Ng subunits is related, then this gives 0.8 > r3/r1 > 0.4. The ISHA is

Ng subunits is related, then this gives 0.8 > r3/r1 > 0.4. The ISHA is an excellent approximation to the actual holoenzyme at the upper end of this range, and also in the lower end the approximation continues to be adequate. CaM-bound CaMKII has the exact same activity toward exogenous substrates regardless of the phosphorylation state of T286 [25]. The simplest assumption is that the activity towards neighboring CaMKII subunits can also be independent of CaMKII phosphorylation, in which case r2 = r1 plus the ISHA is in fantastic agreement with exact models. Actually, all of the readily available experimental data could be match with this assumption. Nevertheless, fits to different experiments give autophosphorylation prices that differ by greater than an order of magnitude [45]. In Ref. [16] it was proposed the discrepancies inside the measurements may be resolved if r2 < r1. By fitting autophosphorylation time courses with a hybrid deterministic-stochastic model, they found a best fit required r2/r1 0.08 [16]. However, the noisy data could also bePhys Biol. Author manuscript; available in PMC 2013 June 08.Michalski and LoewPagereasonably well fit with r2 = r1 [16], and thus it is difficult to draw any firm conclusions from this study. De Koninck and Schulman [25] showed that Bisindolylmaleimide I site autonomous activity after a 6 second autophosphorylation reaction is about 80 of maximal CaM-stimulated activity. There are two ways to interpret this result: either the phosphorylated CaMKII has the same activity as CaM-bound CaMKII, but only 80 of the subunits are phosphorylated in a 6 second reaction, or all of the subunits are phosphorylated but an autonomous subunit only has 80 of the activity of a CaM-bound subunit. More recent data would favor the latter explanation [44]. Either way, the data clearly show that r2 0. In fact, the data require r2 > 0.1 s-1, but usually do not put an upper bound on r1, and therefore usually are not useful for determining r2/r1. Bradshaw, et al. [45] offers time courses of phosphate incorporation and autonomous activity, as well as the experimental circumstances are such that a two state model of CaMKII is proper. In Fig. S4 we show that the autonomous activity information (from Fig. 2(a) in Ref. [45]) is very best fit by r2 = r1, and frequently needs r2/r1 > 0.six for any very good match. In Fig. S4 we also show that the phosphate incorporation information (from Fig. 2(b) in Ref. [45]) is greatest match with r2/r1 = 0.66, though this data is noisier and admits acceptable fits even with r2/r1 0.03. Nonetheless, the preponderance of data from Ref. [45] suggest that r2/r1 > 0.five. Therefore, it can be reasonable to assume that r2/r1 0.5 and r3/r1 0.five, in which case we can count on the ISHA to become valid. It is actually worth noting that these autophosphorylation prices are dependent on both ATP concentration and temperature [45], and hence their ratios may not be constant. If it turns out that our assumption is incorrect and r2 r1, then not only PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/21114274 does the ISHA fail but additionally a dimer model poorly describes CaMKII dynamics (see Fig. four). In this case the best approximation would be to take into account a modified ISHA exactly where the subunits are grouped into dimers on the infinite lattice and the dynamic variables will be the joint probability distributions of these dimers. In this scheme half in the subunits have exact facts in regards to the state of their neighbor and half of the subunits have only probabilistic information and facts about their neighbor, as in the original ISHA. This model needs as a lot of species because the dimer model, but provides phosphorylation levels which might be inside 7.