Whole-genome high-fidelity sequencing: A novel approach to detecting and characterization of mutagenicity in vivo.

Direct DNA sequencing can be used for characterizing mutagenicity in simple and complex biological models. Recently we described a method of whole-genome sequencing for detecting mutations in simple models of cultured bacteria, mammalian cells, and nematode. In the current proof-of-concept study, we expand and improve our method for evaluating a more complex mammalian biological model in outbred mice. We detail the method by applying it to a small set of animals treated with a mutagen with known mutagenicity profiles, N-ethyl-N-nitrosourea (ENU), for consistency with the known data. Whole-genome high-fidelity sequencing (HiFi Sequencing) showed frequencies and spectra of background mutations in tissues of untreated mice that were consistent with normal ageing and characterized by spontaneous or enzymatic deamination of 5-methylcytosine. In mice treated with a single 40 mg/kg dose of ENU, the frequency of mutations in the genomic DNA of solid tissues increased up to 7-fold, with the greatest increase observed in the spleen and the smallest increase in the liver. The most common mutations detected in ENU-treated mice were T > A transitions and T > C transversions, consistent with the types of mutations caused by alkylating agents. The data suggest that HiFi Sequencing may be useful for characterizing mutagenicity of novel compounds in various biological models.

Unbiased whole genome detection of ultrarare off-target mutations in genome-edited cell populations by HiFi sequencing.

DNA base editors (BEs) composed of a nuclease-deficient Cas9 fused to a DNA-modifying enzyme can achieve on-target mutagenesis without creating double-strand DNA breaks (DSBs). As a result, BEs generate far less DNA damage than traditional nuclease-proficient Cas9 systems, which do rely on the creation of DSBs to achieve on-target mutagenesis. The inability of BEs to create DSBs makes the detection of their undesired off-target effects very difficult. PacBio HiFi sequencing can efficiently detect ultrarare mutations resulting from chemical mutagenesis in whole genomes with a sensitivity ~1 × 10-8 mutations per base pair. In this proof-of-principle study, we evaluated whether this technique could also detect the on- and off-target mutations generated by a cytosine-to-thymine (C>T) BE targeting the LacZ gene in Escherichia coli (E. coli). HiFi sequencing detected on-target mutant allele fractions ranging from ~7% to ~63%, depending on the single-guide RNA (sgRNA) used, while no on-target mutations were detected in controls lacking the BE. The presence of the BE resulted in a ~3-fold increase in mutation frequencies compared to controls lacking the BE, irrespective of the sgRNA used. These increases were mostly composed of C:G>T:A substitutions distributed throughout the genome. Our results demonstrate that HiFi sequencing can efficiently identify on- and off-target mutations in cell populations that have undergone genome editing.

Evaluation of the mutagenic effects of Molnupiravir and N4-hydroxycytidine in bacterial and mammalian cells by HiFi sequencing.

Molnupiravir (MOV) is used to treat COVID-19. In cells, MOV is converted to the ribonucleoside analog N4-hydroxycytidine (NHC) and incorporated into the SARS-CoV-2 RNA genome during its replication, resulting in RNA mutations. The widespread accumulation of such mutations inhibits SARS-CoV-2 propagation. Although safety assessments by many regulatory agencies across the world have concluded that the genotoxic risks associated with the clinical use of MOV are low, concerns remain that it could induce DNA mutations in patients, particularly because numerous in vitro studies have shown that NHC is a DNA mutagen. In this study, we used HiFi sequencing, a technique that can detect ultralow-frequency substitution mutations in whole genomes, to evaluate the mutagenic effects of MOV in E. coli and of MOV and NHC in mouse lymphoma L5178Y cells and human lymphoblastoid TK6 cells. In all models, exposure to these compounds increased genome-wide mutation frequencies in a dose-dependent manner, and these increases were mainly composed of A:T → G:C transitions. The NHC exposure concentrations used for mammalian cells were comparable to those observed in the plasma of humans who received clinical doses of MOV.

Genome-wide detection of ultralow-frequency substitution mutations in cultures of mouse lymphoma L5178Y cells and Caenorhabditis elegans worms by PacBio sequencing.

