In silico sequence analysis
Human Nek6 and Nek7 amino acid sequences were used as queries in five different secondary structure prediction databases: PredictProtein/Prof , PSIPRED , SSpro , SOPMA  and GOR4 . Comparison of their outputs resulted in a consensus of predicted secondary structure, where each amino acid was assigned a score ranging from 1 to 5. Our Nek7 consensus of predicted secondary structure was compared to the author-approved secondary structure in PDB (2WQM) as a measure to validate our analysis. We also performed disordered regions analysis for both protein sequences using nine different predictors: FoldIndex , GlobPlot Russell/Linding , PONDR VL-XT , DISpro , IUPred , DisEMBL Hot-loops, DisEMBL Remark-465, DisEMBL Loops/coils , and VSL2B . From this, a consensus of predicted disordered regions was generated with a consensus score ranging from 0 to 9, where a score above 4 represents disorder. Additionally, NetPhosK  and NetPhos  databases were used to predict phosphorylation sites for human Nek6 and Nek7. The conserved glycine-rich sequence, the HRD and DLG motifs, the conserved residues K74 (β3 strand) and E93 (αC helix), the putative nuclear export signal LGDLGL based on la Cour et al., 2004 , the putative WW domain binding motifs PY and pSP based on Ingham et al., 2005 , as long as the PPLP motif, experimentally described for hNek6 by Lee et al., 2007 , were also assigned to both protein sequences.
All plasmid constructions were developed accordingly to Meirelles et al., 2010 .
The hNek6 activation loop mutation S206A was introduced by PCR-based mutagenesis accordingly to Meirelles et al., 2010 .
Protein Expression and Purification
Soluble full-length hNek6 wild-type - 6xHis-hNek6wt - and mutant - 6xHis-hNek6(S206A) - or truncated hNek6 wild-type kinase domain - 6xHis-hNek6(Δ1-44) - fused to a 6xHis tag were expressed and purified accordingly to Meirelles et al., 2010 .
In order to obtain dephosphorylated wild-type and mutant hNek6, plasmids encoding either 6xHis-hNek6wt or 6xHis-hNek6(S206A) and λ phosphatase were transformed into E. coli BL21 (DE3/pRARE) cells that were induced and purified as described by Meirelles et al., 2010 . Lambda phosphatase cloned into pCDF-Duet (Novagen) was kindly provided by Dr. Richard Bayliss (Section of Structural Biology, Institute of Cancer Research, London, UK).
Circular dichroism (CD) spectra were recorded in a JASCO model J-810 CD spectropolarimeter equipped with Peltier-type system PFD 425S. Data were collected from 260 to 200 nm at 4°C using a quartz cuvette of 1 mm pathlength. Thirty-two spectra of purified 6xHis-hNek6wt at 5.8 μM, in 50 mM Phosphate buffer pH 7.5, 300 mM NaCl, were averaged and corrected from the baseline for buffer solvent contribution. Experimental data were analyzed using CDNN  and K2d  softwares.
Comparative/Homology Molecular Modeling
The comparative/homology molecular modeling and model validation were performed in a similar way to that described in Bodade et al., 2010 . Briefly, several comparative/homology modeling tools were used: I-TASSER [69–71], Geno3D , 3D-JIGSAW [73–75], SWISS-MODEL  and MODELLER 9v8 . The NCBI Basic Local Alignment Search Tool (BLAST, http://www.ncbi.nlm.nih.gov/BLAST/) was used to search the crystal structure of the closest homologue available in the Protein Data Bank (PDB, http://www.rcsb.org/pdb/). The input was the amino acid sequence of hNek6(S206A). The NCBI results revealed that the structure of hNek7, deposited under the code 2WQM in the PDB, was a very suitable template (identity score of 81% and E-value 3 × 10-141). This structure was used as a single template for the modeling approach. The overall stereochemical quality of the models was assessed by PROCHECK software . The quality of the models was also evaluated by PROSA [51, 52] and by the standard validation procedures included in the automated mode of the SWISS-MODEL server .
Small Angle X-Ray Scattering Analysis
The sample was first inspected by dynamic light scattering (DLS) to test for its monodispersity and then ultracentrifuged at 200.000 × g for 40 min at 4°C to remove any possible aggregates. The SAXS experiments were performed at the D02A-SAXS2 beam line at the LNLS, and data treatment and analyses were done following standard procedures similar to those described in Trindade et al., 2009 . Briefly, the measurements were performed at 4°C and the sample-to-detector distance was 902 mm, covering a scattering vector range of 0.015Å-1 < q <0.25 Å-1 (q is the magnitude of the q-vector defined by q = (4π/λ)sinθ and 2θ is the scattering angle) using a wavelength of λ = 1.488 Å. The measurements were performed using two different protein concentrations in HEPES buffer (50 mM HEPES pH 7.5, 5 mM sodium phosphate, 300 mM NaCl, 5% glycerol, 200 mM imidazole): 0.5 and 1.0 mg/mL. A 8 mg/ml BSA (66 kDa) solution in the same sample buffer was used as a standard sample to estimate the molecular mass of 6xHis-hNek6(S206A) making use of the ratio of the extrapolated values of the intensity at the origin, I(0) [78, 79]. The radius of gyration (Rg) was calculated from the Guinier approximation (valid for qRg < 1.3) [80–82] and also from the pair distance distribution function, p(r), which was obtained using the program GNOM . The maximum dimension (Dmax) of the molecule was obtained from the p(r) function. The Kratky plot (q2I(q) vs. q) [81, 82] was used to analyze the compactness of the protein conformation.
