Genesis of Acetate and Methane by Gut Bacteria of Nutritionally Diverse Termites Alain Brauman, Matthew D
Total Page：16
File Type：pdf, Size：1020Kb
Genesis of Acetate and Methane by Gut Bacteria of Nutritionally Diverse Termites Alain Brauman, Matthew D. Kane, Marc Labat, John A. Breznak* The evolution of differentfeeding guilds in termites is paralleled by differences in the activity of their gut microbiota. In wood-feeding termites, carbon dioxide-reducing acetogenic bacteria were found to generally outprocess carbon dioxide-reducing methanogenic bac- teria for reductant (presumably hydrogen) generated during microbial fermentation in the hindgut. By contrast, acetogenesis from hydrogen and carbon dioxide was of little signif- icance in fungus-growing and soil-feeding termites, which evolvedmore methane than their wood- and grass-feeding counterparts. Given the large biomass of termites on the earth and especially in the tropics, these findings should help refine global estimates of carbon dioxide reduction in anoxic habitats and the contribution of termite emissions to atmo- spheric methane concentrations. Although generally recognized for their 3 CH3COOH + 6 02+ 6 COZ + 6 HzO ability to thrive on a diet of wood, the feeding behavior and nutritional ecology of. Both H, and CH, (the latter formed by termites is quite diverse and not limited to reduction of CO, by methanogenic bacte- xylophagy. Some species forage for grass ria) are also emitted by termites, and the and leaves, whereas others feed exclusively extent to which termites contribute to glob- on soil, presumably deriving nutrition from al increases in atmospheric CH, has been the humic compounds therein (1, 2). Still hotly debated (8,14). However, emission others cultivate and consume cellulolytic of these gases represents only a small part of fungi, which, when ingested with plant reduction equivalents (H+ + e-) generated materials, augment the digestive enzymes of‘ by microbial fermentation in the hindgut of the insect (3). Nevertheless, all known R. jhips (7). In limited studies of other termites have a dense and diverse hindgut wood-eating “lower” and “higher” termites microbial community, which aids in diges- (the latter of which contain only bacteria in tion and which is the source of fermenta- their hindguts), a similar pattem was ob- tion products such as acetate, methane served (6). This in itself was surprising, (CH,), and H, (4). because, in most anoxic habitats low in The symbiotic hindgut microflora of sulfate and nitrate, CO2 reduction to CH, wood-eating, “lower” tekites (for exam- (not acetate) is the dominant H,-consum- ple, Reticulitennes flauips) includes protozoa ing process (15). Therefore, to determine and bacteria and effects an essentially ho- whether bacterial acetogenesis, rather than moacetic fermentation of wood polysaccha- methanogenesis, was the major H,-con- ride (principally cellulose) consumed by the suming “electron sink” reaction of hindgut insect. Cellulolytic protozoa first hydrolyze fermentation of termites in general, we cellulose and ferment each glucose mono- examined a variety of tropical species rep- mer to acetate, carbon dioxide (CO,), and resenting different feeding guilds (different H, (5): pattems of food resource preference)- Be- cause opportunities to collect fresh speci- C6H1206 2 HzO + + mens of many of the species (especially 2 CH,COOH + 2 COZ+ 4 H2 those from remote regions) were rare, our then CO,-reducing acetogenic bacteria sampling strategy in the time available was convert H, and CO, to an additional ace- to maximize species diversity within a par- tate molecule (6): ticular feeding guild rather than to sample repeatedly a given species from different 4 H2 2 COZ + CH3COOH 2 H2O + + sites for replicate analyses. Included in this The three net acetates formed per glucose study were wood-feeding members of three monomer are absorbed from the hindgut and families of “lower”.termites (Hodo-, Kalo- oxidized by the termite to support up to 100% and Rhinotermitidae), and wood-, grass-, of the insect’s respiratory requirement (7): and soil-feeding and fungus-growing repre- sentatives of the higher termite family Ter- A. Brauman and M. Labat, Laboratoire de Microbiol- ogie, Office de la Recherche Scientifique et Technique mitidae, which includes about three-quar- d‘outre-Mer (ORSTOM), Université de Provence, 1331 ters of all known termite species (1, 2). Marseille Cedex 3, France, and Laboratoire de Micro- We quantified acetogenesis from CO, by biologie, ORSTOM, Brazzaville, Republic of Congo. M. D. Kane and J. A. Breznak, Department of Micro- measuring the reduction of 14C0, to I4C- biology and Center for Microbial Ecology, Michigan acetate by anoxic suspensions of termite gut State University, East Lansing, MI 48824. contents. This was done under two condi- ‘To whom correspondenceshould be addressed. tions: (i) in the presence of exogenously 1384 SCIENCE * VOL. 257 4 SEPTEMBER 1992 supplied H, and (ii) with reductant (pre- the case (7). Moreover, we have measured ly, because of limited time and supplies in sumably H,) produced endogenously