Methane Oxidation in Termite Hindguts: Absence of Evidence and Evidence of Absence

A steep oxygen gradient and the presence of methane render thehindgut internal periphery of termites a potential habitat foraerobic methane-oxidizing bacteria. However, methane emissionsof various termites increased, if at all, only slightly whentermites were exposed to an anoxic (nitrogen) atmosphere, and¹⁴CH₄ added to the air headspace over live termites was notconverted to ¹⁴CO₂. Evidence for the absence of methane oxidationin living termites was corroborated by the failure to detectpmoA, the marker gene for particulate methane monooxygenase,in hindgut DNA extracts of all termites investigated. This addsrobustness to our concept of the degradation network in thetermite hindgut and eliminates the gut itself as a potentialsink of this important greenhouse gas.

INTRODUCTION

Top

Introduction

Methane is a metabolic end product in the hindgut of most termites(4, 26, 42), and it has been estimated that these insects contribute2 to 4% to the global emissions of this important greenhousegas (36, 41). Methanogenic archaea, which are easily identifiedby their coenzyme F₄₂₀ autofluorescence, have been located inseveral microhabitats within the hindgut. Depending on the termitespecies, these organisms can be associated either with the hindgutwall (13, 21, 22) or with filamentous prokaryotes attached tothe latter (21), or they can occur as ectosymbionts or endosymbiontsof certain intestinal flagellates (23, 47).

Oxygen penetrating the gut wall forms a steep gradient in thehindgut internal periphery (50 to 200 µm) (7), where itinevitably overlaps with the methane emanating from within thegut (6). These are ideal conditions for methane-oxidizing bacteria(methanotrophs), which are typically found at the oxic-anoxicinterface of, e.g., lake sediments, rice field soil, and peatbogs (8).

Since it is known that the mound walls of termites and the soilsurrounding the nest harbor methanotrophic activity (41), aconstant inoculation of termites with methanotrophs is easilyenvisaged. However, previous studies of termite hindgut metabolismeither yielded no evidence for the presence of aerobic methaneoxidation (27) or simply assumed that net emission rates ofmethane from living termites equal the gross rate of methanogenesiswithin the hindgut (e.g., 4, 37, 45). Only in one case has itbeen pointed out that the potential methane production ratecalculated for the methanogenic population in Reticulitermesflavipes exceeds the methane emission rates of the termite bymore than threefold (21), suggesting that methane oxidationmight be present.

In this study, we aimed to resolve the issue of methane oxidationwithin termite hindguts by experimentation with various representativesfrom the most species-rich termite families (18). Investigatedtermites comprised the lower wood-feeding termites Reticulitermessantonensis (Feytaud), Reticulitermes flavipes (Kollar), Zootermopsisnevadensis (Hagen), Hodotermopsis sjoestedti (Holmgren), Cryptotermessecundus (Hill), Incisitermes marginipennis (Latreille), andNeotermes castaneus (Burmeister) as well as the higher soil-feedingtermite Cubitermes orthognathus Emerson and the higher fungus-cultivatingtermite Macrotermes michaelseni Sjöstedt.

Influence of oxygen on methane emission.

Top

Influence of oxygen on...

Methane emission from living termites incubated in rubber-stopperedglass vials was measured by gas chromatography (37). All termitespecies tested emitted methane at a constant rate when incubatedunder air. Generally, flushing the headspace with nitrogen evokeda biphasic kinetic of methane emission (exemplified in Fig.1A): (i) methane accumulation rates increased slightly immediatelyafter flushing, and (ii) a continuous increase in methane emissionoccurred after approximately 15 to 20 min.

View larger version (13K):
[in this window]
[in a new window]

FIG. 1. Methane emission by living termites. (A) Time course of methane emission from Reticulitermes santonensis incubated under an atmosphere of air, which was replaced by flushing with N₂ for 1 min at the time point indicated by the arrow. Similar curves were obtained for other species (data not shown). Rates of methane emission were calculated using the slopes of the linear regression curves under air and N₂. (B) Methane emission rates by different termites under air and during the first 15 min after changing the headspace to N₂. To illustrate the contribution to the carbon mineralization flux through the termite (45), the values in Table 1 were normalized to the combined CO₂ and CH₄ emission rates under air. Bars represent the average of three independent assays; standard deviations are given.

