Creatine Kinase (1CRK) from Gallus gallus
Created by: Ashley Hirt
Creatine kinase (CK) is an enzyme whose various isoforms are essential to the phosphocreatine circuit, an energy storage system used in cells with dynamic energy needs (e.g. muscle, brain). Different CK isoforms are present in the cytosol and the mitochondria, where they perform different functions. Mitochondrial CK (MtCK) facilitates the formation of phosphocreatine (PCr) from mitochondrial ATP; PCr is then exported into the cytosol, where cytosolic CK catalyzes its breakdown into creatine (Cr) with the transfer of the phosphate group to ADP, producing ATP (Figure 1). CKs are important to maintain highly localized ATP levels. Near ATPases, cytosolic CK maintains a high [ATP]/[ADP] ratio, thus increasing the ΔG of ATP breakdown (because Keq decreases) (1). Conversely, in the mitochondria MtCK maintains a low [ATP]/[ADP] ratio to spur production of ATP through oxidative phosphorylation (1). Finally, PCr is more diffusible than ATP, so it is a more efficient shuttle of ATP from the mitochondria into specific locations in the cytosol (1).
The isoform investigated here (PDB ID: 1CRK) is mitochondrial and sarcomeric (sMtCK), originating from chicken cardiac tissue (2). As a sarcomeric enzyme, sMtCK is particularly important to maintain a supply of PCr that can be broken down into ATP for muscle contraction. As a result, both sMtCK and cytosolic CK are abundant in muscle tissue, accounting for up to 20% of the soluble proteins in some muscles (3). Also, since most of the heart’s ATP production comes from mitochondria (which, accordingly, make up 40% of cardiac cellular volume), sMtCKs are abundant in the heart (1). This allows blood CK levels to be used as diagnostic markers for myocardial infarctions and muscle deterioration, during which damage to the respective tissue causes the release of CK into the bloodstream (3). sMtCK is extremely sensitive to reactive oxygen/nitrogen species, a problem which is exacerbated due to its localization in the mitochondria, where the majority of these reactive species are produced via the respiratory chain (1). These species reduce octamerization and cause mostly irreversible inactivation of sMtCK (1). If this occurs in the heart, loss of sMtCK and its maintenance of cellular ATP levels impairs cardiac ion pumps and thus leads to cardiac dysfunction (1).
sMtCK was purified from the myocardium of Gallus gallus via affinity chromatography with Blue Sepharose, followed by fast liquid chromatography (2,4). The purified protein was then crystallized at room temperature using the hanging drop technique in a buffer of 25 mM sodium phosphate, 1 mM sodium azide, 2 mM β-mercaptoethanol, and .2 mM sodium EDTA, at pH 6.7. 17% PEG 1000 was used as the precipitant. Also, a second group of co-crystals were grown with 5 mM NaATP, producing sMtCK with bound ATP. Finally, the crystals were freed from unbound ATP and their structures were obtained through X-ray diffraction (2).
sMtCK is an octamer localized in the intermembrane space and cristae of the mitochondria (1). Physiologically, it is often found bound to ATP (specifically the phosphate group and the magnesium associated with ATP) and creatine to catalyze the formation of phosphocreatine. sMtCK is made up of 4 subunits that dimerize to form an octamer, which stabilizes the protein by decreasing the surface area available for solvent interactions (a lower surface area:volume ratio). Each subunit’s primary structure is identical and each has a binding site for one phosphate group of ATP. In terms of secondary structure, each of the four chains have a small α-helical domain (five α-helices and a 310 helix) and a larger α/β domain (one 8-stranded β-sheet surrounded by seven α-helices, as well as a short 3-stranded β-sheet) with random coils in between these domains (2). All but two of the β strands are hydrophobic, so we would expect them to be in the interior of the protein. We also see that there are some amphiphilic α-helices, which we would expect to be on the outside of the protein with the polar face towards the channel and the nonpolar face towards the intermembrane space. For an integral protein, there seem to be excessive hydrophilic residues (6/8 β-strands and 7/12 α-helices), but these are necessary for the polar internal channel that sMtCK forms to bind to and traffick charged, polar ATP (see Figure 2) (2,5).
There are a total of 1520 residues in each half of the octamer (380 per chain). The protein was fully crystallized, but amino acid assignments for residues 316-326, 60-66, and 1-5 are still uncertain due to weak electron density. Also, like some other protein-kinases (PKA, MAPK, and Cdk2), sMtCK has a P-loop consensus sequence matching ΦGXGXXG(A,S)V, where X is any amino acid and Φ is a hydrophobic amino acid (see Figure 3) (2,6).
Specific residues are important to sMtCK’s stability. Since the chains are identical, residues are repeated in each of the 8 chains in the octamer. The positively charged, basic residues between Lys-360 and Lys-380 are involved in binding the negatively charged head groups of the mitochondrial membrane, stabilizing the protein to the outer membrane. These groups also bind cardiolipin in the inner membrane, allowing sMtCK to span the intermembrane space and tether the two membranes together. Also, Trp-264 is thought to greatly stabilize the octamer by participating in hydrophobic interactions between dimer pairs. Finally, there are four contact regions in sMtCK that form interfaces between dimers and the octamer to help with folding; these include Lys-5 to Lys-20 and Pro-31; Tyr-34, Ser-47 to Cys-51, Ile-52 to Asn-58; Asn-44 and Gly-45, Ser-142 to Arg-147, Thr-172, Lys-191 and Arg-204 to Trp-206; Gly-134 to Ser-136 and Glu-261 to Trp-264 (2). Also, the N-terminal sequence (Thr-1 to Phe-9) has a profound effect on octamer formation. Substitution of any one of the charged sites (Glu-4 to Lys-7) with uncharged amino acids results in a 50-fold increase in the Keq and a 13-fold increase in the rate of the dissociation of octamers to dimers, with an exacerbation of these effects if multiple charged sites are substituted (7).
