Genetic Discovery Enabled by A Large Language Model.

Artificial intelligence (AI) has been used in many areas of medicine, and recently large language models (LLMs) have shown potential utility for clinical applications. However, since we do not know if the use of LLMs can accelerate the pace of genetic discovery, we used data generated from mouse genetic models to investigate this possibility. We examined whether a recently developed specialized LLM (Med-PaLM 2) could analyze sets of candidate genes generated from analysis of murine models of biomedical traits. In response to free-text input, Med-PaLM 2 correctly identified the murine genes that contained experimentally verified causative genetic factors for six biomedical traits, which included susceptibility to diabetes and cataracts. Med-PaLM 2 was also able to analyze a list of genes with high impact alleles, which were identified by comparative analysis of murine genomic sequence data, and it identified a causative murine genetic factor for spontaneous hearing loss. Based upon this Med-PaLM 2 finding, a novel bigenic model for susceptibility to spontaneous hearing loss was developed. These results demonstrate Med-PaLM 2 can analyze gene-phenotype relationships and generate novel hypotheses, which can facilitate genetic discovery.

Neuron Navigator 1 (Nav1) regulates the response to cocaine in mice.

Genetic variation accounts for much of the risk for developing a substance use disorder, but the underlying genetic factors and their genetic effector mechanisms are mostly unknown. Inbred mouse strains exhibit substantial and heritable differences in the extent of voluntary cocaine self-administration. Computational genetic analysis of cocaine self-administration data obtained from twenty-one inbred strains identified Nav1, a member of the neuron navigator family that regulates dendrite formation and axonal guidance, as a candidate gene. To test this genetic hypothesis, we generated and characterized Nav1 knockout mice. Consistent with the genetic prediction, Nav1 knockout mice exhibited increased voluntary cocaine intake and had increased motivation for cocaine consumption. Immunohistochemistry, electrophysiology, and transcriptomic studies were performed as a starting point for investigating the mechanism for the Nav1 knockout effect. Nav1 knockout mice had a reduced inhibitory synapse density in their cortex, increased excitatory synaptic transmission in their cortex and hippocampus, and increased excitatory neurons in a deep cortical layer. Collectively, our results indicate that Nav1 regulates the response to cocaine, and we identified Nav1 knockout induced changes in the excitatory and inhibitory synaptic balance in the cortex and hippocampus that could contribute to this effect.

Analysis of structural variation among inbred mouse strains.

BACKGROUND: 'Long read' sequencing methods have been used to identify previously uncharacterized structural variants that cause human genetic diseases. Therefore, we investigated whether long read sequencing could facilitate genetic analysis of murine models for human diseases. RESULTS: The genomes of six inbred strains (BTBR T + Itpr3tf/J, 129Sv1/J, C57BL/6/J, Balb/c/J, A/J, SJL/J) were analyzed using long read sequencing. Our results revealed that (i) Structural variants are very abundant within the genome of inbred strains (4.8 per gene) and (ii) that we cannot accurately infer whether structural variants are present using conventional short read genomic sequence data, even when nearby SNP alleles are known. The advantage of having a more complete map was demonstrated by analyzing the genomic sequence of BTBR mice. Based upon this analysis, knockin mice were generated and used to characterize a BTBR-unique 8-bp deletion within Draxin that contributes to the BTBR neuroanatomic abnormalities, which resemble human autism spectrum disorder. CONCLUSION: A more complete map of the pattern of genetic variation among inbred strains, which is produced by long read genomic sequencing of the genomes of additional inbred strains, could facilitate genetic discovery when murine models of human diseases are analyzed.

Optimizing a therapy for opiate use disorders: Characterizing ondansetron pharmacokinetics in blood and brain.

Administration of a widely used 5-hydroxytryptamine receptor (5HT3A R) antagonist (ondansetron) potently inhibited the development of experimentally induced opioid dependence and withdrawal responses in mice and humans. However, in several studies examining withdrawal symptoms in subjects with chronic opioid use disorders (OUDs), ondansetron exhibited reduced or absent efficacy. Because attenuation of opioid withdrawal symptomatology is mediated within the brain, this study examined single-dose ondansetron pharmacokinetics in the blood and brain of mice. We demonstrate that ondansetron concentrations in the brain (Cbrain ng/mg) are 1000-fold lower than the blood concentrations (Cblood ng/ml) and decrease rapidly after ondansetron administration; and that a large percentage of brain ondansetron remains in the ventricular fluid. These results indicate that the ondansetron dose, and the time window between ondansetron and opioid administration, and when withdrawal is assessed are critical considerations for clinical studies involving subjects with chronic OUD. The pharmacokinetic results and the dosing considerations discussed here can be used to improve the design of subsequent clinical trials, which will test whether a more prolonged period of ondansetron administration can provide a desperately needed therapy that can prevent the development of neonatal opioid withdrawal syndrome in babies born to mothers with chronic OUD.

