Comparison of k-mer-based de novo comparative metagenomic tools and approaches.

Aim: Comparative metagenomic analysis requires measuring a pairwise similarity between metagenomes in the dataset. Reference-based methods that compute a beta-diversity distance between two metagenomes are highly dependent on the quality and completeness of the reference database, and their application on less studied microbiota can be challenging. On the other hand, de-novo comparative metagenomic methods only rely on the sequence composition of metagenomes to compare datasets. While each one of these approaches has its strengths and limitations, their comparison is currently limited. Methods: We developed sets of simulated short-reads metagenomes to (1) compare k-mer-based and taxonomy-based distances and evaluate the impact of technical and biological variables on these metrics and (2) evaluate the effect of k-mer sketching and filtering. We used a real-world metagenomic dataset to provide an overview of the currently available tools for de novo metagenomic comparative analysis. Results: Using simulated metagenomes of known composition and controlled error rate, we showed that k-mer-based distance metrics were well correlated to the taxonomic distance metric for quantitative Beta-diversity metrics, but the correlation was low for presence/absence distances. The community complexity in terms of taxa richness and the sequencing depth significantly affected the quality of the k-mer-based distances, while the impact of low amounts of sequence contamination and sequencing error was limited. Finally, we benchmarked currently available de-novo comparative metagenomic tools and compared their output on two datasets of fecal metagenomes and showed that most k-mer-based tools were able to recapitulate the data structure observed using taxonomic approaches. Conclusion: This study expands our understanding of the strength and limitations of k-mer-based de novo comparative metagenomic approaches and aims to provide concrete guidelines for researchers interested in applying these approaches to their metagenomic datasets.

An extended catalog of integrated prophages in the infant and adult fecal microbiome shows high prevalence of lysogeny.

Background and aims: The acquisition and gradual maturation of gut microbial communities during early childhood is central to an individual's healthy development. Bacteriophages have the potential to shape the gut bacterial communities. However, the complex ecological interactions between phages and their bacterial host are still poorly characterized. In this study, we investigated the abundance and diversity of integrated prophages in infant and adult gut bacteria by detecting integrated prophages in metagenome assembled genomes (MAGs) of commensal bacteria. Methods: Our study included 88 infants sampled at 3 weeks, 3 months, 6 months, and 12 months (n = 323 total samples), and their parents around delivery time (n = 138 total samples). Fecal DNA was extracted and characterized by using shotgun metagenomic sequencing, and a collection of prokaryotic MAGs was generated. The MAG collection was screened for the presence of integrated bacteriophage sequences, allowing their taxonomic and functional characterization. Results: A large collection of 6,186 MAGs from infant and adult gut microbiota was obtained and screened for integrated prophages, allowing the identification of 7,165 prophage sequences longer than 10 kb. Strikingly, more than 70% of the near-complete MAGs were identified as lysogens. The prevalence of prophages in MAGs varied across bacterial families, with a lower prevalence observed among Coriobacteriaceae, Eggerthellaceae, Veillonellaceae and Burkholderiaceae, while a very high prevalence of lysogen MAGs were observed in Oscillospiraceae, Enterococcaceae, and Enterobacteriaceae. Interestingly for several bacterial families such as Bifidobacteriaceae and Bacteroidaceae, the prevalence of prophages in MAGs was higher in early infant time point (3 weeks and 3 months) than in later sampling points (6 and 12 months) and in adults. The prophage sequences were clustered into 5,616 species-like vOTUs, 77% of which were novel. Finally, we explored the functional repertoire of the potential auxiliary metabolic genes carried by these prophages, encoding functions involved in carbohydrate metabolism and degradation, amino acid metabolism and carbon metabolism. Conclusion: Our study provides an enhanced understanding of the diversity and prevalence of lysogens in infant and adult gut microbiota and suggests a complex interplay between prophages and their bacterial hosts.

Sources of gut microbiota variation in a large longitudinal Finnish infant cohort.

