Abstract: Using CaMK2B gene cds of human and mouse to blast Sus scrofa 9 genome, the porcine CaMK2B gene sequence was obtained. Homologous protein sequence between four isoforms in 9 species was analyzed. The results indicate that some key residues differences are responsible for different enzymatic properties of different isoforms.

INTRODUCTION

CaM/Ca2+dependent protein Kinase II (CaMKII) is an autonomous activity protein. Activated CaMKII protein can stimulate many pathways by phsophorylation substrate such as CREB. CaMKII has been considered to relate to neuron memory (Lisman et al., 2002) immune memory (Bui et al., 2000; Ishiguro et al., 2006) and events of egg activation (Ducibella and Fissore, 2008). There are alpha, beta, gamma and delta isoforms of CaMKII in mammals. The protein structures of these isoforms are very similar: an N-terminal catalytic domain and C-teriminal associated domain which used to form oligomeric complex are connected by regulatory domain and variable domain (Fig. 1).

The catalytic domain regulatory domain and associated domain is conserved in four isoforms. The experiment prove that CaMK2G is more sensitive and CaMK2A is less sensitive to changes in CaM2+ concentration than the other isoforms.

Gaertner et al. (2004)’s study also suggest that the rank order of rates of autophosphorylation is CaMK2D>CaMK2B>CaMK2A>CaMK2G. What reason should be responsible for these differences of enzymatic properties of CaMKII? In this study, the protein sequence for four isoforms was compared and find several residues differences could be responsible for these enzymatic properties difference. The gene homology and gene evolution of CaMKII was analyze.

Fig. 1:	The diagram of CaMKII functional domain

CaMK2D gene should be the earliest diverging vertebrate gene was found. This result is agreement with previous report (Tombes et al., 2003).

MATERIALS AND METHODS

CaMK2B cds of human was used and mouse to blast Sus scrofa 9 genome (Altschul et al., 1990) then analysis blast result and construct possible gene fragment of Sus scrofa. The cds (protein coding sequence) and protein sequences of pig CaMK2B were predicted by Genewise (Birney et al., 2004). All process is worked by Perl program (details Perl program available upon request to researchers). Four isoforms of CaMKII protein between nine species (Table 1) were analyzed using clustraw2 (Thompson et al., 1994) to analyze homology to find the reason for enzymatic properties differences. In order to find the conservation properties of four isoforms of CaMKII, the genome sequences of four isoforms of CaMKII was compared between nine species by blast software.

Table 1:	CaMKII gene information
awhich is predicted by blast and genewise software.*which is the position of prediction start and end on genome

Then blast results were parsed (Parsed standard: alignment length >100 bp identity probability >80%). Sum the alignment length of pairwise homologous genes. Then the gene homology between species and assigned a conservation index to every gene was analyzed. For example CaMK2D: hsa gene is homologous to other eight species, so 9 was assigned to CaMK2D: hsa gene.

RESULTS AND DISCUSSION

CaMKII prediction: CaMK2B gene of Sus scrofa was predicted by Genewise software. It locates on plus strand of chromosome 18 of Sus scrofa 9. The length of sus CaMK2B gene is around 84 kb (Table 1). Protein sequences of CaMK2B have been compared and shown in Fig. 2.

The predicted CaMK2B_ssc_new protein sequence is similar to other species except a fragment in kinase domain (red region in Fig. 2). There should have two possible reasons for this result, the predicted CaMK2B is correct, the porcine coding sequence is different with other species. The genome of Sus scrofa is not complete, so Genewise predicated a wrong protein sequence.

CaMKII protein conservation: The regulatory domain is conserved between CaMK2D: hsa, CaMK2A: hsa, CaMK2B: hsa and CaMKII: dme (Fig. 3). There are three residue differences (red mark in Fig. 3) for CaMK2G: hsa. VS residues located at the pocket of calmodulin complex (Fig. 4). The property difference between VS and AT residues must led calmodulin to has different binding affinity for regulatory domain. Thus, it can deduce that CaMK2G’s special residues are responsible for its sensitivity to the calmodulin stimulation.

The protein sequence in the activation loop and Serine/Threonine protein kinases active-site (Fig. 5) was compared. Beside Serine/Threonine protein kinases active-site, there is a NGI peptide which interacts with activation loop (Fig. 5).

