Synthesis of 2-Vinyl sphingolipids as S1PL inhibitors

Abstract

Sphingolipids (SLs) are components of eukaryotic cell membranes that in the last 30 years have shown not only a structural role but also be involved in signaling processes and cellular responses. Therefore, the development of compounds capable of specifically targeting the enzymes of SL metabolism represents an attractive area of research for the treatment of multiple disorders.

The metabolic pathway of SLs has a single exit point, mediated by sphingosine 1-phosphate lyase (S1PL), a pyridoxal 5'-phosphate-dependent enzyme (PLP), which irreversibly degrades sphingosine 1-phosphate (S1P) to 2-hexadecenal and phosphoethanolamine (PE). Together with sphingosine kinase (SK) and S1P phosphatases (SPPases) is responsible for the called "sphingolipid rheostat", a system that regulates cell fate modifying the ratio of S1P, which is considered a proliferative molecule, and apoptogenic sphingosine (So) and ceramide (Cer).

In 1994, Boumendjel and Miller described the inhibitor 2VS1P as a racemic mixture, with a promising IC50 of 2.4 μM in rat liver microsomal preparations as a source of enzyme. One objective of this thesis was the synthesis of the 4 stereoisomers of a truncated chain of 2VS1P analog, phosphate 24. To this end, a synthetic route was devised via the already described prochiral intermediate 15 and diastereomers 23 could be separated by flash chromatography. Two racemic mixtures of 24 were obtained.

The enantiomeric resolution was firstly carried out by derivatizing 23 with the chiral derivatizing agent, 2-methoxyphenylacetic acid (MPA), giving 32. Although the enantiomers were resolved, the conditions for hydrolyzing selectively MPA from phosphate group were not found. The same derivatization was attempted on 21, resulting in 25. Separation of the 4 enantiomers was achieved by TLC but not by flash chromatography, where only pure column heads and tails were obtained. Even so, the access to the 4 enantiomers is possible using both (R or S)-MPA separately.

Finally, a third synthetic route via vinyl alcohol 44 was accomplished, where the syn/anti mixtures of 45 were easily separated by flash chromatography. The resolution of enantiomers and the determination of the absolute configuration of C3(OH) was carried out by using the Riguera’s et al. methodology on 46 and for C2 by NOEs of structures 51.

The synthesis until 24 was performed through 22.

Unfortunately, an excess of DMAP was necessary to derivatize 45, which ended up with a partial racemization as shown by HPLC and VCD: enantiomerically pure (-)-syn-45, enantioenriched (+)-syn-45 (94% ee) and (±)-anti-45 as a racemic mixture.

The products obtained (So, S1P and Cers analogs) were obtained and were tested on inhibition studies of S1PL, cell viability and sphingolipidimics. (2S,3S)-24 (IC50 = 33.4 μM) was demonstrated to be the eutomer in hS1PL and ESI-MS and dilution assays showed the typical

behaviour of irreversible inhibitor. Instead, JG and 31 showed a moderate inhibitory activity, suggesting that phosphate is not necessary.

Lipidomic experiments suggest that (2S,3R)-JG and (±)-anti-31 are substrates of Ceramide 2 (CerS2), but not of the CerS5 or CerS6. In addition, vinylCer 48 and 52 are hydrolyzed by ceramidases (CDases) and only the natural configuration of the free aminodiols is acylated again.

Regarding the effects on natural sphingolipidoma, all the diastereomers of vinylCer 48 and 52 appear to inhibit de novo synthesis, being serine palmitoyltransferase (SPT) a potential target.

Computational studies suggest that the nucleophile that could be responsible for the irreversible behavior of these molecules could be lysine L353 for intermediates coming from the aldimines C2S and lysines L353 and L359 for the C2(R).

El desarrollo de compuestos capaces de dirigirse específicamente a las enzimas del metabolismo de los esfingolípidos representa un área de investigación atractiva para el tratamiento de varios trastornos. En este contexto, la inhibición de la esfingosina 1-fosfato liasa (S1PL) se ha relacionado con trastornos autoinmunes, enfermedades inflamatorias o con la desregulación celular en varios tipos de cáncer, por lo que se considera una diana prometedora para el desarrollo de nuevos fármacos.

Basándonos en los datos descritos en la literatura para el inhibidor 2VS1P1, descrito como mezcla racémica, se sintetizaron los diferentes estereoisómeros de su análogo 24, así como los análogos vinílicos de la esfingosina (So) (31 y JG) y los de la ceramida (Cer) (48 y 52).

Al tener dos centros quirales, uno de ellos cuaternario, la configuración absoluta de los anteriores compuestos se determinó por técnicas de RMN mediante el empleo de (R) o de (S)-MPA, un agente de derivatización entioméricamente puro,2 junto con experimentos de NOE. Desafortunadamente, la esterificación con MPA dio lugar a una racemización parcial de los compuestos 24, 31 y 48, confirmada por HPLC quiral y estudios de VCD, por lo que, de los cuatro estereoisómeros posibles en cada caso, solamente se pudo obtener el (2S,3S) en forma enantioméricamente pura, el (2R,3R) enantioenriquecido y una mezcla racémica de los de configuración anti (2RS,3SR).

Los ensayos de inhibición in vitro de 24, 31 y JG sobre la S1PL humana indicaron una actividad moderada como inhibidores y una toxicidad relativamente alta en células A549.

Los análisis de esfingolipidómica sugieren que los aminodioles (2S,3R)-JG y (±)-anti-31 son sustratos de la ceramida sintasa 2 (CerS2) y que las ceramidas 48 and 52 inhiben la síntesis de novo de Cer, lo que permitiría postular una posible inhibición de la serina palmitoiltransferasa (SPT) por parte de estos compuestos, extremo que no se ha estudiado en la presente tesis doctoral.