BACKGROUND: Although the infant gut microbiota has been extensively studied, comprehensive assessment on the microbiota determinants including technical variables has not been performed in large infant cohorts. METHODS: We studied the effect of 109 variables on the 16S rRNA gene amplicon-based gut microbiota profiles of infants sampled longitudinally from three weeks to two years of life in the Finnish HELMi birth cohort. Spot faecal samples from both parents were included for intra-family analyses, totalling to 7657 samples from 985 families that were evaluated for beta-diversity patterns using permutational multivariate analysis on Bray-Curtis distances, and differential abundance testing and alpha-diversity for variables of interest. We also assessed the effect of different taxonomic levels and distance methods. FINDINGS: In time point-specific models, the largest share of variation explained, up to 2-6%, were seen in decreasing order for the DNA extraction batch, delivery mode and related perinatal exposures, defecation frequency and parity/siblings. Variables describing the infant gastrointestinal function were continuously important during the first two years, reflecting changes in e.g., feeding habits. The effect of parity/siblings on infant microbiota was modified by birth mode and exposure to intrapartum antibiotics, exemplifying the tight interlinkage of perinatal factors relevant for infant microbiota research. In total, up to 19% of the biological microbiota variation in the infant gut could be explained. Our results highlight the need to interpret variance partitioning results in the context of each cohort's characteristics and microbiota processing. INTERPRETATION: Our study provides a comprehensive report of key factors associated with infant gut microbiota composition across the two first years of life in a homogenous cohort. The study highlights possible important future research areas and confounding factors to be considered. FUNDING: This research was supported by Business Finland, Academy of Finland, Foundation for Nutrition Research and the Doctoral Program in Microbiology and Biotechnology, University of Helsinki, Finland.

Evaluation of computational phage detection tools for metagenomic datasets.

Introduction: As new computational tools for detecting phage in metagenomes are being rapidly developed, a critical need has emerged to develop systematic benchmarks. Methods: In this study, we surveyed 19 metagenomic phage detection tools, 9 of which could be installed and run at scale. Those 9 tools were assessed on several benchmark challenges. Fragmented reference genomes are used to assess the effects of fragment length, low viral content, phage taxonomy, robustness to eukaryotic contamination, and computational resource usage. Simulated metagenomes are used to assess the effects of sequencing and assembly quality on the tool performances. Finally, real human gut metagenomes and viromes are used to assess the differences and similarities in the phage communities predicted by the tools. Results: We find that the various tools yield strikingly different results. Generally, tools that use a homology approach (VirSorter, MARVEL, viralVerify, VIBRANT, and VirSorter2) demonstrate low false positive rates and robustness to eukaryotic contamination. Conversely, tools that use a sequence composition approach (VirFinder, DeepVirFinder, Seeker), and MetaPhinder, have higher sensitivity, including to phages with less representation in reference databases. These differences led to widely differing predicted phage communities in human gut metagenomes, with nearly 80% of contigs being marked as phage by at least one tool and a maximum overlap of 38.8% between any two tools. While the results were more consistent among the tools on viromes, the differences in results were still significant, with a maximum overlap of 60.65%. Discussion: Importantly, the benchmark datasets developed in this study are publicly available and reusable to enable the future comparability of new tools developed.

Bacteroides abundance drives birth mode dependent infant gut microbiota developmental trajectories.

Background and aims: Birth mode and other early life factors affect a newborn's microbial colonization with potential long-term health effects. Individual variations in early life gut microbiota development, especially their effects on the functional repertoire of microbiota, are still poorly characterized. This study aims to provide new insights into the gut microbiome developmental trajectories during the first year of life. Methods: Our study comprised 78 term infants sampled at 3 weeks, 3 months, 6 months, and 12 months (n = 280 total samples), and their mothers were sampled in late pregnancy (n = 50). Fecal DNA was subjected to shotgun metagenomic sequencing. Infant samples were studied for taxonomic and functional maturation, and maternal microbiota was used as a reference. Hierarchical clustering on taxonomic profiles was used to identify the main microbiota developmental trajectories in the infants, and their associations with perinatal and postnatal factors were assessed. Results: In line with previous studies, infant microbiota composition showed increased alpha diversity and decreased beta diversity by age, converging toward an adult-like profile. However, we did not observe an increase in functional alpha diversity, which was stable and comparable with the mother samples throughout all the sampling points. Using a de novo clustering approach, two main infant microbiota clusters driven by Bacteroidaceae and Clostridiaceae emerged at each time point. The clusters were associated with birth mode and their functions differed mainly in terms of biosynthetic and carbohydrate degradation pathways, some of which consistently differed between the clusters for all the time points. The longitudinal analysis indicated three main microbiota developmental trajectories, with the majority of the infants retaining their characteristic cluster until 1 year. As many as 40% of vaginally delivered infants were grouped with infants delivered by C-section due to their clear and persistent depletion in Bacteroides. Intrapartum antibiotics, any perinatal or postnatal factors, maternal microbiota composition, or other maternal factors did not explain the depletion in Bacteroides in the subset of vaginally born infants. Conclusion: Our study provides an enhanced understanding of the compositional and functional early life gut microbiota trajectories, opening avenues for investigating elusive causes that influence non-typical microbiota development.

