An RNAi screen for genes that affect nuclear morphology in Caenorhabditis elegans reveals the involvement of unexpected processes

Abstract

Aberration in nuclear morphology is one of the hallmarks of cellular transformation. However, the processes that, when mis-regulated, result aberrant nuclear morphology are poorly understood. In this study, we carried out a systematic, high-throughput RNAi screen for genes that affect nuclear morphology in Caenorhabditis elegans embryos. The screen employed over 1700 RNAi constructs against genes required for embryonic viability. Nuclei of early embryos are typically spherical, and their NPCs are evenly distributed. The screen was performed on early embryos expressing a fluorescently tagged component of the nuclear pore complex (NPC), allowing visualization of nuclear shape as well as the distribution of NPCs around the nuclear envelope. Our screen uncovered 182 genes whose downregulation resulted in one or more abnormal nuclear phenotypes, including multiple nuclei, micronuclei, abnormal nuclear shape, anaphase bridges, and abnormal NPC distribution. Many of these genes fall into common functional groups, including some that were not previously known to affect nuclear morphology, such as genes involved in mitochondrial function, the vacuolar ATPase, and the CCT chaperonin complex. The results of this screen add to our growing knowledge of processes that affect nuclear morphology and that may be altered in cancer cells that exhibit abnormal nuclear shape.

Keywords: C. elegans; CCT chaperonin; ER morphology; anaphase bridges; cytokinesis; micronuclei; nuclear envelope; paired nuclei; vacuolar ATPase.

Figure 1

Classes of abnormal nuclear phenotypes. Shown are representative images of C. elegans embryos expressing NPP-1 fused to GFP (NPP-1::GFP) from worms treated with RNAi against the indicated genes. (A) Wild type phenotype on control RNAi: examples of 1-, 2-, 4-, and multi-cell embryos (some nuclei are on an out-of-focus plane).(B) The multi-nucleated cell phenotype: (i) paired nuclei (some nuclei in this particular case are also deformed), (ii) multiple nuclei in a 1-cell embryo, (iii) Multi-nucleated cells in a $\ge \mathrm{ }$2 cell embryo, and (iv) micronuclei (along with nuclei). (C) The abnormal nuclear shape phenotype: (i) anaphase bridges (accompanied here by micronuclei, arrowheads) and (ii) deformed nuclei. (D) NPC distribution defect phenotypes: abnormal distribution of NPP-1::GFP (i) on the nuclear envelope, (ii) in the cytoplasm, or (iii) inside the nucleus. Scale bar (for all images) = 10 μm.

Figure 2

Distribution of abnormal nuclear phenotypes among the 182 RNAi treatments. (A) Bar graph showing the number of RNAi treatments that resulted in the indicated number of phenotypes. (B) Connectivity of phenotypic categories. The thickness of each connecting line corresponds to the number of genes that, when down-regulated, result in both phenotypes. See Supplementary Table S2 for the exact number of shared phenotypes.

Figure 3

Functional categories of genes whose downregulation results in one or more abnormal nuclear or NPC phenotype. (A) The relative abundance (in percentage) of each functional category within the 182 hits is indicated. (B–D) Statistical overrepresentation analysis of GO terms using the PANTHER Classification System. The top statistically significant overrepresented subclasses based on Biological Process (B), Cellular Component (C), and Molecular Function classes (D) are shown. The number on the right represents fold enrichment. See Supplementary Tables S3–S5 for the complete list and P-values.

Figure 4

Functional categories of genes whose downregulation results in one or more abnormal nuclear or NPC phenotype. The prevalence of each of the 9 phenotypes observed within each functional category: Phenotypic categories are on the x-axis, and the percent of RNAis within a functional category displaying a particular phenotype is on the y-axis.

Figure 5

Different phenotypes are a result of the downregulation of different sets of genes. The contribution of each functional group to a given phenotype, as determined by the number of genes from a particular functional category that, when down-regulated, resulted in the indicated phenotypes. Functional groups are arranged clockwise, starting with cell cycle genes. For example, 37% of genes that lead to abnormal NPC NE distribution (top right) are involved in mRNA processing (red sector). The breakdown for the “Nuclear NPC” category is not shown as it encompasses only 2 genes.

Figure 6

Inactivation of dynein/dynacin proteins has minimal effect on ER structure. Examples of C. elegans embryos expressing SP12::GFP (an ER marker) and histone H2B:: mCherry (chromatin), treated with control RNAi against smd-1 (top two rows) or RNAi against dynacin (dnc-1) or dynein (dyci-1) subunits (bottom two rows). Scale bar = 10 µm. See Supplementary Image Collections S1–S3 for additional examples and confocal microscopy data.

Figure 7

The effects of down-regulating translation machinery, chaperonin, and the vacuolar ATPase on ER morphology. Representative images of C. elegans embryos as in Figure 6 following RNAi treatment against genes involved in translation (A), chaperonin (B), or vacuolar ATPase (C). Scale bar = 10 µm. Supplementary Image Collections S4–S9 for additional examples and confocal microscopy data.