Skip to main content
Download PDF

Unveiling lignocellulolytic potential: a genomic exploration of bacterial lineages within the termite gut

Microbiome volume 12, Article number: 201 (2024) Cite this article

Abstract

Background

The microbial landscape within termite guts varies across termite families. The gut microbiota of lower termites (LT) is dominated by cellulolytic flagellates that sequester wood particles in their digestive vacuoles, whereas in the flagellate-free higher termites (HT), cellulolytic activity has been attributed to fiber-associated bacteria. However, little is known about the role of individual lineages in fiber digestion, particularly in LT.

Results

We investigated the lignocellulolytic potential of 2223 metagenome-assembled genomes (MAGs) recovered from the gut metagenomes of 51 termite species. In the flagellate-dependent LT, cellulolytic enzymes are restricted to MAGs of Bacteroidota (Dysgonomonadaceae, Tannerellaceae, Bacteroidaceae, Azobacteroidaceae) and Spirochaetota (Breznakiellaceae) and reflect a specialization on cellodextrins, whereas their hemicellulolytic arsenal features activities on xylans and diverse heteropolymers. By contrast, the MAGs derived from flagellate-free HT possess a comprehensive arsenal of exo- and endoglucanases that resembles that of termite gut flagellates, underlining that Fibrobacterota and Spirochaetota occupy the cellulolytic niche that became vacant after the loss of the flagellates. Furthermore, we detected directly or indirectly oxygen-dependent enzymes that oxidize cellulose or modify lignin in MAGs of Pseudomonadota (Burkholderiales, Pseudomonadales) and Actinomycetota (Actinomycetales, Mycobacteriales), representing lineages located at the hindgut wall.

Conclusions

The results of this study refine our concept of symbiotic digestion of lignocellulose in termite guts, emphasizing the differential roles of specific bacterial lineages in both flagellate-dependent and flagellate-independent breakdown of cellulose and hemicelluloses, as well as a so far unappreciated role of oxygen in the depolymerization of plant fiber and lignin in the microoxic periphery during gut passage in HT.

Video Abstract

Background

Termites are eusocial insects that effectively digest wood and other lignocellulosic plant matter [1]. Their degradation mechanism involves a dual system. The termite comminutes the wood particles and degrades amorphous regions of the cellulose with endoglucanases and β-glucosidases secreted in the salivary glands or the midgut [2, 3], and symbiotic microorganisms in the dilated hindgut subsequently degrade the wood particles that pass from the midgut into the hindgut largely intact [4]. The contributions of microbial symbionts to fiber digestion are essential as the host lacks both exoglucanases acting on crystalline cellulose and any hemicellulolytic activities.

The microbial community in the hindgut of termites is highly specialized and differs between lineages [5]. Flagellates dominate the microbiota of most termite families (collectively referred to as lower termites, LT), where they sequester wood particles from the hindgut fluid into digestive vacuoles containing both endoglucanases and exoglucanases [55], suggesting that bacteria are of little relevance in the digestion of plant fiber for these termites [5]. In the family Termitidae (higher termites, HT), the loss of cellulolytic flagellates is compensated by the cellulolytic activity of their bacterial microbiota [6, 7]. In the subfamily Macrotermitinae, which cultivate basidiomycete fungi of the genus Termitomyces that degrade (hemi)cellulose and depolymerize lignin [8], the need for degradation during gut passage is alleviated by the pretreatment of lignocellulose in fungal gardens [9].

Therefore, the breakdown of recalcitrant fibers in the hindgut of most HT is an entirely bacterial process [2]. Metagenomic studies revealed the presence of carbohydrate-active enzymes (CAZymes) that are often closely related to those of cellulolytic clades from the rumen of cattle or other environments [10,11,12]. However, these studies covered only a few species of higher termites, and from the arsenal of glycoside hydrolases (GHs), only a few cellulases (GH5) of unidentified hindgut bacteria and hemicellulases (GH10 and GH11) of Treponematales have been characterized [12,13,14].

Apart from a recent study, which analyzed the CAZymes in a wide range of metagenomes from many termite families and documented considerable differences in the CAZyme repertoire of LT and HT [15], little is known about the fiber-digesting potential of specific bacterial lineages of the gut microbiota in LT. Most importantly, only few of the CAZymes identified in the metagenomic datasets have been pinpointed to specific bacterial lineages using metagenome-assembled genomes (MAGs) [10, 16, 18]. Moreover, most previous analyses focused on GHs but did not discuss other CAZyme families acting on lignocellulose, such as carbohydrate-binding modules (CBM) and auxiliary activities (AA). CBMs are enzyme domains involved in the attachment of GHs to their respective substrate, whereas some AAs participate in redox reactions that depolymerize (hemi)celluloses and lignin and are directly or indirectly dependent on molecular oxygen as co-substrate [17]. Certain carbohydrate esterases (CE) remove methyl and acetyl groups from the polysaccharide backbone of hemicelluloses and, together with polysaccharide lyases (PL), also contribute to pectin degradation. Glycosyl transferases (GT) are mostly involved in the biosynthesis of glycoproteins and glycolipids [18,19,20].

Here, we present a comprehensive genome-centric analysis of the CAZymes encoded by more than 2000 bacterial MAGs reconstructed from gut metagenomes of 51 termite species [21, 22]. Using a vast genomic library, covering nearly all termite families, we identify the distribution of CAZymes among all major bacterial lineages previously detected in termite guts. We focus on GHs, AAs, and CBMs with lignocellulolytic activity among MAGs from individual bacterial lineages from different groups of termites, for the first time including also oxygen-dependent activities.

Methods

Metagenome-assembled genomes and phylogenomics

MAGs were generated in previous metagenomic studies from our group [21, 22]. Metadata associated with the corresponding metagenomes can be found in Supplementary Table S1. The quality of the reconstructed genomes was estimated with CheckM v1.1.6 [23]. MAGs that were at least 50% complete and less than 10% contaminated were retained for analyses. Relative abundances were calculated as previously described [22]. MAGs were classified using the Genome Taxonomy Database (GTDB) with the GTDB-Tk v2 toolkit [24], based on an alignment of 120 ubiquitous marker genes retrieved from the MAGs, and phylogenies were reconstructed using IQ-TREE [25]. To validate genome classifications, taxonomies from Pplacer v1.1.17 [26] and CheckM were taken into consideration. Metadata associated with each MAG can be found in Supplementary Table S2. The termite phylogeny is based on previous work [15, 27, 28], using a combination of rRNA genes and mitochondrial genomes. All maximum-likelihood phylogenetic trees were curated using the packages ape [29] and PAML [30]. Genomic data were mapped to phylogenies using iTOL [31] and R package ggtree [32].

Gene prediction and annotation

Genomes were subjected to gene prediction using Prodigal [33], with 4,480,144 genes predicted. Coding densities were also estimated by the software, by averaging the total CDS bases as a function of the total genome bases and expressed in percentages, and can be found in Table S2. Protein domain annotations were added using the TIGRFAM [34], Pfam [35], SUPERFAMILY [36], Panther [37], and CDD [38] databases in the INTERPROSCAN [39] classifier, with default parameters. CAZymes were identified using Hidden Markov Model (HMM) searches (e-value < 1e-15, score > 100, and coverage > 0.35) against the dbCAN3 and dbCAN-sub metaservers [40], using the function run_dbcan. The classification was confirmed with protein domain annotations and DIAMOND [41] searches (e-value < 1e-15, id > 50%) against the CAZy database [42]. Signal peptides were identified with SignalP V.6 [43] and Phobius [44].

Multivariate analyses and cluster visualization

A dataset including all high-quality CAZyme genes with secretion signals was subjected to uniform manifold approximation and projection (UMAP), using the R package umap [45]. The k-nearest neighbors (KNN) algorithm was used with parameters for 15 neighbors and 500 epochs for training. For the input, the raw-count data were square-root transformed and Wisconsin double standardized with the function decostand, and a Bray–Curtis dissimilarity matrix was calculated using the function vegdist; both functions are from the R package vegan [46]. The analysis included host family and subfamily, termite diet, gut compartment, and bacterial phylum as metadata. The contribution of these variables to CAZyme composition was further estimated with the functions ADONIS and ANOSIM, also based on Bray–Curtis dissimilarities, each with 1000 permutations and bootstrapped 100 × for best model fit (max stress = 0.12) in the vegan package. Principal component analyses (PCA) were used as implemented in the function prcomp of the R package stats [47]. Explanatory PCs were employed in rarefactions for the relationship between CAZyme family distributions in relation to either the per-MAG number of genes predicted or metagenomic relative abundance.

