Calcium affinity of human α-actinin 1

Cloning, expression and purification

The gene fragment coding for residues 712–914 (base pair 2444–3055) of human α-actinin 1 (accession NM_001130004) was obtained from GenScript, USA. This fragment contains exons 18–21 of human α-actinin 1 and codes for the brain isoform. The non-muscle isoform, lacking exon 19b, and the smooth muscle isoform, lacking exon 19a, were obtained by mutagenesis of the original gene fragment (Fig. 1). Mutagenesis and control of fidelity of gene fragments were performed by GenScript, USA.

The gene fragments were excised from pUC57 by BamHI and XhoI digestion and ligated into pET-TEV (a modified pET-19b) vector containing an N-terminal 10xHis-tag and a TEV protease cleavage site. This produced three plasmids: pET-BR-ACTN1-EF (brain), pET-NM-ACTN1-EF (non-muscle) and pET-SM-ACTN1-EF (smooth muscle).

E. coli BL21(DE3) cells were transformed (by heat-shock) with the purified plasmids containing the different constructs. The transformed cells were cultured at 37 °C in Luria-Betani medium containing 100 µM carbenicillin until an optical density of 0.6–0.8 at 600 nm was reached. Protein expression was induced by addition of isopropyl thio-β-D-galactoside to a final concentration of 0.5 mM and cells were grown overnight at 16 °C. Cells were harvested by centrifugation (50,000 × g for 30 min), resuspended in 25 mM sodium phosphate buffer, pH 7.6, 150 mM NaCl and stored at −20 °C. Frozen cells were thawed and then lysed by sonication. Triton X-100 and polyethyleneimine were added to final concentrations of 1% and 0.05%, respectively. After 30 min on ice, cell debris was removed by centrifugation at 50,000 × g for 30 min. Imidazole was added to the clarified supernatant to give 10 mM final concentration and loaded onto a HiTrap™ Chelating HP column (GE Healthcare Life Sciences, Little Chalfont, UK) charged with nickel. Unbound proteins were eluted with 25 mM sodium phosphate buffer, pH 7.6, 150 mM NaCl, 10 mM imidazole. Bound proteins were eluted with an imidazole gradient ranging from 10 to 510 mM imidazole in the same buffer. Imidazole in eluted fractions was removed by gel filtration on a HiPrep 26/10 desalting column (GE Healthcare Life Sciences, Little Chalfont, UK). Collected fractions were incubated at 4 °C for 16 h in the presence of Tobacco Etch Virus (TEV) protease (kindly provided by Dr. David S. Waugh). The released 10 × His-tag and the 6 × His-tagged TEV protease were removed by affinity chromatography as before. Finally, purified proteins were transferred into 50 mM Tris–HCl, pH 7.6, 200 mM KCl by gel filtration on HiPrep desalting columns (GE Healthcare Life Sciences, Little Chalfont, UK) and concentrated by Amicon Ultra 3 K centrifugal filter devices (Millipore Corporation, Billerca, Massachusetts, USA).

Protein concentration of native protein was determined from the absorbance at 280 nm using the molar absorptivity, as calculated from the amino acid sequence (using ProtParam at the ExPASy proteomics server). Purity of the expressed peptides were routinely determined under denaturating conditions by SDS-polyacrylamide gel electrophoresis (Laemmli, 1970).

CD and fluorescence spectroscopy

Far-UV circular dichroism (CD) spectra were acquired on a Jasco-J-810 spectrophotometer using a quartz cuvette with 2 mm path length. Spectra were collected between 200 and 260 nm with a data interval of 0.025 nm. The mean residue molar ellipticity ([Θ]_MRW) was determined from three accumulated spectra.

The web service MLRC (Guermeur et al., 1999) at Pôle Bioinformatique Lyonnaise was used to predict the secondary structure. The relation $\frac{\left(\left[{-\Theta }_{222}\right]+3000\right)}{39000}$ (Morrow et al., 2000) was used to estimate the α-helical content.

Intrinsic fluorescence was determined on a Jasco FP-6500 spectrofluorometer (JASCO, Easton, Maryland, USA). Samples were excited at 280 nm and emission spectra between 310 and 600 nm were collected.

Calcium binding

The calcium affinity was determined by isothermal calorimetry using a MicroCal iTC₂₀₀ (GE Healthcare Life Sciences, Little Chalfont, UK). For this the measuring cell was loaded with 0.3–0.6 mM protein in 50 mM Tris–HCl, pH 7.6, 200 mM KCl, and the syringe was filled with 5–10 mM calcium chloride in the same buffer.

Residual calcium content in the concentrated protein samples was estimated using Quin2 (Linse et al., 1987). It was found to be less than 0.02 mol of Ca²⁺ per mol of protein, independent of whether the protein solutions had be decalcificated by passing them over a Chelex 100 column (Bio-Rad Laboratories, Hercules, California, USA). The calcium concentration in the measuring buffer was around 1 µM.

Injection profiles were analysed using the standard single-site model of Origin (MicroCal, GE Healthcare Life Sciences, Little Chalfont, UK) as well as of Sedphat (Houtman et al., 2007).

Fluorescence and CD spectroscopy were used to study the effect of calcium ions on the spectral properties of the peptides.

Cloning, expression and purification

Transformation of E. coli with either of the plasmids, followed by induction and expression produced large amounts of protein. After purification on nickel-columns, cleavage by TEV protease and gel filtration all three proteins were purified to nearly homogeneity as determined by denaturing gel electrophoresis.