The NGI peptide of four isoforms was compared to cel and dme and then find that the rank order of similarity to CaMKII:cel is CaMK2D>CaMK2B> = CaMK2A >CaMK2G. This similarity order is the same as phosphorylation rate order which reported by Gaertner et al. (2004).

Activation loop control the phosphorylation rate of CaMKII (Fig. 6, Adams, 2003). So it can deduce these residue differences lead to that CaMK2D has a higher phosphorylation rate.

Fig. 2:	CaMK2B homology analysis. CaMK2B_ssc_new is predicted by Genewise. Red region means different residues

Fig. 3:	Homology analysis of CaM/Cz2+binding domain sequence

Fig. 4:	3D structure of CaMK2D binding with calmodulin/Ca2+

CaMKII gene evolution: Human CaMK2D:hsa has many homologous segment to other three genes (Fig. 7). Other species such as ssc (Sus scrofa), oan (Ornithorhynchus anatinus) and dre (Danio rerio) have similar results. The CaMK2D gene has three to four times more intronic sequences than other three genes. So we can deduce that CaMK2D and other three genes have the same ancient gene origin.

Fig. 5:	Homology analysis of kinase domain of CaMKII four gene production

Fig. 6:	The structure of peptide NGI and active center of CaMK2D

If CaMKII gene in cel (C. elegant) and dme (D. melanogaster) could be considered as the earliest gene, the gene which is most similar to cel or dme CaMKII gene could be considered as earliest gene. CaMK2D: hsa is similar to CaMKII: cel and CaMKII: dme. CaMK2D: mmu, CaMK2D: ssc and CaMK2D: gga is similar to CaMKII: dme. CaMK2D is the only gene which is similar to CaMKII: cel or CaMKII: dme (Fig. 8). CaMK2A, CaMK2B and CaMK2G gene can only find homologous genes in vertebrate species. CaMK2D has so many homology segment to CaMKII: cel and CaMKII: dme, this fact indicate that CaMK2D, CaMKII: cel and CaMKII: dme have the same ancient gene origin. CaMK2A, CaMK2B and CaMK2G gene emerge later than CaMK2D. This result is agreement with previous report (Tombes et al., 2003).

CaMK2B: sus has been released in genebank (gi: 194042930) but homology analysis indicate the released version is wrong (details available upon request to authors), so we predicted it by blast human and mouse CaMK2B CDs to Sus scrofa 9 genome. It have successfully predicted the CaMK2A protein of Sus scrofa. So we think the predicted CaMK2B: sus is correct on chromosome 18 of Sus scrofa but due to uncompleted genome, Genewise could not predicate protein sequence correctly (Fig. 2).

Fig. 7:	Pairwise alignment of human four CaMKII gene and nine species

Fig. 8:	Homology analysis of CaMKII in nine species

Then we analyzed the relationship between protein sequence differences and enzymatic properties. It is known that activation loop of CaMKII control the phosphorylation rate (Adams, 2003). CaMK2D’s activation loop and key residues which interact with activation loop is most similar to CaMKII: cel (Fig. 5 and 6). C. elegans CaMKII has maximum phosphorylation rate (Griffith et al., 2003; Chao et al., 2010). So it can deduce that CaMK2D should has the maximum phosphrylation rate. CaMK2D have been reports to have maximum phosphorylation rate. The result is in agreement with previous reports (Gaertner et al., 2004). KKFN residues in regulatory domain of CaMKII were found to be sufficient to maintain an inhibited state in a truncate form of the kinase (Cruzalegui et al., 1992). CaMK2G mutant from KKFN-RKFN (Fig. 3), this mutation must influence the capability of KKFN segment to inhibit kinase activity. So RKFN lead CaMK2G was deduced to be more sensitive to changes of Ca2+/CaM concentration. This result is in agreement with previous reports (Gaertner et al., 2004). The homology analyses of four isoforms indicate that CaMK2D is the earliest emergence gene (Fig. 8) and CaMK2D and other three genes have the same ancient gene origin (Fig. 7). This result is agreement with previous report (Tombes et al., 2003).

CONCLUSION

By homology analysis of nine species, it prove that CaMK2D is the earliest emergence gene than other three genes in the CaMKII family.

ACKNOWLEDGEMENTS

This research was financially supported by the National High Science and Technology Foundation of China 863 (2006AA10Z195 and 2008AA101008), the NSFC-Guangdong Key Project (U0631005), the Key Project of National Basic Research and Developmental Plan (2006CB102105) of China, the creative team project of education ministry.