Wolbachia is a genus of obligate, intracellular bacteria in the Anaplasmatacea family of the α-Proteobacteria [37], which also includes the intracellular parasites Ehrlichia, Anaplasma, Neorickettsia, and Rickettsia [14]. Despite the phylogenetic proximity to these mammalian pathogens, Wolbachia evolved a distinct endosymbiotic lifestyle within a broad Ecdyzoan host range that spans a diversity of taxa within the phylum Arthropoda [33] and filarial nematodes [3]. Because of the limitations of studying nonculturable microbes such as obligate bacterial endosymbionts, the majority of Wolbachia research has focused on Wolbachia distribution [20, 38], cytology [10, 23, 25, 34], and the phenotypic effects of infection throughout their hosts [9, 18]. Findings from these studies indicate that Wolbachia might be among the most widespread obligate bacterial endosymbionts on the planet and that the symbiotic associations formed by these bacteria span the symbiosis spectrum from reproductive parasitism to mutualism. Much less is known about the phylogenomics and molecular evolution of this important endosymbiont. This gap in knowledge is evident by a lack of robust molecular markers for fine-scale strain typing and numerous sequence analyses of genetic drift to which other insect endosymbionts have been subjected [12, 22]. With two Wolbachia genome sequences complete [16, 40] and several other partial genomes published [28], it will become increasingly important to combine the perspectives of both phenotypic and molecular evolution approaches under a solid taxonomic framework.

Recent efforts reveal an increasing level of taxonomic diversity. In the absence of a formal species identification system, the Wolbachia community currently refers to the different lineages as supergroups and, today, the genus is divided into seven taxonomic supergroups (denoted A through G) based on phylogenetic studies of up to six chromosomal genes [21, 27]. These seven supergroups are labeled alphabetically and include A and B of various arthropods [39], C and D of filarial nematodes [1, 7], E of springtails [11, 21, 35], F of termites, weevils, true bugs, and filarial nematodes [21, 24], and G of Australian spiders [27]. Additional taxonomic diversity has recently been discovered in filarial nematodes and fleas (Casiraghi et al., unpublished data), although these have not officially been labeled supergroups. Supergroups are defined loosely in the Wolbachia literature as distinct clades that typically adhere to the criteria of greater than 3% divergence in 16S rDNA sequences [32]. However, it is becoming increasingly clear that 16S rDNA sequences do not provide an accurate measure of phylogenetic relationships and that multilocus approaches are needed to better resolve phylogenetic relationships in prokaryotes [26, 31]. We have taken a multilocus approach to characterize a putative new supergroup of Wolbachia from termites.

The order Isoptera (termites) has been relatively undersampled for Wolbachia endosymbionts in comparison to other major insect orders such as the Hymenoptera, Diptera, and Lepidoptera. The few reports of Isopteran Wolbachia infection are from Kalotermes flavicollis (family Kalotermitidae), Microcerotermes sp. (Termitidae), and Coptotermes (Rhinotermitidae) [2, 21], whose Wolbachia nucleotide sequences cluster together in supergroup F. Although termites form intimate symbiotic relationships with diverse partners such as fungi [29], protozoans [19], and bacteria [5, 6], more comprehensive surveys for obligate bacterial endosymbionts are needed. Here, we report on the detection of a new Wolbachia supergroup infecting the Pacific dampwood termites Zootermopsis angusticollis and Z. nevadensis (family Termopsidae). Termopsidae represents one of the earlier-originating lineages within the order Isoptera [15, 30]. The geographical distribution of these two social insect species is in the Neartic region, specifically in the forested areas along the Pacific coast, including the Cascade and Sierra Nevada Mountains of the United States [36]. Their mating behavior, social organization, and ecology have striking similarities; young colonies of both species can be found nesting in the same log.

Materials and Methods

Isolation, amplification, and sequencing of genomic DNA

Geno- mic DNA was extracted using the DNeasy Tissue Kit (Qiagen) from abdomens of single alate females of Zootermopsis angusticollis (collected in Huddart Park, San Mateo County, California) and Z. nevadensis (collected in Pebble Beach, Monterey, California, provided by Dr. Colin Brent). Wolbachia and host chromosomal DNA was initially amplified using polymerase chain reaction (PCR) in a volume of 10 μL (2.5 mM MgCl2 [Promega], 0.25 mM of each dNTP [Invitrogen], 0.4 μM of the forward and reverse primer, 0.2U Taq polymerase [Promega], and 1 μL DNA, 1 μL of 10X PCR buffer [Promega], and 4.36 μL water [ICN Biomedicals]). Primers and PCR product amplification are described in [21] for the Wolbachia 16S rDNA and ftsZ sequences, and in [8] for dnaA sequences. gltA and groEL amplifications were obtained with forward (WgltAF1, 5′-TACG ATCCAGGGTTTGTTTCTAC-3′, and WgroF1, 5′’-GGTGAGCAGTT GCAAGAAGC-3′) and reverse (WgltARev1, 5′-CTCATTAGCTCCA CCGTGTG-3′, and WgroRev1, 5′-AGATCTTCCATCTTGATTCC-3′) primers under the following conditions: 95°C for 2 min, followed by 95°C for 30 s, 50°C for 1 min, and 72°C for 1 min. Sequencing was performed directly and bidirectionally using purified PCR products (Promega Wizard System) and appropriate primers on an ABI 3730 automated sequencer with Big Dye v3.0 (Applied Biosystems). All new sequences were deposited in GenBank under accession numbers AY764275-AY764284.

