Recently we discovered two novel, deeply branching lineages in the domainHere, we report on the specific detection of these bacteria, the candidate phylum TG3 (Termite Group 3) and a subphylum in the phylum Fibrobacteres, by fluorescence in situ hybridization in the guts of the wood-feeding termites Microcerotermes sp. and Nasutitermes takasagoensis. Both bacterial groups were detected almost exclusively from the luminal fluid of the dilated portion in the hindgut. Each accounted for approximately 10% of the total prokaryotic cells, constituting the second-most dominant groups in the whole-gut microbiota. The detected cells of both groups were in undulate or vibroid forms and apparently resembled small spirochetes. The cell sizes were 0.2 to 0.4 by 1.3 to 6.0 m and 0.2 to 0.3 by 1.3 to 4.9 m in the TG3 and Fibrobacteres, respectively. Using PCR screenings with specific primers, we found that both groups are distributed among various termites. The obtained clones formed monophyletic clusters that were delineated by the host genus rather than by the geographic distance, implying a robust association between these bacteria and host termites. TG3 clones were also obtained from a cockroach gut, lake sediment, rice paddy soil, and deep-sea sediments. Our results suggest that the TG3 and Fibrobacteres bacteria are autochthonous gut symbionts of various termites and that the TG3 members are also widely distributed among various other environments.Termites harbor an abundance and diversity of gut bacteria, which are thought to play essential roles in the carbon and nitrogen metabolism of their host termites (4, 26). Recent culture-independent analyses have revealed that the bacterial gut microbiota comprises many termite-specific lineages that are as yet uncultured (11-13, 27, 33, 35, 39). Among them, the candidate phylum Termite Group I (TG1) was first recognized in our previous study as a novel, deeply branching lineage specific to termites (27) and later found to constitute a new phylum, together with clones from various environments (13, 15). Now, the termite-specific cluster in this candidate phylum has been partly characterized as endosymbionts of gut protists in various lower termites (28, 38), whereas no isolate exists so far from this phylum. In higher termites, which generally lack gut symbiotic protists and harbor only prokaryotes (in contrast to lower termites that harbor both), there have been found other novel, deeply branching lineages in the domain Bacteria.Using clonal analyses of 16S rRNA, we recently discovered a novel phylum-level cluster, temporarily named TG3 (Termite Group 3), and a novel subphylum-level cluster in the phylum Fibrobacteres (designated Fibrobacteres subphylum 2 in this study) from the guts of the wood-feeding higher termites Microcerotermes spp. (11). Each group accounted for approximately 10% of the analyzed clones, constituting the second-most dominant groups, together with the orders Bacteroidales and Clostridiales, following the predominant genus, Treponema. The candidate phylum...