Exploration of the cell-cycle genes found within the RIKEN FANTOM2 data set

Abstract

The cell cycle is one of the most fundamental processes within a cell. Phase-dependent expression and cell-cycle checkpoints require a high level of control. A large number of genes with varying functions and modes of action are responsible for this biology. In a targeted exploration of the FANTOM2-Variable Protein Set, a number of mouse homologs to known cell-cycle regulators as well as novel members of cell-cycle families were identified. Focusing on two prototype cell-cycle families, the cyclins and the NIMA-related kinases (NEKs), we believe we have identified all of the mouse members of these families, 24 cyclins and 10 NEKs, and mapped them to ENSEMBL transcripts. To attempt to globally identify all potential cell cycle-related genes within mouse, the MGI (Mouse Genome Database) assignments for the RIKEN Representative Set (RPS) and the results from two homology-based queries were merged. We identified 1415 genes with possible cell-cycle roles, and 1758 potential paralogs. We comment on the genes identified in this screen and evaluate the merits of each approach.

Figure 1

Sequence relationships between cyclins in mouse. (A) Phylogeny of mouse cyclins, based on multiple alignment of the conserved cyclin box domain. Values are shown for 1000 bootstraps. (B) Alignment of the core cyclin box domain of mouse cyclins. Shaded regions and underlying trace show regions of highest sequence similarity. CYCJ-Dm—Drosophila melanogaster cyclin J (AAC47017) is provided as a guide. UNG—Uracil-DNA glycosylase 2 has been recognized as a likely cyclin by a number of workers (Murray and Marks 2001). #: ANIA-6A has been assigned the symbol CCNL. ANIA-6B, which is clearly a paralog, has been assigned the symbol Pcee-pending. ^*-L-like and J-like cyclins identified in the present study.

Figure 1

Sequence relationships between cyclins in mouse. (A) Phylogeny of mouse cyclins, based on multiple alignment of the conserved cyclin box domain. Values are shown for 1000 bootstraps. (B) Alignment of the core cyclin box domain of mouse cyclins. Shaded regions and underlying trace show regions of highest sequence similarity. CYCJ-Dm—Drosophila melanogaster cyclin J (AAC47017) is provided as a guide. UNG—Uracil-DNA glycosylase 2 has been recognized as a likely cyclin by a number of workers (Murray and Marks 2001). #: ANIA-6A has been assigned the symbol CCNL. ANIA-6B, which is clearly a paralog, has been assigned the symbol Pcee-pending. ^*-L-like and J-like cyclins identified in the present study.

Figure 2

Distribution of candidate cell-cycle genes identified in the global screen. (A) Here, 1415 nonredundant sequences were identified as likely cell-cycle homologs by combining the best-hit predictions from the two BLAST searches and the MGI assignments. BAIT1–AMIGO corresponds to the (698) sequences identified by the bait sequences identified by the Gene Ontology Consortium as cell cycle-related. BAIT2-SWALL/IPI corresponds to the (1042) sequences identified by the bait sequences identified in a keyword search of SWALL and IPI. The 328 MGI assignments are those with good supporting evidence (InterPro and SCOP predictions were excluded). (B) Here, 3173 nonredundant sequences were identified when all significant hits were considered as well as the InterPro-based MGI assignments.