CAZyme family selection and statistical analyses

Outliers were removed, and homogeneity of variance and normality was checked using Levene’s and Shapiro Test from the R packages car [48] and stats. Since neither assumption was applicable, a nonparametric Kruskal–Wallis test was used on the entire pre-selection CAZyme dataset to determine significantly different CAZyme families across the 15 bacterial phyla, with the function kruskal.test and with post hoc Mann–Whitney U tests in the function wilcox_test of the package rstatix [49], followed by Bonferroni multiple testing correction at a threshold of p < 0.0001, using the p.adjust function of the stats package. A total of 146 CAZyme families tested positive for association with the phyla, of which those with lignocellulolytic activities (38 CAZyme families containing 8 CBMs) were selected for downstream comparison.

The classification of individual CAZymes into the major lignocellulolytic categories (cellulase, hemicellulase, ligninase, chitinase, and pectinase) was based on the gene-wise presence and complementary activities using the substrate predictions of DBCAN-sub (see Table S3). Enzymes with only endo-β-1,4-glucanase activities or other classically cellulolytic activities, such as cellobiohydrolases, were classified as cellulases, whereas enzymes with endo-β-1,4-glucanase activity and additional xylanolytic or xyloglucan-specific activities were classified as hemicellulases. Enzymes from a given CAZyme family with hemicellulolytic activities were considered hemicellulases even if they comprised endo-β-1,4-glucanase activity. Enzymes from family GH3 were not included in this categorization given the complexity of their predicted substrates. CBMs that were not connected to CAZymes assigned to specific GH families were removed from the dataset. CAZymes from families with peroxidase, aryl alcohol oxidase, and/or laccase activities were classified as ligninases. Pectate lyases and pectin methylesterases were classified as pectinases. Enzymes with acetylglucosaminosidase, chitin deacetylase, and chitinase activities were classified as chitinases.

To confirm bacterial phylogeny as a covariate, we employed a local indicator of phylogenetic association (lipaMoran) analysis coupled with an Abouheif test. CAZyme families containing (hemi)cellulases and lignin-modification enzymes were used as traits, and the phylogenetic signal was considered at p-values < 0.05. Correlograms were plotted with the R package phylosignal [50]. Since there was no considerable coding density variation between MAGs, gene densities (GD) were estimated by normalizing the numbers of genes in the lignocellulolytic CAZyme classifications by per-MAG number of predicted genes. Specific Wilcoxon’s signed-rank tests were used to assess the association of GD in HT and LT, and the effect sizes were estimated using the function wilcox_effsize of the package rstatix. Correlation between continuous variables was tested with Spearman correlation rank, as implemented by the function cor.test in the R package stats.

Results

Representation of bacterial MAGs and phyla in termite microbiomes

The initial dataset comprised 2223 bacterial MAGs that were recovered from 51 termite species, spanning seven families of lower termites (LT) and seven subfamilies of higher termites (HT) (Fig. 1). Archaeal MAGs were excluded because preliminary analyses revealed negligible numbers of CAZymes. The MAGs represented 22 bacterial phyla, with the majority classified among Bacillota (24.1%), Bacteroidota (15.6%), Spirochaetota (14.2%), and Pseudomonadota (13.4%).

Fig. 1
figure 1

Distribution of MAGs from major bacterial phyla among metagenomes from different host groups. The ordinate shows the number of MAGs in each phylum as a heatmap that is based on the total number of MAGs (2,223) in the dataset. The abscissa shows the number of metagenomes from which the MAGs were recovered, summarized for each host group, as a heatmap based on the total number of metagenomes (51). Relative abundance values indicate the proportion of reads in a metagenome that mapped against the corresponding MAGs, expressed as averages per termite family (lower termites) or subfamily (higher termites, Termitidae). The termite species contained in each host group, the relative abundances of individual MAGs, and their classification down to genus level are given in Tables S1 and S2

Full size image

The MAGs from different bacterial phyla differed strongly in relative abundance, ranging between 0.004 and 19.8% (average 0.31%) of the reads in the corresponding metagenomes (Fig. 1). Certain phyla (e.g., Elusimicrobiota and Verrucomicrobiota) were more prevalent among lower termites (LT), whereas others were more prevalent (Pseudomonadota) in or even restricted to (Fibrobacterota) higher termites (HT). Many phyla were represented only in certain (sub)families of LT and HT but virtually absent in others (e.g., Actinobacteriota, Bacteroidota, Desulfobacterota, Spirochaetota, and Patescibacteria). The bacterial phyla Chlamydiota, Cyanobacteriota, Myxococcota, Cloacimonadota, Chloroflexota, Fusobacteriota, and Deferribacteria presented less than 5 MAGs and were removed from subsequent analyses.

The remaining 2204 MAGs from 15 bacterial phyla were analyzed for CAZyme distribution. From the full dataset (Table S5), we selected 38 CAZyme families with predicted secretion signals that are involved in lignocellulose degradation (Table S3) and were differentially distributed across phyla (p ≤ 0.001) (Fig. 2). They comprised 18 families with cellulolytic, 13 families with hemicellulolytic, and 7 families with lignin-modifying activities (Table S3), possessing additionally 8 CBM domain families. CAZyme families that are not directly involved in lignocellulolytic processes, such as glycosyltransferases (GT) and polysaccharide lyases (PL), were excluded from the analysis. Carbohydrate esterases (CE) of subfamily CE4 were represented in almost all MAGs (Table S5), but since their predicted substrates did not include acetylxylan, they were also excluded.

Fig. 2
figure 2

The distribution of carbohydrate-active enzymes with lignocellulolytic activities (cellulase, hemicellulase, auxiliary) and cellulose-binding modules (CBM) among the MAGs of major bacterial families recovered from termite gut metagenomes. The heatmap indicates the mean gene abundance in selected CAZyme families for the MAGs from the respective bacterial family. CAZYme families were clustered using a hierarchical clustering algorithm; values were scaled from 0 to 1 to facilitate visualization. An interactive spreadsheet with details for all MAGs is given in Table S4; the complete dataset is given in Table S5

Full size image

There were clear patterns in the distribution of the selected CAZymes among bacterial lineages (Fig. 2). Both cellulolytic and hemicellulolytic activities were prominent among the MAGs of Bacteroidota, especially Sphingobacteriaceae, Chitinophagaceae, and most families of Bacteroidales. MAGs of the Fibrobacterota possessed a high abundance of cellulolytic enzymes, with specific sets of CAZyme families that were not observed in other phyla. Large numbers of (hemi)cellulases were present also in MAGs of the Bacillota and Actinobacteriota. Among Bacillota, MAGs with considerable numbers of cellulases were found mostly among Clostridia (e.g., Lachnospiraceae, Acutalibacteraceae, and Clostridiceae), whereas the only prominent family with cellulases among Bacilli was Paenibacillaceae. Clear differences were apparent also in the distribution of auxiliary activities and CBMs (see below).

Factors driving bacterial CAZyme distribution

Based on the obvious patterns in CAZyme distribution among MAGs from different bacterial lineages, we went back to the entire pre-selection dataset of CAZymes (256,212 genes across 534 CAZyme families; Table S5) to test whether the distribution patterns of CAZymes in the termite gut reflect bacterial taxonomy or host factors. Dimensionality reduction using all CAZyme families revealed that the bacterial phylum is an almost exclusive driver of CAZyme distribution (Fig. 3A), followed to a lesser extent by the (sub)family of its termite host (Fig. 3B). A separation of the MAGs from LT and HT reveals similar distribution patterns (Fig. S1A-B). Other factors, such as host diet or gut compartment, did not affect the ordination (Fig. S1C-D). A fitted phylogenetic signal model confirmed the association of bacterial taxonomy to the 38 selected CAZyme families. There was phylogenetic signal in distribution of cellulases in the phyla Bacillota, Actinomycetota, Bacteroidota, Fibrobacterota, and Planctomycetota (Figs. S2 and S3). The phylum Spirochaetota presented a phylogenetic signal for the distribution of hemicellulases. Actinomycetota, Pseudomonadota, and Acidobacteriota were the only phyla with a significant phylogenetic signal for the ligninases.

Fig. 3
figure 3

Ordination analysis of CAZyme composition in 2,204 MAGs of termite-associated bacteria. The UMAP analysis (stress = 0.12) used the entire dataset of CAZymes (256,212 genes across 534 CAZyme families; see Table S5) to visualize differences in the general content of CAZymes among phyla. The biplot displays MAGs color-coded according to their respective phyla (A) and according to the termite (sub)families from which they were recovered (B). The theoretical distances were calculated based on the KNN score, and the statistics are based on ANOSIM (bacterial phyla R = 0.34; termite (sub)families R = 0.04; p < 0.001) and ADONIS (bacterial phyla R2 = 0.24; termite (sub)families R2 = 0.03; p < 0.001) tests

Full size image

Division of fibrolytic niches among bacterial clades

Using the ratio of (hemi)cellulase genes in the selected 38 CAZyme families to genes per genome (gene density, GD) to allow for abundance comparison between clades, we found that MAGs with the highest GD of cellulases were Fibrobacterota, followed by Bacteroidota, Bacillota, and Spirochaetota (Fig. 4A). The bacterial phyla with the highest hemicellulolytic potential were Bacteroidota, Bacillota, and Spirochaetota, followed by Fibrobacterota and Verrucomicrobiota (Fig. 4B).

