Biology subjects

Feder, A.

Publications and source records attributed to Feder, A..

8 recordsLinked to original sources

bioRxiv2023

Parallel Evolution of Linezolid Resistant Staphylococcus aureus in Patients with Cystic Fibrosis

BackgroundLinezolid is an antibiotic used to treat serious Staphylococcus aureus infections. Resistance to linezolid is considered rare but could emerge with repeated dosing. We recently reported widespread prescription of linezolid for a cohort of patients with cystic fibrosis (CF). ObjectivesThe goals of this study were to determine the incidence of linezolid resistance in CF and identify molecular mechanisms for linezolid resistance. MethodsWe identified patients with S. aureus resistant to linezolid (MIC > 4) at the University of Iowa CF Center between 2008 and 2018. We obtained isolates from these patients and retested susceptibility to linezolid using broth microdilution. We used whole genome sequencing to perform phylogenetic analysis of linezolid resistant isolates and examine sequences for mutations or accessory genes that confer linezolid resistance. Main ResultsBetween 2008 and 2018, 111 patients received linezolid and 4 of these patients cultured linezolid resistant S. aureus. We sequenced 11 resistant and 21 susceptible isolates from these 4 subjects. Phylogenetic analysis indicated that linezolid resistance developed in ST5 or ST105 backgrounds. Three individuals had linezolid resistant S. aureus with a G2576T mutation in 23S rRNA. One of these subjects additionally had a mutS- mutL- hypermutating S. aureus that produced 5 resistant isolates with multiple ribosomal subunit mutations. In one subject, the genetic basis for linezolid resistance was unclear. ConclusionsLinezolid resistance evolved in 4 of 111 patients in this study. Linezolid resistance occurred by multiple genetic mechanisms. All resistant strains developed in ST5 or ST105 MRSA backgrounds. Key PointLinezolid resistance arises through multiple genetic mechanisms and could be facilitated by mutator phenotypes. Linezolid resistance was transient, possibly due to growth disadvantage.

microbiology↗

bioRxiv2022

Correspondence between the layered structure of deep language models and temporal structure of natural language processing in the human brain

Deep language models (DLMs) provide a novel computational paradigm for how the brain processes natural language. Unlike symbolic, rule-based models described in psycholinguistics, DLMs encode words and their context as continuous numerical vectors. These "embeddings" are constructed by a sequence of computations organized in "layers" to ultimately capture surprisingly sophisticated representations of linguistic structures. How does this layered hierarchy map onto the human brain during natural language comprehension? In this study, we used electrocorticography (ECoG) to record neural activity in language areas along the superior temporal gyrus and inferior frontal gyrus while human participants listened to a 30-minute spoken narrative. We supplied this same narrative to a high-performing DLM (GPT2-XL) and extracted the contextual embeddings for each word in the story across all 48 layers of the model. We next trained a set of linear encoding models to predict the temporally-evolving neural activity from the embeddings at each layer. We found a striking correspondence between the layer-by-layer sequence of embeddings from GPT2-XL and the temporal sequence of neural activity in language areas. In addition, we found evidence for the gradual accumulation of recurrent information along the linguistic processing hierarchy. However, we also noticed additional neural processes in the brain, but not in DLMs, during the processing of surprising (unpredictable) words. These findings point to a connection between human language processing and DLMs where the layer-by-layer accumulation of contextual information in DLM embeddings matches the temporal dynamics of neural activity in high-order language areas.

neuroscience↗

bioRxiv2022

Brain embeddings with shared geometry to artificial contextual embeddings, as a code for representing language in the human brain

Contextual embeddings, derived from deep language models (DLMs), provide a continuous vectorial representation of language. This embedding space differs fundamentally from the symbolic representations posited by traditional psycholinguistics. Do language areas in the human brain, similar to DLMs, rely on a continuous embedding space to represent language? To test this hypothesis, we densely recorded the neural activity in the Inferior Frontal Gyrus (IFG, also known as Brocas area) of three participants using dense intracranial arrays while they listened to a 30-minute podcast. From these fine-grained spatiotemporal neural recordings, we derived for each patient a continuous vectorial representation for each word (i.e., a brain embedding). Using stringent, zero-shot mapping, we demonstrated that brain embeddings in the IFG and the DLM contextual embedding space have strikingly similar geometry. This shared geometry allows us to precisely triangulate the position of unseen words in both the brain embedding space (zero-shot encoding) and the DLM contextual embedding space (zero-shot decoding). The continuous brain embedding space provides an alternative computational framework for how natural language is represented in cortical language areas.

neuroscience↗

bioRxiv2020

Thinking ahead: prediction in context as a keystone of language in humans and machines

Departing from traditional linguistic models, advances in deep learning have resulted in a new type of predictive (autoregressive) deep language models (DLMs). Using a self-supervised next-word prediction task, these models are trained to generate appropriate linguistic responses in a given context. We provide empirical evidence that the human brain and autoregressive DLMs share three fundamental computational principles as they process natural language: 1) both are engaged in continuous next-word prediction before word-onset; 2) both match their pre-onset predictions to the incoming word to calculate post-onset surprise (i.e., prediction error signals); 3) both represent words as a function of the previous context. In support of these three principles, our findings indicate that: a) the neural activity before word-onset contains context-dependent predictive information about forthcoming words, even hundreds of milliseconds before the words are perceived; b) the neural activity after word-onset reflects the surprise level and prediction error; and c) autoregressive DLM contextual embeddings capture the neural representation of context-specific word meaning better than arbitrary or static semantic embeddings. Together, our findings suggest that autoregressive DLMs provide a novel and biologically feasible computational framework for studying the neural basis of language.

