A procedure combining molecular docking and semiempirical method PM7 for identification of selective Shp2 inhibitors

Abstract

Shp2 and Shp1 make up a small family of protein tyrosine phosphatases. Finding selective inhibitors for Shp2 is useful because although its inhibition is advantageous for the treatment of some types of cancer, inhibition of Shp1 may have the opposite effect, since it acts as a suppressor of tumors. We combined molecular docking and semiempirical molecular orbital-based calculations to produce data that were effective for the identification of selective inhibitors for Shp2. After definition of the interaction modes of the inhibitors with Shp2 and Shp1 by molecular docking, the resulting interaction enthalpy values were calculated following refinement of the enzyme/inhibitor complexes' geometries with the PM7 semiempirical molecular orbital method. Despite the complexity of the thermodynamics involved in the enzyme/inhibitor interaction, the selectivity for Shp2 of a series of 76 inhibitors, divided in two groups, could be effectively correlated with the difference in their interaction enthalpy values with both enzymes. For the first group, composed by 52 Shp2-selective indoline inhibitors for which only Shp2 inhibition activity data are available, we demonstrated that the interaction enthalpy can be used as a criterion for identification of selective Shp2 inhibitors, as it is significantly more favorable for Shp2 than Shp1 at a 99% confidence level. For the second group, composed by 24 oxindole derivatives with available Shp2 and Shp1 inhibition activity data, a satisfactory correlation (R = 0.70) could be obtained between the selectivity, based on the IC 50 data, and the relative percentage difference of the calculated interaction enthalpies with the two enzymes. K E Y W O R D S molecular docking, PM7 method, selectivity, Shp2 inhibitors

Emerging chemical scaffolds with potential SHP2 phosphatase inhibitory capabilities – A comprehensive review

“…In this F I G U R E 3 6 SHP2 inhibitors (109-115) discovered by computer-aided drug design tools F I G U R E 3 7 SHP2 inhibitor (116) identified using fragmentbased ligand design, de novo design, ADMET profiling and molecular docking studies respect, Rocha and team (2019) used a combination of several computational methods, including molecular docking and PM7 semi empirical molecular orbital based calculations to recognize some novel and selective inhibitors for SHP2 from two series of 76 SHP2 inhibitors, viz. 52 SHP2-selective indoline inhibitors and 24 oxindole derivatives with SHP2 and SHP1 inhibitory activity (Rocha & Sant'Anna, 2019). The results confirmed that enthalpy contributes as an important factor in determining and governing the selectivity since the greater favorable binding enthalpy is probably because of the specific intermolecular interactions.…”

Target–ligand binding affinity from single point enthalpy calculation and elemental composition

“…43–46 Calculated Δ H b is also promising in prediction of binding selectivity. It has been shown 47 that relative differences in Δ H b values of a set of ligands calculated by a PM7/MOZYME/COSMO combination correlated well ( R = 0.7) with relative IC 50 values measured on Src homology region 2 domain-containing phosphatase targets. This combination of the semi-empirical QM approaches with COSMO implicit solvent model for the calculation of Δ H b has been a choice in various studies with PM3, 28,30 PM6 48 and PM7 33,49 parametrizations among which PM7 may be considered as the most robust one.…”

Reposicionamento de fármacos para o tratamento da doença de Chagas: desenvolvimento de modelos computacionais baseados em ligantes ativos fenotipicamente contra o >i<Trypanosoma cruzi>/i< e na inibição da diidroorotato desidrog

“…4.3.5 Desenvolvimento de um modelo de reclassificação por energia livre de Gibbs para a identificação de inibidores de TcDHODH Os descritores utilizados para o desenvolvimento do modelo de energia livre de Gibbs para a reclassificação dos resultados da ancoragem molecular foram a entalpia de interação do ligante com a proteína (ΔHint, calculada pela equação 17), assim como as perdas entrópicas devido a restrições conformacionais do ligante durante a complexação com o alvo (Etor) e a lipofilicidade (clogP) do ligante. Os descritores foram selecionados com base em trabalhos que descreveram modelos preditivos para diversos alvos terapêuticos, como proteínas kinase (PI4KIIIβ e PK-C), fosfodiesterase (PDE4) e tirosina fosfatases (Shp2), empregando abordagens similares (COLODETTE et al, 2020;OLIVEIRA et al, 2006;ROCHA;SANT'ANNA, 2019;WANG et al, 2004). Como o método de cálculo semi-empírico do ΔHint necessita das coordenadas do complexo P:L, utilizou-se a função de pontuação ASP para ancorar na TcDHODH os 28 ligantes que não possuíam co-cristais experimentais disponíveis na literatura.…”

“…Por esta razão, outros descritores específicos foram incluídos para o desenvolvimento do modelo de reclassificação neste trabalho: a lipofilicidade do ligante, estimada pelo valor de clogP, e efeitos entrópicos associados à perda de movimentos torcionais do ligante após a complexação, estimados com o parâmetro Etor da função de pontuação ChemPLP. Esta estratégia foi baseada em trabalhos que haviam reportado modelos de alta capacidade preditiva, utilizando abordagens similares(COLODETTE et al, 2020;OLIVEIRA et al, 2006;ROCHA;SANT'ANNA, 2019). O presente trabalho apresenta como diferencial o fato de ter empregado um algoritmo de aprendizado de máquinas, capaz de identificar correlações não lineares entre os dados.…”