Phosphorylation of the carboxyl terminus of inner centromere protein (INCENP) by the Aurora B Kinase stimulates Aurora B kinase activity

Abstract

How the events of mitosis are coordinated is not well understood. Intriguing mitotic regulators include the chromosomal passenger proteins. Loss of either of the passengers inner centromere protein (INCENP) or the Aurora B kinase results in chromosome segregation defects and failures in cytokinesis. Furthermore, INCENP and Aurora B have identical localization patterns during mitosis and directly bind each other in vitro. These results led to the hypothesis that INCENP is a direct substrate of Aurora B. Here we show that the Caenorhabditis elegans Aurora B kinase AIR-2 specifically phosphorylated the C. elegans INCENP ICP-1 at two adjacent serines within the carboxyl terminus. Furthermore, the full length and a carboxyl-terminal fragment of ICP-1 stimulated AIR-2 kinase activity. This increase in AIR-2 activity required that AIR-2 phosphorylate ICP-1 because mutation of both serines in the AIR-2 phosphorylation site of ICP-1 abolished the potentiation of AIR-2 kinase activity by ICP-1. Thus, ICP-1 is directly phosphorylated by AIR-2 and functions in a positive feedback loop that regulates AIR-2 kinase activity. Since the Aurora B phosphorylation site within INCENP and the functions of INCENP and Aurora B have been conserved among eukaryotes, the feedback loop we have identified is also likely to be evolutionarily conserved.

FIG.1

GST-AIR-2 specifically phosphorylates GST-INCENP in vitro. In vitro kinase reactions with MBP as a substrate for GST-AIR-2 (lane 1), GST-AIR-2ts (lane 2), and GST-AIR-1 (lane 3) and with GST-ICP-1 alone (lane 4) or as a substrate for GST-AIR-2 (lane 5), GST-AIR-2ts (lane 6), and GST-AIR-1 (lane 7) are shown. Top panel, ³²P incorporation. Bottom panel, protein loading as determined by Ponceau S staining (MBP and GST-AIR-1), anti-AIR-2 Western (GST-AIR-2), and anti-ICP-1 Western (GST-ICP-1).

FIG.2

The carboxyl terminus of INCENP is strongly phosphorylated by GST-AIR-2. A, schematic of the GST-ICP-1 truncation proteins and relative phosphorylation by GST-AIR-2. B, in vitro kinase reactions with GST-ICP-1 (ICP-1) (lanes 1 and 2), GST-ICP-1N (N) (lanes 3 and 4), GST-ICP-1M (M) (lanes 5 and 6), or GST-ICP-1C (C) (lanes 7 and 8) as substrates for GST-AIR-2 (lanes 1, 3, 5, and 7) and GST-AIR-2ts (lanes 2, 4, 6, and 8). Top panel, ³²P incorporation. Bottom panel, protein loading as determined by Ponceau S staining.

FIG.3

Serines 598 and 599 are additively phosphorylated by GST-AIR-2. A, the carboxyl-terminal IN box is highly conserved across species (with accession numbers from GenBank™ (AF) and NCBI (gi) in parentheses): chicken (GgINCENP1, gi 2134313), Xenopus (XlINCENP, gi 2072290), human (HsINCENP, gi 9438179), mouse (MmINCENP, gi 4886899), C. elegans (CeICP-1, AF300704), Drosophila melanogaster (DmINCENP, gi 17647535) and S. cerevisiae (Sli15p, gi 6319632). As-terisks denote the conserved serines that are phosphorylated by AIR-2. B, in vitro kinase reactions with GST-ICP-1C (lanes 1 and 2), GST-ICP-1C (S598A) (lanes 3 and 4), GST-ICP-1C (S599A) (lanes 5 and 6), or GST-ICP-1C (S598A/S599A) (lanes 7 and 8) as substrates for GST-AIR-2 (lanes 1, 3, 5, and 7) and GST-AIR-2ts (lanes 2, 4, 6, and 8). Top panel, ³²P incorporation. Bottom panel, protein loading as determined by Ponceau S staining.

FIG.4

Wild-type GST-ICP-1 and GST-ICP-1C, but not the S598A/S599A mutants, stimulate GST-AIR-2 kinase activity. A, lanes 1-8, in vitro phosphorylation of MBP by either GST-ICP-1 alone (lane 1), GST-AIR-2 alone (lane 2), or GST-AIR-2 supplemented with GST-ICP-1 (lane 3), GST-ICP-1 (S598A/S599A) (lane 4), GST-ICP-1N (lane 5), GST-ICP-1M, (lane 6), GST-ICP-1C (lane 7), or GST-ICP-1C (S598A/S599A) (lane 8). Lanes 9 and 10, in vitro phosphorylation of MBP by either GST-AIR-1 alone (lane 9) or GST-AIR-1 supplemented with GST-ICP-1 (lane 10). Top panel, ³²P incorporation. Bottom panel, protein loading as determined by Ponceau S staining (MBP, GST-AIR-2, and GST-AIR-1) or anti-ICP-1 Western. B, ³²P incorporation into MBP by GST-AIR-2 in the presence of GST-ICP-1 (n = 7), GST-ICP-1 (S598A/S599A) (n = 3), GST-ICP-1C (n = 4), and GST-ICP-1C (S598A/S599A) (n = 3) was quantified by phosphorimaging. Error bars indicate one standard deviation.