Recently serum TK1 activity has been used as biomarker for the diagnosis and prognosis of canine malignant lymphoma and leukemia [11
]. However, the canine TK1 enzyme has not been characterized. Therefore, the full-length canine TK1 cDNA was cloned and expressed in E. coli.
Recombinant canine TK1 was purified and characterized, and compared with human TK1, which was cloned and purified using the same procedures.
The kinetic properties of canine and human TK1 with their natural substrate Thd, ATP, and the anti-HIV nucleoside analog (e.g., AZT) were investigated. A reason for testing AZT is that it is used in the commercial TK1 Liaison assay. Canine TK1 had higher Vmax
values for all tested substrates than did human TK1. Therefore, the overall efficiency of canine TK1 was higher than that of human TK1. The high stability and efficiency of canine TK1 when using AZT as a substrate explains why the TK Liaison assay is suitable for canine lymphoma and leukemia studies [15
The main focus of this study was to compare the quaternary structures of serum TK1 with those of cellular and recombinant enzymes, and to investigate the effect of reducing agents on the respective subunit compositions. Earlier studies have shown that native and recombinant human TK1 occur as tetramers in the presence of ATP or at high concentrations, and as dimers in the absence of ATP or at low enzyme concentrations [24
]. Using similar techniques, we observed that recombinant human TK1 is mainly present in high MW complexes in addition to dimers and tetramers, and that pretreatment with DTE increased the extent of dimer and tetramer forms. Recombinant canine TK1, on the other hand, appeared to require reducing agents for proper folding, since in the absence of DTE only a minor fraction of the analyzed protein was recovered in high MW fractions, whereas in samples pre-treated with DTE, about 85% of the activity was recovered in the high MW form. Both canine and human serum TK1 eluted mainly as high MW complexes, and the dimer and tetramer forms had very low activity. Pretreatment with DTE resulted in >3-fold lower activity. In the case of human serum TK1, pretreatment with DTE also increased the proportion of active dimer and tetramer forms. We found that serum TK1 activity is associated with the TK1 oligomer, and there was no apparent correlation between serum TK1 activity and protein levels.
However, cytosolic TK1 from cultured canine and human cells was found mainly in dimer and tetramer forms, similar to previous reports [24
]. These results indicate that the discrepancy regarding recombinant TK1 quaternary structures in our study compared with previous reports is not due to technical reasons but may be due to the conditions used.
Human TK1 contains 11 cysteines and canine TK1 contains 8 cysteines, four of which coordinate with Zn, which leaves 7 and 4 cysteines, respectively, as free thiol groups. The structures of TK1-like enzymes, from human, bacterial, and viral origins, are all in tetrameric forms and there are no intramolecular disulfide bonds observed, since the enzymes were crystalized in reducing conditions (10
mM DTT) [7
]. In solution, in the absence of reducing agent, it is possible that surface cysteines form S-S bridges between the monomers, thereby forming oligomers. However, the oxidation of other residues like tyrosine or methionine, which lead to the formation of high molecular weight aggregates, is also possible. If oligomers were formed entirely through disulfide bonds, they should be completely reduced to dimer or tetramer in the presence of high concentrations of reducing agent. The fact that both serum and recombinant TK1 are persistently present as oligomeric forms irrespective of the presence of reducing agents suggested that other mechanisms are involved. In the case of human TK1, disulfide bonds are most likely involved, since pre-incubation with DTE increased the extent of dimer and tetramer formation. Furthermore, recombinant dog TK1 activity increased upon pre-incubation with DTE, but there was no reduction in the oligomeric form, suggesting that the reduction of oxidized residues, not necessarily cysteines, helped the enzyme to fold into the active form.
In the case of serum TK1, pre-incubation of the serum samples with DTE resulted in lower overall recovery of TK1 activity, suggesting that serum TK1 may be covalently associated through disulfide bonds with other stimulating or stabilizing factors. However, elucidation of the exact mechanism requires further study.