1 August 2017 in Cell Reports
Bacterial genes that change in expression upon environmental disturbance have commonly been seen as those that must also phenotypically matter. However, several studies suggest that differentially expressed genes are rarely phenotypically important. We demonstrate, for Gram-positive and Gram-negative bacteria, that these seemingly uncoordinated gene sets are involved in responses that can be linked through topological network analysis. However, the level of coordination is stress dependent. While a well-coordinated response is triggered in response to nutrient stress, antibiotics trigger an uncoordinated response in which transcriptionally and phenotypically important genes are neither linked spatially nor in their magnitude. Moreover, a gene expression meta-analysis reveals that genes with large fitness changes during stress have low transcriptional variation across hundreds of other conditions, and vice versa. Our work suggests that cellular responses can be understood through network models that incorporate regulatory and genetic relationships, which could aid drug target predictions and genetic network engineering.
31 August 2020 in Nature Communications
Current approaches explore bacterial genes that change transcriptionally upon stress exposure as diagnostics to predict antibiotic sensitivity. However, transcriptional changes are often specific to a species or antibiotic, limiting implementation to known settings only. While a generalizable approach, predicting bacterial fitness independent of strain, species or type of stress, would eliminate such limitations, it is unclear whether a stress-response can be universally captured. By generating a multi-stress and species RNA-Seq and experimental evolution dataset, we highlight the strengths and limitations of existing gene-panel based methods. Subsequently, we build a generalizable method around the observation that global transcriptional disorder seems to be a common, low-fitness, stress response. We quantify this disorder using entropy, which is a specific measure of randomness, and find that in low fitness cases increasing entropy and transcriptional disorder results from a loss of regulatory gene-dependencies. Using entropy as a single feature, we show that fitness and quantitative antibiotic sensitivity predictions can be made that generalize well beyond training data. Furthermore, we validate entropy-based predictions in 7 species under antibiotic and non-antibiotic conditions. By demonstrating the feasibility of universal predictions of bacterial fitness, this work establishes the fundamentals for potentially new approaches in infectious disease diagnostics.
14 October 2018 in bioRxiv
The emergence of fluoroquinolone resistance in nosocomial pathogens has restricted the clinical efficacy of this antibiotic class. In Acinetobacter baumannii , the majority of clinical isolates now show high-level resistance due to mutations in gyrA (DNA gyrase) and parC (Topo IV). To investigate the molecular basis for fluoroquinolone resistance, an exhaustive mutation analysis was performed in both drug sensitive and resistant strains to identify loci that alter the sensitivity of the organism to ciprofloxacin. To this end, parallel fitness tests of over 60,000 unique insertion mutations were performed in strains with various alleles in genes encoding the drug targets. The spectrum of mutations that altered drug sensitivity was found to be similar in the drug sensitive and double mutant gyrAparC background having resistance alleles in both genes. In contrast, introduction of a single gyrA resistance allele, resulting in preferential poisoning of Topo IV by ciprofloxacin, led to extreme alterations in the insertion mutation fitness landscape. The distinguishing feature of preferential Topo IV poisoning was induction of DNA synthesis in the region of two endogenous prophages, which appeared to occur in situ . Induction of the selective DNA synthesis in the gyrA background was also linked to enhanced activation of SOS response and heightened transcription of prophage genes relative to that observed in either the WT or gyrAparC double mutants. Therefore, the accumulation of mutations that result in the stepwise evolution of high ciprofloxacin resistance is tightly connected to suppression of hyperactivation of the SOS response and endogenous prophage DNA synthesis. Importance: Fluoroquinolones have been extremely successful antibiotics. Their clinical efficacy derives from the ability to target multiple bacterial enzymes critical to DNA replication, the topoisomerases DNA gyrase and Topo IV. Unfortunately, mutations lowering drug affinity for both enzymes are now widespread, rendering these drugs ineffective for many pathogens. To undermine this form of resistance, we sought to understand how bacteria with target alterations differentially cope with fluoroquinolone exposures. We studied this problem in the nosocomial pathogen A. baumannii , which causes resistant, life-threating infections. Employing genome-wide approaches, we uncovered numerous pathways that could be exploited to lower fluoroquinolone resistance independently of target alteration. Remarkably, fluoroquinolone targeting of Topo IV in specific mutants caused dramatic prophage hyperinduction, a response that was muted in strains with DNA gyrase as the primary target. This work demonstrates that resistance evolution via target modification can profoundly modulate the antibiotic stress response, revealing potential resistance-associated liabilities.
