Manganese-Dependent Serine/Threonine/Tyrosine Kinase From Arabidopsis Thaliana : Role Of Serine And Threonine Residues In The Regulation Of Kinase Activity

Abstract

Protein phosphorylation is an important post-translational modification of proteins, which can either activate or inhibit the function of a given protein. The enzymes, protein kinases and protein phosphatases catalyze the phosphorylation and dephosphorylation of target proteins, respectively. Protein kinases catalyze the transfer of γ-phosphate from ATP to serine, threonine or tyrosine residues in target proteins. They are traditionally classified as protein serine/threonine kinases and protein tyrosine kinases based on the amino acid to which they transfer the phosphate group. Protein tyrosine kinases play vital roles in numerous pathways that regulate growth, development and oncogenesis in animals. However, no protein tyrosine kinase has been cloned so far from plants. The sequence motif, CW(X)6RPXF of sub-domain XI is well conserved among biochemically characterized protein tyrosine kinases from human, rat, mice, worm, fruitfly and Dictyostelium. To seek plant genes encoding tyrosine kinase, we have performed extensive genome-wide analysis of Arabidopsis thaliana using the delineated tyrosine kinase from animal systems. Repetitive database mining with CW(X)6RPXF sequence motif revealed the presence of 57 different protein kinases that have tyrosine kinase motifs. Myosin light chain protein kinase was identified as false positive with this motif. Multiple sequence alignment of all the 57 kinases indicated the presence of twelve conserved sub-domains in their kinase catalytic domain. Out of the 12 sub-domains present in protein kinases, sub-domain VIb confers serine/threonine kinase Specificity and sub-domains VIII and XI confer tyrosine kinase specificity. All the 57 kinases were Verified to contain CW(X) 6RPXF in sub-domain XI. However, the catalytic domain of all the catalogued kinases contain KXXN motif in sub-domain VIb, which is indicative of serine/threonine Kinase specificity. None of the kinases had the tyrosine kinase consensus motif RAA or ARR in sub-domain VIb. Thus, the catalytic domains of all the identified Arabidopsis protein kinases have motifs for serine/threonine specificity in sub-domain VIb and tyrosine kinase motif in sub-domain XI. These results indicate that perhaps all the kinases belong to the dual-specificity kinase family.

Hence, we have tentatively named these protein sequences as STY (serine/threonine/tyrosine) protein kinases. To examine the relationships of Arabidopsis STY protein kinases, a topographic cladogram was constructed. Casein kinase 1 was used as an outgroup to perceive the true class of STY protein kinase family. Neighbor joining tree was constructed with the full-length protein sequences following the alignments. Dendrogram of STY protein kinases suggested that the kinases are mainly clustered into four groups. Group I includes kinases related to ATN1-like kinases, peanut STY related kinases, soybean GmPK6-like kinases and ATMRK1-like kinases. Group II consists of MAP3K-like kinases, CTR1 and EDR1 related kinases. Group III includes protein kinases that harbor ankyrin domain repeat motifs. These kinases are related to Medicago sativa ankyrin kinase, MsAPK1. Group IV consists of light sensory kinases that are related to Ceratodon purpureus phytochrome kinase. C. purpureus light sensory kinase has both phytochrome and protein kinase domains. However, the protein kinases of group IV do not have a phytochrome domain.

BLAST analysis was performed using CW(X)6RPXF motif against all the available plant sequences in the database. We retrieved 11 rice protein kinases and 14 Dictyostelium kinases. In addition, we obtained STY protein kinases from wheat, barley, soybean, tomato, beech and alfalfa. Dendrogram analysis indicated that the plant STY protein kinases are clustered in similar manner as observed for Arabidopsis. Large number of sequences were retrieved when the tyrosine kinase motif CW(X)6RPXF was used to perform BLAST analysis against all the known sequences from animals. As large numbers of protein tyrosine kinases are available in animals, we have used representative kinases of each family towards the construction of phylogenetic tree. The main difference between the animal and plant tyrosine kinases is in the consensus motif conferring the tyrosine and serine/threonine specificity in the sub-domain VIb. Animal tyrosine kinases have consensus ARR/RAA in sub-domain VIb and plant kinases have KXXN which is indicative of serine/threonine specificity.

Expression analysis of Arabidopsis STY protein kinases was performed using Genevestigator online search tool Meta-Analyzer. Genes were grouped based on their relative expression levels during a specific growth stage, in a particular organ or following different environmental stresses. Various kinases are highly expressed in stamens and seeds while some kinases are expressed ubiquitously. A number of biotic and abiotic factors upregulated plant STY protein kinases. Gene expression data suggests the importance of STY protein kinases in plant growth and development.

