ARTHROPOD BIOLOGY Changes in Composition of Culturable Bacteria Community in the Gut of the Formosan Subterranean Termite Depending on Rearing Conditions of the Host 1 2 3 C. HUSSENEDER, J. M. BERESTECKY, AND J. K. GRACE Ann. Entomol. Soc. Am. 102(3): 498Ð507 (2009) Downloaded from https://academic.oup.com/aesa/article/102/3/498/8634 by guest on 23 September 2021 ABSTRACT The hindgut of feeding termites that feed on wood and litter contains a diverse population of bacteria and protists that contribute to the carbon, nitrogen, and energy requirements of the termite. For understanding the ecological balance in the termite gut, detailed knowledge about the composition of the microbial gut ßora is imperative, i.e., the numbers and relative proportions of the microbial taxa and the variability in the microbial composition among different termite colonies and living conditions of termites should be described. Therefore, we isolated and enumerated eight bacterial morphotypes from the gut of the Formosan subterranean termite, Coptotermes formosanus Shiraki. Five morphotypes (three isolates of lactic acid bacteria, isolates of the family Enterobacte- riaceae and isolates belonging to the genus Dysgonomonas) were found frequently in all termite colonies. Three additional morphotypes were found sporadically and were considered to be transient ßora. We compared the proportions of the three lactic acid bacteria isolates and the Enterobacte- riaceae among three different termite colonies. Furthermore, we investigated the shift in proportions of these four major morphotypes depending on whether bacteria were isolated from freshly collected termites or from termites reared in the laboratory under seminatural conditions (in arenas on wood) or artiÞcial conditions (in petri dishes on Þlter paper). Differences in the culturable microbial composition were not signiÞcant among termite colonies, or between Þeld-collected termites and termites reared under seminatural conditions in the laboratory. However, we found signiÞcant shifts in the microbial composition between Þeld-collected termites and termites reared on Þlter paper. KEY WORDS Coptotermes, Isoptera, Rhinotermitidae, insect gut, gut symbionts Wood and litter-feeding termites (Isoptera) are of ers et al. 1982) indicate that symbionts vital for the global economic and ecological importance as decom- survival of the termite host species are within the posers of lignocellulose matter (Kambhampati and microbial community. Their roles include nitrogen Eggleton 2000). To be able to digest lignocellulose Þxation, acetogenesis, cellulose degradation, mainte- efÞciently and use lignocellulose as sole source of nance of pH and redox potential in the gut, as well as nutrition, termites harbor a morphologically and bio- preventing foreign microbes from invading (Veivers chemically diverse microbial ßora in their intestines. et al. 1982, Bauer et al. 2000, Breznak 2000). Addition- Densities of microbial populations in termite intes- ally, the bacteria composition might inßuence nest- tines are as high as 1012 per ml gut ßuid, and thus mate recognition (Matsuura 2001). similar to other herbivorous and detrivorous inverte- Because of the vital role of the gut ßora for the brates and even vertebrates (Bignell 2000). In addition termite hostÕs survival, it could be assumed that se- to protists and fungi, there is a signiÞcant community lective pressures ensure a comparable microbial com- of prokaryotes from the domains of Archaea and Eu- munity among termites of the same species, with the bacteria with densities of 109 to 1011 per ml gut ßuid most important microbe groups always present (Breznak 2000). (Schmitt-Wagner et al. 2003, Hongoh et al. 2005, Yang The majority of microbes in the termite gut are yet et al. 2005). However, not all of the gut inhabiting uncultured and unidentiÞed, and their role in termite microbes are necessarily symbionts sensu strictu (pro- nutrition is not well understood. Studies eradicating viding vital advantages to the termite host); some the termitesÕ intestinal ßora through antibiotics (Eu- microorganisms might be transient and subject to en- tick et al. 1978b, Mauldin et al. 1978) or oxygen (Veiv- vironmental factors, such as nutrition, which could lead to a considerable variation of the composition of 1 Corresponding author: Department of Entomology, Louisiana gut ßora within the same termite species. State University Agricultural Center, Baton Rouge, LA 70803. Because termites are social insects living in colonies, 2 Kapiolani Community College, Honolulu, HI 96816. 