Sequence nomenclature, alignments, and analyses

Sequences are identified by the name of the host species. Accession numbers from previously published sequences are shown adjacent to each taxon in brackets. Sequences were assembled in Sequencher 4.1.2, checked manually, and any ambiguous base calls were changed to N and were treated as missing data. Translated amino acid sequences were aligned in Clustal X, and the corresponding nucleotide sequences were manually edited in MacClade 4.05. Maximum likelihood (ML) inference methods based on nucleotide sequences were used to infer phylogenetic relationships. Prior to ML analyses, a DNA substitution model for each dataset was selected using Modeltest v3.06 and the Akaike information criterion. The following models were selected for each of the single-gene and concatenated analyses: 16S (TrN + Γ), concatenated (GTR + I + Γ), dnaA (TIM + Γ), ftsZ (TrN + I + Γ), gltA (TIM + Γ), and groEL (GTR + I + Γ). ML heuristic searches were performed using 100 random taxon addition replicates with tree bisection and reconnection (TBR) branch swapping. ML bootstrap support was determined using 100 bootstrap replicates, each using 10 random taxon addition replicates with TBR branch swapping, except for the concatenated dataset. In this case, bootstrap support was determined using 500 bootstrap replicates, each using 20 random taxon addition replicates with TBR branch swapping. Searches were performed in parallel on a Beowulf cluster using a clusterpaup program and PAUP version 4.0b10.

Results

From 3458 nucleotide base pairs spanning five Wolbachia genes per species of Zootermopsis, we describe a new taxonomic lineage that is highly divergent from the other Isopteran Wolbachia in supergroup F. We tentatively denote this lineage as supergroup H and we verify previous findings of the other six supergroups by inferring the global Wolbachia phylogeny based on a concatenated dataset of four protein-coding genes positioned across the length of the chromosome.

An initial screen of genomic DNAs from queens, kings, pseudergates (false workers), and nymphs of Zootermopsis angusticollis (n = 18) and a queen of Z. nevadensis (n = 1) yielded PCR products of the correct band size for two different Wolbachia genes (16S rDNA and ftsZ). PCR products of genomic DNA from an alate female of each species were then directly sequenced for portions of five genes spanning the length of the endosymbiont chromosome: 16S rDNA (875 bp), dnaA (356 bp), gltA (636 bp), groEL (865), and ftsZ (718 bp). These genes were selected because of the abundant sequence information that already exists for these genes in GenBank.

Maximum likelihood analyses of the 16S rDNA sequences indicated that the Wolbachia from Zootermopsis are divergent from the published Wolbachia sequences of other Isopteran species (i.e., Kalotermes, Microcerotermes, Coptotermes) [21]. The lineage grouped more closely to supergroup E (springtail hosts) than to supergroup F (termite hosts), as shown in the unrooted 16S rDNA tree of (Fig. 1) This is the first report of a termite Wolbachia sequence that is divergent from supergroup F. To confirm this evolutionary relationship, we sequenced portions of conserved orthologs from the four additional protein-coding genes listed earlier. ML analyses of these genes indicated the same pattern of divergence as the 16S rDNA analyses did (data not shown), with some incongruency for the positioning of supergroup H among the various single-gene phylogenies. Nonetheless, the branch is typically positioned close to supergroups A and E for all five genes.

Fig. 1
figure 1

Maximum likelihood phylogenetic tree constructed from portions of 16S rDNA nucleotide sequences of Wolbachia. The tree is unrooted and ML bootstrap values are shown. Names of the host arthropod species followed by an accession number denote the specific Wolbachia taxa. All supergroups (A–H) are represented by at least two strains and the divergent genetic lineages from nematodes (Dipetalonema gracile) and fleas (Ctenocephalides felis and Orchopeas leucopus) are shown. Taxa in bold lettering denote the Wolbachia lineages from termites.

