Cellular & Molecular Imaging

In multicellular organisms, tissue architecture reflects not only dedicated patterning genes but the integrated state of core cellular physiology. Whether the same holds for the communities that bacteria build has been difficult to test at scale. To this end, we developed µPULLI, a high-throughput platform that pairs low-magnification, label-free brightfield timelapse microscopy with computer vision. Using µPULLI, we screened a genome-wide Vibrio cholerae transposon library. Our screen revealed that community architecture is governed as much by core physiology as by dedicated biofilm genes: perturbations to central metabolism, cofactor biosynthesis, and cell-envelope architecture each leave distinctive, pathway-specific fingerprints on how a community develops. These fingerprints arise through both transcriptional and non-transcriptional mechanisms, including cell-surface changes invisible to RNA sequencing. Small molecules drive communities predictably through this phenotype space, and the principle holds across taxonomically diverse pathogens. Emergent community architecture thus encodes the integrated physiological state of its constituent cells, revealing a rich, largely untapped layer of biological information that can be read from simple, low-magnification brightfield videos of growing cultures.

The Darwin Tree of Life project has the goal to sequence the genomes of 70,000 species of eukaryotic organisms in Britain and Ireland. This is a collection of photographs of the samples included in the study, provided by the National History Museum (NHM).

Mini-immunofluorescence assay that builds on a custom neural network-based image analysis pipeline for automated and multiplexed detection of immunoglobulins (IgG, IgA, and IgM) in patient samples.

Plasmodium oocysts within the mosquito midgut epithelium produce sporozoites, the stage which is infective for the vertebrate host. The formation process of these sporozoites is termed sporogony and is still poorly understood. During sporogony a 30 nm thick longitudinal fiber, the rootlet fiber, connects the centriolar plaque of the nucleus to the apical polar ring at the tip of the forming sporozoite. It is hypothesized to play a role in trafficking secretory organelles and nuclear uptake into the newly forming sporozoite bud. A similar fiber was previously observed in Toxoplasma gondii which is formed by two striated fiber assemblins (SFAs). Here we characterize the two SFA homologs in P. berghei (PbSFA1/PBANKA_1227000, PbSFA2/PBANKA_1012000) via fluorescent tagging, confirming a fiber-like localization at the tips of nascent sporozoites in oocysts. Knockout of either SFA gene lead to significant reduction in oocyst numbers and a complete failure to produce any sporozoites. Upon investigation by electron microscopy and electron tomography, initiation of sporozoite formation could be observed but the rootlet fiber was absent in sporozoite buds. Strikingly, nuclear uptake was strongly affected, as no nuclei could be observed in elongated sporozoite buds. Further, elongated sporozoite buds failed to bud off, ultimately creating a network of incompletely formed ‘sporozoites’. These findings emphasize the role of the rootlet fiber in nuclear uptake und organization of sporozoite formation, making them essential for the progression through the mosquito.

Ribosomes are huge ribonucleoprotein particles that mediate protein synthesis in all organisms. The synthesis of ribosomes is a complex process that involves hundreds of supporting factors in mammalian cells, including proto-oncogenes and tumor suppressors. Dysregulation of ribosome biogenesis can contribute to tumorigenesis and the increased production of ribosomes in cancer cells is known to promote proliferative cell growth. Therefore, ribosome biogenesis represents an attractive vulnerability of cancer cells that ought to be exploited for the development of anti-cancer drugs. Despite the large number of trans-acting factors promoting in ribosome assembly including potentially druggable enzymes, only few chemical inhibitors that act on ribosome biogenesis, especially downstream of pre-rRNA transcription, have been identified to date. We have established a pipeline to perform single-cell, imaging-based screening campaigns using four different readouts, including fluorescent ribosomal protein reporters (RPS2-YFP, RPL29-GFP) and immunofluorescence analyses of the ribosome biogenesis factor ENP1(BYSL), in HeLa cells, a human cancer line. This dataset contains the raw data and segmentation analysis of our pilot screen, using a library comprising more than 1000 FDA-approved drugs with known targets in other pathways. This pilot screen obtained excellent quality scores and identified altogether ten compounds as hits.

Image data of Spatial distribution of three ARGONAUTEs regulates the anther phasiRNA pathway

The nitrogen nutrition status affects the main factors of rice yield. In traditional rice nitrogen nutrition monitoring methods, most experts enter the farmland to observe leaf color and growth and apply an appropriate amount of nitrogen fertilizer according to the results. However, this method is laborand time-consuming. To realize automatic rice nitrogen nutrition monitoring, we constructed the Jiangxi rice nitrogen nutrition monitoring model based on a convolution neural network (CNN) using the same region rice canopy image in different generation periods. The results show that the same CNN model could distinguish the rice nitrogen nutrition status in different periods, which can completely realize the automatic discrimination of nitrogen nutrition status so as to guide the scientific nitrogen application of rice in this area.

Time-lapse images of the seed germination procedure for 21 wheat genotypes, 12 barley genotypes, and 6 rice genotypes.

Image data of Spatial distribution of three ARGONAUTEs regulates the anther phasiRNA pathway

The nitrogen nutrition status affects the main factors of rice yield. In traditional rice nitrogen nutrition monitoring methods, most experts enter the farmland to observe leaf color and growth and apply an appropriate amount of nitrogen fertilizer according to the results. However, this method is laborand time-consuming. To realize automatic rice nitrogen nutrition monitoring, we constructed the Jiangxi rice nitrogen nutrition monitoring model based on a convolution neural network (CNN) using the same region rice canopy image in different generation periods. The results show that the same CNN model could distinguish the rice nitrogen nutrition status in different periods, which can completely realize the automatic discrimination of nitrogen nutrition status so as to guide the scientific nitrogen application of rice in this area.