neuroscience↗

bioRxiv2020

A Solanum lycopersicoides reference genome facilitates biological discovery in tomato

Wild relatives of tomato are a valuable source of natural variation in tomato breeding, as many can be hybridized to the cultivated species (Solanum lycopersicum). Several, including Solanum lycopersicoides, have been crossed to S. lycopersicum for the development of ordered introgression lines (ILs). Despite the utility of these wild relatives and their associated ILs, limited finished genomes have been produced to aid genetic and genomic studies. We have generated a chromosome-scale genome assembly for Solanum lycopersicoides LA2951 using PacBio sequencing, Illumina, and Hi-C. We identified 37,938 genes based on Illumina and Isoseq and compared gene function to the available cultivated tomato genome resources, in addition to mapping the boundaries of the S. lycopersicoides introgressions in a set of cv. VF36 x LA2951 introgression lines (IL). The genome sequence and IL map will support the development of S. lycopersicoides as a model for studying fruit nutrient/quality, pathogen resistance, and environmental stress tolerance traits that we have identified in the IL population and are known to exist in S. lycopersicoides.

genomics↗

bioRxiv2020

Ptr1 evolved convergently with RPS2 and Mr5 to mediate recognition of AvrRpt2 in diverse solanaceous species

The Ptr1 (Pseudomonas tomato race 1) locus in Solanum lycopersicoides confers resistance to strains of Pseudomonas syringae pv. tomato expressing AvrRpt2 and Ralstonia pseudosolanacearum expressing RipBN. Here we describe the identification and phylogenetic analysis of the Ptr1 gene. A single recombinant among 585 F2 plants segregating for the Ptr1 locus was discovered that narrowed the Ptr1 candidates to eight nucleotide-binding leucine-rich repeat protein (NLR)-encoding genes. From analysis of the gene models in the S. lycopersicoides genome sequence and RNA-Seq data, two of the eight genes emerged as the strongest candidates for Ptr1. One of these two candidates was found to encode Ptr1 based on its ability to mediate recognition of AvrRpt2 and RipBN when it was transiently expressed with these effectors in leaves of Nicotiana glutinosa. The ortholog of Ptr1 in tomato and in Solanum pennellii is a pseudogene. However, a functional Ptr1 ortholog exists in N. benthamiana and potato and both mediate recognition of AvrRpt2 and RipBN. In apple and Arabidopsis, recognition of AvrRpt2 is mediated by the Mr5 and RPS2 proteins, respectively. Phylogenetic analysis places Ptr1 in a distinct clade compared to Mr5 and RPS2 and it therefore appears to have arisen by convergent evolution for recognition of AvrRpt2.

plant biology↗

bioRxiv2020

Melon ethylene-mediated transcriptome and methylome dynamics provide insights to volatile production

During climacteric ripening large-scale transcriptional modifications are governed by ethylene. While ripening-related chromatin modifications are also known to occur, a direct connection between these factors has not been demonstrated. We characterized ethylene-mediated transcriptome modification, genome methylation dynamics, and their relation to organoleptic modifications during fruit ripening in the climacteric melon and an ethylene repressed line where the fruit-specific ACC oxidase 1 (ACO1) gene was targeted by antisense. The ACO1 antisense line exhibited mainly reduced transcriptional repression of ripening-related genes associated with DNA CHH hypomethylation at the onset of ripening. Additionally, transcription of a small set of ethylene-induced genes, including known ripening-associated genes, was inhibited by ACO1 repression and this inhibition was associated with CG hypermethylation. In the ACO1 antisense line, the accumulation of aromatic compounds, which are mainly derived from the catabolism of amino acids, is known to be inhibited. One of the ethylene-mediated transcriptionally up-regulated genes, CmTHA1, encoding a threonine aldolase, exhibited differential cytosine methylation. Threonine aldolase catalyzes the conversion of L-threonine/L-allo threonine to glycine and acetaldehyde and thus is likely involved in threonine-dependent ethyl ester biosynthesis. Yeast mutant complementation and incubation of melon discs with labeled threonine verified CmTHA1 threonine aldolase activity, revealing an additional ethylene-dependent amino acid catabolism branch involved in climacteric melon ripening.

plant biology↗

bioRxiv2019

The Ptr1 locus of Solanum lycopersicoides confers resistance to race 1 strains of Pseudomonas syringae pv. tomato and to Ralstonia pseudosolanacearum by recognizing the type III effectors AvrRpt2/RipBN

Race 1 strains of Pseudomonas syringae pv. tomato, which causes bacterial speck disease of tomato, are becoming increasingly common and no simply-inherited genetic resistance to such strains is known. We discovered that a locus in Solanum lycopersicoides, termed Pseudomonas tomato race 1 (Ptr1), confers resistance to race 1 Pst strains by recognizing the type III effector AvrRpt2. In Arabidopsis, AvrRpt2 degrades the RIN4 protein thereby activating RPS2-mediated immunity. Ptr1 also recognized homologs of AvrRpt2 from diverse bacteria including one in Ralstonia pseudosolanacearum and this correlated with the ability of AvrRpt2 to degrade RIN4. Using site-directed mutagenesis of AvrRpt2 we found that Ptr1 and RPS2 recognize identical features of AvrRpt2. However, the genome sequence of S. lycopersicoides revealed no RPS2 homolog in the Ptr1 region. Ptr1 could play an important role in controlling bacterial speck disease and its future cloning may shed light on an example of convergent evolution for recognition of a widespread type III effector.

plant biology↗