Fig. 4
figure 4

Gene density (GD) of cellulases (A) and hemicellulases (B) in the termite-associated MAGs of different bacterial phyla, and the relationships between the respective GDs and the relative abundance (RA) of the MAGs in the metagenomes of lower termites and higher termites (C, D). The number of (hemi)cellulolytic genes of a given MAG is normalized by its total number of predicted genes and expressed as percentages. Statistical tests were performed both globally and pairwise, against the mean values (dashed line). Significance values: ns = non-significant; * ≤ 0.05; ** ≤ 0.01; *** ≤ 0.001; **** ≤ 0.0001. Linear regressions are shown for all phyla but a positive correlation (Spearman ϱ > 0.1 and p < 0.05) was present only for LT-associated MAGs of Bacteroidota and HT-associated MAGs of Fibrobacterota, Bacteroidota, Spirochaetota, Bacillota and Actinomycetota

Full size image

Among Fibrobacterota, a cluster of cellulase-containing CAZymes was highly abundant in MAGs of all families (Fibrobacteraceae, Chitinivibrionaceae, and Chitinispirillaceae). It is composed of GH9, GH5 (subfamilies 5, 2, and 39), GH8, and GH94 (Figs. 2 and 5A). The GH9s from our bacterial dataset comprise mainly endo-β-1,4-glucanases but only very few exo-β-1,4-glucanases, while GH8 and the specified GH5 subfamilies contain only endo-β-1,4-glucanases (Table S3). GH94 is composed mainly of enzymes with cellodextrin phosphorylase (CDP) and cellobiose phosphorylase (CBP) activity (Table S3). Although the enzymes characterized to date are generally cytoplasmic, one-third of the GH94 enzymes in our dataset have a signal peptide (Table S6). Also, MAGs of the Oscillospiraceae, Lachnospiraceae, and Clostridiaceae (Bacillota) possessed CDPs and CBPs, albeit in lower abundance (Figs. 2 and 5A).

Fig. 5
figure 5

Abundance and proportion of selected CAZymes from GH families with cellulolytic (A) and hemicellulolytic (B) functions in the MAGs of selected bacterial families with a high fibrolytic potential. Each bar shows the total number of CAZymes classified as cellulase or hemicellulase that are encoded by a particular MAG (gray and black color indicates the origin from lower or higher termites) and the proportion of genes in the respective GH families (visualized with different colors). Only MAGs that encode more than one cellulase or hemicellulase were included. The families GH5 and GH30 appear twice but refer to different subfamilies classified as cellulases or hemicellulases (see Table S3 for details)

Full size image

The most abundant GH family in all lineages of Bacteroidota was GH3 (1,139 genes), corresponding to 3.3 genes per genome (GPG). Members of GH3 have β-glucosidase activity with wide substrate specificity, precluding classification as cellulases or hemicellulases (Table S3). Additionally, the orders Bacteroidales, Chitinophagales, and Sphingobacteriales displayed consistent abundances of GH16 (subfamilies 4, 5, and 3) and GH30 (subfamilies 1, 6, and 3; Figs. 5A and 6A). These subfamilies contained lichenases/laminarinases with endo-β-1,3- and endo-β-1,4-glucanase activity (Table S3) [98].

Fig. 6
figure 6

Principal component analysis (PCA) of the abundance of cellulases (A) and hemicellulases (B) in the MAGs from selected bacterial families with a high fibrolytic potential (see Fig. 5). The biplots show MAGs color-coded either according to bacterial phylum (left) or according to their respective host group (lower or higher termite; right)

Full size image

Among Spirochaetota, MAGs of Breznakiellaceae, Spirochaetaceae, and Treponemataceae had high abundances of GH11, GH10, GH5_22, GH5_4, and GH5_8, which contribute strongly to their divergence in hemicellulolytic functions from other phyla (Fig. 6B). Enzymes from these families have a variety of substrates, acting mainly on xyloglucans and xylomannans (Table S3). By contrast, MAGs of Bacteroidota showed a high abundance of GH2 (1,734 genes, 5 GPG on average). Enzymes in this family have β-galactosidase and α-arabinosidase activity (Table S3). Additionally, the MAGs from many families in the orders Bacteroidales, Sphingobacteriales, and Chitinophagales had GH43 genes (Figs. 2 and 5B), which encode endo-xylanase and α-arabinofuranosidase activities (Table S3). The high abundance of GH2 and GH43 genes drives the hemicellulase composition of the Bacteroidota MAGs (Fig. 6B), accounting for most of the observed variation. High numbers of GH2 and GH43 genes were present also among MAGs of Verrucomicrobiota (Opitutaceae and UBA953) and Bacillota (Oscillospiraceae, Lachnospiraceae, Peptostreptococcaceae, Ruminococcaceae) (Figs. 2 and 5B). However, Bacillota MAGs displayed even higher proportions of GH26, with activity on xylomannan (Fig. 5B, Table S3). While none of the CE associated with our dataset were annotated as acetylxylan esterases, we found pectin lyases and CE involved in pectin degradation (pectin acetyl esterases and pectin methyl esterases) abundantly represented in many MAGs (Tables S3, S5), particularly from the phyla Bacteroidota and Planctomycetota (Figure S6).

Differences in (hemi)cellulolytic lineages between lower and higher termites

MAGs of Fibrobacterota were recovered exclusively from HT, and MAGs of Spirochaetota MAGs from HT had a higher gene density of cellulases than those from LT (p < 0.001 and r = 0.6; Fig. 4A). There was a clear correlation between (hemi)cellulase density and metagenomic relative abundances in MAGs of Fibrobacterota and Spirochaetota from HT (both with p < 0.001 and ϱ = 0.5; Fig. 4C–D). Both Fibrobacterota (Fibrobacteraceae, Chitinivibrionaceae, and Chitinispirillaceae) and Spirochaetota (Breznakiellaceae) MAGs from HT had high proportions of GH5, GH8, and GH9 (Fig. 5A). MAGs of Bacteroidota and Bacillota differed only slightly between LT and HT (both with p < 0.001 and effect size r < 0.1; Fig. 3). However, LT-derived genomes of Dysgonomonadaceae, Azobacteroidaceae, Bacteroidaceae, Tannerellaceae (Bacteroidota), and Breznakiellaceae (Spirochaetota) had a higher proportion of GH30, GH16, and GH116, which act on cellodextrins (Fig. 5A).

The situation was different in the case of hemicellulases. Here, the GDs were higher in MAGs of Spirochaetota and Bacillota derived from HT (both with p < 0.001 and approx. r = 0.3; Fig. 4B), with large proportions of GH5, GH10, and GH11 (Fig. 5B). By contrast, MAGs of Bacteroidota from LT had higher densities of hemicellulases than those from HT (p < 0.001 and r = 0.2). LT-derived genomes of Dysgonomonadaceae, Azobacteroidaceae, Bacteroidaceae, Tannerellaceae (Bacteroidota), and Breznakiellaceae (Spirochaetota) had a higher proportion of GH2 and GH43, which comprise CAZymes with endoxylanase, arabinofuranosidase and β-galactosidase activities (Fig. 5B). LT-derived Bacteroidota MAGs with high GDs of (hemi)cellulases were abundant in the metagenomes, although the correlation between GD and relative abundance was not strong (p < 0.1, Spearman ϱ < 0.1; Fig. 4C–D). The ordination analysis reveals that the prevalence of CAZymes with different activities was driven not only by differences in taxonomy but also by their colonization patterns of LT and HT (Fig. 6).

Carbohydrate-binding modules in CAZymes from (hemi)cellulolytic clades

Carbohydrate-binding modules (CBMs) are domains responsible for the adherence of certain CAZymes to their ligand, assuring that secreted enzymes do not diffuse away from their substrates. Among MAGs of the phylum Fibrobacterota, CBM4 and CBM6 were highly abundant in CAZymes from GH8, GH9, and GH5 (62 genes, average 2 GPG; 23 genes, average 0.7 GPG, respectively). Among MAGs of Bacteroidota, CBM6 and CBM62 were represented particularly among CAZymes from GH3 and GH2 from Prolixibacteraceae, Bacteroidaceae, Tannerellaceae, Dysgonomonadaceae, and Azobacteroidaceae (Fig. 2). MAGs of Bacillota showed a low content of CBM9, CBM22, and CBM35 (on average, max. 0.4 GPG) in CAZymes from GH2, GH43, and the GH5 subfamilies with hemicellulase activity. Cohesins, which are characteristic of the cellulosomes of Bacillota, were generally present only in Oscillospiraceae (average 1.12 GPG) but rare in Ruminococcaceae (average 0.3 GPG) (Table S4).

