Evolution of T cells in the cancer-resistant naked mole-rat.

Naked mole-rats (NMRs) are best known for their extreme longevity and cancer resistance, suggesting that their immune system might have evolved to facilitate these phenotypes. Natural killer (NK) and T cells have evolved to detect and destroy cells infected with pathogens and to provide an early response to malignancies. While it is known that NMRs lack NK cells, likely lost during evolution, little is known about their T-cell subsets in terms of the evolution of the genes that regulate their function, their clonotypic diversity, and the thymus where they mature. Here we find, using single-cell transcriptomics, that NMRs have a large circulating population of γδT cells, which in mice and humans mostly reside in peripheral tissues and induce anti-cancer cytotoxicity. Using single-cell-T-cell-receptor sequencing, we find that a cytotoxic γδT-cell subset of NMRs harbors a dominant clonotype, and that their conventional CD8 αßT cells exhibit modest clonotypic diversity. Consistently, perinatal NMR thymuses are considerably smaller than those of mice yet follow similar involution progression. Our findings suggest that NMRs have evolved under a relaxed intracellular pathogenic selective pressure that may have allowed cancer resistance and longevity to become stronger targets of selection to which the immune system has responded by utilizing γδT cells.

Fully Phased Sequence of a Diploid Human Genome Determined de Novo from the DNA of a Single Individual.

In recent years, improved sequencing technology and computational tools have made de novo genome assembly more accessible. Many approaches, however, generate either an unphased or only partially resolved representation of a diploid genome, in which polymorphisms are detected but not assigned to one or the other of the homologous chromosomes. Yet chromosomal phase information is invaluable for the understanding of phenotypic trait inheritance in the cases of compound heterozygosity, allele-specific expression or cis-acting variants. Here we use a combination of tools and sequencing technologies to generate a de novo diploid assembly of the human primary cell line WI-38. First, data from PacBio single molecule sequencing and Bionano Genomics optical mapping were combined to generate an unphased assembly. Next, 10x Genomics linked reads were combined with the hybrid assembly to generate a partially phased assembly. Lastly, we developed and optimized methods to use short-read (Illumina) sequencing of flow cytometry-sorted metaphase chromosomes to provide phase information. The final genome assembly was almost fully (94%) phased with the addition of approximately 2.5-fold coverage of Illumina data from the sequenced metaphase chromosomes. The diploid nature of the final de novo genome assembly improved the resolution of structural variants between the WI-38 genome and the human reference genome. The phased WI-38 sequence data are available for browsing and download at wi38.research.calicolabs.com. Our work shows that assembling a completely phased diploid genome de novo from the DNA of a single individual is now readily achievable.

Solo: Doublet Identification in Single-Cell RNA-Seq via Semi-Supervised Deep Learning.

Single-cell RNA sequencing (scRNA-seq) measurements of gene expression enable an unprecedented high-resolution view into cellular state. However, current methods often result in two or more cells that share the same cell-identifying barcode; these "doublets" violate the fundamental premise of single-cell technology and can lead to incorrect inferences. Here, we describe Solo, a semi-supervised deep learning approach that identifies doublets with greater accuracy than existing methods. Solo embeds cells unsupervised using a variational autoencoder and then appends a feed-forward neural network layer to the encoder to form a supervised classifier. We train this classifier to distinguish simulated doublets from the observed data. Solo can be applied in combination with experimental doublet detection methods to further purify scRNA-seq data to true single cells. It is freely available from https://github.com/calico/solo. A record of this paper's transparent peer review process is included in the Supplemental Information.

Single-cell transcriptomics of the naked mole-rat reveals unexpected features of mammalian immunity.

The immune system comprises a complex network of specialized cells that protects against infection, eliminates cancerous cells, and regulates tissue repair, thus serving a critical role in homeostasis, health span, and life span. The subterranean-dwelling naked mole-rat (NM-R; Heterocephalus glaber) exhibits prolonged life span relative to its body size, is unusually cancer resistant, and manifests few physiological or molecular changes with advancing age. We therefore hypothesized that the immune system of NM-Rs evolved unique features that confer enhanced cancer immunosurveillance and prevent the age-associated decline in homeostasis. Using single-cell RNA-sequencing (scRNA-seq) we mapped the immune system of the NM-R and compared it to that of the short-lived, cancer-prone mouse. In contrast to the mouse, we find that the NM-R immune system is characterized by a high myeloid-to-lymphoid cell ratio that includes a novel, lipopolysaccharide (LPS)-responsive, granulocyte cell subset. Surprisingly, we also find that NM-Rs lack canonical natural killer (NK) cells. Our comparative genomics analyses support this finding, showing that the NM-R genome lacks an expanded gene family that controls NK cell function in several other species. Furthermore, we reconstructed the evolutionary history that likely led to this genomic state. The NM-R thus challenges our current understanding of mammalian immunity, favoring an atypical, myeloid-biased mode of innate immunosurveillance, which may contribute to its remarkable health span.

