A Globally Distributed Bacteroides caccae Strain Is the Most Prevalent Mother-Child Shared Bacteroidaceae Strain in a Large Scandinavian Cohort.

Bacteroides and Phocaeicola, members of the family Bacteroidaceae, are among the first microbes to colonize the human infant gut. While it is known that these microbes can be transmitted from mother to child, our understanding of the specific strains that are shared and potentially transmitted is limited. In this study, we aimed to investigate the shared strains of Bacteroides and Phocaeicola in mothers and their infants. We analyzed fecal samples from pregnant woman recruited at 18 weeks of gestation from the PreventADALL study, as well as offspring samples from early infancy, including skin swab samples taken within 10 min after birth, the first available fecal sample (meconium), and fecal samples at 3 months of age. We screened 464 meconium samples for Bacteroidaceae, with subsequent selection of 144 mother-child pairs for longitudinal analysis, based on the presence of Bacteroidaceae, longitudinal sample availability, and delivery mode. Our results showed that Bacteroidaceae members were mainly detected in samples from vaginally delivered infants. We identified high prevalences of Phocaeicola vulgatus, Phocaeicola dorei, Bacteroides caccae, and Bacteroides thetaiotaomicron in mothers and vaginally born infants. However, at the strain level, we observed high prevalences of only two strains: a B. caccae strain and a P. vulgatus strain. Notably, the B. caccae strain was identified as a novel component of mother-child shared strains, and its high prevalence was also observed in publicly available metagenomes worldwide. Our findings suggest that mode of delivery may play a role in shaping the early colonization of the infant gut microbiota, in particular the colonization of Bacteroidaceae members. IMPORTANCE Our study provides evidence that Bacteroidaceae strains present on infants' skin within 10 min after birth, in meconium samples, and in fecal samples at 3 months of age in vaginally delivered infants are shared with their mothers. Using strain resolution analyses, we identified two strains, belonging to Bacteroides caccae and Phocaeicola vulgatus, as shared between mothers and their infants. Interestingly, the B. caccae strain showed a high prevalence worldwide, while the P. vulgatus strain was less common. Our findings also showed that vaginal delivery was associated with early colonization of Bacteroidaceae members, whereas cesarean section delivery was associated with delayed colonization. Given the potential for these microbes to influence the colonic environment, our results suggest that understanding the bacterial-host relationship at the strain level may have implications for infant health and development later in life.

Metagenome-mining indicates an association between bacteriocin presence and strain diversity in the infant gut.

BACKGROUND: Our knowledge about the ecological role of bacterial antimicrobial peptides (bacteriocins) in the human gut is limited, particularly in relation to their role in the diversification of the gut microbiota during early life. The aim of this paper was therefore to address associations between bacteriocins and bacterial diversity in the human gut microbiota. To investigate this, we did an extensive screening of 2564 healthy human gut metagenomes for the presence of predicted bacteriocin-encoding genes, comparing bacteriocin gene presence to strain diversity and age. RESULTS: We found that the abundance of bacteriocin genes was significantly higher in infant-like metagenomes (< 2 years) compared to adult-like metagenomes (2-107 years). By comparing infant-like metagenomes with and without a given bacteriocin, we found that bacteriocin presence was associated with increased strain diversities. CONCLUSIONS: Our findings indicate that bacteriocins may play a role in the strain diversification during the infant gut microbiota establishment.

Prediction of high fecal propionate-to-butyrate ratios using 16S rRNA-based detection of bacterial groups with liquid array diagnostics.

Butyrate and propionate represent two of three main short-chain fatty acids produced by the intestinal microbiota. In healthy populations, their levels are reportedly equimolar, whereas a deviation in their ratio has been observed in various diseased cohorts. Monitoring such a ratio represents a valuable metric; however, it remains a challenge to adopt short-chain fatty acid detection techniques in clinical settings because of the volatile nature of these acids. Here we aimed to estimate short-chain fatty acid information indirectly through a novel, simple quantitative PCR-compatible assay (liquid array diagnostics) targeting a limited number of microbiome 16S markers. Utilizing 15 liquid array diagnostics probes to target microbiome markers selected by a model that combines partial least squares and linear discriminant analysis, the classes (normal vs high propionate-to-butyrate ratio) separated at a threshold of 2.6 with a prediction accuracy of 96%.

HumGut: a comprehensive human gut prokaryotic genomes collection filtered by metagenome data.