Auxiliary activities involved in fiber digestion and lignin modification

CAZyme families classified as “Auxiliary Activities” (AA) comprise different classes of oxidative enzymes that, among other substrates, can act on lignin or (hemi)cellulose (Table S3). They were less abundant than the GHs and restricted to MAGs from selected lineages (Figs. 2 and 7). Lytic cellulose/chitin monooxygenases (LPMO; AA10, 92 genes) and cellooligosaccharide dehydrogenases (CDH; AA7, 1068 genes) were most abundant among MAGs of Pseudomonadota (0.5 GPG on average) and Actinobacteriota (0.7 GPG), particularly in the orders Burkholderiales, Propionibacteriales, Actinomycetales, and Microccocales (Fig. 7).

Fig. 7
figure 7

Distribution of gene densities in CAZymes with lignin-modifying activities in MAGs from different bacterial phyla. The insets break down selected phyla to the order level and show the proportion in percentages of genes from different enzyme families: POL (phenol-oxidizing laccases, AA1), LMP (lignin-modifying peroxidases, AA2), and AAO (aryl alcohol oxidases, AA3_2)

Full size image

CAZymes acting on lignin or its derivatives were also present among MAGs from Pseudomonadota and Actinomycetota (Fig. 7). They comprised phenol-oxidizing laccases (POL; AA1), lignin-modifying peroxidases (LMP; AA2), and aryl alcohol oxidases (AAO; AA3_2). CAZymes of these families were also found in MAGs from several orders of Bacteroidota (mainly LMP) and Bacillota (mainly POL) (Fig. 7). Notably, most MAGs that encode lignin-modifying activities were derived from the guts of higher termites (p < 0.05 and r = 0.25).

Discussion

Two events in the evolutionary history of termites involved fundamental changes in the strategy of symbiotic digestion: The first event was the acquisition of cellulolytic flagellates by the common ancestor of termites and wood roaches (Cryptocercidae), which are essential for the degradation of plant fibers in lower termites (LT). The second event was the loss of these flagellates in a common ancestor of higher termites (HT), which led to functional changes that shifted lignocellulose digestion to the prokaryotic gut microbiota [5, 51].

The finding that most of the cellulolytic activity in the hindgut of Nasutitermes spp. (Nasutitermitinae) is associated with wood particles implicated in fiber-associated bacterial lineages in the degradation of (hemi)celluloses [6, 52]. Wood particles are colonized predominantly by members of the phyla Fibrobacterota (Fibrobacterales and Chitinivibrioniales) and Spirochaetota (mostly Breznakiellaceae) [14, 52]. Our results corroborate previous metagenomic and metatranscriptomic analyses of Nasutitermes, Amitermes, Cortaritermes, and Microcerotermes species [10, 14, 16, 53], extending the cellulolytic potential of Fibrobacteraceae and Chitinivibrionaceae and the hemicellulolytic potential of Breznakiellaceae to MAGs from almost all HT investigated, and documenting a cellulolytic potential also in MAGs of the family Chitinispirillaceae. The consistent presence of endo-glucanases and β-glucosidases (GH8, GH9, and certain subfamilies of GH5) — accounting for the highest cellulase gene densities of all phyla — indicates that all lineages of Fibrobacterota in the hindgut of HT are specialized in cellulose degradation. Conversely, high gene densities of endo-xylanases, β-xylosidases, and endo-mannanases (GH11, GH10, and certain subfamilies of GH5) identify MAGs of Breznakiellaceae (Spirochaetota) as important hemicellulose degraders in the hindgut of HT. This expands previous evidence for high expression of GH11 and GH5 transcripts in Nasutitermes and Cortaritermes sp. [10, 14], suggesting a functional specialization among the lineages of Breznakiellaceae associated with HT.

Notably, the (hemi)cellulolytic functions in the genomes of fiber-associated bacteria from HT resemble the activities revealed by metatranscriptomic analyses of the cellulolytic flagellates in LT [54, 55]. While the cellulolytic functions of the flagellates comprise exo-glucanases (GH7) and endo-glucanases (GH45) that were not prominent in the bacterial MAGs, the hemicellulolytic functions of the flagellates, i.e., arabinosidases (GH43), endomannases (GH26), and endo-β-1,4-xylanases (GH8, GH10 and GH11), were the same as those encoded by the fiber-associated bacteria in HT. Apparently, the loss of the (hemi)cellulolytic functions of the gut flagellates selected for bacterial lineages with different cellulolytic strategies but a similar arsenal of hemicellulolytic enzymes.

Obviously, the prediction of substrate specificity has clear limitations, and not all CAZymes in our dataset could be attributed to a specific function. This may obfuscate the results of quantitative analyses, such as phylogenetic signal or functional variability, which turned out to be remarkably similar in all instances for cellulases and hemicellulases. Furthermore, the classification of CAZymes involves simplifications, especially when enzymes that act exclusively on cellulose or hemicellulose fall into the same subfamily, or when the enzyme in question is multifunctional. Nonetheless, the fact that our qualitative analyses revealed clear differences in the subfamilies of CAZymes between particular bacterial lineages corroborates that the classifications used in our study were productive and appropriate for this dataset.

The large number of CBMs in cellulases of Fibrobacterota MAGs and the presence of hemicellulose-binding CBM9 in Breznakiellaceae MAGs are consistent with the localization of the cellulolytic and hemicellulolytic activities with the particulate fraction both in the termite gut [14, 52] and in the rumen [56]. The organization and mechanisms by which these activities are secreted are, nevertheless, still elusive in both phyla. Moreover, CBMs were encoded also by MAGs from other phyla, such as Verrucomicrobiota, Bacteroidota, and Actinomycetota (Fig. 2). By contrast, CBMs were rare among the diverse lineages of Clostridia with (hemi)cellulolytic potential.

The dominant cellulases in the MAGs of most clostridial lineages were cellobiose and cellodextrin phosphorylases (CBP and CDP; GH94). The CBPs and CDPs characterized to date are cytosolic enzymes that help anaerobic bacteria conserve metabolic energy by an intracellular cleavage of cellodextrins or cellobiose into glucose and glucose 1-phosphate [57, 58]. It was therefore quite unexpected that about a third of GH94 enzymes encoded by termite-associated MAGs carried a signal peptide. So far, extracellular GH94 with CBP/CDP activities have been reported only in a single proteomic study of the clostridial Caldicellulosiruptor bescii [59]. The presence of potentially secreted CBP/CDPs in MAGs from both LT and HT suggests that these bacteria utilize cellobiose and dextrins produced by other, cellulolytic microbiota and may explain the low number of CBMs and virtual lack of cellulosomes among the clostridial lineages.

An analysis of lignocellulose digestion by solid-state NMR has revealed significant differences in the fate of polysaccharides between wood-feeding LT and HT [60]. While both Cryptotermes (LT) and Nasutitermes (HT) efficiently degrade cellulose, the latter significantly outperforms the former in the breakdown of hemicelluloses. Apparently, the breakdown of hemicelluloses, whose structural complexity requires the concerted action of numerous enzymes [61], is accomplished more efficiently by the consortium of bacterial species that colonize the wood particles in the hindgut of HT than by the hemicellulases produced by the gut flagellates of LT [62,63,64]. In LT, the wood particles are sequestered into the digestive vacuoles of the flagellates as soon as they pass the enteric valve [65], precluding a primary role of gut bacteria in the breakdown of (hemi)cellulose. However, the consistent presence of β-glucosidase and cellodextrinases with CBMs in the MAGs of Bacteroidales associated with LT (Bacteroidaceae, Dysgonomonadaceae, Tannerellaceae, and Azobacteroidaceae) suggests that these lineages are involved in the breakdown of cellobiose and cellodextrins, which either stem from the partial degradation of amorphous cellulose by the endogenous cellulases in the midgut [66] or represent products of this incomplete digestion of cellulose by the flagellates. Such task partitioning between bacteria and flagellates had been postulated already to explain the considerable differences in the repertoire of bacterial CAZymes between LT and HT [15].

The utilization of partially digested hemicelluloses by Bacteroidales in LT is consistent with the compositional differences in hemicellulolytic activities to HT-associated Breznakiellaceae (Spirochaetota) and the presence of polysaccharide utilization loci (PUL) that encode both carbohydrate capture and uptake systems and CAZymes for the degradation of cellulose, xylans, and mannans in all lineages of Bacteroidota from termite guts [15, 67]. Notably, many of the MAGs from LT represent bacterial lineages associated with gut flagellates [22]. While the ectosymbiotic members of Azobacteroidaceae (Candidatus Symbiothrix dinenymphae), Dysgonomonadaceae, and Bacteroidaceae show the same (hemi)cellulolytic potential as their free-living relatives [68,69,70], the situation is different for the intracellular symbionts (Candidatus Azobacteroides pseudotrichonymphae) present in almost all Rhinotermidae [71, 72], whose genomes lack (hemi)cellulolytic potential [73] (Table S4).