Ontology-driven analysis of marine metagenomics: what more can we learn from our data?

BACKGROUND: The proliferation of metagenomic sequencing technologies has enabled novel insights into the functional genomic potentials and taxonomic structure of microbial communities. However, cyberinfrastructure efforts to manage and enable the reproducible analysis of sequence data have not kept pace. Thus, there is increasing recognition of the need to make metagenomic data discoverable within machine-searchable frameworks compliant with the FAIR (Findability, Accessibility, Interoperability, and Reusability) principles for data stewardship. Although a variety of metagenomic web services exist, none currently leverage the hierarchically structured terminology encoded within common life science ontologies to programmatically discover data. RESULTS: Here, we integrate large-scale marine metagenomic datasets with community-driven life science ontologies into a novel FAIR web service. This approach enables the retrieval of data discovered by intersecting the knowledge represented within ontologies against the functional genomic potential and taxonomic structure computed from marine sequencing data. Our findings highlight various microbial functional and taxonomic patterns relevant to the ecology of prokaryotes in various aquatic environments. CONCLUSIONS: In this work, we present and evaluate a novel Semantic Web architecture that can be used to ask novel biological questions of existing marine metagenomic datasets. Finally, the FAIR ontology searchable data products provided by our API can be leveraged by future research efforts.

Ontology-Enriched Specifications Enabling Findable, Accessible, Interoperable, and Reusable Marine Metagenomic Datasets in Cyberinfrastructure Systems.

Marine microbial ecology requires the systematic comparison of biogeochemical and sequence data to analyze environmental influences on the distribution and variability of microbial communities. With ever-increasing quantities of metagenomic data, there is a growing need to make datasets Findable, Accessible, Interoperable, and Reusable (FAIR) across diverse ecosystems. FAIR data is essential to developing analytical frameworks that integrate microbiological, genomic, ecological, oceanographic, and computational methods. Although community standards defining the minimal metadata required to accompany sequence data exist, they haven't been consistently used across projects, precluding interoperability. Moreover, these data are not machine-actionable or discoverable by cyberinfrastructure systems. By making 'omic and physicochemical datasets FAIR to machine systems, we can enable sequence data discovery and reuse based on machine-readable descriptions of environments or physicochemical gradients. In this work, we developed a novel technical specification for dataset encapsulation for the FAIR reuse of marine metagenomic and physicochemical datasets within cyberinfrastructure systems. This includes using Frictionless Data Packages enriched with terminology from environmental and life-science ontologies to annotate measured variables, their units, and the measurement devices used. This approach was implemented in Planet Microbe, a cyberinfrastructure platform and marine metagenomic web-portal. Here, we discuss the data properties built into the specification to make global ocean datasets FAIR within the Planet Microbe portal. We additionally discuss the selection of, and contributions to marine-science ontologies used within the specification. Finally, we use the system to discover data by which to answer various biological questions about environments, physicochemical gradients, and microbial communities in meta-analyses. This work represents a future direction in marine metagenomic research by proposing a specification for FAIR dataset encapsulation that, if adopted within cyberinfrastructure systems, would automate the discovery, exchange, and re-use of data needed to answer broader reaching questions than originally intended.

Cyanolichen microbiome contains novel viruses that encode genes to promote microbial metabolism.