Formation and stabilization of the octamer is critical for sMtCK function. Indeed, mutant sMtCK that is enzymatically active but lacks the ability to octamerize cannot bind the mitochondrial membrane, nor stimulate oxidative phosphorylation in the mitochondria by driving a low [ATP]/[ADP] ratio (1). Also, the specific mechanism by which reactive oxygen species are so dangerous for sMtCK is through their oxidation of Met, Trp, and Tyr residues. As these residues enable the dimer-dimer interface, their oxidation prevents octamerization (1).
Various residues facilitate the binding of sMtCK’s ligands. Glu-227, Glu-226, and Asp-228 (all acidic, negatively charged residues) can interact with the positive magnesium ion bound to ATP to coordinate its binding. Multiple residues (Met-235, His-186, Ser-123, His-291, Gly-289, Arg-125, Arg-287, Asp-330) form a pocket that associates with the adenine group of ATP; three in particular (Arg-336, Arg-127, and Arg-287) form hydrogen bonds with the phosphate group of ATP. Also, Trp-223 is necessary for the enzyme to bind ATP; its replacement results in complete inactivation of the enzyme. For sMtCK’s other ligand, Cys-278 marks the creatine-binding site, and His-92 and His-186 assist in the binding of creatine (2,8). In response to ligand binding, various “flexible loop” residues (60-66 and 316-326) move closer to the active site, causing a conformational change that excludes water (2). Also, a histidine near the active site (His-61) may function as an acid-base catalyst in the reaction due to its relatively high pKa of 7. Thus, in the forward reaction (intermembrane space, pH=8), it would be deprotonated and able to gain a proton from Cr, making the latter able to nucleophilically attack the gamma-phosphate of ATP; in the backward reaction (cytosol, pH=7), it would be protonated and able to donate a proton to PCr after PCr donates its phosphate to ADP. However, this model is not yet confirmed (8).
Only one drug that interacts with CK has a known mechanism of action. Acyclovir is a nucleotide analog used in the treatment of chickenpox and various types of herpes infections; its final form is preferentially targeted by viral DNA polymerase, when it is incorporated into viral DNA and halts replication. To achieve its final form (acyclovir triphosphate), acyclovir must be acted upon by any of various kinases, one of which is creatine kinase (9,10). Unfortunately, it is not known which isoform of CK is involved in this process, but one can assume that it would not be sMtCK due to its localization in the mitochondrial membrane.
Multiple databases were used to compare sMtCK to similar proteins. sMtCK (including the mitochondrial targeting sequence) has a molecular weight of about 47 kDa and a predicted isoelectric point of 8.86 (5,11). Without the mitochondrial targeting sequence, it has a molecular weight of about 43 kDa and an isoelectric point of 8.50 (11,12). PSI-BLAST was used to find proteins with similar primary structure by aligning the sequences and finding gaps; an E value below the threshold of .05 indicates similar primary structure. Also, the Dali Server was used to find proteins with similar tertiary structure by calculating the difference in their intramolecular distances; a Z-score above 2 indicates similar tertiary structure. From these databases, we see that a similar protein is Lombricine kinase (PDB ID: 3JPZ), which has an E value of 0.0 (indicating very similar primary sequence) and a Z-score of 51.2 (indicating similar tertiary structure) (13,14). Lombricine kinase (LK) originates from the innkeeper worm, Urechis caupo, and was chosen as a comparison protein because it serves a similar purpose as sMtCK.
Based on the PSI-BLAST alignment, we see that the primary sequences of sMtCK and LK differ. Only 58% of the amino acids are identical, with a long stretch of identical or conservative amino acid substitutions near sMtCK residues 190-250, marking the conserved catalytic core (13,2). From the Dali alignment, we see that there is similarity between the secondary structure of the two proteins (14). Also, the tertiary structures superimpose well, with the exception of the location of the substrate binding site and a few random coils at the beginning, middle, and end of the sMtCK protein that are not in LK (14). These similarities are to be expected, as lombricine kinase performs a similar function as sMtCK. Rather than using creatine as a substrate, LK uses lombricine, but the use of these molecules is the same: they are phosphorylated in the mitochondria and used to shuttle ATP (in the form of phosphorylated storage molecules) to specific parts of the cytosol. Further, the sites that each kinase interacts with are similar: both kinases bind the phosphate moiety of ATP, and they both bind the structurally similar guanidino end of lombricine or creatine (see Figure 4). However, due to the increased size of lombricine compared to creatine, LK has a slightly modified active site conformation to allow for the presence of these extra atoms (15).
In conclusion, sarcomeric mitochondrial creatine kinase (sMtCK) is important to study because of its critical role in providing rapid access to ATP and in maintaining proper localized ATP concentrations. These functions are particularly important in tissues with fluctuating energy needs, like the heart and skeletal muscle. In the future, studies could investigate how to prevent the loss of sMtCK (and subsequent cardiac deterioration) due to reactive oxygen species, as well as confirming whether His-61 functions as an acid-base catalyst. Currently, an alternative model to the one stated previously is that His-61 simply accelerates a spontaneous process without serving as an acid-base catalyst. This model proposes that, by excluding water from the protein channel, His-61 enables close interactions between creatine and ATP that may result in spontaneous transfer of the phosphate group (8).