Since the polyphenolic component of lignocellulose provides a major obstacle to the enzymatic attack of cellulose and hemicelluloses, the fate of lignin during termite gut passage has been disputed for decades [74]. The mechanistic challenges of lignin degradation are rooted in the stability of the inter-monomeric linkages, which cannot be hydrolyzed but require oxidative attack by (per)oxidases [75]. Meanwhile, there is compelling evidence that lignin is modified during gut passage in all termites investigated [76,77,78,79]. Nonetheless, there are fundamental differences between LT and HT. While LT disrupts the covalent bonds between hemicelluloses and lignin, leaving the latter largely intact, HT efficiently degrade the lignin polymer, as evidenced by the depletion of major inter-unit linkages and methoxyls [78, 79].

The capacity for the breakdown of lignin is found in many soil bacteria [81], which possess the same types of extracellular (per)oxidases that are present in lignin-degrading fungi [75, 81, 82]. Many of these lineages were represented among our termite MAGs, particularly within Actinomycetota (Micrococcales, Actinomycetales, and Mycobacteriales) and Pseudomonadota (Burkholderiales and Pseudomonadales). The MAGs of each lineage encoded a diverse set of laccases, lignin peroxidases, and aryl-alcohol oxidases. Laccases (phenol oxidases) are the most important agents of bacterial lignin degradation that not only oxidize phenolic substrates but — in the presence of lignin degradation products as mediator — also cleave the inter-unit bonds of the lignin polymer [83]. The same bonds are cleaved also by lignin peroxidases, which require hydrogen peroxide produced by aryl-alcohol oxidases as co-substrate [75].

It is noteworthy that almost all MAGs with lignin-modifying capacity were recovered from HT (Fig. 7). In LT, the situation is reminiscent of lignocellulose degradation by brown rot fungi, which employ Fenton chemistry to selectively metabolize carbohydrates without significant lignin removal, where the extracellular production of Fe2+ and H2O2 leads to the formation of hydroxyl radicals [84]. It has been proposed that laccases produced in the salivary glands of the host also contribute to delignification during midgut passage in LT [85].

Several lineages of MAGs assigned to Actinobacteriota and Pseudomonadota encode lytic cellulose monooxygenase (LPMO) and cellobiose dehydrogenase (CDH). LPMOs are highly effective in the depolymerization of recalcitrant polysaccharides because they cleave crystalline cellulose, generating free ends that are accessible to exoglucanases [87, 88]. CDHs shuttle electrons from the oxidation of cellobiose to the catalytic site of LPMO [89]. A metatranscriptomic analysis of a Labiotermes sp. has documented that bacterial LPMOs are expressed mainly in the anterior hindgut [11]. It is possible that the oxygen-dependent activities of Actinomycetales and Pseudomonadales in the hindgut of HT enhance the degradation of wood fibers by polysaccharide-digesting bacteria. However, the small proportion of exoglucanases in the hindgut microbiota of higher termites may have a different explanation. Many endoglucanases, particularly in GH5 and GH9, are processive and have been shown to attack both microcrystalline and amorphous regions of cellulose, producing cellobiose, cellotriose, and cellotetraose as final products [80]. Hence, it is plausible that fiber-associated bacteria can effectively hydrolyze cellulose fibers with the synergistic action of their diverse endoglucanases, as already demonstrated for Fibrobacter succinogenes, the most important fiber-degrading bacterium in the rumen [86]. A recent report on a GH5_4 cellulase with high activities on crystalline cellulose from the metagenome of the camel rumen [99] illustrates that further insights into the mechanism of crystalline cellulase degradation by the hindgut bacteria of termites will be gained by integrating both bioinformatics and enzymology.

Both oxidases and oxygenases, and indirectly also peroxidases, require molecular oxygen as co-substrate. Microsensor measurements have revealed a strong influx of oxygen across the termite gut epithelium, creating microoxic habitats in the periphery of an otherwise anoxic hindgut [90]. Notably, MAGs encoding lignin-modifying enzymes belonged to the same bacterial lineages that were localized at the internal surface of the hindgut wall. Fluorescence in situ hybridization (FISH) has revealed that the microoxic periphery of the hindgut of Mastotermes darwiniensis (LT) is colonized by Alphaproteobacteria and Betaproteobacteria (Pseudomonadota) [91]. Clone libraries of the hindgut contents of Reticulitermes santonensis and Reticulitermes speratus (LT) and a Nasutitermes spp. (HT) have identified Actinomycetota, Bacteroidota, Pseudomonadota, and Spirochaetota phylotypes in the wall fraction [92,93,94]. They include the same lineages of Actinomycetales, Corynebacteriales, and Propionibacteriales that are represented by our MAGs with lignin- and cellulose-oxidizing potential (Fig. 7).

Conclusions

Previous studies revealed unique patterns of bacterial community structure of termites that were attributed to differences in host phylogeny, diet, and microenvironmental factors [95,96,97]. Our results altogether indicate that this distribution of specific bacterial lineages determines the CAZyme patterns among different host groups. Many members of the bacterial gut microbiota, such as major lineages in the phyla Elusimicrobiota and Patescibacteria, have a very low content of GHs, indicating they do not contribute to the degradation of polysaccharides and possess only CAZymes necessary for assimilatory functions, such as cell wall maintenance. Many lineages with pronounced (hemi)cellulytic capacities, such as certain Bacteroidales families, occur only in LT, whereas others, such as Fibrobacterota and Actinomycetota, were present only in HT. In Spirochaetota, the fibrolytic capacities differ strongly between lineages from LT and HT, and Actinomycetota and Pseudomonadota with fibrolytic functions occur only in HT.

Although the reasons for these patterns and their implications for the nutritional symbioses in termite guts seem to be multifactorial and complex, they are ultimately driven by the presence of flagellates in LT and their subsequent loss in HT. In LT, where lignin modification seems to be restricted to host activities in the midgut, the flagellates efficiently depolymerize cellulose but accomplish only incomplete digestion of hemicelluloses, and free-living or flagellate-associated Bacteroidales and Breznakiellaceae exploit residual cellodextrins and hemicelluloses. In HT, however, the degradation of lignocellulose is partitioned among a consortium of bacterial lineages with different capabilities. Pseudomonadota and Actinomycetota that colonize the microoxic periphery of the hindgut partially depolymerize lignin and generate free-ends in the crystalline regions of cellulose using oxygen-dependent activities. This increases the accessibility of the plant fibers for Fibrobacterota and HT-specific Breznakiellaceae, which possess sophisticated fiber-adhesion mechanisms and effectively depolymerize high-molecular-weight cellulose and hemicelluloses. Finally, the low-molecular remains of (hemi)celluloses released by the activities of the fiber-associated microbiota are processed by various bacterial lineages, including several families of Clostridia, which seem to employ hitherto unrecognized strategies involving extracellular cellobiose and cellodextrin phosphorylases.

Data availability

The MAGs used in this study are available in the NCBI Sequence Read Archive (SRA) under the BioProject accession numbers PRJNA560329 and PRJNA732531. All data generated or analyzed during this study are included in this published article and its supplementary information files.

References

  1. Chouvenc T, Šobotník J, Engel MS, Bourguignon T. Termite evolution: mutualistic associations, key innovations, and the rise of Termitidae. Cell Mol Life Sci. 2021;78:2749–69.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Ni J, Tokuda G. Lignocellulose-degrading enzymes from termites and their symbiotic microbiota. Biotechnol Adv. 2013;31:838–50.

    Article  CAS  PubMed  Google Scholar 

  3. Tokuda, G. Plant cell wall degradation in insects: recent progress on endogenous enzymes revealed by multi-omics technologies. In: Jurenka, R. editor. Advances in Insect Physiology. Cambridge: Academic Press; 2019. vol. 57. p. 97–136

  4. Brune A. Symbiotic digestion of lignocellulose in termite guts. Nat Rev Microbiol. 2014;12:168–80.

    Article  CAS  PubMed  Google Scholar 

  5. Brune A, Dietrich C. The gut microbiota of termites: digesting the diversity in the light of ecology and evolution. Annu Rev Microbiol. 2015;69:145–66.

    Article  CAS  PubMed  Google Scholar 

  6. Tokuda G, Watanabe H. Hidden cellulases in termites: revision of an old hypothesis. Biol Lett. 2007;3:336–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Warnecke F, Luginbühl P, Ivanova N, Ghassemian M, Richardson TH, Stege JT, et al. Metagenomic and functional analysis of hindgut microbiota of a wood-feeding higher termite. Nature. 2007;450:560–5.