BACKGROUND: A major bottleneck in the use of metagenome sequencing for human gut microbiome studies has been the lack of a comprehensive genome collection to be used as a reference database. Several recent efforts have been made to re-construct genomes from human gut metagenome data, resulting in a huge increase in the number of relevant genomes. In this work, we aimed to create a collection of the most prevalent healthy human gut prokaryotic genomes, to be used as a reference database, including both MAGs from the human gut and ordinary RefSeq genomes. RESULTS: We screened > 5,700 healthy human gut metagenomes for the containment of > 490,000 publicly available prokaryotic genomes sourced from RefSeq and the recently announced UHGG collection. This resulted in a pool of > 381,000 genomes that were subsequently scored and ranked based on their prevalence in the healthy human metagenomes. The genomes were then clustered at a 97.5% sequence identity resolution, and cluster representatives (30,691 in total) were retained to comprise the HumGut collection. Using the Kraken2 software for classification, we find superior performance in the assignment of metagenomic reads, classifying on average 94.5% of the reads in a metagenome, as opposed to 86% with UHGG and 44% when using standard Kraken2 database. A coarser HumGut collection, consisting of genomes dereplicated at 95% sequence identity-similar to UHGG, classified 88.25% of the reads. HumGut, half the size of standard Kraken2 database and directly comparable to the UHGG size, outperforms them both. CONCLUSIONS: The HumGut collection contains > 30,000 genomes clustered at a 97.5% sequence identity resolution and ranked by human gut prevalence. We demonstrate how metagenomes from IBD-patients map equally well to this collection, indicating this reference is relevant also for studies well outside the metagenome reference set used to obtain HumGut. All data and metadata, as well as helpful code, are available at http://arken.nmbu.no/~larssn/humgut/ . Video Abstract.

Assessing time dependent changes in microbial composition of biological crime scene traces using microbial RNA markers.

Current body fluid identification methods do not reveal any information about the time since deposition (TsD) of biological traces, even though determining the age of traces could be crucial for the investigative process. To determine the utility of microbial RNA markers for TsD estimation, we examined RNA sequencing data from five forensically relevant body fluids (blood, menstrual blood, saliva, semen, and vaginal secretion) over seven time points, ranging from fresh to 1.5 years. One set of samples was stored indoors while another was exposed to outdoor conditions. In outdoor samples, we observed a consistent compositional shift, occurring after 4 weeks: this shift was characterized by an overall increase in non-human eukaryotic RNA and an overall decrease in prokaryotic RNA. In depth analyses showed a high fraction of tree, grass and fungal signatures, which are characteristic for the environment the samples were exposed to. When examining the prokaryotic fraction in more detail, three bacterial phyla were found to exhibit the largest changes in abundance, namely Actinobacteria, Proteobacteria and Firmicutes. More detailed analyses at the order level were done using a Lasso regression analysis to find a predictive subset of bacterial taxa. We found 26 bacterial orders to be indicative of sample age. Indoor samples did not reveal such a clear compositional change at the domain level: eukaryotic and prokaryotic abundance remained relatively stable across the assessed time period. Nonetheless, a Lasso regression analysis identified 32 bacterial orders exhibiting clear changes over time, enabling the prediction of TsD. For both indoor and outdoor samples, a larger number (around 60%) of the bacterial orders identified as indicative of TsD are part of the Actinobacteria, Proteobacteria and Firmicutes. In summary, we found that the observed changes across time are not primarily due to changes associated with body fluid specific bacteria but mostly due to accumulation of bacteria from the environment. Orders of these environmental bacteria could be evaluated for TsD prediction, considering the location and environment of the crime scene. However, further studies are needed to verify these findings, determine the applicability across samples, replicates, donors, and other variables, and also to further assess the effect of different seasons and locations on the samples.

Reduced metagenome sequencing for strain-resolution taxonomic profiles.