Rack1 function in intestinal epithelia: regulating crypt cell proliferation and regeneration and promoting differentiation and apoptosis.

Previously, we showed that receptor for activated C kinase 1 (Rack1) regulates growth of colon cells in vitro, partly by suppressing Src kinase activity at key cell cycle checkpoints, in apoptotic and cell survival pathways and at cell-cell adhesions. Here, we generated mouse models of Rack1 deficiency to assess Rack1's function in intestinal epithelia in vivo. Intestinal Rack1 deficiency resulted in proliferation of crypt cells, diminished differentiation of crypt cells into enterocyte, goblet, and enteroendocrine cell lineages, and expansion of Paneth cell populations. Following radiation injury, the morphology of Rack1-deleted small bowel was strikingly abnormal with development of large polypoid structures that contained many partly formed villi, numerous back-to-back elongated and regenerating crypts, and high-grade dysplasia in surface epithelia. These abnormalities were not observed in Rack1-expressing areas of intestine or in control mice. Following irradiation, apoptosis of enterocytes was strikingly reduced in Rack1-deleted epithelia. These novel findings reveal key functions for Rack1 in regulating growth of intestinal epithelia: suppressing crypt cell proliferation and regeneration, promoting differentiation and apoptosis, and repressing development of neoplasia. NEW & NOTEWORTHY Our findings reveal novel functions for receptor for activated C kinase 1 (Rack1) in regulating growth of intestinal epithelia: suppressing crypt cell proliferation and regeneration, promoting differentiation and apoptosis, and repressing development of neoplasia.

Rack1 maintains intestinal homeostasis by protecting the integrity of the epithelial barrier.

Previously, we generated mouse models of Rack1 deficiency to identify key functions for Rack1 in regulating growth of intestinal epithelia: suppressing crypt cell proliferation and regeneration, promoting differentiation and apoptosis, and repressing development of neoplasia. However, other than low body weight, we did not detect an overt phenotype in mice constitutively deleted of Rack1 in intestinal epithelia ( vil-Cre: Rack1fl/fl mice), presumably because Rack1 was deleted in <10% of the total surface area of the epithelia. To assess the effect of Rack1 loss throughout the entire intestinal epithelia, we generated another mouse model of Rack1 deficiency, vil-Cre-ERT2: Rack1fl/fl. Within 5-10 days of the initial tamoxifen treatment, the mice lost over 20% of their body weight, developed severe diarrhea that for some was bloody, became critically ill, and died, if not euthanized. Necropsies revealed mildly distended, fluid-, gas-, and sometimes blood-filled loops of small and large bowel, inguinal lymphadenopathy, and thrombocytosis. Rack1 was deleted in nearly 100% of the epithelia in both the small intestine and colon when assessed by immunofluorescent or immunoblot analyses. Rack1 expression in other tissues and organs was not different than in control mice, indicating tissue specificity of the recombination. Histopathology revealed a patchy, erosive, hemorrhagic, inflammatory enterocolitis with denuded, sloughed off surface epithelium, and crypt hyperplasia. These results suggest a protective function for Rack1 in maintaining the integrity of intestinal epithelia and for survival. NEW & NOTEWORTHY Our findings reveal a novel function for Rack1 in maintaining intestinal homeostasis by protecting the epithelial barrier. Rack1 loss results in a patchy, erosive, hemorrhagic, inflammatory enterocolitis, which resembles that of inflammatory bowel diseases (IBD) in humans. Understanding mechanisms that protect barrier function in normal intestine and how loss of that protection contributes to the pathogenesis of IBD could lead to improved therapies for these and other erosive diseases of the gastrointestinal tract.

Intestinal Enteroendocrine Lineage Cells Possess Homeostatic and Injury-Inducible Stem Cell Activity.