    Article  CAS  PubMed  Google Scholar 

  8. Schalk, F., Gostinčar, C., Kreuzenbeck, N. B., Conlon, B. H., Sommerwerk, E., Rabe, P., et al. The termite fungal cultivar Termitomyces combines diverse enzymes and oxidative reactions for plant biomass conversion. mBio. 2021;12:e0355120.

  9. Li H, Young SE, Poulsen M, Currie CR. Symbiont-mediated digestion of plant biomass in fungus-farming insects. Annu Rev Entomol. 2021;66:297–316.

    Article  CAS  PubMed  Google Scholar 

  10. Calusinska M, Marynowska M, Bertucci M, Untereiner B, Klimek D, Goux X, et al. Integrative omics analysis of the termite gut system adaptation to Miscanthus diet identifies lignocellulose degradation enzymes. Commun Biol. 2020;3:1–12.

    Article  Google Scholar 

  11. Marynowska M, Sillam-Dussès D, Untereiner B, Klimek D, Goux X, Gawron P, et al. A holobiont approach towards polysaccharide degradation by the highly compartmentalised gut system of the soil-feeding higher termite Labiotermes labralis. BMC Genomics. 2023;24:115.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Romero Victorica M, Soria MA, Batista-García RA, Ceja-Navarro JA, Vikram S, Ortiz M, et al. Neotropical termite microbiomes as sources of novel plant cell wall degrading enzymes. Sci Rep. 2020;10:3864.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Guerrero EB, de Villegas RMD, Soria MA, Santangelo MP, Campos E, Talia PM. Characterization of two GH5 endoglucanases from termite microbiome using synthetic metagenomics. Appl Microbiol Biotechnol. 2020;104:8351–66.

    Article  CAS  PubMed  Google Scholar 

  14. Tokuda G, Mikaelyan A, Fukui C, Matsuura Y, Watanabe H, Fujishima M, et al. Fiber-associated spirochetes are major agents of hemicellulose degradation in the hindgut of wood-feeding higher termites. Proc Natl Acad Sci. 2018;115:E11996–2004.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Arora J, Kinjo Y, Šobotník J, Buček A, Clitheroe C, Stiblik P, et al. The functional evolution of termite gut microbiota. Microbiome. 2022;10:78.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Rahman, N., Parks, D. H., Vanwonterghem, I., Morrison, M., Tyson, G. W. & Hugenholtz, P. A phylogenomic analysis of the bacterial phylum Fibrobacteres. Front. Microbiol. 2016;6; https://doi.org/10.3389/fmicb.2015.01469.

  17. Levasseur A, Drula E, Lombard V, Coutinho PM, Henrissat B. Expansion of the enzymatic repertoire of the CAZy database to integrate auxiliary redox enzymes. Biotechnol Biofuels. 2013;6:41.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Armendáriz-Ruiz M, Rodríguez-González JA, Camacho-Ruíz RM, Mateos-Díaz JC. Carbohydrate esterases: an overview. Methods Mol Biol Clifton NJ. 2018;1835:39–68.

    Article  Google Scholar 

  19. Garron M-L, Cygler M. Structural and mechanistic classification of uronic acid-containing polysaccharide lyases. Glycobiology. 2010;20:1547–73.

    Article  CAS  PubMed  Google Scholar 

  20. Lairson LL, Henrissat B, Davies GJ, Withers SG. Glycosyltransferases: structures, functions, and mechanisms. Annu Rev Biochem. 2008;77:521–55.

    Article  CAS  PubMed  Google Scholar 

  21. Hervé V, Liu P, Dietrich C, Sillam-Dussès D, Stiblik P, Šobotník J, et al. Phylogenomic analysis of 589 metagenome-assembled genomes encompassing all major prokaryotic lineages from the gut of higher termites. PeerJ. 2020;8: e8614.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Mies, U. S., Hervé, V., Kropp, T., Platt, K., Sillam-Dussès, D., Šobotník, J., et al. Genome reduction and horizontal gene transfer in the evolution of Endomicrobia—rise and fall of an intracellular symbiosis with termite gut flagellates. mBio. 2024;0: e00826–24.

  23. Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 2015;25:1043–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Chaumeil P-A, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics. 2022;38:5315–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol. 2015;32:268–74.

    Article  CAS  PubMed  Google Scholar 

  26. Matsen FA, Kodner RB, Armbrust EV. Pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree. BMC Bioinformatics. 2010;11:538.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Romero Arias J, Boom A, Wang M, Clitheroe C, Šobotník J, Stiblik P, et al. Molecular phylogeny and historical biogeography of Apicotermitinae (Blattodea: Termitidae). Syst Entomol. 2021;46:741–56.

    Article  Google Scholar 

  28. Wang M, Hellemans S, Šobotník J, Arora J, Buček A, Sillam-Dussès D, et al. Phylogeny, biogeography and classification of Teletisoptera (Blattaria: Isoptera). Syst Entomol. 2022;47:581–90.

    Article  Google Scholar 

  29. Paradis E, Claude J, Strimmer K. APE: analyses of phylogenetics and evolution in R language. Bioinformatics. 2004;20:289–90.

    Article  CAS  PubMed  Google Scholar 

  30. Yang Z. PAML: a program package for phylogenetic analysis by maximum likelihood. Bioinformatics. 1997;13:555–6.

    Article  CAS  Google Scholar 

  31. Letunic I, Bork P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 2021;49:W293–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Xu S, Li L, Luo X, Chen M, Tang W, Zhan L, et al. Ggtree: a serialized data object for visualization of a phylogenetic tree and annotation data. iMeta. 2022;1(e56).

  33. Hyatt D, Chen G-L, LoCascio PF, Land ML, Larimer FW, Hauser LJ. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics. 2010;11:119.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Haft DH, Selengut JD, White O. The TIGRFAMs database of protein families. Nucleic Acids Res. 2003;31:371–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, et al. Pfam: the protein families database in 2021. Nucleic Acids Res. 2021;49:D412–9.

    Article  CAS  PubMed  Google Scholar 

  36. Pandurangan AP, Stahlhacke J, Oates ME, Smithers B, Gough J. The SUPERFAMILY 2.0 database: a significant proteome update and a new webserver. Nucleic Acids Res. 2019;47:D490–4.

    Article  CAS  PubMed  Google Scholar 

  37. Mi H, Muruganujan A, Casagrande JT, Thomas PD. Large-scale gene function analysis with the PANTHER classification system. Nat Protoc. 2013;8:1551–66.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Marchler-Bauer A, Derbyshire MK, Gonzales NR, Lu S, Chitsaz F, Geer LY, et al. CDD: NCBI’s conserved domain database. Nucleic Acids Res. 2015;43:D222–6.

    Article  CAS  PubMed  Google Scholar 

  39. Jones P, Binns D, Chang H-Y, Fraser M, Li W, McAnulla C, et al. InterProScan 5: genome-scale protein function classification. Bioinformatics. 2014;30:1236–40.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Zheng J, Ge Q, Yan Y, Zhang X, Huang L, Yin Y. dbCAN3: automated carbohydrate-active enzyme and substrate annotation. Nucleic Acids Res. 2023;51:W115–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Buchfink B, Xie C, Huson DH. Fast and sensitive protein alignment using DIAMOND. Nat Methods. 2015;12:59–60.

    Article  CAS  PubMed  Google Scholar 

  42. Drula E, Garron M-L, Dogan S, Lombard V, Henrissat B, Terrapon N. The carbohydrate-active enzyme database: functions and literature. Nucleic Acids Res. 2022;50:D571–7.

    Article  CAS  PubMed  Google Scholar 

  43. Teufel F, Almagro Armentero JJ, Johansen AR, Gíslason MH, Pihl SI, Tsirigos KD. SignalP 6.0 predicts all five types of signal peptides using protein language models. Nat Biotechnol. 2022;40:1023–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Käll L, Krogh A, Sonnhammer ELL. Advantages of combined transmembrane topology and signal peptide prediction—the Phobius web server. Nucleic Acids Res. 2007;35:W429–32.

    Article  PubMed  PubMed Central  Google Scholar 

  45. McInnes L, Healy J, Saul N, Großberger L. UMAP: uniform manifold approximation and projection. J Open Source Softw. 2018;3:861.

    Article  Google Scholar 

  46. Oksanen, J., Blanchet, G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., et al. Vegan: community ecology package. 2019; https://cran.r-project.org/package=vegan

  47. R Core Team. R: A language and environment for statistical computing. 2013.

  48. Fox, J. & Weisberg, S. An R companion to applied regression. SAGE Publications Inc. 2024;https://us.sagepub.com/en-us/nam/an-r-companion-to-applied-regression/book246125.

  49. Kassambara, A. rstatix: pipe-friendly framework for basic statistical tests. 2023; https://rpkgs.datanovia.com/rstatix/.