Systems biology-guided identification of synthetic lethal gene pairs and its potential use to discover antibiotic combinations.

Mathematical models of metabolism from bacterial systems biology have proven their utility across multiple fields, for example metabolic engineering, growth phenotype simulation, and biological discovery. The usefulness of the models stems from their ability to compute a link between genotype and phenotype, but their ability to accurately simulate gene-gene interactions has not been investigated extensively. Here we assess how accurately a metabolic model for Escherichia coli computes one particular type of gene-gene interaction, synthetic lethality, and find that the accuracy rate is between 25% and 43%. The most common failure modes were incorrect computation of single gene essentiality and biological information that was missing from the model. Moreover, we performed virtual and biological screening against several synthetic lethal pairs to explore whether two-compound formulations could be found that inhibit the growth of Gram-negative bacteria. One set of molecules was identified that, depending on the concentrations, inhibits E. coli and S. enterica serovar Typhimurium in an additive or antagonistic manner. These findings pinpoint specific ways in which to improve the predictive ability of metabolic models, and highlight one potential application of systems biology to drug discovery and translational medicine.

Reconciling a Salmonella enterica metabolic model with experimental data confirms that overexpression of the glyoxylate shunt can rescue a lethal ppc deletion mutant.

The in silico reconstruction of metabolic networks has become an effective and useful systems biology approach to predict and explain many different cellular phenotypes. When simulation outputs do not match experimental data, the source of the inconsistency can often be traced to incomplete biological information that is consequently not captured in the model. To address this problem, general approaches continue to be needed that can suggest experimentally testable hypotheses to reconcile inconsistencies between simulation and experimental data. Here, we present such an approach that focuses specifically on correcting cases in which experimental data show a particular gene to be essential but model simulations do not. We use metabolic models to predict efficient compensatory pathways, after which cloning and overexpression of these pathways are performed to investigate whether they restore growth and to help determine why these compensatory pathways are not active in mutant cells. We demonstrate this technique for a ppc knockout of Salmonella enterica serovar Typhimurium; the inability of cells to route flux through the glyoxylate shunt when ppc is removed was correctly identified by our approach as the cause of the discrepancy. These results demonstrate the feasibility of our approach to drive biological discovery while simultaneously refining metabolic network reconstructions.

Exploiting adaptive laboratory evolution of Streptomyces clavuligerus for antibiotic discovery and overproduction.

Adaptation is normally viewed as the enemy of the antibiotic discovery and development process because adaptation among pathogens to antibiotic exposure leads to resistance. We present a method here that, in contrast, exploits the power of adaptation among antibiotic producers to accelerate the discovery of antibiotics. A competition-based adaptive laboratory evolution scheme is presented whereby an antibiotic-producing microorganism is competed against a target pathogen and serially passed over time until the producer evolves the ability to synthesize a chemical entity that inhibits growth of the pathogen. When multiple Streptomyces clavuligerus replicates were adaptively evolved against methicillin-resistant Staphylococcus aureus N315 in this manner, a strain emerged that acquired the ability to constitutively produce holomycin. In contrast, no holomycin could be detected from the unevolved wild-type strain. Moreover, genome re-sequencing revealed that the evolved strain had lost pSCL4, a large 1.8 Mbp plasmid, and acquired several single nucleotide polymorphisms in genes that have been shown to affect secondary metabolite biosynthesis. These results demonstrate that competition-based adaptive laboratory evolution can constitute a platform to create mutants that overproduce known antibiotics and possibly to discover new compounds as well.

Genetic basis of growth adaptation of Escherichia coli after deletion of pgi, a major metabolic gene.

Bacterial survival requires adaptation to different environmental perturbations such as exposure to antibiotics, changes in temperature or oxygen levels, DNA damage, and alternative nutrient sources. During adaptation, bacteria often develop beneficial mutations that confer increased fitness in the new environment. Adaptation to the loss of a major non-essential gene product that cripples growth, however, has not been studied at the whole-genome level. We investigated the ability of Escherichia coli K-12 MG1655 to overcome the loss of phosphoglucose isomerase (pgi) by adaptively evolving ten replicates of E. coli lacking pgi for 50 days in glucose M9 minimal medium and by characterizing endpoint clones through whole-genome re-sequencing and phenotype profiling. We found that 1) the growth rates for all ten endpoint clones increased approximately 3-fold over the 50-day period; 2) two to five mutations arose during adaptation, most frequently in the NADH/NADPH transhydrogenases udhA and pntAB and in the stress-associated sigma factor rpoS; and 3) despite similar growth rates, at least three distinct endpoint phenotypes developed as defined by different rates of acetate and formate secretion. These results demonstrate that E. coli can adapt to the loss of a major metabolic gene product with only a handful of mutations and that adaptation can result in multiple, alternative phenotypes.