BACKGROUND: Studies of shifts in microbial community composition has many applications. For studies at species or subspecies levels, the 16S amplicon sequencing lacks resolution and is often replaced by full shotgun sequencing. Due to higher costs, this restricts the number of samples sequenced. As an alternative to a full shotgun sequencing we have investigated the use of Reduced Metagenome Sequencing (RMS) to estimate the composition of a microbial community. This involves the use of double-digested restriction-associated DNA sequencing, which means only a smaller fraction of the genomes are sequenced. The read sets obtained by this approach have properties different from both amplicon and shotgun data, and analysis pipelines for both can either not be used at all or not explore the full potential of RMS data. RESULTS: We suggest a procedure for analyzing such data, based on fragment clustering and the use of a constrained ordinary least square de-convolution for estimating the relative abundance of all community members. Mock community datasets show the potential to clearly separate strains even when the 16S is 100% identical, and genome-wide differences is < 0.02, indicating RMS has a very high resolution. From a simulation study, we compare RMS to shotgun sequencing and show that we get improved abundance estimates when the community has many very closely related genomes. From a real dataset of infant guts, we show that RMS is capable of detecting a strain diversity gradient for Escherichia coli across time. CONCLUSION: We find that RMS is a good alternative to either metabarcoding or shotgun sequencing when it comes to resolving microbial communities at the strain level. Like shotgun metagenomics, it requires a good database of reference genomes and is well suited for studies of the human gut or other communities where many reference genomes exist. A data analysis pipeline is offered, as an R package at https://github.com/larssnip/microRMS . Video abstract.

Fecal Microbiota Nutrient Utilization Potential Suggests Mucins as Drivers for Initial Gut Colonization of Mother-Child-Shared Bacteria.

The nutritional drivers for mother-child sharing of bacteria and the corresponding longitudinal trajectory of the infant gut microbiota development are not yet completely settled. We therefore aimed to characterize the mother-child sharing and the inferred nutritional utilization potential for the gut microbiota from a large unselected cohort. We analyzed in depth gut microbiota in 100 mother-child pairs enrolled antenatally from the general population-based Preventing Atopic Dermatitis and Allergies in Children (PreventADALL) cohort. Fecal samples collected at gestational week 18 for mothers and at birth (meconium), 3, 6, and 12 months for infants were analyzed by reduced metagenome sequencing to determine metagenome size and taxonomic composition. The nutrient utilization potential was determined based on the Virtual Metabolic Human (VMH, www.vmh.life) database. The estimated median metagenome size was ∼150 million base pairs (bp) for mothers and ∼20 million bp at birth for the children. Longitudinal analyses revealed mother-child sharing (P < 0.05, chi-square test) from birth up to 6 months for 3 prevalent Bacteroides species (prevalence, >25% for all age groups). In a multivariate analysis of variance (ANOVA), the mother-child-shared Bacteroides were associated with vaginal delivery (1.7% explained variance, P = 0.0001). Both vaginal delivery and mother-child sharing were associated with host-derived mucins as nutrient sources. The age-related increase in metagenome size corresponded to an increased diversity in nutrient utilization, with dietary polysaccharides as the main age-related factor. Our results support host-derived mucins as potential selection means for mother-child sharing of initial colonizers, while the age-related increase in diversity was associated with dietary polysaccharides.IMPORTANCE The initial bacterial colonization of human infants is crucial for lifelong health. Understanding the factors driving this colonization will therefore be of great importance. Here, we used a novel high-taxonomic-resolution approach to deduce the nutrient utilization potential of the infant gut microbiota in a large longitudinal mother-child cohort. We found mucins as potential selection means for the initial colonization of mother-child-shared bacteria, while the transition to a more adult-like microbiota was associated with dietary polysaccharide utilization potential. This knowledge will be important for a future understanding of the importance of diet in shaping the gut microbiota composition and development during infancy.

Butyrate Levels in the Transition from an Infant- to an Adult-Like Gut Microbiota Correlate with Bacterial Networks Associated with Eubacterium Rectale and Ruminococcus Gnavus.

Relatively little is known about the ecological forces shaping the gut microbiota composition during infancy. Therefore, the objective of the present study was to identify the nutrient utilization- and short-chain fatty acid (SCFA) production potential of gut microbes in infants during the first year of life. Stool samples were obtained from mothers at 18 weeks of pregnancy and from infants at birth (first stool) at 3, 6, and 12-months of age from the general population-based PreventADALL cohort. We identified the taxonomic and SCFA composition in 100 mother-child pairs. The SCFA production and substrate utilization potential of gut microbes were observed by multiomics (shotgun sequencing and proteomics) on six infants. We found a four-fold increase in relative butyrate levels from 6 to 12 months of infant age. The increase was correlated to Eubacterium rectale and its bacterial network, and Faecalibacterium prausnitzii relative abundance, while low butyrate at 12 months was correlated to Ruminococcus gnavus and its associated network of bacteria. Both E. rectale and F. prausnitzii expressed enzymes needed for butyrate production and enzymes related to dietary fiber degradation, while R. gnavus expressed mucus-, fucose, and human milk oligosaccharides (HMO)-related degradation enzymes. Therefore, we believe that the presence of E. rectale, its network, and F. prausnitzii are key bacteria in the transition from an infant- to an adult-like gut microbiota with respect to butyrate production. Our results indicate that the transition from an infant- to an adult-like gut microbiota with respect to butyrate producing bacteria, occurs between 6 and 12 months of infant age. The bacteria associated with the increased butyrate ratio/levels were E. rectale and F. prausnitzii, which potentially utilize a variety of dietary fibers based on the glycoside hydrolases (GHs) expressed. R. gnavus with a negative association to butyrate potentially utilizes mucin, fucose, and HMO components. This knowledge could have future importance in understanding how microbial metabolites can impact infant health and development.