Lichen thalli are formed through the symbiotic association of a filamentous fungus and photosynthetic green alga and/or cyanobacterium. Recent studies have revealed lichens also host highly diverse communities of secondary fungal and bacterial symbionts, yet few studies have examined the viral component within these complex symbioses. Here, we describe viral biodiversity and functions in cyanolichens collected from across North America and Europe. As current machine-learning viral-detection tools are not trained on complex eukaryotic metagenomes, we first developed efficient methods to remove eukaryotic reads prior to viral detection and a custom pipeline to validate viral contigs predicted with three machine-learning methods. Our resulting high-quality viral data illustrate that every cyanolichen thallus contains diverse viruses that are distinct from viruses in other terrestrial ecosystems. In addition to cyanobacteria, predicted viral hosts include other lichen-associated bacterial lineages and algae, although a large fraction of viral contigs had no host prediction. Functional annotation of cyanolichen viral sequences predicts numerous viral-encoded auxiliary metabolic genes (AMGs) involved in amino acid, nucleotide, and carbohydrate metabolism, including AMGs for secondary metabolism (antibiotics and antimicrobials) and fatty acid biosynthesis. Overall, the diversity of cyanolichen AMGs suggests that viruses may alter microbial interactions within these complex symbiotic assemblages.

Planet Microbe: a platform for marine microbiology to discover and analyze interconnected 'omics and environmental data.

In recent years, large-scale oceanic sequencing efforts have provided a deeper understanding of marine microbial communities and their dynamics. These research endeavors require the acquisition of complex and varied datasets through large, interdisciplinary and collaborative efforts. However, no unifying framework currently exists for the marine science community to integrate sequencing data with physical, geological, and geochemical datasets. Planet Microbe is a web-based platform that enables data discovery from curated historical and on-going oceanographic sequencing efforts. In Planet Microbe, each 'omics sample is linked with other biological and physiochemical measurements collected for the same water samples or during the same sample collection event, to provide a broader environmental context. This work highlights the need for curated aggregation efforts that can enable new insights into high-quality metagenomic datasets. Planet Microbe is freely accessible from https://www.planetmicrobe.org/.

The dynamic wound microbiome.

BACKGROUND: Diabetic foot ulcers (DFUs) account for the majority of all limb amputations and hospitalizations due to diabetes complications. With 30 million cases of diabetes in the USA and 500,000 new diagnoses each year, DFUs are a growing health problem. Diabetes patients with limb amputations have high postoperative mortality, a high rate of secondary amputation, prolonged inpatient hospital stays, and a high incidence of re-hospitalization. DFU-associated amputations constitute a significant burden on healthcare resources that cost more than 10 billion dollars per year. Currently, there is no way to identify wounds that will heal versus those that will become severely infected and require amputation. MAIN BODY: Accurate identification of causative pathogens in diabetic foot ulcers is a critical component of effective treatment. Compared to traditional culture-based methods, advanced sequencing technologies provide more comprehensive and unbiased profiling on wound microbiome with a higher taxonomic resolution, as well as functional annotation such as virulence and antibiotic resistance. In this review, we summarize the latest developments in defining the microbiology of diabetic foot ulcers that have been unveiled by sequencing technologies and discuss both the future promises and current limitations of these approaches. In particular, we highlight the temporal patterns and system dynamics in the diabetic foot microbiome monitored and measured during wound progression and medical intervention, and explore the feasibility of molecular diagnostics in clinics. CONCLUSION: Molecular tests conducted during weekly office visits to clean and examine DFUs would allow clinicians to offer personalized treatment and antibiotic therapy. Personalized wound management could reduce healthcare costs, improve quality of life for patients, and recoup lost productivity that is important not only to the patient, but also to healthcare payers and providers. These efforts could also improve antibiotic stewardship and control the rise of "superbugs" vital to global health.

Identification and quantitation of clinically relevant microbes in patient samples: Comparison of three k-mer based classifiers for speed, accuracy, and sensitivity.

Infections are a serious health concern worldwide, particularly in vulnerable populations such as the immunocompromised, elderly, and young. Advances in metagenomic sequencing availability, speed, and decreased cost offer the opportunity to supplement or even replace culture-based identification of pathogens with DNA sequence-based diagnostics. Adopting metagenomic analysis for clinical use requires that all aspects of the workflow are optimized and tested, including data analysis and computational time and resources. We tested the accuracy, sensitivity, and resource requirements of three top metagenomic taxonomic classifiers that use fast k-mer based algorithms: Centrifuge, CLARK, and KrakenUniq. Binary mixtures of bacteria showed all three reliably identified organisms down to 1% relative abundance, while only the relative abundance estimates of Centrifuge and CLARK were accurate. All three classifiers identified the organisms present in their default databases from a mock bacterial community of 20 organisms, but only Centrifuge had no false positives. In addition, Centrifuge required far less computational resources and time for analysis. Centrifuge analysis of metagenomes obtained from samples of VAP, infected DFUs, and FN showed Centrifuge identified pathogenic bacteria and one virus that were corroborated by culture or a clinical PCR assay. Importantly, in both diabetic foot ulcer patients, metagenomic sequencing identified pathogens 4-6 weeks before culture. Finally, we show that Centrifuge results were minimally affected by elimination of time-consuming read quality control and host screening steps.