Full size image

Incongruence between single-gene phylogenies are most often overcome by an analysis of concatenated single genes into one large dataset [26]. This is especially useful for bacterial phylogenetics in which recombination and lateral gene transfer can cause conflicting phylogenetic signals among single-gene analyses. We inferred a ML phylogeny based on the concatenated dataset of four protein-coding genes, spanning a total length of 2.6 kb of the Wolbachia chromosome from each Zootermopsis species. Orthologs of the Anaplasmatacea relatives Anaplasma marginale and Ehrlichia ruminantium were also included as outgroups. As shown in Fig. 2, the position of supergroup H in this analysis is between A (various arthropod hosts) and E (springtail hosts), and separate from F (other termite hosts). When taken together with the similar positioning in the 16S rDNA single-gene analysis (Fig. 1), it likely reflects that the evolutionary relationship of this branch. ML bootstrap support is below 50% in the concatenated analysis because of the varied support for this branch in the single-gene phylogenies, and resolving why the four different protein-coding genes give conflicting bootstrap support for supergroup H should be a topic of future investigation. In general, ML bootstrap percentages are typically considered the most conservative estimates of branch support, and as a result, ML bootstrap analyses are less prone to supporting a false phylogenetic history than other methods such as Bayesian posterior probabilities [13].

Fig. 2
figure 2

Maximum likelihood phylogenetic tree constructed from concatenated nucleotide sequences of four protein-coding genes, spanning 2594 base pairs. The tree is rooted with Ehrlichia and Anaplasma as outgroups; ML bootstrap values are shown. Names of the host arthropod species denote the specific Wolbachia taxa. Taxa representing each supergroup are included when at least two of the four orthologs are available. Taxa shaded in gray boxes denote the Wolbachia lineages from termites.

Full size image

Discussion

The discovery of a novel lineage of Wolbachia from termites has two implications for the biology of Wolbachia. First, it expands the scope of known genetic diversity across the Wolbachia clade and raises future interest in comprehensively surveying the order Isoptera for obligate endosymbiont infections, especially in basal species that might be undersampled and host to novel phylotypes. We are currently engaged in a systematic sampling of Wolbachia in the order Isoptera. Second, it indicates for the first time that termites as a group, like other groups of arthropods [38], can tolerate infection by more than one supergroup of Wolbachia, although individual species can harbor only single infections. This finding provides some evidence for horizontal transmission of Wolbachia into the Isoptera because the infection of two divergent Wolbachia lineages in different termite species can be parsimoniously explained by independent acquisitions of these two lineages into termites, rather than a single ancient ancestral infection in Isoptera, with subsequent divergence and/or widespread loss. However, more sampling for Wolbachia infection in termite species is needed to fully resolve this issue. The termite families of Kalotermes flavicollis (Kalotermitidae), Microcerotermes sp. (Termitidae), and Coptotermes (Rhinotermitidae) probably share a more recent common ancestor with each other than they do with the family of Zootermopsis (Termopsidae) [15], which is a pattern consistent with vertical transmission of Wolbachia through these termites.

The fast pace of supergroup discovery in the Wolbachia clade raises several important evolutionary questions, including the following: How much genetic diversity exists? What are the evolutionary relationships of the supergroups? In addition, in which hosts will new supergroups be discovered? The recent surge in supergroup discovery from arthropods such as springtails [35], termites [2], and spiders [27] strongly suggests that the major insect orders (i.e., Hymenoptera, Diptera, Lepidoptera), which have long been the subject of Wolbachia studies, might offer diminishing returns for discoveries of novel genetic lineages. Rather, undersampled groups of arthropods like the aforementioned one will be more promising to reveal further Wolbachia genetic diversity for the simple reason that they have been studied far less than the other orders.

Why is the discovery of new genetic lineages important? First, it might facilitate current phylogenetic challenges in determining the root of the Wolbachia tree. Most global trees are depicted as unrooted trees since the genetic diversity of the clade is too divergent from that of its Anaplasmatacea relatives (e.g., Ehrlichia, Anaplasma, Neorickettsia) to confidently place the outgroup taxa [21]. The discovery of the taxonomic “missing links” on the branch between the Wolbachia clade and Anaplasmatacea relatives could allow a more reliable determination of the position of the root and answer important evolutionary questions. Second, with the publications of the Wolbachia wMel and wBm genome sequences [16, 40], comparative genomic studies of Wolbachia diversity are underway to assess the genomic factors that underlie phenotypic diversity and chromosomal plasticity of this endosymbiont [4]. Information on new taxonomic lineages is a particularly helpful resource for molecular evolution and comparative genomic analyses, such as comparing gene gain and loss events, the timing of those events during the clade’s radiation, and the association of those events with the appearance of new phenotypes. Finally, the discovery of new supergroups in this dynamic symbiosis will provide fertile ground for expanding the outlook of the field and the scope and range of its research. Only a decade ago, Wolbachia was deemed a relatively obscure bacterium occurring in just a few insect species. The discovery of wider diversity (in the form of new supergroups) has had a reinvigorating effect on the research and global importance of this widespread and abundant, heritable endosymbiont.