  50. Keck F, Rimet F, Bouchez A, Franc A. Phylosignal: an R package to measure, test, and explore the phylogenetic signal. Ecol Evol. 2016;6:2774–80.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Ohkuma, M. & Brune, A. Diversity, structure, and evolution of the termite gut microbial community. In: Bignell, D. E., Roisin, Y. & Lo, N., editors. Biology of Termites: a Modern Synthesis. Dordrecht: Springer; 2011; https://doi.org/10.1007/978-90-481-3977-4_15.

  52. Mikaelyan A, Strassert JFH, Tokuda G, Brune A. The fibre-associated cellulolytic bacterial community in the hindgut of wood-feeding higher termites (Nasutitermes spp.). Environ Microbiol. 2014;16:2711–22.

    Article  CAS  Google Scholar 

  53. He S, Ivanova N, Kirton E, Allgaier M, Bergin C, Scheffrahn RH. Comparative metagenomic and metatranscriptomic analysis of hindgut paunch microbiota in wood- and dung-feeding higher termites. Plos one. 2013;8:e61126.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Todaka N, Inoue T, Saita K, Ohkuma M, Nalepa CA, Lenz M, et al. Phylogenetic analysis of cellulolytic enzyme genes from representative lineages of termites and a related cockroach. PLoS ONE. 2010;5: e8636.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Todaka N, Moriya S, Saita K, Hondo T, Kiuchi I, Takasu H, et al. Environmental cDNA analysis of the genes involved in lignocellulose digestion in the symbiotic protist community of Reticulitermes speratus. FEMS Microbiol Ecol. 2007;59:592–9.

    Article  CAS  PubMed  Google Scholar 

  56. Hess M, Sczyrba A, Egan R, Kim T-W, Chokhawala H, Schroth G, et al. Metagenomic discovery of biomass-degrading genes and genomes from cow rumen. Science. 2011;331:463–7.

    Article  CAS  PubMed  Google Scholar 

  57. Liu N, Fosses A, Kampik C, Parsiegla G, Denis Y, Vita N, et al. In vitro and in vivo exploration of the cellobiose and cellodextrin phosphorylases panel in Ruminiclostridium cellulolyticum: implication for cellulose catabolism. Biotechnol Biofuels. 2019;12:208.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Lou J, Dawson KA, Strobel HJ. Role of phosphorolytic cleavage in cellobiose and cellodextrin metabolism by the ruminal bacterium Prevotella ruminicola. Appl Environ Microbiol. 1996;62:1770–3.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  59. Poudel S, Giannone RJ, Basen M, Nookaew I, Poole FL, Kelly RM, et al. The diversity and specificity of the extracellular proteome in the cellulolytic bacterium Caldicellulosiruptor bescii is driven by the nature of the cellulosic growth substrate. Biotechnol Biofuels. 2018;11:80.

    Article  PubMed  PubMed Central  Google Scholar 

  60. Xue Y, Li H, Kang X. Molecular unraveling of polysaccharide digestion in wood-feeding termites: a solid-state NMR perspective. Carbohydr Polym. 2024;331: 121843.

    Article  CAS  Google Scholar 

  61. Houfani AA, Anders N, Spiess AC, Baldrian P, Benallaoua S. Insights from enzymatic degradation of cellulose and hemicellulose to fermentable sugars–a review. Biomass Bioenergy. 2020;134: 105481.

    Article  CAS  Google Scholar 

  62. Franco Cairo JPL, Mandelli F, Tramontina R, Cannella D, Paradisi A, Ciano L., et al. Oxidative cleavage of polysaccharides by a termite-derived superoxide dismutase boosts the degradation of biomass by glycoside hydrolases. Green Chem. Int J Green Chem Res GC 2022;24,4845–58.

  63. Xie L, Zhang L, Zhong Y, Liu N, Long Y, Wang S, et al. Profiling the metatranscriptome of the protistan community in Coptotermes formosanus with emphasis on the lignocellulolytic system. Genomics. 2012;99:246–55.

    Article  CAS  PubMed  Google Scholar 

  64. Nishimura Y, Otagiri M, Yuki M, Shimizu M, Inoue J, Moriya S, et al. Division of functional roles for termite gut protists revealed by single-cell transcriptomes. ISME J. 2020;14:2449–60.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  65. Yoshimura T, Fujino T, Itoh T, Tsunoda K, Takahashi M. Ingestion and decomposition of wood and cellulose by the protozoa in the hindgut of Coptotermes formosanus Shiraki (Isoptera: Rhinotermitidae) as evidenced by polarizing and transmission electron microscopy. Holzforschung. 1996;50:99–104.

    Article  CAS  Google Scholar 

  66. Lo, N., Tokuda, G. & Watanabe, H. Evolution and function of endogenous termite cellulases. In: Bignell, D. E., Roisin, Y. & Lo, N., editors. Biology of Termites: a Modern Synthesis. Dordrecht: Springer; 2011; https://doi.org/10.1007/978-90-481-3977-4_3.

  67. Arnal G, Bastien G, Monties N, Abot A, Anton Leberre V, Bozonnet S, et al. Investigating the function of an arabinan utilization locus isolated from a termite gut Community. Appl Environ Microbiol. 2015;81:31–9.

    Article  PubMed  Google Scholar 

  68. Noda S, Hongoh Y, Sato T, Ohkuma M. Complex coevolutionary history of symbiotic Bacteroidales bacteria of various protists in the gut of termites. BMC Evol Biol. 2009;9:158.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Treitli SC, Kolisko M, Husník F, Keeling PJ, Hampl V. Revealing the metabolic capacity of Streblomastix strix and its bacterial symbionts using single-cell metagenomics. Proc Natl Acad Sci. 2019;116:19675–84.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  70. Yuki M, Kuwahara H, Shintani M, Izawa K, Sato T, Starns D, et al. Dominant ectosymbiotic bacteria of cellulolytic protists in the termite gut also have the potential to digest lignocellulose. Environ Microbiol. 2015;17:4942–53.

    Article  CAS  PubMed  Google Scholar 

  71. Noda S, Iida T, Kitade O, Nakajima H, Kudo T, Ohkuma M. Endosymbiotic Bacteroidales bacteria of the flagellated protist Pseudotrichonympha grassii in the gut of the termite Coptotermes formosanus. Appl Environ Microbiol. 2005;71:8811–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  72. Noda S, Kitade O, Inoue T, Kawai M, Kanuka M, Hiroshima K, et al. Cospeciation in the triplex symbiosis of termite gut protiststheir hosts, and their bacterial endosymbionts. Mol Ecol. 2007;16:1257–66.

    Article  CAS  PubMed  Google Scholar 

  73. Hongoh Y, Sharma VK, Prakash T, Noda S, Toh H, Taylor TD, et al. Genome of an endosymbiont coupling N2 fixation to cellulolysis within protist cells in termite gut. Science. 2008;322:1108–9.

    Article  CAS  PubMed  Google Scholar 

  74. Breznak JA, Brune A. Role of microorganisms in the digestion of lignocellulose by termites. Annu Rev Entomol. 1994;39:453–87.

    Article  CAS  Google Scholar 

  75. Janusz G, Pawlik A, Sulej J, Świderska-Burek U, Jarosz-Wilkołazka A, Paszczyński A. Lignin degradation: microorganisms, enzymes involved, genomes analysis and evolution. FEMS Microbiol Rev. 2017;41:941–62.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  76. Geib, S. M., Filley, T. R., Hatcher, P. G., Hoover, K., Carlson, J. E., Jimenez-Gasco, M. del M., et al. Lignin degradation in wood-feeding insects. Proc. Natl. Acad. Sci. 2008;105:12932–7.

  77. Li H, Yelle DJ, Li C, Yang M, Ke J, Zhang R, et al. Lignocellulose pretreatment in a fungus-cultivating termite. Proc Natl Acad Sci. 2017;114:4709–14.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  78. Li, H., Kang, X., Yang, M., Kasseney, B. D., Zhou, X., Liang, S., et al. Molecular insights into the evolution of woody plant decay in the gut of termites. Sci. Adv. 2023;9: eadg1258.

  79. Tarmadi D, Tobimatsu Y, Yamamura M, Miyamoto T, Miyagawa Y, Umezawa T, et al. NMR studies on lignocellulose deconstructions in the digestive system of the lower termite Coptotermes formosanus Shiraki. Sci Rep. 2018;8:1290.

    Article  PubMed  PubMed Central  Google Scholar 

  80. Wu S, Wu S. Processivity and the mechanisms of processive endoglucanases. Appl Biochem Biotechnol. 2020;190:448–63.

    Article  CAS  PubMed  Google Scholar 

  81. Tian J-H, Pourcher A-M, Bouchez T, Gelhaye E, Peu P. Occurrence of lignin degradation genotypes and phenotypes among prokaryotes. Appl Microbiol Biotechnol. 2014;98:9527–44.