by gut rates of L4C0,reduction to ',CH, by an- the field, we were unable to determine rates microbes present in the suspension (6). By oxic gut contents from the wood-feeding R. of ''C0, reduction to ',CH, by gut con- contrast, CO,-reducing methanogenesis by flaplipes, 2. angusticollis, and N. nigriceps (6) tents from other soil-feeding species or from the gut microbiota was usually estimated as (see also Table l), as well .as from M. fungus-growing species. CH, emission from live termites during purplus, N. lujae, and the soil-feeding C. It is not surprising that animals with brief (2 to 4 hours) incubation in stoppered speciosus (20). In all cases, such rates were anaerobic, fermentative microbial commu- bottles (7). This latter technique is sensi- less than or equal to CH, emission by live nities in their alimentary tract evolve CH,. tive and noninvasive and minimizes disrup- termites, even when the gut contents were A classic example is a bovine animal, tion of the insects and their gut microbiota, supplied with exogenous H,. Unfortunate- whose rumen microbiota evolves up to 200 and it was well suited to measurements in the field (16). Rates of acetogenesis from CO, (with Table 1. Rates of H,P4C0, acetogenesis by termite gut contents and CH, emission by live termites endogenous H,) for 14 wood-feeding ter- of different feeding guilds. Units are micromoles of product per gram of termite per hour. The origin mites and one grass-feeding species were, and condition of termites before the assay are as indicated in (77). The first six species listed are on average, three times those of CH, emis- "lower" termites, the others are "higher" termites [see (7)]. The standard assay system has been sion (Table 1) (17, 18). However, this described in detail (6) and is only summarized here. Guts from 20 to 60 worker termites were removed in an anaerobic chamber and were pooled in an anoxic, buffered salt solution before assay condition may seriously underestimate homogenization. Reaction vials (8-ml) had a final liquid volume of 0.5 ml and contained 1.2 pmol of in situ rates of acetogenesis from CO,, NaHI4CO, (specific activity, -6.5 x 1O4 dpmlpmol) and the equivalent of two to four homogenized because homogenization and dilution of gut termite guts. The atmosphere in the reaction vials consisted of 100% N, (for determination of rates contents probably disrupt important physi- of I4C-acetateformation from 14C0, by endogenously produced H2) or 100% H,. After termination cal interactions between H,-producing mi- of the reaction, the supernatant fluid was analyzed for 14C-labeledproducts by high-performance crobes and H,-utilizing acetogenic bacteria liquid chromatography. Modified assays, performed with gut homogenates of R. flavipes incubated with 52 mM NaH14C03in the liquid phase and 20% 14C0,/80% N, (or 80% H,) in the gas phase, that would otherwise occur in situ (19). gave results virtually identical to those tabulated for the standard assay system. Results are mean Not surprisingly then, rates of acetogenesis values of duplicate reactions of samples from the same pooled gut homogenate for n = 1 from CO, by wood-feeding termites usually homogenate, except for the following species (for which the data are mean values of duplicate increased to more than ten times those of reactions for nas indicated): R. flavipes, n = 20; Z. angusticollis, n = 3; M. parvus, n = 3; N. lujae, CH, emission when acetogenesis was mea- n = 2; C. albotarsalis, n = 2; C. speciosus, n = 3. Values for R. flavipes, P. simplex, Z. angusticollis, sured in the presence of exogenously sup- N. costalis, and N. nigriceps were published as portions of a separate study (6) and are included plied By contrast, for both fungus- here for comparison. The rate of H,/CO, acetogenesis reported here for R. flavipes is slightly lower H,. than the value reported previously, which was based on~n= 6 (6).Assays of C. cavifronswere done growing and soil-feeding termites, rates of by J. Klenz with J.A.B. during a summer course in microbial diversity at the Marine Biological CH, emission were always greater than Laboratory, Woods Hole, Massachusetts. For results where n 2 3, results are mean values c rates of acetogenesis from CO,, even when standard deviation (78). the latter process was measured in the presence of exogenously supplied H,. 14C-acetate Differences in acetogenesis and methan- Termite CH, ogenesis activity between termites of differ- Exogenous Endogenous emitted* ent feeding guilds were also apparent. Rates H, i H, of CO, reduction to acetate by gut contents Wood-feeding termites from wood- and grass-feeding termites (.uti& Coptotermes formosanus 1.66 0.10 0.01 or without exogenously supplied H,) were Cryptotermes cavifrons 1.34 0.58 0.00 greater than those of fungus-growing or Prorhinotermes simplex 1.18 0.57 0.45t Pterotermes occidentis 2.07 0.48 0.00 ' soil-feeding termites (Table 1) (18). By Reticulitermes flavipes 0.93 + 0.43 0.09 2 0.06 0.10 contrast, rates of CH, emission by soil- Zootermopsis angusticollis 0.33 2 0.25 0.07 2 0.02 1.30 feeding and, to a lesser extent, fungus- Amitermes sp.