Many conventional genetic toxicology assays require specialized cell cultures or animals and can only detect mutations that inactivate the function of a reporter gene. These limitations make such assays incompatible with many toxicological models but could be overcome by the development of techniques capable of directly detecting genome-wide somatic mutations through DNA sequencing. PacBio sequencing can generate almost error-free consensus reads by repeatedly inspecting both DNA strands from circularized molecules (a method known as PacBio HiFi). In this study, we show that PacBio HiFi can detect genome-wide ultralow-frequency substitution mutations in cultures of mouse lymphoma L5178Y cells and Caenorhabditis elegans worms. The mutation frequencies (MFs) of unexposed samples in both models were ~1 × 10-7 mutations per base pair. Compared to these controls, PacBio HiFi detected MF increases of 23-fold in cultures of L5178Y cells exposed to 5 mM ethyl methanosulfonate (EMS) for 4 h, and 5-, 12-, and 29-fold in cultures of C. elegans worms exposed to 12.5, 25, and 50 mM EMS for 4 h, respectively. In both models, the mutation spectra of controls were diverse, while those derived from EMS-exposed samples were dominated by C:G → T:A transitions. To validate these results, clone sequencing analyses were performed on the same cultures of L5178Y cells. The results obtained by clone sequencing and PacBio HiFi were almost identical. Our results suggest that PacBio sequencing could be used for the detection, quantitation, and characterization of mutations in any DNA-containing sample, including those that are not compatible with conventional mutation detection approaches.

PacBio sequencing detects genome-wide ultra-low-frequency substitution mutations resulting from exposure to chemical mutagens.

Genetic toxicology uses several assays to identity mutagens and protects the public. Most of these assays, however, rely on reporter genes, can only measure mutation indirectly based on phenotype, and often require specific cell lines or animal models-features that impede their integration with existing and emerging toxicological models, such as organoids. In this study, we show that PacBio Single-Molecule, Real-Time (PB SMRT) sequencing identified substitution mutations caused by chemical mutagens in Escherichia coli by generating nearly error-free consensus reads after repeatedly inspecting both strands of circular DNA molecules. Using DNA from E. coli exposed to ethyl methanosulfonate (EMS) or N-ethyl-N-nitrosourea (ENU), PB SMRT sequencing detected mutation frequencies (MFs) and spectra comparable to those obtained by clone-sequencing from the same exposures. The optimized background MF of PB SMRT sequencing was ≤ 1 × 10-7 mutations per base pair (mut/bp).

Mutational signatures in T-lymphocytes of rats treated with N-propyl-N-nitrosourea and procarbazine.

We have used whole genome sequencing (WGS) to determine mutational signatures induced in the T-cells of rats treated in vivo with N-propyl-N-nitrosourea (PNU) or procarbazine (PCZ). The signatures from the treated rats were different from the signature of background mutations. The main component of the spontaneous T-cell mutational signature was C➔T transition with all other single base substitutions evenly distributed. The PNU-induced mutational signature showed relatively equal contributions from C➔T and T➔C transitions, and T➔A transversions. The PCZ-induced signature was characterized by T➔C transitions, T➔A and, to a smaller extent, T➔G transversions. C➔G transversions were infrequent in either the PNU or PCZ signatures. WGS not only allowed mutational signature detection, but also measured quantitative responses to mutagen treatment: 10-40× increases in the number of mutations per clone were detected in T-cell clones from treated rats. The overall strand specificity of induced mutations for annotated rat genes was comparable to the strand specificity of mutations determined previously for the endogenous X-linked Pig-a gene. Our results provide valuable reference data for future applications of WGS in safety research and risk assessment.

Pig-a gene mutations in bone marrow granulocytes of procarbazine-treated F344 rats.

It was previously demonstrated that procarbazine (PCZ) is positive in the rat erythrocyte Pig-a gene mutation assay. However, since mammalian erythrocytes lack genomic DNA, it was necessary to analyze nucleated bone-marrow erythroid precursor cells to confirm that PCZ induces mutations in the Pig-a gene (Revollo et al., Environ Mol Mutagen, 2020). In this study, the association between Pig-a mutation and loss of GPI anchors was further strengthened and the genesis of Pig-a mutation in PCZ-dosed rats was evaluated by analyzing bone-marrow granulocytes. Erythrocytes and granulocytes both originate from myeloid progenitor cells, but granulocytes contain DNA throughout their developmental stages. F344 rats were treated with three doses of 150 mg/kg PCZ; 2 weeks later, CD48-deficient mutant phenotype bone-marrow granulocytes (BMGs [CD11b+ ]) were isolated by flow-cytometric sorting. Sequencing data showed that the CD48-deficient mutant phenotype BMGs contained mutations in the Pig-a gene while wild-type BMGs did not. PCZ-induced mutations included missense, nonsense and splice site variants; the majority of mutations were A > T, A > C, and A > G, with the mutated A on the nontranscribed DNA strand. The PCZ-induced mutational analysis in BMGs supports the association between the phenotype measured in the Pig-a assay and mutation in the Pig-a gene. Also, PCZ mutation spectra were similar in bone-marrow erythroids and BMGs, but none of the mutations detected in BMGs were the same as the erythroid precursor cell mutations from the same rats. Thus, mutations induced in the Pig-a assay appear to be induced after commitment of myeloid progenitor cells to either the granulocyte or erythroid pathway.

