Technique hinges on computational method to rank importance of functional genetic links.
Researchers have developed a new technique they claim has allowed them to map the interactive network of some 818 essential genes in the model worm Caenorhabditis elegans. Rather than carry out high-content screening in single cells, the new approach profiled genetic function and interaction by analyzing the effects of gene knockdowns on the architecture of the worm’s gonad.
The scientists, led by a team at the University of California, San Fransisco’s (UCSF) Ludwig Institute for Cancer Research and Department of Cellular and Molecular Medicine, report their methods and results in Cell. The paper is titled “A High-Resolution C. elegans Essential Gene Network Based on Phenotypic Profiling of a Complex Tissue.”
High-content screening is the primary method for mapping functional gene networks in single animal cells and has been used to successfully classify genes based on cell morphology following gene knockdown, explain Ludwig Institute’s Karen Oegema, Ph.D., and New York University’s Kristin C. Gunsalus, Ph.D., and colleagues. However, while such approaches have identified genes that contribute to cell division and morphology, they have been less successful at assessing protein function across the breadth of cellular processes that has been achieved using genetic interaction profiling in fungi.
Drs. Oegema and Gunsalas’ teams have now developed an approach that analyzes the effects of gene knockdowns on complex tissue architecture in the C. elegans gonad at a single time point. The researchers focused on profiling the organism’s 554 sterile genes so called because their inhibition leads to sterility.
A number of the 554 genes are known to control fundamental cellular processes such as membrane trafficking, translation, proteasome function, and cortical remodeling, 166 remain unnamed, indicating no prior characterization, and only 50 have known characterized orthologs that predict their potential function in C. elegans. These genes have remained largely uninvestigated to date because existing techniques for assessing the effects of gene knockdown in C. elegans have involved time-lapse filming the first cellular divisions of the embryo, the authors note.
To profile the genes the team developed a high content assay based on 3-D two-color fluorescence confocal imaging of the C. elegans hermaphrodite gonad. They scored their images according to 94 phenotypic parameters resulting from gene knockdown. The genes were initially partitioned manually into classes to generate a reference that could be used to develop computational methods. The resulting Connection Specificity Index (CSI) was developed as a network context-dependent measure that ranks the significance of functional links between genes.
The authors claim this allowed them to tease out meaningful functional links for network construction from the ‘hairball’ of connections that phenotypic scans churn out. “CSI is ideally suited for integrating these datasets because it filters out low specificity phenotypic links, leading to a network that combines only the significant relationships,” they write.
Moreover, in comparison with the Pearson’s Correlation Coefficient, constructing networks using CSI “reduces nonspecific connection noise, improves network clarity, and allows connections of a similar functional significance to be simultaneously viewed across the entire network at a single threshold,” the team claims.
The authors generated a final network by combining the results from their study of the 554 sterile genes in gonad architecture with those of time-lapse embryo filming-based screens for 661 genes, which were also evaluated using CSI. 330 genes were profiled using both methods.
They claim the resulting network represents the functional connections between some 818 essential C. elegans genes that can be viewed at multiple levels of functional resolution. Importantly, the work generated functional predictions for 106 of the 116 previously uncharacterized sterile genes. The accuracy of the predictions was confirmed by carrying out validation studies for eight of the genes.
The authors stress that the means of inferring function using a single point morphology assay is distinct from that used in embryo filming screens based on assessing the phenotype during the early divisions of embryos following inhibition of specific C. elegans genes. In the embryo screen gene function is inferred directly from phenotype.
So, for example, defects in spindle assembly, chromosome segregation, or polarity establishment lead to the responsible gene’s assignment to classes implicated in the corresponding processes. In contrast, the link between catalogued phenotypic features in the gonad architecture assay and gene function is not direct.
“We do not know why knockdown of genes with specific functions lead to a specific spectrum of phenotypic features,” the authors note. “Instead, protein function is inferred by comparing the profile of phenotypic features to those of other genes in the dataset, a process conceptually analogous to how function is inferred from genetic interaction profiles in budding yeast.
“Our results demonstrate that the biological complexity of the gonad translates into phenotypes that enable profiling across a wide spectrum of essential cellular processes at a resolution approaching that of genetic interaction profiling in yeasts,” the team concludes. “A major finding of our work is that profiles based on complex tissue architecture acquired at a single time point following gene knockdown have greater information depth than profiles derived from time-lapse imaging of the first two divisions of the early embryo (a single cell).”
Although the gonad was used as a “test tube” to functionally profile the sterile genes, subsequent work in early and later stage embryos does indicate that the majority of the essential genes in this collection are broadly important, the researchers claim. Of the 554 sterile genes, about 60% have predicted homologs in higher organisms, and more than 50% of the 106 uncharacterized genes have a predicted homolog or conserved domain.
“Thus, in addition to filling a large gap in the analysis of the C. elegans essential gene set, our findings provide a starting point to address the functions of related genes in humans,” the researchers note.