Predicting the origin of stains from whole miRNome massively parallel sequencing data.

In this study, we have screened the six most relevant forensic body fluids / tissues, namely blood, semen, saliva, vaginal secretion, menstrual blood and skin, for miRNAs using a whole miRNome massively parallel sequencing approach. We applied partial least squares (PLS) and linear discriminant analysis (LDA) to predict body fluids based on the expression of the miRNA markers. We estimated the prediction accuracy for models including different subsets of miRNA markers to identify the minimum number of markers needed for sufficient prediction performance. For one selected model consisting of 9 miRNA markers we calculated their importance for prediction of each of the six different body fluid categories.

Optimizing body fluid recognition from microbial taxonomic profiles.

In forensics the DNA-profile is used to identify the person who left a biological trace, but information on body fluid can also be essential in the evidence evaluation process. Microbial composition data could potentially be used for body fluid recognition as an improved alternative to the currently used presumptive tests. We have developed a customized workflow for interpretation of bacterial 16S sequence data based on a model composed of Partial Least Squares (PLS) in combination with Linear Discriminant Analysis (LDA). Large data sets from the Human Microbiome Project (HMP) and the American Gut Project (AGP) were used to test different settings in order to optimize performance. From the initial cross-validation of body fluid recognition within the HMP data, the optimal overall accuracy was close to 98%. Sensitivity values for the fecal and oral samples were ≥0.99, followed by the vaginal samples with 0.98 and the skin and nasal samples with 0.96 and 0.81 respectively. Specificity values were high for all 5 categories, mostly >0.99. This optimal performance was achieved by using the following settings: Taxonomic profiles based on operational taxonomic units (OTUs) with 0.98 identity (OTU98), Aitchisons simplex transform with C = 1 pseudo-count and no regularization (r = 1) in the PLS step. Variable selection did not improve the performance further. To test for robustness across sequencing platforms, we also trained the classifier on HMP data and tested on the AGP data set. In this case, the standard OTU based approach showed moderately decline in accuracy. However, by using taxonomic profiles made by direct assignment of reads to a genus, we were able to nearly maintain the high accuracy levels. The optimal combination of settings was still used, except the taxonomic level being genus instead of OTU98. The performance may be improved even further by using higher resolution taxonomic bins.

Comparison of reduced metagenome and 16S rRNA gene sequencing for determination of genetic diversity and mother-child overlap of the gut associated microbiota.

Use of the 16S rRNA gene in microbiota studies is limited by the lack of taxonomic and functional resolution. High resolution analyses are particularly important for understanding transmission and persistence of bacteria. The aim of our work was therefore to compare a novel reduced metagenome sequencing (RMS) approach with 16S rRNA gene sequencing to determine both the metagenome genetic diversity and the mother-to-child sharing of the microbiota in a cohort of 17 mother-child pairs. We found that although both approaches gave comparable results with respect to sample separation and taxonomy, RMS gave higher resolution and the potential for genomic-/functional assignment. Using RMS we estimated that the metagenome size increased from about 60 Mbp for 4-day-old children to about 225 Mbp for mothers. The 4-day-old children shared 7% of the metagenome sequences with the mothers, while the metagenome sequence sharing was >30% among the mothers. We found 15 genomes shared across >50% of the mothers, of which 10 belonged to Clostridia. Only Bacteroides showed a direct mother-child association, with B. vulgatus being abundant in both 4-day-old children and mothers. For the functional assignments, we identified a significant association between antibiotic usage during labor, and quantity of Fosfomycin resistance genes. In conclusion, our results show a higher functional and taxonomic resolution for RMS compared to 16S rRNA gene sequencing, where RMS enabled a detailed description of mother to child gut microbiota transmission - supporting a late recruitment of most gut bacteria and an effect of antibiotic treatment during labor on infant antibiotic resistance gene patterns.

