To help learn how phytopathogens feed from their hosts, genes for nutrient transporters from the hemibiotrophic potato and tomato pest Phytophthora infestans were annotated. This identified 453 genes from 19 families. Comparisons with a necrotrophic oomycete, Pythium ultimum var. ultimum, and a hemibiotrophic fungus, Magnaporthe oryzae, revealed diversity in the size of some families although a similar fraction of genes encoded transporters. RNA-seq of infected potato tubers, tomato leaves, and several artificial media revealed that 56 and 207 transporters from P. infestans were significantly up- or down-regulated, respectively, during early infection timepoints of leaves or tubers versus media. About 17 were up-regulated >4-fold in both leaves and tubers compared to media and expressed primarily in the biotrophic stage. The transcription pattern of many genes was host-organ specific. For example, the mRNA level of a nitrate transporter (NRT) was about 100-fold higher during mid-infection in leaves, which are nitrate-rich, than in tubers and three types of artificial media, which are nitrate-poor. The NRT gene is physically linked with genes encoding nitrate reductase (NR) and nitrite reductase (NiR), which mobilize nitrate into ammonium and amino acids. All three genes were coregulated. For example, the three genes were expressed primarily at mid-stage infection timepoints in both potato and tomato leaves, but showed little expression in potato tubers. Transformants down-regulated for all three genes were generated by DNA-directed RNAi, with silencing spreading from the NR target to the flanking NRT and NiR genes. The silenced strains were nonpathogenic on leaves but colonized tubers. We propose that the nitrate assimilation genes play roles both in obtaining nitrogen for amino acid biosynthesis and protecting P. infestans from natural or fertilization-induced nitrate and nitrite toxicity.

BackgroundAn important feature of eukaryotic evolution is metabolic compartmentalization, in which certain pathways are restricted to the cytosol or specific organelles. Glycolysis in eukaryotes is described as a cytosolic process. The universality of this canon has been challenged by recent genome data that suggest that some glycolytic enzymes made by stramenopiles bear mitochondrial targeting peptides.ResultsMining of oomycete, diatom, and brown algal genomes indicates that stramenopiles encode two forms of enzymes for the second half of glycolysis, one with and the other without mitochondrial targeting peptides. The predicted mitochondrial targeting was confirmed by using fluorescent tags to localize phosphoglycerate kinase, phosphoglycerate mutase, and pyruvate kinase in Phytophthora infestans, the oomycete that causes potato blight. A genome-wide search for other enzymes with atypical mitochondrial locations identified phosphoglycerate dehydrogenase, phosphoserine aminotransferase, and phosphoserine phosphatase, which form a pathway for generating serine from the glycolytic intermediate 3-phosphoglycerate. Fluorescent tags confirmed the delivery of these serine biosynthetic enzymes to P. infestans mitochondria. A cytosolic form of this serine biosynthetic pathway, which occurs in most eukaryotes, is missing from oomycetes and most other stramenopiles. The glycolysis and serine metabolism pathways of oomycetes appear to be mosaics of enzymes with different ancestries. While some of the noncanonical oomycete mitochondrial enzymes have the closest affinity in phylogenetic analyses with proteins from other stramenopiles, others cluster with bacterial, plant, or animal proteins. The genes encoding the mitochondrial phosphoglycerate kinase and serine-forming enzymes are physically linked on oomycete chromosomes, which suggests a shared origin.ConclusionsStramenopile metabolism appears to have been shaped through the acquisition of genes by descent and lateral or endosymbiotic gene transfer, along with the targeting of the proteins to locations that are novel compared to other eukaryotes. Colocalization of the glycolytic and serine biosynthesis enzymes in mitochondria is apparently necessary since they share a common intermediate. The results indicate that descriptions of metabolism in textbooks do not cover the full diversity of eukaryotic biology.Electronic supplementary materialThe online version of this article (10.1186/s12862-017-1087-8) contains supplementary material, which is available to authorized users.

The use of host nutrients to support pathogen growth is central to disease. We addressed the relationship between metabolism and trophic behavior by comparing metabolic gene expression during potato tuber colonization by two oomycetes, the hemibiotroph Phytophthora infestans and the necrotroph Pythium ultimum . Genes for several pathways including amino acid, nucleotide, and cofactor biosynthesis were expressed more by Ph . infestans during its biotrophic stage compared to Py . ultimum . In contrast, Py . ultimum had higher expression of genes for metabolizing compounds that are normally sequestered within plant cells but released to the pathogen upon plant cell lysis, such as starch and triacylglycerides. The transcription pattern of metabolic genes in Ph . infestans during late infection became more like that of Py . ultimum , consistent with the former's transition to necrotrophy. Interspecific variation in metabolic gene content was limited but included the presence of γ-amylase only in Py . ultimum . The pathogens were also found to employ strikingly distinct strategies for using nitrate. Measurements of mRNA, 15 N labeling studies, enzyme assays, and immunoblotting indicated that the assimilation pathway in Ph . infestans was nitrate-insensitive but induced during amino acid and ammonium starvation. In contrast, the pathway was nitrate-induced but not amino acid-repressed in Py . ultimum . The lack of amino acid repression in Py . ultimum appears due to the absence of a transcription factor common to fungi and Phytophthora that acts as a nitrogen metabolite repressor. Evidence for functional diversification in nitrate reductase protein was also observed. Its temperature optimum was adapted to each organism's growth range, and its K m was much lower in Py . ultimum . In summary, we observed divergence in patterns of gene expression, gene content, and enzyme function which contribute to the fitness of each species in its niche.

Introduction Healthcare insurance claims data contain an unrecognized wealth of structured data that can be leveraged to investigate epidemiologic and economic relationships in health and disease. We studied the feasibility for machine learning algorithms to improve upon screening for obstructive and central sleep apnea (SA) at the population health level using existing health insurance claims data. Methods A logistic regression model was trained to predict the presence or absence of SA from an aggregated healthcare insurance claims dataset. The dataset was composed of medical and pharmacy claims between the years 2016 and 2020 from the Wisconsin All-Payor claims database which included coverage of >4,000,000 patients, >10,000 ICD codes, and >$50 billion in medical spending. A total of 1,870,000 patients and 39,712 unique federal drug identification codes were included within 91.5 million pharmacy claims in the dataset. Input features were constructed by counting the total number of claims for each unique drug in each subject resulting in a patient-level feature vector of 39,712 drug frequencies. The positive SA population was defined by individuals who had both at least one medical claim for sleep apnea diagnosis (ICD codes G4733/G4731) and an appropriate sleep test (CPT codes 9580*/9581*). The logistic regression model was evaluated using randomized 10-fold cross-validation and performance reported using ROC-AUC statistics and top-10 feature importance analysis. Results The logistic regression model detecting SA based solely on observed medication frequencies produced a ROC-AUC of 0.77. In a feature importance analysis, three of the Top-10 most discriminative features were medications for the treatment of diabetes, hypertension, and hyperlipidemia. We hypothesize this drug-frequency based model functions by exploiting the strong correlation of SA with specific clusters of known co-morbid conditions and corresponding medication regimens. Conclusion We demonstrate health insurance claims records contain predictive information that can aid in more systematic screening of undiagnosed conditions like SA. Furthermore, in a statistical analysis of feature importance, we observed medications indicative of comorbidities with known association to SA. These findings are useful to clinicians and payers in identifying undiagnosed SA populations, including those responsible for value-based payment models. Support (If Any)