22 October 2019 in bioRxiv
Genes implicated in bacterial stress responses have been used to construct models that infer the growth outcome of a bacterium in the presence of antibiotics with the objective to develop novel diagnostic methods in the clinic. Current models are trained on data specific to a species or type of stress, making them potentially limited in their application. It is unclear if a generalizable response-signature exists that can predict bacterial fitness independent of strain, species or type of stress. Here we present a substantial RNA-Seq and experimental evolution dataset for 9 strains and species, under multiple antibiotic and non-antibiotic stress conditions. We show that gene panel-based models can accurately predict antibiotic mechanism of action, as well as the fitness outcome of Streptococcus pneumoniae in the presence of antibiotics or under nutrient depletion. However, these models quickly become species-specific as gene homology is limited. Instead, we define a new concept, transcriptomic entropy, which we use to quantify the amount of transcriptional disruption that occurs in a bacterium when responding to the environment. With entropy at the center, we train a suite of predictive (machine learning) models enabling generalizable fitness and antibiotic sensitivity predictions. These entropy-based models that predict bacterial fitness are validated for 7 Gram-positive and -negative species under antibiotic and non-antibiotic conditions indicating that transcriptomic entropy can be used as a generalizable stress signature. Moreover, rather than being a binary indicator of fitness, an entropy-based model was developed and validated to predict the minimum inhibitory concentration of an antibiotic. Lastly, we show that the inclusion of a varied-set of multi-omics features of a bacterial stress response further enhances fitness predictions by reducing ambiguity. By demonstrating the feasibility of generalizable predictions of bacterial fitness, this work establishes the fundamentals for potentially new approaches in infectious disease diagnostics, including antibiotic susceptibility testing. Significance statement: Accurate predictions of bacterial fitness outcome could potentially have clinical diagnostic value, such as predicting optimum antibiotic choice and dosage for treating infectious diseases. Existing models of fitness predictions rely mainly on gene panel approaches, which may be species- and stress-specific due to a lack of gene and response conservation. In order to overcome this limitation, we generated a substantial experimental dataset and identified entropy as a universal stress response signature that quantifies the level of transcriptional disruption that is indicative of fitness outcome under a stressful condition. We present and validate for Gram-positive and negative species a suite of entropy-based models that enable accurate predictions of fitness outcome and the level of antibiotic sensitivity in a species and stress-type independent manner.
26 August 2016 in bioRxiv
Background: Bacteria modulate subcellular processes to handle stressful environments. Genome-wide profiling of gene expression (RNA-Seq) and fitness (Tn-Seq) allows two views of the same genetic network underlying these responses. However, it remains unclear how they combine, enabling a bacterium to overcome a perturbation. Results: Here we generate RNA-Seq and Tn-Seq profiles in three strains of S. pneumoniae in response to stress defined by different levels of nutrient depletion. These profiles show that genes that change their expression and/or become phenotypically important come from a diverse set of functional categories, and genes that are phenotypically important tend to be highly expressed. Surprisingly, we find that expression and fitness changes rarely occur on the same gene, which we confirmed by over 140 validation experiments. To rationalize these unexpected results we built the first genome-scale metabolic model of S. pneumoniae showing that differential expression and phenotypic importance actually correlate between nearest neighbors, although they are distinctly partitioned into small subnetworks. Moreover, a meta-analysis of 234 S. pneumoniae gene expression studies reveals that essential genes and phenotypically important subnetworks rarely change expression, indicating that they are shielded from transcriptional fluctuations and that a clear distinction exists between transcriptional and phenotypic response networks. Conclusions: We present a genome-wide computational/experimental approach that contextualizes changes that occur on transcriptomic and phenomic levels in response to stress. Importantly, this highlights the need to connect disparate response networks, for instance in antibiotic target identification, where preferred targets are phenotypically important genes that would be overlooked by transcriptomic analyses alone.