    Article  CAS  PubMed  Google Scholar 

  82. Silva JP, Ticona ARP, Hamann PRV, Quirino BF, Noronha EF. Deconstruction of lignin: from enzymes to microorganisms. Molecules. 2021;26:2299.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  83. Munk L, Sitarz AK, Kalyani DC, Mikkelsen JD, Meyer AS. Can laccases catalyze bond cleavage in lignin? Biotechnol Adv. 2015;33:13–24.

    Article  CAS  PubMed  Google Scholar 

  84. Martinez D, Challacombe J, Morgenstern I, Hibbett D, Schmoll M, Kubicek CP, et al. Genome, transcriptome, and secretome analysis of wood decay fungus Postia placenta supports unique mechanisms of lignocellulose conversion. Proc Natl Acad Sci. 2009;106:1954–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  85. Coy MR, Salem TZ, Denton JS, Kovaleva ES, Liu Z, Barber DS, et al. Phenol-oxidizing laccases from the termite gut. Insect Biochem Mol Biol. 2010;40:723–32.

    Article  CAS  PubMed  Google Scholar 

  86. Qi, M., Jun, H., Forsberg, C.W. Characterization and synergistic interactions of Fibrobacter succinogenes glycoside hydrolases. Appl Environ Microbiol. 2007;73. https://doi.org/10.1128/AEM.01037-07

  87. Franco Cairo JPL, Cannella D, Oliveira LC, Gonçalves TA, Rubio MV, Terrasan CRF, et al. On the roles of AA15 lytic polysaccharide monooxygenases derived from the termite Coptotermes gestroi. J Inorg Biochem. 2021;216: 111316.

    Article  CAS  PubMed  Google Scholar 

  88. Vaaje-Kolstad G, Westereng B, Horn SJ, Liu Z, Zhai H, Sørlie M, et al. An oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides. Science. 2010;330:219–22.

    Article  CAS  PubMed  Google Scholar 

  89. Tan T-C, Kracher D, Gandini R, Sygmund C, Kittl R, Haltrich D, et al. Structural basis for cellobiose dehydrogenase action during oxidative cellulose degradation. Nat Commun. 2015;6:7542.

    Article  PubMed  Google Scholar 

  90. Brune A, Emerson D, Breznak JA. The termite gut microflora as an oxygen sink: microelectrode determination of oxygen and pH gradients in guts of lower and higher termites. Appl Environ Microbiol. 1995;61:2681–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  91. Berchtold M, Chatzinotas A, Schönhuber W, Brune A, Amann R, Hahn D, et al. Differential enumeration and in situ localization of microorganisms in the hindgut of the lower termite Mastotermes darwiniensis by hybridization with rRNA-targeted probes. Arch Microbiol. 1999;172:407–16.

    Article  CAS  PubMed  Google Scholar 

  92. Nakajima H, Hongoh Y, Usami R, Kudo T, Ohkuma M. Spatial distribution of bacterial phylotypes in the gut of the termite Reticulitermes speratus and the bacterial community colonizing the gut epithelium. FEMS Microbiol Ecol. 2005;54:247–55.

    Article  CAS  PubMed  Google Scholar 

  93. Yang H, Schmitt-Wagner D, Stingl U, Brune A. Niche heterogeneity determines bacterial community structure in the termite gut (Reticulitermes santonensis). Environ Microbiol. 2005;7:916–32.

    Article  CAS  PubMed  Google Scholar 

  94. Köhler, T., Dietrich, C., Scheffrahn, R. H. & Brune, A. high-resolution analysis of gut environment and bacterial microbiota reveals functional compartmentation of the gut in wood-feeding higher termites (Nasutitermes spp.). Appl. Environ. Microbiol. 2012;78:4691–701.

  95. Rossmassler K, Dietrich C, Thompson C, Mikaelyan A, Nonoh JO, Scheffrahn RH, et al. Metagenomic analysis of the microbiota in the highly compartmented hindguts of six wood- or soil-feeding higher termites. Microbiome. 2015;3:56.

    Article  PubMed  PubMed Central  Google Scholar 

  96. Mikaelyan, A., Meuser, K. & Brune, A. Microenvironmental heterogeneity of gut compartments drives bacterial community structure in wood- and humus-feeding higher termites. FEMS Microbiol. Ecol. 2017;93:fiw210.

  97. Mikaelyan A, Dietrich C, Köhler T, Poulsen M, Sillam-Dussès D, Brune A. Diet is the primary determinant of bacterial community structure in the guts of higher termites. Mol Ecol. 2015;24:5284–95.

    Article  CAS  PubMed  Google Scholar 

  98. Huang Z, Ni G, Wang F, Zhao X, Chen Y, Zhang L, Qu M. Characterization of a thermostable lichenase from Bacillus subtilis B110 and its effects on β-glucan hydrolysis. J Microbiol Biotechnol. 2022;32:484–92. https://doi.org/10.4014/jmb.2111.11017.

    Article  CAS  PubMed  Google Scholar 

  99. Kholousi Adab, F., Mehdi Yaghoobi, M. & Gharechahi, J. Enhanced crystalline cellulose degradation by a novel metagenome-derived cellulase enzyme. Sci Rep. 2024;14:8560. https://doi.org/10.1038/s41598-024-59256-4.

Download references

Acknowledgements

We thank Sebastian Treitli (Max Planck Institute for Terrestrial Microbiology) for critical discussions and comments on the manuscript.

Funding

Open Access funding enabled and organized by Projekt DEAL. This study was funded by the Max Planck Society and a grant of the Deutsche Forschungsgemeinschaft (DFG) in the Collaborative Research Center SFB 987. JFMS and MAGV received doctoral fellowships of the International Max Planck Research Schools Principles of Microbial Life (IMPRS-μLife) and Environmental, Cellular, and Molecular Microbiology (IMPRS-Mic) respectively. The funding bodies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Author information

Authors and Affiliations

  1. RG Insect Microbiology and Symbiosis, Max Planck Institute for Terrestrial Microbiology, 35043, Marburg, Germany

    João Felipe M. Salgado, Vincent Hervé, Manuel A. G. Vera & Andreas Brune

  2. Tropical Biosphere Research Center, Center of Molecular Biosciences, University of the Ryukyus, Nishihara, Okinawa, 903-0213, Japan

    Gaku Tokuda

Authors
  1. João Felipe M. Salgado
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Vincent Hervé
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Manuel A. G. Vera
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Gaku Tokuda
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Andreas Brune
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

JFMS and AB conceptualized the study; VH and MAGV performed preliminary analyses; JMFS and VH analyzed the genomic data; JFMS performed statistical analyses; JFMS and AB wrote the manuscript with input from VH and GT. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Andreas Brune.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

40168_2024_1917_MOESM1_ESM.xlsx

Additional file 1: Table S1. Termite gut metagenomes metadata. Table S2. MAG classification, metagenomic relative abundance, and quality statistics. Table S3. Functional classification and activity of CAZyme (sub)families represented in the MAGs analyzed in this study. Table S4. Genomic abundance of CAZymes with lignocellulolytic activity in each MAG of the bacterial lineages depicted in Figure 2. Table S5. Genomic abundance of all CAZymes across all MAGs included in this study. Table S6. Prediction of secretion fate for all GH94 genes and the taxonomy of the respective MAGs.

40168_2024_1917_MOESM2_ESM.pdf

Additional file 2: Figure S1. UMAP biplot of the distribution of CAZymes of termite-associated bacteria color-coded according to phylum in (A) lower termites, (B) higher termites, (C) termite diet, and (D) gut compartment. Figure S2. Distribution of phylogenetic signal across termite-associated bacterial phyla tested against random effects. Figure S3. The drivers of uniqueness in the distribution patterns of (hemi)cellulases and lignin-modification enzymes in 2,223 termite-associated bacterial MAGs. Figure S4. The gene density of cellulases (A) and hemicellulases (B) among the MAGs of Firmicutes, Spirochaetota,Fibrobacterota, and Bacteroidota from lower and higher termites. Figure S5. Biplot of the PCA based on the distribution of lignin-modifying enzymes, color coded according to the phylum to which each genome belongs (A) and according to the gut compartment sampled for the metagenomes (B). Figure S6. Distribution of gene densities in CAZymes with pectinase activities in MAGs from different bacterial phyla. The insets break down the two most abundant phyla, Bacteroidota and Planctomycetota, to the order level.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Salgado, J.F.M., Hervé, V., Vera, M.A.G. et al. Unveiling lignocellulolytic potential: a genomic exploration of bacterial lineages within the termite gut. Microbiome 12, 201 (2024). https://doi.org/10.1186/s40168-024-01917-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s40168-024-01917-7

Keywords

Microbiome

ISSN: 2049-2618

Contact us