Genome-wide analysis is supported by phosphoproteomics in Arabidopsis seedlings. Evidence for tyrosine phosphorylated proteins is provided by alkaline hydrolysis, phosphoamino acid analysis and peptide mass fingerprinting. Alkaline treatment detected two proteins corresponding to 46 and 37.5 kD. Phosphoamino acids analysis confirmed their dual-specificity nature. MALDI mass spectrometry and peptide mass fingerprinting analysis identified these two proteins as ATN1 and peanut serine/threonine/tyrosine protein kinase like protein from Arabidopsis. To further support the in silico approach, we have overexpressed one of the identified Arabidopsis thaliana serine/threonine/tyrosine protein kinases (AtSTYPK) in E. coli. The recombinant kinase was induced with IPTG and purified by using nickel-nitrilotriacetic acid affinity chromatography. AtSTYPK exhibited a strong preference for manganese over magnesium for kinase activity. The autophosphorylation activity of AtSTYPK was inhibited by the addition of calcium to reaction buffer containing manganese. The rate of autophosphorylation reaction was linear with increasing time and protein concentration. The AtSTYPK phosphorylated histone H1 (type III-S), and myelin basic protein (MBP) in substrate phosphorylation reaction and it did not phosphorylate casein or enolase. To see whether calcium or magnesium inhibits phosphorylation of MBP, we have performed the reaction in the presence of combination of different metal ions. The MBP phosphorylation reaction is more efficient in the presence of Mg2++ Mn2+ than Ca2++ Mn2+ under the same conditions. The recombinant kinase autophosphorylated on serine, threonine and tyrosine residues and phosphorylated myelin basic protein on threonine and tyrosine residues. The AtSTYPK harbors a manganese-dependent serine/threonine kinase domain, COG3642. H248 and S265 on COG3642 are conserved in AtSTYPK and the site-directed mutation of H248 to alanine resulted in loss of serine/threonine kinase activity, but the mutation of S265 to alanine showed a slight increase in its kinase activity.

The protein kinase activity is regulated by various mechanisms that include autophosphorylation, protein phosphorylation by other kinases and by the action of regulatory domains or subunits. The role of tyrosine residues in the regulation of peanut dual-specificity kinase activity is well documented, but the importance of serine and threonine residues in the regulation of dual-specificity protein kinase is not studied so far. The kinase activity is generally regulated by phosphorylation of one or more residues within the kinase activation loop. The role of threonine residues in the kinase activation loop and the TEY motif of AtSTYPK were investigated in the present study. Four threonine residues in the activation loop and a T208 in the TEY sequence motif were converted to alanine to study their role in the regulation of kinase activity. The protein kinase activity was abolished when T208 and T293 of the activation loop were converted to alanine. Interestingly, the conversion of T284 in the activation loop to alanine resulted in an increase in both auto- and substrate phosphorylations. The mutation of T288 and T291 to alanine had no effect on kinase activity.

There are eight serine residues in the kinase catalytic domain of AtSTYPK and surprisingly there is no serine residue in the kinase activation loop. So it is worthwhile to see how phosphorylation of serine residues regulates the dual-specificity protein kinase activity. The role of each serine residue was studied individually by substituting them with alanine. Serines at positions 215, 259, 269 and 315 regulate the kinase activity both in terms of autophosphorylation and substrate phosphorylation of myelin basic protein. The mutation of serine 265 to alanine resulted in slight increase in auto- and substrate phosphorylations, suggesting that it could be autoinhibitory in function. The other serine residues at positions 165, 181 and 360 did not show any change in the phosphorylation status when compared to wild-type AtSTYPK. In conclusion, this data suggests the importance of serine and threonine residues in the regulation of dual-specificity protein kinase activity and emphasizes the complexity of regulation of dual-specificity protein kinases in plants.

To summarise, this study suggests that tyrosine phosphorylation in plants could be brought about only by dual-specificity protein kinases that phosphorylate on serine, threonine and tyrosine residues. This raises an interesting possibility that plants lack classical tyrosine kinases. The results provide a first report of manganese-dependent dual-specificity kinase from plant systems. This data also suggests that plant dual-specificity kinases undergo phosphorylation at multiple amino acid residues and their activity is regulated by various mechanisms, suggesting that they could be regulated by different mechanisms at different stages of plant growth and development.

However, the role of dual-specificity kinases in planta has to be understood by mutant analysis in order to assign the physiological roles to these kinases. Further studies are needed to identify the upstream kinase(s) and downstream targets. Determination of physiological functions for dual-specificity protein kinases raises an important challenge in future in the area of plant signal transduction.