Cloning, expression, and promoter structure of a mammalian Inner Centromere Protein (INCENP) Richard Saffery, Danielle V. Irvine, Benjamin T. Kile, Damien F. Hudson, Suzanne M. Cutts, K.H. Andy Choo The Murdoch Institute, Royal Children’s Hospital, Flemington Road, Parkville, 3052, Victoria, Australia Received: 28 September 1998 / Accepted: 2 December 1998 The term ‘chromosomal passenger proteins’ was first coined to describe a group of cellular proteins that tightly associate with chromosomes during the early stages of mitosis, localize at the centromere at metaphase, and subsequently relocate to other sub- cellular regions at the commencement of anaphase (Earnshaw and Bernat 1990). One of these proteins, the Inner Centromere Protein or INCENP, was first identified using antibodies raised against chicken chromosomal scaffold proteins (Cooke et al. 1987). Like other passenger proteins, chicken INCENP abruptly dissociates from the chromosomes at the metaphase/anaphase transition and associates with the overlapping microtubules of the central spindle. In addition, a subpopulation of protein concentrates in the pre- sumptive cleavage furrow prior to the commencement of furrow- ing, suggesting a crucial role for INCENP in cytokinesis (Earn- shaw and Cooke 1991). Chicken cells contain two INCENP proteins (designated class I and II) of 133 and 145 kDa respectively. Molecular analysis has revealed that both classes arise from a single, differentially spliced primary transcript with class II, differing from class I by the in- sertion of 114 additional nucleotides (Mackay et al. 1993). Chicken INCENPs are highly basic proteins containing long coiled-coil domains, five putative nuclear localization sites, and many putative phosphorylation sites. They share little homology to other known proteins (Mackay et al. 1993). Both chicken INCENP cDNAs have been expressed in mammalian cells and shown to localize identically to the endogenous mammalian INCENP pro- tein, suggesting conservation of nuclear, chromosomal, spindle, and cell cortex factors that interact with INCENP during the cell cycle (Mackay et al. 1993). Expression of an amino terminal mu- tant of chicken INCENP (INCENP 43-839 ) in mammalian cells has identified an essential role for the first 42 residues in centromere targeting and transfer of the protein from chromosomes to the spindle and cell cortex during the metaphase/anaphase transi- tion (Mackay et al. 1993, 1998). Deletion of the coiled-coil domain prevented microtubule association, but had no effect on spin- dle association (Mackay et al. 1993). A truncation derivative (INCENP 1-405 ) that targets correctly to centromeres does not re- distribute at anaphase and acts as a dominant-negative mutant, disrupting both metaphase chromosome alignment and cytokinesis in mammalian cells (Mackay et al. 1998). Recently, a Xenopus homolog of INCENP was identified fol- lowing a screen to identify mitotic phosphoproteins (Stukenburg et al. 1997). INCENP in cell extracts isolated from interphase Xeno- pus embryos migrates as two bands of approximately 130 kDa, with the smaller band representing an unphosphorylated form. With progression through the cell cycle, the appearance of the slower migrating form correlates with increased kinase activity in the cell extract. Direct phosphorylation of Xenopus INCENP by purified cyclin B-Cdc2 was demonstrated using in vitro expressed Xenopus INCENP protein. Subsequent sequencing analysis of the cDNA revealed the presence of 3 Cdc2 kinase sites and 3 MAP kinase sites within the Xenopus protein (Stukenburg et al. 1997). To date, no mammalian homolog of INCENP has been de- scribed. This report describes the cloning, sequencing, tissue ex- pression, and promoter analysis of mouse INCENP. BLAST searches were carried out using chicken INCENP se- quences against the EST database. Several mouse EST clones with significant homologies to both the carboxy and amino portions of chicken INCENP were identified which formed two contiguous stretches of DNA from the putative mouse INCENP cDNA. Over- lapping clones from these contigs were obtained from Genome Systems and subjected to DNA sequencing on both strands. Oli- gonucleotide primers A (5'-AAGAAGAGGCGGGTGTCTAAC) and B (5'-GAAGCCTAGATAAGAGGGTGA) were used to am- plify intervening sequences by RT-PCR, with both mouse ES cell and F9 cell RNA as template. As expected, based on the size of the chicken INCENP, a product of approximately 2.3 kb was identi- fied in both cell types and was then subcloned and sequenced on both strands to complete the cDNA sequence. The final sequence (Genbank accession No. AF117610) comprises 3104 bp with an open reading frame extending from bases 143 to a stop codon at position 2772, followed by a 3' untranslated region containing two AATAAA polyadenylation signals at 2864 and 2932. This se- quence encodes a protein of 876 amino acids with predicted size of 101 kDa and extensive homology to chicken and Xenopus INCENPs (Fig. 1). Like the chicken INCENP, mouse INCENP is extremely basic with an isoelectric point of 9.6 and a charge of 35.41 at pH 7.0. Genomic Southern hybridization analysis with various mouse INCENP cDNAs as probes confirmed the presence of a single-copy INCENP gene within the mouse genome (data not shown) which we have recently mapped to mouse chromosome location 19.05-19.4 (Fowler et al. 1998). The mouse INCENP protein contains two sequences resem- bling bipartite nuclear localization signals (RRX 12 RKR, KKX 11 RRK; Dingwall and Laskey 1991), three putative MAP kinase sites (PKTP, PPSP, PQSP; Stukenburg et al. 1997), four putative cdc2 kinase sites (KTPS, TPKK, KTPS, TPLR; Mackay et al. 1993; Stukenburg et al. 1997), two putative N-glycosylation sites (NGSR, NMTV; Miletich 1990), and many other putative phosphorylation sites including 12 putative casein kinase II (Pinna 1990), nine cAMP-dependent protein kinase (Glass et al. 1986) and 19 protein kinase C (Woodgett et al. 1986) phosphorylation sites. Several of these sites are conserved in Xenopus and chicken INCENP proteins (Fig. 1). Overall identity is 43.8% with 55.5% similarity between mouse INCENP and chicken INCENP II, and 41.8% identity and 54.9% similarity between mouse and Xenopus INCENPs. This compares with 44.4% identity and 56.1% similar- ity between the Xenopus and chicken INCENP II proteins. A direct comparison of the primary sequence of the three INCENPs (Fig. 1) reveals several very highly conserved domains. An amino terminal region of 11 amino acids (residues 53–63 in the mouse) shows 100% conservation among the three proteins and contains putative Correspondence to: R. Saffery Mammalian Genome 10, 415–418 (1999). © Springer-Verlag New York Inc. 1999 Incorporating Mouse Genome