Substrate prediction

Figure 10. Field distributions for two WG modes in a microdisk resonator on a substrate predicted by the Spectral Index method (Greedy, 2000).

P-glycoprotein (P-gp) substrate predictions are evaluated using Almond 3.2, a pharmacophore pattern analysis software that combines MIF distances with energies [27], using a model initially derived from P-gp ATP-ase activity [28]. The model contains about 100 drugs (60 known substrates, and 40 nonbinders), each evaluated over 100 diverse conformations (G. Cruciani, personal communication). 

Activity profile. A histidine residue in the active site of aspartate transcarhamoylase is thought to be important in stabilizing the transition state of the bound substrates. Predict the pH dependence of the catalytic rate, assuming that this interaction is essential and dominates the pH-activity profile of the enzyme. 

Prior to MS-based substrate specificity assays, certain NRPS substrate specificities can be predicted by bioinformatics. Adenylation domain substrates can be predicted based on their 10 letter code 99,100 by substrate prediction tools such as the NRPS predictor.101 Methyltransferases can be predicted in their substrates and methylation sites by bioinformatic analysis too.102 In addition, substrates of catalytic NRPS domains and tailoring enzymes can be predicted by the structure of the known NRP natural product. Either way, predicted substrates of NRPS domains need to be experimentally verified. A traditional technique to determine substrate specificity of an A domain is the adeonsine triphosphate-pyrophosphate (ATP-PP ) exchange assay. The ATP-PP exchange assay characterizes substrates indirectly by observing the radioactive pyrophosphate incorporation into ATP from a reverse reaction with pyrophosphate and the acyl-adenylate of the substrate.103 Because the PP exchange measures the back exchange of pyrophosphate into ATP, the determined substrate can deviate from the true substrate as it may be only the kinetically most competent substrate of the reverse adenylation reaction. In contrast to this assay, MS has become a more reliable tool to identify NRPS substrates because it determines the true substrate specificity by detection of the complete adenylation reaction product, that is, the substrate tethered on a T domain. 

Verras, A., Kuntz, I.D., Ortiz de MonteUano, P.R. Cytochrome P450 enzymes Computational approaches to substrate prediction. Ann. Reports Comp. Chem. 2005,1,171-95. 

Figure 113.2 Plots of values of the total masses of cells and substrate predicted using the -rate parameters determined in Illustration 13.1, together with the corresponding data points. The smooth curves are the values predicted using the model the triangles and diamonds correspond to the synthetic data points for biomass and substrate, respectively.

N-acetyl cysteamine (SNAC) thioester substrate mimics were used to test the specificity of four KS domains with a range of predicted biosynthetic intermediates (Table 3.1). In the case of PsyD KS3° and BaeL KS5, the synthesised SNAC mimic are shorter in length than the predicted biosynthetic intermediate. However, the phylogenetic KS-based substrate prediction would suggest that the specificity does not extend past the P-position [1], and therefore SNAC substrate mimics synthesised to this length may be sufficient to reveal KS specificity. 

---