NCBI's Virus Discovery Hackathon: Engaging Research Communities to Identify Cloud Infrastructure Requirements.

A wealth of viral data sits untapped in publicly available metagenomic data sets when it might be extracted to create a usable index for the virological research community. We hypothesized that work of this complexity and scale could be done in a hackathon setting. Ten teams comprised of over 40 participants from six countries, assembled to create a crowd-sourced set of analysis and processing pipelines for a complex biological data set in a three-day event on the San Diego State University campus starting 9 January 2019. Prior to the hackathon, 141,676 metagenomic data sets from the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) were pre-assembled into contiguous assemblies (contigs) by NCBI staff. During the hackathon, a subset consisting of 2953 SRA data sets (approximately 55 million contigs) was selected, which were further filtered for a minimal length of 1 kb. This resulted in 4.2 million (Mio) contigs, which were aligned using BLAST against all known virus genomes, phylogenetically clustered and assigned metadata. Out of the 4.2 Mio contigs, 360,000 contigs were labeled with domains and an additional subset containing 4400 contigs was screened for virus or virus-like genes. The work yielded valuable insights into both SRA data and the cloud infrastructure required to support such efforts, revealing analysis bottlenecks and possible workarounds thereof. Mainly: (i) Conservative assemblies of SRA data improves initial analysis steps; (ii) existing bioinformatic software with weak multithreading/multicore support can be elevated by wrapper scripts to use all cores within a computing node; (iii) redesigning existing bioinformatic algorithms for a cloud infrastructure to facilitate its use for a wider audience; and (iv) a cloud infrastructure allows a diverse group of researchers to collaborate effectively. The scientific findings will be extended during a follow-up event. Here, we present the applied workflows, initial results, and lessons learned from the hackathon.

iMicrobe: Tools and data-dreaiven discovery platform for the microbiome sciences.

BACKGROUND: Scientists have amassed a wealth of microbiome datasets, making it possible to study microbes in biotic and abiotic systems on a population or planetary scale; however, this potential has not been fully realized given that the tools, datasets, and computation are available in diverse repositories and locations. To address this challenge, we developed iMicrobe.us, a community-driven microbiome data marketplace and tool exchange for users to integrate their own data and tools with those from the broader community. FINDINGS: The iMicrobe platform brings together analysis tools and microbiome datasets by leveraging National Science Foundation-supported cyberinfrastructure and computing resources from CyVerse, Agave, and XSEDE. The primary purpose of iMicrobe is to provide users with a freely available, web-based platform to (1) maintain and share project data, metadata, and analysis products, (2) search for related public datasets, and (3) use and publish bioinformatics tools that run on highly scalable computing resources. Analysis tools are implemented in containers that encapsulate complex software dependencies and run on freely available XSEDE resources via the Agave API, which can retrieve datasets from the CyVerse Data Store or any web-accessible location (e.g., FTP, HTTP). CONCLUSIONS: iMicrobe promotes data integration, sharing, and community-driven tool development by making open source data and tools accessible to the research community in a web-based platform.

The Promises and Pitfalls of Machine Learning for Detecting Viruses in Aquatic Metagenomes.

Tools allowing for the identification of viral sequences in host-associated and environmental metagenomes allows for a better understanding of the genetics and ecology of viruses and their hosts. Recently, new approaches using machine learning methods to distinguish viral from bacterial signal using k-mer sequence signatures were published for identifying viral contigs in metagenomes. The promise of these content-based approaches is the ability to discover new viruses, with no or few known relatives. In this perspective paper, we examine the use of the content-based machine learning tool VirFinder for the identification of viral sequences in aquatic metagenomes and explore the possibility of using ecosystem-focused models targeted to marine metagenomes. We discuss the impact of the training set composition on the tool performance and the current limitation for the retrieval of low abundance viral sequences in metagenomes. We identify potential biases that could arise from machine learning approaches for viral hunting in real-world datasets and suggest possible avenues to overcome them.

Libra: scalable k-mer-based tool for massive all-vs-all metagenome comparisons.