Several cell populations have been reported to possess intestinal stem cell (ISC) activity during homeostasis and injury-induced regeneration. Here, we explored inter-relationships between putative mouse ISC populations by comparative RNA-sequencing (RNA-seq). The transcriptomes of multiple cycling ISC populations closely resembled Lgr5+ ISCs, the most well-defined ISC pool, but Bmi1-GFP+ cells were distinct and enriched for enteroendocrine (EE) markers, including Prox1. Prox1-GFP+ cells exhibited sustained clonogenic growth in vitro, and lineage-tracing of Prox1+ cells revealed long-lived clones during homeostasis and after radiation-induced injury in vivo. Single-cell mRNA-seq revealed two subsets of Prox1-GFP+ cells, one of which resembled mature EE cells while the other displayed low-level EE gene expression but co-expressed tuft cell markers, Lgr5 and Ascl2, reminiscent of label-retaining secretory progenitors. Our data suggest that the EE lineage, including mature EE cells, comprises a reservoir of homeostatic and injury-inducible ISCs, extending our understanding of cellular plasticity and stemness.

An important role for RNase R in mRNA decay.

mRNA decay is a major determinant of gene expression. In Escherichia coli, message degradation initiates with an endoribonucleolytic cleavage followed by exoribonuclease digestion to generate 5'-mononucleotides. Although the 3' to 5' processive exoribonucleases, PNPase and RNase II, have long been considered to be mediators of this digestion, we show here that another enzyme, RNase R, also participates in the process. RNase R is particularly important for removing mRNA fragments with extensive secondary structure, such as those derived from the many mRNAs that contain REP elements. In the absence of RNase R and PNPase, REP-containing fragments accumulate to high levels. RNase R is unusual among exoribonucleases in that, by itself, it can digest through extensive secondary structure provided that a single-stranded binding region, such as a poly(A) tail, is present. These data demonstrate that RNase R, which is widespread in prokaryotes and eukaryotes, is an important participant in mRNA decay.

Quality control of ribosomal RNA mediated by polynucleotide phosphorylase and RNase R.

Despite their overall accuracy, errors in macromolecular processes, such as rRNA synthesis and ribosome assembly, inevitably occur. However, whether these errors are remediated and how this might be accomplished is not known. In previous work, we showed that a double mutant strain lacking both polynucleotide phosphorylase (PNPase) and RNase R activities is inviable. In the course of examining the molecular basis for this phenotype, we found that shifting a temperature-sensitive mutant strain to 42 degrees C led to cessation of growth and loss of cell viability. Northern analysis of RNA isolated from such cells after the temperature shift revealed that fragments of 16S and 23S rRNA accumulated to a high level, and that the amount of ribosomes and ribosomal subunits decreased due to defects in ribosome assembly. rRNA fragments were not detected at 31 degrees C or when single mutant strains were grown at 42 degrees C. Pulse-chase analysis showed that the rRNA fragments appeared within 5 min at 42 degrees C, and that they accumulated before the loss of cell viability. The data are consistent with a model in which PNPase and RNase R mediate a previously unknown quality control process that normally removes defective rRNAs as soon as they are generated. In the absence of these RNases, rRNA fragments accumulate, leading to interference with ribosome maturation and ultimately to cell death.

Purification and characterization of the Escherichia coli exoribonuclease RNase R. Comparison with RNase II.

Escherichia coli RNase R, a 3' --> 5' exoribonuclease homologous to RNase II, was overexpressed and purified to near homogeneity in its native untagged form by a rapid procedure. The purified enzyme was free of nucleic acid. It migrated upon gel filtration chromatography as a monomer with an apparent molecular mass of approximately 95 kDa, in close agreement with its expected size based on the sequence of the rnr gene. RNase R was most active at pH 7.5-9.5 in the presence of 0.1-0.5 mm Mg(2+) and 50-500 mm KCl. The enzyme shares many catalytic properties with RNase II. Both enzymes are nonspecific processive ribonucleases that release 5'-nucleotide monophosphates and leave a short undigested oligonucleotide core. However, whereas RNase R shortens RNA processively to di- and trinucleotides, RNase II becomes more distributive when the length of the substrate reaches approximately 10 nucleotides, and it leaves an undigested core of 3-5 nucleotides. Both enzymes work on substrates with a 3'-phosphate group. RNase R and RNase II are most active on synthetic homopolymers such as poly(A), but their substrate specificities differ. RNase II is more active on poly(A), whereas RNase R is much more active on rRNAs. Neither RNase R nor RNase II can degrade a complete RNA-RNA or DNA-RNA hybrid or one with a 4-nucleotide 3'-RNA overhang. RNase R differs from RNase II in that it cannot digest DNA oligomers and is not inhibited by such molecules, suggesting that it does not bind DNA. Although the in vivo function of RNase R is not known, its ability to digest certain natural RNAs may explain why it is maintained in E. coli together with RNase II.