RNA-seq analysis of synchronized developing pollen isolated from a single anther

Abstract

Pollen development, from unicellular microspores to anthesis, is a complex process involving the coordinated specification, differentiation and functions of different cell types. Key to understanding this development is identifying the genes expressed at precise stages of development. However, transcriptomic studies on pollen prior to anthesis are complicated by the inaccessible nature of pollen developing in the anther and the resistant pollen wall. To assist with understanding gene expression during pollen development we have developed a protocol to perform RNA-Seq on pollen isolated from a single anther (SA RNA-Seq). The protocol involves removing pollen from a single anther for analysis and viewing the remaining pollen to determine the developmental stage. The isolated pollen is chemically lysed and mRNA isolated from the lysate using an oligo-dT column before library preparation. Here, we report on the development and testing of our method and the generation of a transcriptome for three stages of pollen development from Arabidopsis (Arabidopsis thaliana) and two stages from male kiwifruit (Actinidia chinensis). This protocol enables the transcriptome of precise developmental stages of pollen to be analyzed, and uses a small number of plants, potentially facilitating studies that require a range of treatments or the analysis of the first generation of transgenic plants.

A foundational large language model for edible plant genomes

“…The prediction performance was relatively consistent across individual tissues with the exception of pollen which had consistently lower prediction performance across plant species (Supplementary Tables 2 – 6 ). This might be related to the inaccessible nature of pollen developing in the anther and the resistant pollen wall that might affect proper pollen isolation, as recently suggested 52 . Considering the absence of models capable of predicting quantitative gene expression levels, and to conduct a comparative analysis for this task, we also trained AgroNT to predict whether genes are expressed or not.…”

A Foundational Large Language Model for Edible Plant Genomes

“…The prediction performance was relatively consistent across individual tissues with the exception of pollen which had consistently lower prediction performance across plant species (Supplementary Tables 2-6). This might be related to the inaccessible nature of pollen developing in the anther and the resistant pollen wall that might affect proper pollen isolation, as recently suggested [52]. Con- sidering the absence of models capable of predicting quantitative gene expression levels, and to conduct a comparative analysis for this task, we also trained AgroNT to predict whether genes are expressed or not.…”