Molecular analysis of GPI-anchor biosynthesis pathway genes in rat strains used for the Pig-a gene mutation assay.

Recent studies support the assumption that mutation of the X-linked Pig-a gene is most likely responsible for the mutant phenotype of the cells deficient in glycosylphosphatidylinositol (GPI)-anchored proteins quantified in the rodent Pig-a gene mutation assay. In humans, however, mutations in both alleles of one of the 30 other genes involved in GPI-anchor synthesis, e.g., PIG-L and PIG-O, cause reduced expression of surface GPI-anchored proteins. Here, we investigated the possibility that the loss of the GPI-anchor detected by the rat Pig-a assay also could be caused by mutation in other GPI-biosynthesis genes. 31 samples were obtained from 8 inbred and outbred rat strains commonly used for genetic toxicology assays. In order to investigate possible sources of variation in the Pig-a assay, variant DNA sequences were evaluated in Cd59 and 24 GPI-biosynthesis genes. In some genes, such as Pig-n and Pig-u, homozygous variations occurred in all animals, suggesting that these variations are due to deviations in the reference genome. Heterozygous Pig-s, Pig-w, Pig-o, Pig-c, Pgap1, Pgap2, Pig-k and Pig-t variations were found, however, indicating that these genes could serve as targets for mutation in the assay. Protein alignment for these altered genes was conducted with possible human, mouse and rat phenotypic mutants from the literature; this analysis demonstrated that many of the variations that we detected were in non-conserved sequences and that no phenotypes for any of these variants could be inferred from known mutants from the literature. All heterozygous variants were in outbred rats. Overall, the findings of this study cannot totally rule out the possibility that mutations in GPI-biosynthesis genes other than Pig-a are detected in the Pig-a assay, but suggest that if it occurs, it must occur only rarely and therefore mutations in genes other than Pig-a have little impact on rat-based experiments.

CD59-deficient bone marrow erythroid cells from rats treated with procarbazine and propyl-nitrosourea have mutations in the Pig-a gene.

Procarbazine (PCZ) and N-propyl-N-nitrosourea (PNU) are rodent mutagens and carcinogens. Both induce GPI-anchored marker-deficient mutant-phenotype red blood cells (RBCs) in the flow cytometry-based rat RBC Pig-a assay. In the present study, we traced the origin of the RBC mutant phenotype by analyzing Pig-a mutations in the precursors of RBCs, bone marrow erythroid cells (BMEs). Rats were exposed to a total of 450 mg/kg PCZ hydrochloride or 300 mg/kg PNU, and bone marrow was collected 2, 7, and 10 weeks later. Using a flow cell sorter, we isolated CD59-deficient mutant-phenotype BMEs from PCZ- and PNU-treated rats and examined their endogenous X-linked Pig-a gene by next generation sequencing. Pig-a mutations consistent with the properties of PCZ and PNU were found in sorted mutant-phenotype BMEs. PCZ induced mainly A > T transversions with the mutated A on the nontranscribed strand of the Pig-a gene, while PNU induced mainly T > A transversions with the mutated T on the nontranscribed strand. The treatment-induced mutations were distributed across the protein coding sequence of the Pig-a gene. The causal relationship between BMEs and RBCs and the agent-specific mutational spectra in CD59-deicient BMEs indicate that the rat RBC Pig-a assay, scoring CD59-deficient mutant-phenotype RBCs in peripheral blood, detects Pig-a gene mutation.

Pig-a mutations in bone marrow erythroblasts of rats treated with 7,12-dimethyl-benz[a]anthracene.