Low resolution SAXS-based modeling
The low resolution model of 6xHis-hNek6(S206A) was obtained from the SAXS data using a combination of ab initio calculation and rigid body modeling methods. Taking advantage of the homology model obtained, we used the program BUNCH  to model the protein. No symmetry restraints were used in the calculation. We would like to mention that no unique solution can be obtained from these calculations. For this reason, 10 independent calculations were run for each sample data. The multiple solutions were analyzed and the reliability and stability of the set of models were estimated. A pairwise comparison and the normalized spatial discrepancy (NSD) evaluation was performed using the DAMAVER program suite  complemented by the SUPCOMB  routine. Analyzing the NSD values (which describe the dissimilarity between pairs of models of the several calculations), the models with common features led to the selection of a representative, low resolution conformation for hNek6(S206A) protein. Models were displayed by the PyMOL program .
For comparison purposes, two other low resolution models were also obtained by using two different ab initio approaches: the dummy atoms method implemented in the program DAMMIN  and the dummy residues method implemented in GASBOR . The procedures were similar to those described in Trindade et al. .
We used Analytical S ize-E xclusion C hromatography coupled to M ulti-A ngle L ight S cattering (SEC-MALS) to estimate the hydrodynamic or Stokes radii (Rs) of recombinant hNek6wt, hNek6(S206A), hNek6(Δ1-44) and dephosphorylated hNek6wt and hNek6(S206A), all fused to a 6xHis tag. SEC was performed with an analytical Superdex 200 10/300 GL column using an ÄKTA FPLC system (GE Healthcare) equilibrated with two column volumes of 50 mM HEPES pH 7.5, 5 mM sodium phosphate, 600 mM NaCl, 5% glycerol, at a flow rate of 0.5 ml/min, at 20°C. Recombinant hNek6 variants at concentrations ranging from 0.2 to 0.7 mg/ml and a mixture of standard proteins with known Stokes radii (conalbumin: 3.64 nm, 3.2 mg/ml; ovalbumin: 3.05 nm, 4.2 mg/ml; carbonic anhydrase: 2.30 nm, 3.0 mg/ml; and ribonuclease: 1.64 nm, 3.4 mg/ml) (Sigma) were loaded onto the column and their elution profiles were monitored by absorbance at 280 nm. The Stokes radius of each hNek6 variant was estimated by a linear fit of the Stokes radii of the standard proteins versus the partition coefficient Kav[88, 89] as described by the equation: Kav = Ve - Vo/Vt - Vo , where Ve is the elution volume of the protein, Vo the void volume and Vt is the total volume of the column. The SEC was also coupled to a DAWN TREOS™ MALS instrument (Wyatt Technology). The on-line measurement of the intensity of the Rayleigh scattering as a function of the angle of the eluting peaks in SEC was used to determine the weight average molecular masses (Mw) of the eluted proteins , using the ASTRA™ (Wyatt Technologies) software. SEC-MALS measurements were performed using two different buffers: the first one described above for SEC and a second one also used in SAXS experiments (50 mM HEPES pH 7.5, 5 mM sodium phosphate, 300 mM NaCl, 5% glycerol, 200 mM imidazole). The chromatographic profile of the recombinant hNek6 variants were the same in both measurements and the mean and standard errors of their Mw and Mn were calculated.
Thermal Shift Assays
Thermal shift assays were performed based on a protocol devised by the Structural Genomics Consortium  using a real time PCR machine 7300 (Applied Biosystems). Proteins were buffered in 10 mM HEPES pH 7.5, 150 mM NaCl and assayed at a final concentration of 2.0 μM in 25 μL volume. SYPRO-Orange (Molecular Probes) was added as a fluorescence probe at a dilution of 1 in 1000. The emission filter for the SYPRO-Orange dye was set to 580 nm. Temperature was raised with a step of 1°C per 1.0 min from 10°C to 85°C and fluorescence readings were taken at each interval. OriginPro 8 software was used to fit data to the Boltzmann equation, y = LL+(UL-LL)/1+exp((Tm-x)/a), where LL and UL are the slopes of the native and denatured baselines, Tm is the apparent melting temperature and a describes the slope of the denaturation. Tm values were calculated by determination of the maximum of the first derivative.