Predicting the origin of stains from next generation sequencing mRNA data.

We used our previously published NGS mRNA approach for body fluid identification to analyse 183 body fluids/tissues, including mock casework samples. The resulting data set was used to build a probabilistic model that predicts the origin of a stain. Our approach uses partial least squares followed by linear discriminant analysis to classify samples into six commonly occurring forensic body fluids. The model differs from the ones previously suggested in that it incorporates quantitative information (NGS read counts) rather than just presence/absence of markers. The suggested approach also allows for visualisation of important markers and their correlation with the different body fluids. We compared our model to previously published methods to show that the inclusion of read count information improves the prediction. Finally, we applied the model to mixed body fluid samples to test its ability to identify the individual components in a mixture.

Stable Core Gut Microbiota across the Freshwater-to-Saltwater Transition for Farmed Atlantic Salmon.

Gut microbiota associations through habitat transitions are fundamentally important yet poorly understood. One such habitat transition is the migration from freshwater to saltwater for anadromous fish, such as salmon. The aim of the current work was therefore to determine the freshwater-to-saltwater transition impact on the gut microbiota in farmed Atlantic salmon, with dietary interventions resembling freshwater and saltwater diets with respect to fatty acid composition. Using deep 16S rRNA gene sequencing and quantitative PCR, we found that the freshwater-to-saltwater transition had a major association with the microbiota composition and quantity, while diet did not show significant associations with the microbiota. In saltwater there was a 100-fold increase in bacterial quantity, with a relative increase of Firmicutes and a relative decrease of both Actinobacteria and Proteobacteria Irrespective of an overall shift in microbiota composition from freshwater to saltwater, we identified three core clostridia and one Lactobacillus-affiliated phylotype with wide geographic distribution that were highly prevalent and co-occurring. Taken together, our results support the importance of the dominating bacteria in the salmon gut, with the freshwater microbiota being immature. Due to the low number of potentially host-associated bacterial species in the salmon gut, we believe that farmed salmon can represent an important model for future understanding of host-bacterium interactions in aquatic environments.IMPORTANCE Little is known about factors affecting the interindividual distribution of gut bacteria in aquatic environments. We have shown that there is a core of four highly prevalent and co-occurring bacteria irrespective of feed and freshwater-to-saltwater transition. The potential host interactions of the core bacteria, however, need to be elucidated further.

Rapid Succession of Actively Transcribing Denitrifier Populations in Agricultural Soil During an Anoxic Spell.

Denitrification allows sustained respiratory metabolism during periods of anoxia, an advantage in soils with frequent anoxic spells. However, the gains may be more than evened out by the energy cost of producing the denitrification machinery, particularly if the anoxic spell is short. This dilemma could explain the evolution of different regulatory phenotypes observed in model strains, such as sequential expression of the four denitrification genes needed for a complete reduction of nitrate to N2, or a "bet hedging" strategy where all four genes are expressed only in a fraction of the cells. In complex environments such strategies would translate into progressive onset of transcription by the members of the denitrifying community. We exposed soil microcosms to anoxia, sampled for amplicon sequencing of napA/narG, nirK/nirS, and nosZ genes and transcripts after 1, 2 and 4 h, and monitored the kinetics of NO, N2O, and N2. The cDNA libraries revealed a succession of transcribed genes from active denitrifier populations, which probably reflects various regulatory phenotypes in combination with cross-talks via intermediates ( NO 2 - , NO) produced by the "early onset" denitrifying populations. This suggests that the regulatory strategies observed in individual isolates are also displayed in complex communities, and pinpoint the importance for successive sampling when identifying active key player organisms.

Effects of glucose availability in Lactobacillus sakei; metabolic change and regulation of the proteome and transcriptome.

Effects of glucose availability were investigated in Lactobacillus sakei strains 23K and LS25 cultivated in anaerobic, glucose-limited chemostats set at high (D = 0.357 h-1) and low (D = 0.045 h-1) dilution rates. We observed for both strains a shift from homolactic towards more mixed acid fermentation when comparing high to low growth rates. However, this change was more pronounced for LS25 than for 23K, where dominating products were lactate>formate>acetate≥ethanol at both conditions. A multivariate approach was used for analyzing proteome and transcriptome data from the bacterial cultures, where the predictive power of the omics data was used for identifying features that can explain the differences in the end-product profiles. We show that the different degree of response to the same energy restriction revealed interesting strain specific regulation. An elevated formate production level during slow growth, more for LS25 than for 23K, was clearly reflected in correlating pyruvate formate lyase expression. With stronger effect for LS25, differential expression of the Rex transcriptional regulator and NADH oxidase, a target of Rex, indicated that maintainance of the cell redox balance, in terms of the NADH/NAD+ ratio, may be a key process during the metabolic change. The results provide a better understanding of different strategies that cells may deploy in response to changes in substrate availability.