Flow cytometry-based phenotypic detection of red blood cells (RBCs) deficient in surface markers anchored by glycosylphosphatidylinositol (GPI) is an efficient tool for monitoring somatic mutation in mammalian species. Biochemical considerations suggest that GPI-anchored marker-deficient RBCs found in peripheral blood are due to mutations in the endogenous X-linked phosphatidylinositolglycan, class A gene (Pig-a gene). Yet the linkage between the detected mutant phenotype and the actual mutation in the Pig-a gene is difficult to establish directly in mammalian RBCs that are naturally free of genomic DNA and may have only traces of heavily degraded mRNA. We have traced the origin of the marker-deficient RBC phenotype in the precursors of peripheral RBCs, bone marrow erythroid cells (BMEs, also known as erythroblasts), in rats treated by gavage with 75 mg/kg of the potent mutagen, 7,12-dimethyl-benz[a]anthracene (DMBA). The frequencies of marker-deficient BMEs were significantly increased in DMBA-treated rats. We identified Pig-a mutations in sorted mutant phenotype BMEs. The spectrum of DMBA-induced Pig-a mutations in erythroid lineage cells was identical to the spectra of mutations previously determined for the Pig-a and for another X-linked reporter gene, hypoxanthine-guanine phosphoribosyltransferase gene, in cells of lymphoid lineage, spleen T-lymphocytes. Our observations lend additional support to the hypothesis that GPI-anchored marker-deficient RBCs are true Pig-a mutants.

Evaluation of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) mutagenicity using in vitro and in vivo Pig-a assays.

4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) is a genotoxic carcinogen found in tobacco and tobacco smoke. Several in vitro and in vivo assays have been used for evaluating the genotoxicity of tobacco smoke and tobacco smoke constituents like NNK, yet it is not clear which in vitro assays are most appropriate for extrapolating the in vitro responses of these test agents to animal models and humans. The Pig-a gene mutation assay can be performed in vitro, in laboratory animals, and in humans, a potential benefit in estimating in vivo responses from in vitro data. In the current study we used Pig-a as a reporter of gene mutation both in vitro, in L5178Y/Tk+/- cells, and in vivo, in Sprague-Dawley rats. NNK significantly increased Pig-a mutant frequency in L5178Y/Tk+/- cells, but only at concentrations of 100 µg/ml and greater, and only in the presence of S9 activation. Pig-a mutations in L5178Y/Tk+/- cells were detected in 80% of the NNK-induced mutants, with the predominate mutation being G→A transition; vehicle control mutants contained deletions. In the in vivo study, rats were exposed to NNK daily for 90 days by inhalation, a common route of exposure to NNK for humans. Although elevated mutant frequencies were detected, these responses were not clearly associated with NNK exposure, so that overall, the in vivo Pig-a assays were negative. Thus, while NNK induces mutations in the in vitro Pig-a assay, the in vivo Pig-a assay has limited ability to detect NNK mutagenicity under conditions relevant to NNK exposure in smokers.

Analysis of mutation in the rat Pig-a assay: II. Studies with bone marrow granulocytes.

The in vivo erythrocyte Pig-a gene mutation assay measures the phenotypic loss of GPI-anchored surface markers. Molecular analysis of the marker-deficient erythrocytes cannot provide direct proof that the mutant phenotype is due to mutation in the Pig-a gene because mammalian erythrocytes lack genomic DNA. Granulocytes are nucleated cells that originate from myeloid progenitor cells in bone marrow as is the case for erythrocytes, and thus analysis of Pig-a mutation in bone marrow granulocytes can provide information about the source of mutations detected in the erythrocyte Pig-a assay. We developed a flow cytometric Pig-a assay for bone marrow granulocytes and evaluated granulocyte Pig-a mutant frequencies in bone marrow from male rats treated acutely with N-ethyl-N-nitrosourea (ENU). Bone marrow cells from these rats were stained with anti-CD11b for identifying granulocytes and anti-CD48 for detecting the Pig-a mutant phenotype. The average Pig-a mutant frequency in granulocyte precursors of control rats was 8.42 × 10-6 , whereas in ENU-treated rats it was 567.13 × 10-6 . CD11b-positive/CD48-deficient mutant cells were enriched using magnetic separation and sorted into small pools for sequencing. While there were no Pig-a mutations found in sorted CD48-positive wild-type cells, Pig-a mutations were detected in mutant granulocyte precursors. The most frequent mutation observed was T→A transversion, followed by T→C transition and T→G transversion, with the mutated T on the nontranscribed DNA strand. While the spectrum of mutations in bone marrow granulocytes was similar to that of erythroid cells, different Pig-a mutations were found in mutant-phenotype granulocytes and erythroids from the same bone marrow samples, suggesting that most Pig-a mutations were induced in bone marrow cells after commitment to either the granulocyte or erythroid developmental pathway. Environ. Mol. Mutagen. 59:733-741, 2018. © 2018 Wiley Periodicals, Inc.