Background: Shotgun metagenomics provides powerful insights into microbial community biodiversity and function. Yet, inferences from metagenomic studies are often limited by dataset size and complexity and are restricted by the availability and completeness of existing databases. De novo comparative metagenomics enables the comparison of metagenomes based on their total genetic content. Results: We developed a tool called Libra that performs an all-vs-all comparison of metagenomes for precise clustering based on their k-mer content. Libra uses a scalable Hadoop framework for massive metagenome comparisons, Cosine Similarity for calculating the distance using sequence composition and abundance while normalizing for sequencing depth, and a web-based implementation in iMicrobe (http://imicrobe.us) that uses the CyVerse advanced cyberinfrastructure to promote broad use of the tool by the scientific community. Conclusions: A comparison of Libra to equivalent tools using both simulated and real metagenomic datasets, ranging from 80 million to 4.2 billion reads, reveals that methods commonly implemented to reduce compute time for large datasets, such as data reduction, read count normalization, and presence/absence distance metrics, greatly diminish the resolution of large-scale comparative analyses. In contrast, Libra uses all of the reads to calculate k-mer abundance in a Hadoop architecture that can scale to any size dataset to enable global-scale analyses and link microbial signatures to biological processes.

Molecular response to nitrogen starvation by Frankia alni ACN14a revealed by transcriptomics and functional analysis with a fosmid library in Escherichia coli.

The transcriptome of Frankia alni strain ACN14a was compared between in vitro ammonium-replete (N-replete) and ammonium-free dinitrogen-fixing (N-fixing) conditions using DNA arrays. A Welch-test (p < 0.05) revealed significant upregulation of 252 genes under N-fixing vs. N-replete (fold-change (FC) ≥ 2), as well as significant downregulation of 48 other genes (FC ≤ 0.5). Interestingly, there were 104 Frankia genes upregulated in vitro that were also significantly upregulated in symbiosis with Alnus glutinosa, while the other 148 genes were not, showing that the physiology of in vitro fixation is markedly different from that under symbiotic conditions. In particular,in vitro fixing cells were seen to upregulate genes identified as coding for a nitrite reductase, and amidases that were not upregulated in symbiosis. Confirmatory assays for nitrite reductase showed that Frankia indeed reduced nitrite and used it as a nitrogen source. An Escherichia coli fosmid clone carrying the nirB region was able to grow better in the presence of 5 mM nitrite than without it, confirming the function of the genome region. The physiological pattern that emerges shows that Frankia undergoes nitrogen starvation that induces a molecular response different from that seen in symbiosis.

Phage hunters: Computational strategies for finding phages in large-scale 'omics datasets.

A plethora of tools exist for identifying phage sequences in bacterial genomes, single cell amplified genomes, and host-associated and environmental metagenomes. Yet because the genetics of phages and their hosts are closely intertwined, distinguishing viral from bacterial signal remains an ongoing challenge. Further the size, quantity and fragmentary nature of modern 'omics datasets ushers in a new set of computational challenges. Here, we detail the promises and pitfalls of using currently available gene-centric or k-mer based tools for identifying prophage sequences in genomes and prophage and viral contigs in metagenomes. Each of these methods offers a unique piece of the puzzle to elucidating the intriguing signatures of phage-host coevolution.

Endoplasmic Reticulum Transport of Glutathione by Sec61 Is Regulated by Ero1 and Bip.

In the endoplasmic reticulum (ER), Ero1 catalyzes disulfide bond formation and promotes glutathione (GSH) oxidation to GSSG. Since GSSG cannot be reduced in the ER, maintenance of the ER glutathione redox state and levels likely depends on ER glutathione import and GSSG export. We used quantitative GSH and GSSG biosensors to monitor glutathione import into the ER of yeast cells. We found that glutathione enters the ER by facilitated diffusion through the Sec61 protein-conducting channel, while oxidized Bip (Kar2) inhibits transport. Increased ER glutathione import triggers H2O2-dependent Bip oxidation through Ero1 reductive activation, which inhibits glutathione import in a negative regulatory loop. During ER stress, transport is activated by UPR-dependent Ero1 induction, and cytosolic glutathione levels increase. Thus, the ER redox poise is tuned by reciprocal control of glutathione import and Ero1 activation. The ER protein-conducting channel is permeable to small molecules, provided the driving force of a concentration gradient.