Cecal microbiota association with tumor load in a colorectal cancer mouse model.

Background: Colorectal cancer (CRC) is one of the most common cancer types worldwide. The role of the intestinal microbiota in CRC, however, is not well established. In particular, the co-variation between age, tumor progression and microbiota remains largely unknown. Objective and design: We therefore used a recently developed A/J Min/+ mouse model resembling human CRC to investigate how microbial composition in cecum correlates with tumor progression, butyrate and age. Results: We found that the association between the gut microbiota and tumor load was stronger, by far, than the association with both butyrate and age. The strongest direct tumor association was found for mucosal bacteria, with nearly 60% of the significantly correlating operational taxonomic units being correlated with CRC tumor load alone. Conclusion: We favor a systemic association between tumor load and microbiota, since the correlations are associated with tumor load in gut segments other than the cecum (both small and large intestine).

Body fluid prediction from microbial patterns for forensic application.

The association of a DNA profile with a certain body fluid can be of essential importance in the evaluation of biological evidence. Several alternative methods for body fluid prediction have been proposed to improve the currently used presumptive tests. Most of them measure gene expression. Here we present a novel approach based on microbial taxonomic profiles obtained by standard 16S rRNA gene sequencing. We used saliva deposited on skin as a forensically relevant study model, but the same principle can be applied for predicting other bacteria rich body fluids. For classification we used standard pattern recognition based on principal component analysis in combination with linear discriminant analysis. A cross-validation of the experimental data shows that the new method is able to successfully classify samples from saliva deposited on skin and samples from pure skin in 94% of the cases. We found that there is a person-effect influencing the result, especially from skin, indicating that a reference sample of pure skin microbiota from the same person could improve accuracy. In addition the pattern recognition methods could be further optimized. Although there is room for improvement, this study shows the potential of microbial profiles as a new forensic tool for body fluid prediction.

microclass: an R-package for 16S taxonomy classification.

BACKGROUND: Taxonomic classification based on the 16S rRNA gene sequence is important for the profiling of microbial communities. In addition to giving the best possible accuracy, it is also important to quantify uncertainties in the classifications. RESULTS: We present an R package with tools for making such classifications, where the heavy computations are implemented in C++ but operated through the standard R interface. The user may train classifiers based on specialized data sets, but we also supply a ready-to-use function trained on a comprehensive training data set designed specifically for this purpose. This tool also includes some novel ways to quantify uncertainties in the classifications. CONCLUSIONS: Based on input sequences of varying length and quality, we demonstrate how the output from the classifications can be used to obtain high quality taxonomic assignments from 16S sequences within the R computing environment. The package is publicly available at the Comprehensive R Archive Network.

The nucleotide composition of microbial genomes indicates differential patterns of selection on core and accessory genomes.

BACKGROUND: The core genome consists of genes shared by the vast majority of a species and is therefore assumed to have been subjected to substantially stronger purifying selection than the more mobile elements of the genome, also known as the accessory genome. Here we examine intragenic base composition differences in core genomes and corresponding accessory genomes in 36 species, represented by the genomes of 731 bacterial strains, to assess the impact of selective forces on base composition in microbes. We also explore, in turn, how these results compare with findings for whole genome intragenic regions. RESULTS: We found that GC content in coding regions is significantly higher in core genomes than accessory genomes and whole genomes. Likewise, GC content variation within coding regions was significantly lower in core genomes than in accessory genomes and whole genomes. Relative entropy in coding regions, measured as the difference between observed and expected trinucleotide frequencies estimated from mononucleotide frequencies, was significantly higher in the core genomes than in accessory and whole genomes. Relative entropy was positively associated with coding region GC content within the accessory genomes, but not within the corresponding coding regions of core or whole genomes. CONCLUSION: The higher intragenic GC content and relative entropy, as well as the lower GC content variation, observed in the core genomes is most likely associated with selective constraints. It is unclear whether the positive association between GC content and relative entropy in the more mobile accessory genomes constitutes signatures of selection or selective neutral processes.