Analysis of mutation in the rat Pig-a assay: I) studies with bone marrow erythroid cells.

We have established a flow cytometry-based Pig-a assay for rat bone marrow erythroid cells (BMEs). The BME Pig-a assay uses a DNA-specific stain and two antibodies: one against the transmembrane transferrin receptor (CD71 marker) and the other against the GPI-anchored complement inhibitory protein (CD59 marker). In F344 male rats treated acutely with a total of 120 mg/kg of N-ethyl-N-nitrosourea (ENU) the frequency of CD59-deficient phenotypically mutant BMEs increased approximately 24-fold compared to the rats concurrently treated with the vehicle. Such an increase of mutant BMEs coincides with increases of CD59-deficient reticulocytes measured in rats treated with similar doses of ENU. Sequence analysis of the endogenous X-linked Pig-a gene of CD59-deficient BMEs revealed that they are Pig-a mutants. The spectrum of ENU-induced Pig-a mutations in these BMEs was consistent with the in vivo mutagenic signature of ENU: 73% of mutations occurred at A:T basepairs, with the mutated T on the nontranscribed strand of the gene. T→A transversion was the most frequent mutation followed by T→C transition; no deletion or insertion mutations were present in the spectrum. Since BMEs are precursors of peripheral red blood cells, our findings suggest that CD59-deficient erythrocytes measured in the flow cytometric erythrocyte Pig-a assay develop from BMEs containing mutations in the Pig-a gene. Thus, the erythrocyte Pig-a assay detects mutation in the Pig-a gene. Environ. Mol. Mutagen. 59:722-732, 2018. © 2018 Wiley Periodicals, Inc.

Genome-wide mutation detection by interclonal genetic variation.

Genetic toxicology assays estimate mutation frequencies by phenotypically screening for the activation or inactivation of endogenous or exogenous reporter genes. These reporters can only detect mutations in narrow areas of the genome and their use is often restricted to certain in vitro and in vivo models. Here, we show that Interclonal Genetic Variation (ICGV) can directly identify mutations genome-wide by comparing sequencing data of single-cell clones derived from the same source or organism. Upon ethyl methanesulfonate (EMS) exposure, ICGV detected greater levels of mutation in a dose- and time-dependent manner in E. coli. In addition, ICGV was also able to identify a ∼20-fold increase in somatic mutations in T-cell clones derived from an N-ethyl-N-nitrosourea (ENU)-treated rat vs. a vehicle-treated rat. These results demonstrate that the genetic differences of single-cell clones can be used for genome-wide mutation detection.

Whole genome sequencing of mouse lymphoma L5178Y-3.7.2C (TK+/-) reveals millions of mutations and genetic markers.

The mouse lymphoma L5178Y-3.7.2C (TK+/-) cell line is extensively used in genetic toxicology to conduct the mouse lymphoma assay (MLA). The MLA is used to establish the mutagenic and clastogenic effects of chemicals and pharmaceuticals, and is one of the few genetic tests widely accepted by regulatory agencies throughout the world. Despite the extensive use and regulatory impact of L5178Y-3.7.2C (TK+/-) cells, little is known about their genetic composition or how it affects the outcome of the MLA. To determine the genetic background of this cell line, we sequenced and analyzed its entire genome. Our results confirm the existence of previously described mutations in the Tk1 and Trp53 genes and catalog millions of other mutations, many of which impair the function of genes with key roles in cell physiology and genetic toxicology.

Mutation analysis with random DNA identifiers (MARDI) catalogs Pig-a mutations in heterogeneous pools of CD48-deficient T cells derived from DMBA-treated rats.