Initially, proteins were after the final gel filtration step kept in a Tris-buffered sodium chloride solution. Preliminary experiments by fluorescence spectroscopy indicated that the brain EF-hand peptide unfolded slowly with time at room temperature in this media. However, when the peptide was transferred into a Tris-buffered potassium chloride solution the folding appeared stable as the emission spectra did not change with time. Therefore, the final gel filtration step was used to transfer the purified proteins into this buffer.

CD and fluorescence measurements

To check whether the purified peptides were folded CD measurements in the far-UV range were used. Figure 2 shows that each of the three peptides displayed the typical negative peaks at 208 and 222 nm observed for proteins with high α-helical content. Estimating the helical content from the mean residue molar ellipticity at 222 nm indicated helical contents of 45.1% (brain), 52.9% (non-muscle) and 47.9% (smooth muscle). These values agreed reasonable well with the predicted ones for the brain (49.3%) and non-muscle (51.4%) forms whereas the prediction for the smooth muscle form deviated more (54.5%) from the estimated value. In neither case, addition of calcium did not affect the CD spectra, independent of the concentration.

The π-π^∗ transition of the amide at 208 nm is sensitive to inter-helix coupling. In contrast, the n-π^∗ transition at 222 nm is more or less unaffected by inter-helix coupling but instead responsive to α-helical content (Lau, Taneja & Hodges, 1984; Zhou, Kay & Hodges, 1992). Therefore, the ratio of ellipticity at 222 and 208 nm has been used as a measure for formation of coiled-coils. Values of Θ₂₂₂/Θ₂₀₈ around or greater than 1 has been taken to indicate the presence of coiled-coils whereas a value around 0.83 would indicate non-interacting single-stranded α-helices. As the peptides exhibit ratios ranging from 0.88–0.93, it suggested that the α-helices in the EF-hand motifs are non-interacting. This would be expected assuming that the peptides fold into the typical helix-loop-helix structure of EF-hand motifs, with rather limited contacts between the helices.

Fluorescence spectroscopy is useful for monitoring structural changes (Eftink, 1991). The emission from excited aromatic residues, particular tryptophans, is very sensitive to changes in the surroundings. Thus, upon a conformation change leading to more or less exposure of the excited residue, the intensity of emission and/or the wavelength of emitted light will change. However, independent of the amount of added calcium to any of the α-actinin peptides no change in the emitted light could be detected. This was not particularly unexpected, as the only tryptophan present as well as all tyrosines are located outside the plausible calcium-binding loop (Fig. 4). Although there are several phenyalanine residues in the calcium-binding loop that may experience a change in the surroundings, the quantum yield from these are probably too low to be detected in the presence of other aromatic residues.

Isothermal calorimetry

To analyse the plausible calcium affinity of the α-actinin EF-hand domains in more detail isothermal calorimetry was used. Figure 3 shows typical heat traces and binding isotherms obtained from measurements of the different peptides. The results clearly showed that the brain and non-muscle isoforms both bound calcium in contrast to the smooth muscle isoform. Although measured heat of injection of the brain EF-hand domain was around 10 times as large as that of the non-muscle isoform, the estimated affinity of both were similar. Analysis of the binding isotherms using a one-to-one model gave dissociation constants of 49 ± 9 µM (11 measurements) and 56 ± 18 µM (12 measurements) for the brain and non-muscle isoforms, respectively. In both cases the number of sites were estimated to 0.9 ± 0.1. Interestingly, the entaphy and entropy changes differed substantially; ΔH and TΔS of the brain peptide was −3.7 and 2.2 kcal/mol compared to −0.34 and 5.46 kcal/mol for the non-muscle peptide.

The heat trace obtained for the smooth muscle peptide was nearly identical to that obtained when injecting buffer without calcium, indicating that this EF-hand domain does not bind calcium under these conditions.

Sequence analysis

Analysis by Superfamily (Wilson et al., 2009) predicted that there should be 4 EF-hand motifs in the C-terminal domain. However, the E-values were only significant for the first EF-hand motif of brain and non-muscle α-actinin 1. Likewise, Pfam (Finn et al., 2014) recognized only the first EF-hand motif in the brain and non-muscle sequences as a proper calcium-binding motif. Pfam also identified motifs 3 and 4 as calcium-insensitive EF-hand motifs.

In an EF-hand loop, the calcium ion is coordinated in a pentagonal bipyramidal arrangement comprising 12 residues. In the typical coordination sphere, six residues at positions 1, 3, 5, 7, 9 and 12, denoted X, Y, Z, −Y, −X and −Z, coordinate the calcium ion. Residues X, Y, Z and −Z coordinate the calcium ion through side chain oxygenes whereas the −Y coordinate is provide by a backbone oxygene and the −X coordinate is provide by a water molecule interacting with the residue at this position.

As Fig. 4 shows, manual alignment of the calcium-binding loop of the consensus sequence with the different isoforms indicated that only EF-hand motif 1 of brain and non-muscle α-actinin contained the proper residues to coordinate a calcium ion. At position Y in the smooth muscle isoform, there is a lysine residue instead of a residue with an oxygene-bearing side chain; according to the consensus pattern (Kretsinger et al., 1991; Grabarek, 2006; Gifford, Walsh & Vogel, 2007) there is usually an aspartate residue or less often asparagine or serin residue at this position. In all other motifs, one or more of the oxygene-bearing residues required for coordinating calcium are missing.