Supplemental Information

Preparation of Aurora expression plasmids and protein expression

The gene fragments for either mouse or human Aurora genes were amplified by Tgo polymerase (Roche Diagnostics) with primers containing a BamHI site on the 5’ primer and an XhoI site on the 3’ primer, and cloned into corresponding sites of the pGEX6P-1 plasmid (GE Healthcare) engineered to express -phage phosphatase (36).

The mouse Aurora-A (NCBI Reference Sequence [RefSeq] entry NM_011497) activity construct for expression of residues 120-417, corresponding to human residues 107-403, was described earlier (34). The RefSeq mouse Aurora-A activity sequence corresponds to residues 98-386 of another, shorter variant of the same enzyme in the UniProt database (entry Q8C3H8).

The mouse Aurora-B gene (RefSeq entry NM_011496) was amplified by Tgo polymerase with GACACGGATCCAAACGTCAGTTCACTATTGACAACTTTGAGATTGGGCGTCCTTTGGGCAAAGGC as the 5’ primer, GACGACTCGAGTCAGGACGGCCTCCTTGAGTTGGCCCGGACCCAAGGGTGAGC as the 3’ primer, and 15-day mouse embryonic cDNA (Ambion/Applied Biosystems) as template DNA. The amplified fragment encodes mouse Aurora-B residues 74 (except that Lys-Gln-Pro of the mouse sequence have been replaced with human Lys-Arg-Gln) to 338 (plus a serine residue added to the C-terminus of the sequence to mimic the end of the human Aurora-A crystallography construct). The mouse residues of Aurora-B correspond to residues 125-391 of human Aurora-A.

The human Aurora-C gene (RefSeq entry NM_003160), encoding the full-length Aurora-C variant 3 protein spanning residues 1 through 275 (expected molecular mass of the N-terminally GST-tagged protein = 59,011.6 Da, and of the tagless enzyme plus the linker residues = 32,598.9 Da), was amplified with GACGTGGATCCATGCGGCGCCTCACAGTCGATGACTTTGAAATCGGGCG as the 5’ primer, GATCGACTCGAGTCAGGAAGCCATCTGAGCAGAGGGAGGCAGCACCCTTCGGGAG as the 3’ primer, and a full-length clone from OriGene as template DNA.

Surface-residue substitutions in mouse Aurora B.