Identification of mutations induced by xenotoxins is a common task in the field of genetic toxicology. Mutations are often detected by clonally expanding potential mutant cells and genotyping each viable clone by Sanger sequencing. Such a "clone-by-clone" approach requires significant time and effort, and sometimes is even impossible to implement. Alternative techniques for efficient mutation identification would greatly benefit both basic and regulatory genetic toxicology research. Here, we report the development of Mutation Analysis with Random DNA Identifiers (MARDI), a novel high-fidelity Next Generation Sequencing (NGS) approach that circumvents clonal expansion and directly catalogs mutations in pools of mutant cells. MARDI uses oligonucleotides carrying Random DNA Identifiers (RDIs) to tag progenitor DNA molecules before PCR amplification, enabling clustering of descendant DNA molecules and eliminating NGS- and PCR-induced sequencing artifacts. When applied to the Pig-a cDNA analysis of heterogeneous pools of CD48-deficient T cells derived from DMBA-treated rats, MARDI detected nearly all Pig-a mutations that were previously identified by conventional clone-by-clone analysis and discovered many additional ones consistent with DMBA exposure: mostly A to T transversions, with the mutated A located on the non-transcribed DNA strand.

HES1 is a master regulator of glucocorticoid receptor-dependent gene expression.

Hairy and enhancer of split-1 (HES1) is a basic helix-loop-helix transcription factor that is a key regulator of development and organogenesis. However, little is known about the role of HES1 after birth. Glucocorticoids, primary stress hormones that are essential for life, regulate numerous homeostatic processes that permit vertebrates to cope with physiological challenges. The molecular actions of glucocorticoids are mediated by glucocorticoid receptor-dependent regulation of nearly 25% of the genome. Here, we established a genome-wide molecular link between HES1 and glucocorticoid receptors that controls the ability of cells and animals to respond to stress. Glucocorticoid signaling rapidly and robustly silenced HES1 expression. This glucocorticoid-dependent repression of HES1 was necessary for the glucocorticoid receptor to regulate many of its target genes. Mice with conditional knockout of HES1 in the liver exhibited an expanded glucocorticoid receptor signaling profile and aberrant metabolic phenotype. Our results indicate that HES1 acts as a master repressor, the silencing of which is required for proper glucocorticoid signaling.

The ways and means that fine tune Sirt1 activity.

Sirt1 is the most evolutionarily conserved mammalian sirtuin. It plays a vital role in the regulation of metabolism, stress responses, genome stability, and ultimately aging. Although much attention has focused on the identification of the cellular targets and functional networks controlled by Sirt1, the mechanisms that regulate Sirt1 activity by biological stimuli have only recently begun to emerge. As an enzyme, the activity of Sirt1 can be controlled by the availability of its substrates, post-translational modifications, interactions with other proteins, or changes in its expression levels. In this review, we briefly discuss the ways and means by which the activity of Sirt1 is fine-tuned under different conditions.

Mechanisms generating diversity in glucocorticoid receptor signaling.

Glucocorticoids regulate diverse biological processes throughout the body via the glucocorticoid receptor (GR). Ligand-bound GR translocates into the nucleus and can elicit changes in gene expression by direct contact with the DNA or by protein-protein interactions with other transcription factors. The GR can also mediate rapid nongenomic signaling events initiated in the cytoplasm. In this chapter, we review the biological and physiological implications of glucocorticoids, the GR, and many of the signal transduction mechanisms that mediate their action.

Nampt/PBEF/Visfatin regulates insulin secretion in beta cells as a systemic NAD biosynthetic enzyme.

Intracellular nicotinamide phosphoribosyltransferase (iNampt) is an essential enzyme in the NAD biosynthetic pathway. An extracellular form of this protein (eNampt) has been reported to act as a cytokine named PBEF or an insulin-mimetic hormone named visfatin, but its physiological relevance remains controversial. Here we show that eNampt does not exert insulin-mimetic effects in vitro or in vivo but rather exhibits robust NAD biosynthetic activity. Haplodeficiency and chemical inhibition of Nampt cause defects in NAD biosynthesis and glucose-stimulated insulin secretion in pancreatic islets in vivo and in vitro. These defects are corrected by administration of nicotinamide mononucleotide (NMN), a product of the Nampt reaction. A high concentration of NMN is present in mouse plasma, and plasma eNampt and NMN levels are reduced in Nampt heterozygous females. Our results demonstrate that Nampt-mediated systemic NAD biosynthesis is critical for beta cell function, suggesting a vital framework for the regulation of glucose homeostasis.