Synthesis of Highly Structured and Receptor-Selective Tetradecapeptidic Analogs of Somatostatin: Fine-tuning the non-covalent interactions among their aromatic residues

Abstract

Somatostatin, a peptidic hormone also known as somatotropin release-inhibiting factor (SRIF), is implicated in multiple biological functions, mediated by direct interaction of the peptide with a family of structurally related, G-protein coupled transmembrane receptors (GPCRs), named SSTR1-5. However, the full therapeutic potential of SRIF cannot be further exploited owing to its short half-life of <3 min in plasma and its lack of selectivity against all receptors. With the aim to overcome these limitations, many SRIF analogs containing key modifications have been synthesized over recent decades. Octreotide, lanreotide, vapreotide and pasireotide are the four synthetic analogs that have reached the market so far. They are shorter analogs they contain a more rigid ring than somatostatin. Consequently, they are long-acting analogs with increased receptor selectivity. Also, they share a similar fragment including residues Phe/Thr7-D-Trp8-Lys9-Thr/Val10, which populates betaturn conformations and has been proposed to be the pharmacophore of the hormone.

Early studies uncovered the possibility that an aromatic interaction between Phe6, Phe7 or Phe11 plays a key role in the structural stabilization of the hormone. We considered that exploring the presumed aromatic interactions that stabilize some bioactive conformations in SRIF would be a unique way to obtain analogs with enhanced stability and conformational rigidity, and improved biological activity. For this purpose, we have synthesized and studied 24 SRIF-analogs containing aromatic non-natural amino acids as site-directed modifications. The gram-scale syntheses of these amino acids (3-mesitylalanine (Msa), 3-(3’,4’,5’-trimethylphenyl)-alanine (Tmp) and 3-(3’,5’-difluorophenyl)-alanine (Dfp)) were achieved by different routes, using a Rhodium-Catalyzed Asymmetric Hydrogenation reaction as source of enantioselectivity. These non-natural residues displayed specific hydrophobic, electronic and steric properties. The introduction of Msa, Tmp and Dfp residues in the tetradecapeptidic SRIF scaffold allowed us to prove and modulate the noncovalent interactions between the aromatic residues in the natural hormone somatostatin.

Several of these analogs showed a high conformational rigidity relative to that of SRIF, which allowed us to characterize for the first time the 3D structures of their major conformations in solution using NMR and computational techniques. These investigations allowed us to prove that bulky, hydrophobic and electron-rich residues (i.e. Msa or Tmp) in position 7 shifts the conformational equilibrium in solution toward a major conformation that should be very similar to that of SRIF when it binds to the SST2 receptor. We considered that an aromatic interaction between the Ar6-Ar11 residues is the factor that contributes the most to stabilize this particular conformation. In this regard, we have used the L-Msa7_D-Trp8 fragment to fine-tune the Ar6-Ar11 interaction, thus developing several highly SSTR2-active and –selective 14-residue SRIF analogs. The introduction of Dfp residues allowed us to prove that this electron-poor aromatic amino acid in position 11 engage an aromatic interaction with the cross-strand Phe6. Our studies involving SRIF analogs containing the Cha amino acid indicated that the specific interaction between aromatic residues that contributes to the conformational stabilization in SRIF is unique to aromatic moieties, and is not exclusively driven by a hydrophobic collapse.

Our studies, initially devoted to prove the postulated Phe6-Phe11 interaction in somatostatin, have led us to gain invaluable and unprecedented insight into structure-activity relationships of somatostatin and SRIF analogs. Looking ahead, despite we have been able to exploit specific aromatic interactions to obtain directed SSTR-selectivity, there is still much that we do not yet understand regarding the nature of these non-covalent interactions. Our combination of NMR and computational techniques in the study of 14-residue SRIF analogs has provided a powerful tool to fill this gap. Taking into account the wide distribution of SSTR2 and its role in SRIF-regulated processes, the therapeutic potential of SRIF analogs that bind exclusively to this receptor continue to be of high importance. Our highly potent and SSTR2-selective peptides, which presumably show an appropriate persistence (half-life) in blood, would certainly present a potential alternative to the currently marketed drugs. The success in future design of 14-residue SRIF analogs will rely on the right modifications of its sequence based on the structural information obtained from our studies.

La hormona peptídica somatostatina (SRIF) está implicada en un gran número de funciones biológicas, mediadas por la interacción directa del péptido con una familia de cinco GPCRs, comúnmente denominados SSTR1-5. Sin embargo, no ha sido posible maximizar el potencial terapéutico de la somatostatina debido a su bajo tiempo de vida media en plasma y a la falta de selectividad por sus receptores.

Con el objetivo de superar esas limitaciones se han sintetizado un gran número de análogos de somatostatina. La mayoría de estos análogos son octapéptidos, y por lo tanto son más rígidos que la somatostatina, más estables y más selectivos frente a los cinco receptores distintos. Los recientes avances en la síntesis de péptidos en fase sólida nos llevaron a reconsiderar la estructura completa de la somatostatina (es decir, los catorce residuos) como molde para sintetizar nuevos análogos que tuvieran mayor estabilidad y selectividad. Además, trabajando con tetradecapéptidos estaríamos estructuralmente mucho más cerca de la hormona natural de lo que lo están los octapéptidos. Nosotros consideramos que el estudio de las presuntas interacciones aromáticas que estabilizan ciertas conformaciones bioactivas en SRIF podría funcionar como un método único para obtener análogos con mayor estabilidad y rigidez conformacional, así como una actividad biológica mejorada.

Para ello, estudiamos la sustitución de las Phe6, Phe7 y Phe11 por tres aminoácidos aromáticos no naturales: 3-mesitilalanina (Msa), 3-(3’,4’,5’-trimetilfenil)-alanina (Tmp) y 3-(3’,5’-difluorofenil)-alanina (Dfp). Seleccionamos estos tres residuos por su alta hidrofobicidad y sus peculiares propiedades electrónicas y estéricas.

En esta tesis doctoral probamos que la introducción de estos residuos aromáticos en la secuencia de somatosatina incrementa ciertas interacciones aromáticas entre Ar6-Ar11, lo que aumenta la rigidez conformacional de estos péptidos. Debido a ello, las estructuras de las conformaciones mayoritarias de estos tetradecapéptidos has sido por primera vez caracterizadas por técnicas de RMN y computacionales. La mayoría de los análogos presentaban perfiles de selectividad y afinidad muy interesantes. Identificamos el fragmento L-Msa7-D-Trp8 como clave para incrementar la interacción Ar6-Ar11 y obtener alta afinidad por el receptor SSTR2, y lo hemos usado junto con otros aminoácidos no naturales, para obtener los tetradecapéptidos conformacionalmente más rígidos descritos hasta la fecha, con una actividad y selectividad excepcionales frente a SSTR2.