Biology subjects

Cadena, S.

Publications and source records attributed to Cadena, S..

3 recordsLinked to original sources

bioRxiv2026

Duckweeds as Multiplexable Plant Models for Exploring Abiotic Stress Responses

In nature, plants experience complex combinations of environmental stresses. Prior studies using Arabidopsis thaliana show that transcriptional responses to combinatorial stress are unique and cannot be predicted from responses to individual treatments, making it imperative to study combinatorial stress in the monocotyledonous cereals that comprise the majority of our food supply. However, crop species are not easily amenable for such studies because of their large size, long generation time, and organismal complexity. In contrast, the monocotyledonous aquatic duckweeds are small, grow rapidly, and show reduced organismal and genomic complexity. Here, we explore the physiological responses to combinatorial stress in three duckweed species, S. polyrhiza, L. minor, and W. australiana and examine their potential to serve as multiplexable models. Focusing on S. polyrhiza, the most complex and best annotated species, we investigated transcriptional response to individual and combinatorial stress treatments. While most individual treatments elicit few transcriptional changes, combinatorial stress elicited a non-additive transcriptional response unique to each combination. Exploring the duckweed regulatory landscape, we found that regulatory regions near genes that were differentially expressed in combinatorial stress in S. polyrhiza resemble environmentally responsive regulatory elements found in terrestrial plants. Examined duckweed species show numbers and genomic distributions of regulatory elements similar to those of A. thaliana and maize, with subtle effects of genome size but none of organismal complexity. Taken together, our results establish that duckweeds respond to combinatorial stress in a manner similar to terrestrial plants and can therefore serve as high-throughput, multiplexable models for studying its molecular underpinnings.

plant biology↗

bioRxiv2026

A microbiome assessment of healthy and bleached Halymenia floresii (Clemente) C. Agardh (Rhodophyta, Florideophyceae)

Macroalgae serve as critical habitat-formers and primary producers in coastal ecosystems, functioning across near-subtidal and intertidal zones in three distinct states: substrate-attached, free-floating (drift), and beach-cast. While substrate-attached macroalgae are susceptible to infectious diseases with significant ecological implications, diseases affecting drift macroalgal communities remain virtually unstudied. Here, we investigated bleaching disease - one of the most common macroalgal afflictions - in the drift rhodophyte Halymenia floresii from the Gulf of Mexico. Using 16S rRNA gene high-throughput sequencing and scanning electron microscopy, we characterized the bacterial community structure and composition associated with the free-floating healthy, bleached and degrading H. floresii to understand how bacterial partners respond to host health status. Principal Coordinate Analysis based on UniFrac distance revealed distinct clustering of bacterial communities according to host health condition. Shannon diversity indices showed distinct patterns ranging from 1.14 - 3.15 for healthy, bleached, and degrading samples, while Simpson indices ranged from 0.62 to 0.91, reflecting substantial variation in community evenness. In healthy samples, Cyanobacteria (17 - 52%) and Pseudomonadota (previously, Proteobacteria) (41 - 81%) dominated, and the bleached samples were characterized by elevated Bacteroidota (formerly, Bacteroidetes) (5 - 35%) and Pseudomonadota (41 - 88%). Notably, Novosphingobium (25 - 49%) dominated healthy hosts while showing lower abundance in degrading (13 - 17%) and bleached (18 - 22%) specimens. Conversely, Reinekea emerged as a dominant genus (22.5%) specifically in bleached samples, suggesting a potential role in disease pathogenesis. Microbial network analysis using NetCoMi revealed three distinct bacterial clusters corresponding to health states: a healthy-associated cluster dominated by Novosphingobium and uncultured Cyanobacterial with predominantly positive associations, and two disease-associated clusters enriched in opportunistic genera including Reinekea, Vibrio, Colwellia, and Alteromonas, indicating network reorganization from cooperative to exploitative interactions. This study provides the first descriptive assessment of microbiome transitions associated with bleaching disease in a drift macroalga and highlights the importance of considering free-floating macroalgal diseases and their potential impacts on coastal ecosystem health.

microbiology↗

bioRxiv2020

Generalization in data-driven models of primary visual cortex

AO_SCPLOWBSTRACTC_SCPLOWDeep neural networks (DNN) have set new standards at predicting responses of neural populations to visual input. Most such DNNs consist of a convolutional network (core) shared across all neurons which learns a representation of neural computation in visual cortex and a neuron-specific readout that linearly combines the relevant features in this representation. The goal of this paper is to test whether such a representation is indeed generally characteristic for visual cortex, i.e. gener-alizes between animals of a species, and what factors contribute to obtaining such a generalizing core. To push all non-linear computations into the core where the generalizing cortical features should be learned, we devise a novel readout that reduces the number of parameters per neuron in the readout by up to two orders of magnitude compared to the previous state-of-the-art. It does so by taking advantage of retinotopy and learns a Gaussian distribution over the neurons receptive field po-sition. With this new readout we train our network on neural responses from mouse primary visual cortex (V1) and obtain a gain in performance of 7% compared to the previous state-of-the-art network. We then investigate whether the convolutional core indeed captures general cortical features by using the core in transfer learning to a different animal. When transferring a core trained on thousands of neurons from various animals and scans we exceed the performance of training directly on that animal by 12%, and outperform a commonly used VGG16 core pre-trained on imagenet by 33%. In addition, transfer learning with our data-driven core is more data-efficient than direct training, achieving the same performance with only 40% of the data. Our model with its novel readout thus sets a new state-of-the-art for neural response prediction in mouse visual cortex from natural images, generalizes between animals, and captures better characteristic cortical features than current task-driven pre-training approaches such as VGG16.

neuroscience↗