Protease inhibitors as therapeutic agents

Abstract

Proteases are involved in a high number of diseases, and thus, are relevant targets. For that reason our main goal was the discovery of protease inhibitors as therapeutic agents. We focused our study in four proteases: dipeptidyl peptidase IV (diabetes mellitus type 2), prolyl oligopeptidase (cognitive disorders) and cathepsins L and B (cancer).For the discovery of inhibitors, three strategies were selected: medicinal plant screening, high throughput screening and the characterization of a combinatorial chemistry library.

Once accomplished the DPP IV recombinant expression optimization, the protein was studied by means of nuclear magnetic resonance (NMR) in order to obtain information of its dynamism. Since DPP IV is a large protein, a strategy combining selective labeling and the use of TROSY-HSQC experiments was used. From the 14 methionine residues of the protease, 11 of them were detected in the NMR spectra. Then, a study of the inhibitor effect on the NMR spectra of DPP IV was done. Interestingly, the corresponding spectra of DPP IV / inactivator afforded an extra signal. We believe that it is a consequence of a small structural change that the protease suffers after inhibition.

Afterwards, we planned to find DPP IV from botanical sources. First, we selected plants that were already reported to have antidiabetic action. Common antidiabetic plants were chosen, as well as Brazilian plants and others from the Traditional Chinese Medicine. Besides, a library of Mediterranean plants was also selected. After, extraction and testing of DPP IV inhibitory activity was done. From our tailored collection, the plant AP-3 was selected for further analysis. After fractionation and purification, two molecules were found to be DPP IV inhibitors. Kinetic experiments of the best one, AP-3-a, demonstrated that it was inhibiting DPP IV in a parabolic manner. Then, AP-3-a inhibition of DPP IV was analysed by NMR. The extra signal that was observed with competitive inhibitors was not present. We hypothesized that the lack of appearance of this signal is a result of the parabolic inhibition of AP-3-a.

Then, we planned the identification of POP inhibitors by HTS. Our strategy was based in the use of libraries containing non-toxic compounds and lead-like properties. The assay we selected was FP, which allowed the identification of protein binders by competition with a fluorophore-labelled probe. First, the peptide probes were validated as useful probes for the FP assay. Then, the HTS was carried out. Over the 4,500 tested compounds, 73 hits were found to be POP binders. Later, 37 hits were selected by means of clustering, docking data and FP results in order to be validated as POP inhibitors. The subset of molecules was evaluated by enzymatic assays. Six compounds presenting the highest POP inhibition ration were selected for further study. Finally, two POP inhibitors have been described. HTS-43 is a competitive POP inhibitor and HTS-75 displays a parabolic behaviour. It was the first time that parabolic inhibition is reported for POP. We believe that the existence of a non-competitive site would help in the understanding of the relationship between POP and mental diseases.

Finally, a novel peptidyl aryl vinyl sulfone library was tested for its inhibitory activity against cathepsins L and B. Among all the 20 molecules of the library, a potent covalent irreversible cathepsin L inhibitor has been found, PAVS-20. The progress-curve of the pre-incubation time representation allowed the calculation of its inhibition constants. Furthermore, evaluation of subsite preferences was done by docking analysis. This allowed understanding the experimental differences in inhibition constants obtained for similar compounds. Since cathepsins L and B are targets for cancer, molecules of the PAVS library are promising candidates for the development of new anticancer drugs.

Las proteasas están involucradas en un alto número de enfermedades y por lo tanto, son dianas terapéuticas relevantes. Por este motivo, nuestra principal meta era el descubrimiento de inhibidores de proteasas cómo agentes terapéuticos. Para ello nuestro estudio en cuatro proteasas: la dipeptidil peptidasa IV, la prolil oligopeptidasa y las catepsinas L y B. Para la búsqueda de inhibidores, se seleccionaron tres estrategias: cribado de plantas medicinales, cribado de alto rendimiento y caracterización de una biblioteca proveniente de la química combinatoria.

Una vez se llevó a cabo la expresión recombinante de la DPPIV, la proteína se estudió mediante resonancia magnética nuclear (NMR) para obtener información acerca de su dinamismo. Dado que es una proteína grande, se utilizó una estrategia en la que se combinó el marcaje selectivo y el uso de experimentos TROSY-HSQC. Posteriormente, se realizó el estudio de la DPP IV en presencia de sus inhibidores, para observar como estos afectan a la estructura proteica.

Después, se realizó la búsqueda de inhibidores de la DPP IV a partir de extractos de plantas medicinales. De nuestra colección, se seleccionó el extracto de la planta AP-3 para un análisis en profundidad. Se detectaron dos inhibidores de la proteasa. El más potente, AP-3-a, se caracterizó cómo un inhibidor parabólico. Después se llevó a cabo el estudio del complejo DPP IV/AP-3-a por NMR.

En cuanto a la POP, se realizó la búsqueda de inhibidores mediante cribado de alto rendimiento (HTS). De los 4,500 compuestos testados se obtuvo un total de 73 hits en el ensayo de polarización de la fluorescencia (FP). La validación de estas moléculas mediante docking, clustering y ensayos enzimáticos, permitió identificar seis potentes inhibidores de POP. Uno de ellos, HTS-75, se caracterizó como un inhibidor parabólico. Esta es la primera vez que se describe un inactivador de este tipo para POP.

Finalmente, en cuanto a las catepsinas L y B, se cribó una librería de peptidil aril vinil sulfonas. Entre las 20 moléculas testadas se encontró un potente inhibidor irreversible de la catepsina L, el PAVS-20. Además, se realizó un estudio de docking que permitió evaluar las preferencias de los subsitios de las dos proteasas.