3 Department of Plant and Environmental Protection Sciences, the variation of the gut ßora within a species can vary University of Hawaii at Manoa, Honolulu, HI 96822. at different levels of social organization. Between ter- 0013-8746/09/0498Ð0507$04.00/0 ᭧ 2009 Entomological Society of America May 2009 HUSSENEDER ET AL.: BACTERIAL GUT FLORA OF TERMITES 499 mites within the same colony variation is supposed to Materials and Methods be low (Minkley et al. 2005), because colony members Termite Collection and Rearing. Workers and sol- live under the same conditions and commonly use the diers of C. formosanus were collected from traps made same nutrition source (Matsuura 2001). Additionally, from Douglas Þr, Pseudotsuga menziesii (Mirbel) the isolation of bacterial populations in the guts of Franco, in 2000 and 2001 from three collection sites, individual termites is constantly counteracted by the located on the campus of the University of Hawaii at exchange of gut ßuids containing microbes via proc- Manoa (Gilmore [G], Miller [M], and Publication todeal and stomodeal trophallaxis between colony [P]). Physical distance between collection sites mates and by the obligatory refaunation after molting ranged from 120 to 400 m (for a detailed map, see (McMahan 1969, Thorne 1997). However, termites of Husseneder and Grace 2001). Previous studies, using different colonies usually do not interact with each molecular genetic methods, established that termites other and therefore do not exchange microbial ßora. from these three collection sites belong to three in- In addition, geographically separated termite colonies dependent colonies (Husseneder and Grace 2001). Downloaded from https://academic.oup.com/aesa/article/102/3/498/8634 by guest on 23 September 2021 might be subjected to different ecological conditions From each colony, Ͼ250 workers and at least 20 sol- and use different nutrition sources. Different chemi- diers were collected. The composition of the cultur- cal components in the food might favor different mi- able gut ßora of termites from each of the three col- crobial groups becoming predominant (Mannesmann onies was assessed under three different rearing 1972). conditions. 1) To investigate the natural gut ßora, Þve The presumed variation in the composition of the workers of each colony were dissected within a few microbial gut ßora might explain why studies cultur- hours of removal from their colony in the Þeld. 2) To ing, identifying and enumerating microbial taxa differ investigate the gut ßora under seminatural conditions from each other even when focusing on the same host in the laboratory, we established a laboratory colony species (Mannesmann and Piechowski 1989, Taguchi composed of 100 workers and 10 soldiers from each et al. 1993, Husseneder et al. 2005, Ko¨nig et al. 2006). Þeld colony in arenas (25 by 25 cm) containing 100 g Attempts to track the factors causing variation in the of sand with 25 ml of distilled water added. The worker- microbial gut communities within termite species by to-soldier ratio mimics natural conditions (Haverty reviewing the literature are hindered by the fact that 1977). Termites were maintained on a diet of Douglas most authors did not clearly state whether the inves- Þr wafers, and Þve workers of each colony were dis- tigated termites have been collected from the same sected after seven days to isolate their gut ßora. 3) To colony, nor whether the termites have been freshly investigate the changes of gut ßora under artiÞcial collected from the Þeld or have been laboratory rearing conditions, we kept 100 workers and 10 sol- reared. In some studies, termites were laboratory diers in petri dish (90 mm diameter) arenas on a diet reared and fed either on wood or Þlter paper (Wenzel of Whatman no. 3 (70 mm diameter) Þlter paper et al. 2002). However, it has not yet been established moistened with distilled water. The paper was whether the bacterial composition changes with rear- changed every 3 d and water was added when needed. ing conditions and nutrition. After 7 d, the guts of Þve workers from each colony To put Þndings on microbial variability within the were dissected to isolate and culture the bacteria. same termite species into perspective, we investi- Isolation of Gut Bacteria. Five workers from each of the three colonies and each of the three rearing con- gated the variability of the composition of the cul- ditions were anesthetized by chilling on ice, the mouth turable gut ßora of the Formosan subterranean ter- and anus of the termites were sealed with parafÞn, and mite, Coptotermes formosanus Shiraki (Isoptera: the surface was sterilized with 70% alcohol. Guts were Rhinotermitidae). This termite is an invasive pest removed with sterile forceps. The whole hindgut por- species in the United States, inßicting extensive tion was separated from the midgut and homogenized economical damage in Hawaii and the southeastern in 1 ml of sterile distilled water with a sterile glass rod states. Shinzato et al. 2005 used culture independent in an autoclaved Eppendorf tube and then vortexed methods (16S rRNA gene sequencing) to describe for Ϸ4 min with sterile glass beads. Isolation and ho- the bacterial species composition in the gut of Jap- mogenization were performed under aerobic condi- anese C. formosanus as a Þrst step toward the un- tions. Ten-fold serial dilutions were made from ho- derstanding of the ecology of the gut of this termite. mogenates with distilled H2O and plated out in These authors found 49 phylotypes from 10 bacterial triplicates on Todd-Hewitt agar (BD Biosciences, phyla, including 39 novel phylotypes but did not Cockeysville, MD).