= 7.5; P = .417. response. Animals may be immunized obtained anti-idiotypic antibodies that 15. Outdoor cages were constructed by surrounding against the variable regions of specific bound to rat insulin receptor. individual tree branches with nylon mesh fabric. They were located in a region of the lakeshore antibody molecules (idiotypes) and so We explored the possibility that ani- where adult iguanas were commonly seen feed- produce anti-idiotypes, antibodies di- mals immunized against insulin might ing or sleeping. Food was provided as in (9). 16. Fisher's exact probability test, P < .005. rected against idiotypes (4). Some anti- spontaneously develop antibodies to 17. Hatc4lings were caged individually and fed idiotypes recognize the combining site of their own insulin antibodies, and that weighed amounts of Tetragonia expansa leaves daily. Feces were collected daily, dried, and the idiotype (5). Thus, either aptigen or some of these anti-idiotypes might inter- weighed. Gross digestive efficiency [GDE = (food eaten - feces produced)/(food eaten) = anti-idiotypes may bind to the combining act with the insulin receptor. Our ap- (food retained)/(food consumed)] was estimated sites of idiotypic antibodies. Jerne (6) proach was to immunize mice against frQm dry weight of food eaten and feces pro- duced daily. proposed that the immune system may bovine or porcine insulins and to assay 18. For feces-fed hatchlings, GDE = 0.83 (N = 4); be regulated by a network in which anti- samples of their serum for the appear- for controls, GDE = 0.79 (N = 6). Median test, P = .024. gen induces production of idiotypes ance of two factors: antibodies to insulin 542 0036-8075/82/0430-0542$01.00/0 Copyright 1982 AAAS SCIENCE, VOL. 216, 30 APRIL 1982