The immediate increase of methane emission after the removalof oxygen can be explained by two different processes. First,the removal of oxygen would immediately cause a cessation ofmethane oxidation by methanotrophs, thereby increasing the effluxof methane from the gut. Second, methanogenesis itself may increaseif oxygen is removed. This has been demonstrated for Methanobrevibactercuticularis, an oxygen-consuming methanogen residing in themicrooxic gut periphery of Reticulitermes flavipes (21), withcell suspensions that were exposed to low oxygen fluxes (43).A diversion of reducing equivalents from methanogenesis towardoxygen reduction is substantiated by the finding of an F₄₂₀H₂-oxidasein Methanobrevibacter arboriphilus (38).

It is not possible to differentiate between the two processes,but even if the small rate changes upon removal of oxygen werecompletely due to methanotrophic activity, the respective methaneoxidation rates would account for at most 0.3% of the carbonmineralization flux through the termite (Fig. 1B). The delayed,continuous increase is most likely an indirect effect of methanogenesis,caused by elevated hindgut hydrogen partial pressures in theabsence of oxygen, as previously documented for Reticulitermesflavipes and Cubitermes orthognathus (12, 37), and has no furtherimplications for methane oxidation.

Although methane emission rates increased slightly but not alwayssignificantly after gassing with N₂ for the majority of thetermites investigated (Fig. 1B), the results are somewhat ambiguouswhen data are compared at the level of the individual assay(Table 1). Generally, there was a strong variance between assays,and in the case of Cryptotermes secundus and Hodotermopsis sjoestedti,several batches even showed a decreased methane emission afterthe removal of oxygen, corroborating that methane oxidation,if present at all, is only marginal.

View this table:
[in this window]
[in a new window]

TABLE 1. Respiration rates and methane emission rates of termites incubated under air and under N₂^a

In vivo incubation with ¹⁴CH₄.

Top

In vivo incubation with...

The absence of methane oxidation was further supported by theresults of a radiotracer study with Reticulitermes flavipesand Cubitermes orthognathus, representing lower and higher termites,respectively. Living termites were incubated for several daysunder an atmosphere of air supplemented with ¹⁴CH₄, using thesame setup as above. Shifts in the amounts of ¹⁴CH₄ and ¹⁴CO₂were monitored over time by a gas chromatograph equipped witha methanizer and a radioactivity monitor (44). For both termitespecies, neither a decrease of ¹⁴CH₄ nor a formation of ¹⁴CO₂was observed (Fig. 2), in the case of R. flavipes not even after1 week of incubation, which provides sound evidence that methaneoxidation is not taking place. The successful application ofthis strategy in a similar experiment with Nasutitermes walkeri(Hill), where ¹⁴CO₂ supplemented to the headspace was subsequentlyconverted to [¹⁴C]acetate in the hindgut (49), documents thata diffusion limitation of tracheal gas transport into the hindgutcan be excluded.

View larger version (13K):
[in this window]
[in a new window]

FIG. 2. Incubation of living termites under an atmosphere of air supplemented with ¹⁴CH₄. Data shown for Cubitermes orthognathus (closed symbols) represent means of two independent assays; the bars showing mean deviation are covered by the symbols. Data shown for Reticulitermes flavipes (open symbols) stem from a single experiment.

Screening for methane-oxidizing microorganisms.

Top

Screening for methane-oxidizing...

Evidence for the absence of methane oxidation in living termiteswas corroborated by testing all termites studied for the presenceof methanotrophs, using the pmoA gene, encoding the ${alpha}$ subunitof the particulate methane monooxygenase, as a molecular marker.Primers targeting the pmoA gene of most methanotrophs (14) readilyamplified pmoA genes from littoral lake sediment and the pmoAclones obtained from a previous study (34). However, they didnot yield PCR products of the expected length (531 bp) withDNA extracted from termite hindgut contents (33), using thesame protocol. In the case of Reticulitermes santonensis, aproduct of similar length was obtained, but cloning and sequencingrevealed that it did not represent pmoA genes. Control experiments,where the same DNA extracts and amplification reagents wereused to amplify 16S rRNA genes or fhs genes (PCR conditionswere as described in references 50 and 33, respectively) alwaysresulted in PCR products of the expected length, proving thata general inhibition of the PCRs can be excluded.