Wildtype full-length or truncated versions of mouse Aurora B copurify from E. coli lysates in approximately 1:1 molar ratio with bacterial chaperone protein GroEL [37, 38]. By analyzing the three-dimensional structure of humanized/stabilized mouse Aurora-A and assuming that the structure of mouse Aurora-B would be nearly identical to that of mouse Aurora-A, we initially identified seven hydrophobic surface residues in mouse Aurora-B that correspond to polar residues in mouse and human Aurora-A. We mutated codons for the following residues using the QuikChange Site-Directed Mutagenesis Kit from Stratagene as recommended by the manufacturer, to stabilize the truncated Aurora-B in solution and reduce the amount of copurifying GroEL (this construct is referred to as mAurB 7X): Leu215Ser (human Aurora-A/B residues: Ser266/Leu210; mouse Aurora-A residue: Ser279; 5’ primer: AACCTGCTGTTAGGTTCCCAGGGAGAACTGAAG, 3’ primer CTTCAGTTCTCCCTGGGAACCTAACAGCAGGTT), Leu233Ser (human Aurora-A/B residues: Ser284/Leu228; mouse Aurora-A residue: Ser297; 5’ primer: GTGCATGCCCCATCCTCCAGGAGGAAGACCATGTGC, 3’ primer: GCACATGGTCTTCCTCCTGGAGGATGGGGCATGCAC), Met258Lys (human Aurora-A/B residues: Lys309/Lys253; mouse Aurora-A residue: Lys322; 5’ primer: CGCATGCATAATGAAAAAGTAGATCTATGGTGC, 3’ primer: GCACCATAGATCTACTTTTTCATTATGCATGCG), Val291Ser (human Aurora-A/B residues: Ser342/Val286; mouse Aurora-A residue: Ser355; 5’ primer: GAGACGTATCGTCGGATTTCCAAGGTGGACCTGAAGTTC, 3’ primer: GAACTTCAGGTCCACCTTGGAAATCCGACGATACGTCTC), Pro302Thr (human Aurora-A/B residues: Thr353/Pro297; mouse Aurora-A residue: Thr366; 5’ primer: TTCCCCTCTTCTGTGACCTCGGGCGCCCAGGAC, 3’ primer: GTCCTGGGCGCCCGAGGTCACAGAAGAGGGGAA), Trp318Ser (human Aurora-A/B residues: Ser369/Ser313; mouse Aurora-A residue: Ser382; 5’ primer for Trp318Ser and Pro322Thr: CTCAAACATAACCCCTCCCAACGGCTGACCCTGGCGGAGGTTGCA, 3’ primer: TGCAACCTCCGCCAGGGTCAGCCGTTGGGAGGGGTTATGTTTGAG), and Pro322Thr (human Aurora-A/B residues: Met373/Pro317; mouse Aurora-A residue: Thr386). mAurB 7X was expressed in E. coli, purified as described below, and analyzed by electro-spray mass spectrometry to verify the integrity of the purified protein. Approximately 50% of the protein was found to undergo proteolysis between Ser291 (changed from wild type Val291) and Lys292. We purified a construct missing the Val291Ser and Met258Lys substitutions (referred to as mAurB 5X), and analyzed it by mass spectrometry to determine the extent of proteolysis between Val291 and Lys292. We found that even in this wild type configuration, proteolysis still affected approximately 25% of the total GST-tagged protein. We then introduced two additional conservative substitutions to eliminate the major site of proteolysis: Val291Leu and Lys292Arg referred to as mAurB 8X VKLR (5’ primer CCACAGTGAGACGTATCGTCGGATTCTGCGCGTGGACCTGAAGTTCC, 3’ primer GGAACTTCAGGTCCACGCGCAGAATCCGACGATACGTCTCACTGTGG), and Val291Leu and Lys292Gln referred to as mAurB 8X VKLQ (5’ primer GTGAGACGTATCGTCGGATTCTGCAGGTGGACCTGAAGTTCC, 3’ primer GGAACTTCAGGTCCACCTGCAGAATCCGACGATACGTCTCAC). Both mutant versions of the protein were no longer proteolyzed between Leu291 and Arg292 or Leu291 and Gln292, and both retained full enzyme activity. By analyzing both purified protein preparations, we discovered three additional minor sites of proteolysis between surface residues Val293 and Asp294, Thr322 (mutated from Pro322) and Leu323, as well as Ala324 and Glu325. We therefore introduced three additional mutations to the final GST-tagged mAurB construct used for activity assays (referred to as mAurB 10X VKLR): Asp294Asn (human Aurora-A/B residues: Glu345/Asp289; mouse Aurora-A residue: Glu358; 5’ primer: CGTCGGATTCTGCGCGTGAACCTGAAGTTCCCCTCTTC, 3’ primer: GAAGAGGGGAACTTCAGGTTCACGCGCAGAATCCGACG), Thr322Met (human Aurora-A/B residues: Met373/Pro317; mouse Aurora-A residue: Thr386; 5’ primer for Thr322Met and Glu325Gln: CCCTCCCAACGGCTGATGCTGGCGCAGGTTGCAGCTCACCC, 3’ primer: GGGTGAGCTGCAACCTGCGCCAGCATCAGCCGTTGGGAGGG), and Glu325Gln (human Aurora-A/B residues: Glu376/Gln320; mouse Aurora-A residue: Glu389).

Expression and purification of Aurora proteins.

Expression plasmids were transformed into E. coli BL21(DE3) Star cells (Invitrogen), and grown with vigorous shaking at 37°C in Terrific Broth supplemented with 200 µg/ml ampicillin. When the cultures reached an OD600 of approximately 0.5-0.7, the temperature in the shaker was lowered to 17°C, and the cultures were incubated for 40 min prior to induction of protein expression with 0.2 mM IPTG.

Cells were harvested 16-18 hrs post-induction by centrifugation and resuspension in a lysis buffer composed of 50 mM Tris-HCl pH 7.5, 1.0 M NaCl, 20 mM DTT supplemented with DNAse (Roche Diagnostics) and aprotinin (Sigma-Aldrich). Cells were lysed by passing the harvested culture four times through a microfluidizer (Microfluidics). Lysates were cleared by high-speed centrifugation. Approximately 70-80% of each protein variant was found in the inclusion bodies. Soluble protein was loaded onto glutathione-Sepharose columns pre-equilibrated with 50 mM Tris-HCl pH 7.5 and 400 mM NaCl. Bound protein was eluted with 200 mM Tris-HCl pH 7.5, 400 mM NaCl, 15 mM reduced glutathione and 3 mM DTT. Following elution, the Aurora-A samples were treated with GST-tagged PreScission protease (rhinoviral 3C protease, GE Healthcare) and processed as described previously [36], while several milliliters of the most concentrated Aurora-B and -C samples coming off the top of the glutathione-Sepharose column were aliquoted in small volumes and immediately snap-frozen in liquid nitrogen for storage at -80°C. Protein purity was assessed by Coomassie-stained SDS-PAGE gels, its concentration was estimated by measuring the absorbance of the samples at 280 nm, and protein modifications were analyzed with electrospray mass spectrometry as previously described [39]. Aurora-A was activated by pre-incubation with ATP, while Aurora-B and -C displayed robust enzymatic activity in the absence of such pre-incubation.