The absence of methane-oxidizing bacteria from termite gutsis further supported by the extensive bacterial 16S rRNA geneclone libraries obtained from several Reticulitermes and Microcerotermesspecies (15, 16, 50). Among a total of 2,500 clones obtainedfrom these termites, only a single clone (clone RS-K73 fromReticulitermes speratus) showed a loose phylogenetic affiliationwith methanotrophic isolates (96 to 97% sequence identity withMethylocella species). In addition, none of the transmissionelectron micrographs of hindgut thin sections of lower and highertermites shows bacterial cells with structures resembling thetypical internal membrane systems of methanotrophs (5, 10, 46,51).

To exclude also the possibility of anaerobic methane oxidation,we conducted two PCR assays that specifically amplify the mcrAgenes of ANME-1 or ANME-2 archaea (M. Krüger and M. W.Friedrich, unpublished data). Although DNA extracts of microbialmats, which showed anaerobic methane oxidation activity (25),always resulted in a specific PCR product, no mcrA genes wereamplified from termite hindgut contents. Again, support forthe absence of ANME archaea is provided by numerous archaeal16S rRNA gene clone libraries obtained for various lower andhigher termites (11, 13, 28-31, 39, 47). None of the clonesin these libraries clusters with ANME-1, ANME-2, or ANME-3 archaea(phylogeny is summarized by Meyerdierks et al. [24]) or ANMEarchaea involved in anaerobic methane oxidation coupled to denitrification(35).

Ammonia as a potential inhibitor.

Top

Ammonia as a potential...

There may be numerous reasons for the absence of methanotrophsfrom termite hindguts, e.g., the high nutrient concentrations(27, 45), which are unfavorable for many methanotrophs (48),or even the presence of specific inhibitors. A concrete factorresponsible for this phenomenon, however, may be ammonia, whichis readily cooxidized by methane monooxygenase (1) and is thoughtto affect methane oxidation by competitive inhibition (2).

Ammonia concentrations in termite hindguts were found to correlatewith the feeding guild of the respective termite (Table 2).In the hindgut paunch of all wood-feeding termites, total ammoniaconcentrations were in the lower millimolar range, which isreported to partially inhibit methane oxidation even at highconcentrations of methane (3, 9, 19, 20, 32, 48).

View this table:
[in this window]
[in a new window]

TABLE 2. Total ammonia concentrations in the hindgut paunch of wood-feeding termites and the major hindgut compartments of fungus-cultivating and soil-feeding termites

In the fungus-cultivating termite Macrotermes michaelseni andthe soil-feeding termite Cubitermes orthognathus, total ammoniaconcentrations in the individual gut sections were well abovethe concentrations found with wood-feeding termites (Table 2),with the highest values in the methane-producing posterior hindgutsections of C. orthognathus (37). Such concentrations completelyinhibited methanotrophs in other studies (48). Curiously, thenest material of soil-feeding termites also contains high ammoniaconcentrations (17), and the determination of stable carbonisotope fractionation indicated that about 70% of the methaneproduced in the nest is oxidized within the mound wall (41).

In summary, the comprehensive evidence provided in this studydocuments (i) that methane is not oxidized in the termite hindgutand (ii) that methane-oxidizing bacteria are absent, irrespectiveof the phylogenetic status and feeding guild of the host. Theresults eliminate the gut itself as a potential sink of thisimportant greenhouse gas and add robustness to our concept ofthe degradation network in the termite hindgut.

ACKNOWLEDGMENTS

We thank Judith Korb, Roland Brandl, and Horst Hertel for providingtermites, and we thank David Kamanda Ngugi, Hamadi Boga, SybilleFrankenberg, and Mahesh Desai for DNA preparations. The excellenttechnical assistance of Katja Meuser is gratefully acknowledged.

FOOTNOTES

* Corresponding author. Mailing address: Max Planck Institute for Terrestrial Microbiology, Karl-von-Frisch-Strasse, 35043 Marburg, Germany. Phone: (49) 6421 178701. Fax: (49) 6421 178709. E-mail: brune{at}mpi-marburg.mpg.de.

Published ahead of print on 19 January 2007.

Present address: Altana Pharma Deutschland GmbH, Moltkestrasse4, 78467 Konstanz, Germany.

REFERENCES

Top