Organocatalytic asymmetric mannich-type reactions: an easy approach to optically active amine derivatives

The topics I came across during the period I spent as a Ph.D. student are mainly two. The first concerns new organocatalytic protocols for Mannich-type reactions mediated by Cinchona alkaloids derivatives (Scheme I, left); the second topic, instead, regards the study of a new approach towards the enantioselective total synthesis of Aspirochlorine, a potent gliotoxin that recent studies indicate as a highly selective and active agent against fungi (Scheme I, right). At the beginning of 2005 I had the chance to join the group of Prof. Alfredo Ricci at the Department of Organic Chemistry of the University of Bologna, starting my PhD studies. During the first period I started to study a new homogeneous organocatalytic aza-Henry reaction by means of Cinchona alkaloid derivatives as chiral base catalysts with good results. Soon after we introduced a new protocol which allowed the in situ synthesis of N-carbamoyl imines, scarcely stable, moisture sensitive compounds. For this purpose we used α-amido sulfones, bench stable white crystalline solids, as imine precursors (Scheme II). In particular we were able to obtain the aza-Henry adducts, by using chiral phase transfer catalysis, with a broad range of substituents as R-group and excellent results, unprecedented for Mannich-type transformations (Scheme II). With the optimised protocol in hand we have extended the methodology to the other Mannich-type reactions. We applied the new method to the Mannich, Strecker and Pudovik (hydrophosphonylation of imines) reactions with very good results in terms of enantioselections and yields, broadening the usefulness of this novel protocol. The Mannich reaction was certainly the most extensively studied work in this thesis (Scheme III). Initially we developed the reaction with α-amido sulfones as imine precursors and non-commercially available malonates with excellent results in terms of yields and enantioselections.3 In this particular case we recorded 1 mol% of catalyst loading, very low for organocatalytic processes. Then we thought to develop a new Mannich reaction by using simpler malonates, such as dimethyl malonate.4 With new optimised condition the reaction provided slightly lower enantioselections than the previous protocol, but the Mannich adducts were very versatile for the obtainment of β3-amino acids. Furthermore we performed the first addition of cyclic β-ketoester to α-amido sulfones obtaining the corresponding products in good yield with high level of diastereomeric and enantiomeric excess (Scheme III). Further studies were done about the Strecker reaction mediated by Cinchona alkaloid phase-transfer quaternary ammonium salt derivatives, using acetone cyanohydrin, a relatively harmless cyanide source (Scheme IV). The reaction proceeded very well providing the corresponding α-amino nitriles in good yields and enantiomeric excesses. Finally, we developed two new complementary methodologies for the hydrophosphonylation of imines (Scheme V). As a result of the low stability of the products derived from aromatic imines, we performed the reactions in mild homogeneous basic condition by using quinine as a chiral base catalyst giving the α-aryl-α-amido phosphonic acid esters as products (Scheme V, top).6 On the other hand, we performed the addition of dialkyl phosphite to aliphatic imines by using chiral Cinchona alkaloid phase transfer quaternary ammonium salt derivatives using our methodology based on α-amido sulfones (Scheme V, bottom). The results were good for both procedures covering a broad range of α-amino phosphonic acid ester. During the second year Ph.D. studies, I spent six months in the group of Prof. Steven V. Ley, at the Department of Chemistry of the University of Cambridge, in United Kingdom. During this fruitful period I have been involved in a project concerning the enantioselective synthesis of Aspirochlorine. We provided a new route for the synthesis of a key intermediate, reducing the number of steps and increasing the overall yield. Then we introduced a new enantioselective spirocyclisation for the synthesis of a chiral building block for the completion of the synthesis (Scheme VI).

Asymmetric synthesis of benzylic and heterobenzylic amines

C2-Symmetrical, enantiopure 2,6-di[1-(1-aziridinyl)alkyl]pyridines (DIAZAPs) were prepared by a high-yielding, three-step sequence starting from 2,6-pyridinedicarbaldehyde and (S)-valinol or (S)-phenylglycinol. The new compounds were tested as ligands in palladium-catalyzed allylation of carbanions in different solvents. Almost quantitative yield and up to 99% enantiomeric excess were obtained in the reactions of the enolates derived from malonate, phenyl- and benzylmalonate dimethyl esters with 1,3-diphenyl-2-propenyl ethyl carbonate. Asymmetric synthesis of 2-(2-pyridyl)aziridines from chiral 2-pyridineimines bearing a stereogenic center at the nitrogen atom was development. The envisioned route involves the addition of chloromethyllithium to the imine derived from 2-pyridinealdehyde and (S)-valinol, protected as O-trimethylsilyl ether. The analogous reaction performed on the imine derived from (S)-valine methyl ester gave the product containing the aziridine ring as well as the α-chloro ketone group coming from the attack of chloromethyllithium to the ester function. Other stereogenic alkyl substituents at nitrogen gave less satisfactory results. Moreover, the aziridination protocol did not work on other aromatic imines, e.g. 3-pyridineimine and benzaldimine, which are not capable of bidentate chelation. The N-substituent could not be removed, but aziridine underwent ring-opening by attack of nitrogen, sulfur, and oxygen nucleophiles. Complete or prevalent regioselectivity was obtained using cerium trichloride heptahydrate as a catalyst. In some cases, the N-substituent could be removed by an oxidative protocol. The addition of organometallic (lithium, magnesium, zinc) reagents to 2-pyrroleimines derived from (S)-valinol and (S)-phenylglycinol gave the N-substituted-1-(2-pyrrolyl)alkylamines with high yields and diastereoselectivities. The (S,S)-diastereomers were useful intermediates for the preparation of enantiopure 1-[1-(2-pyrrolyl)alkyl]aziridines by routine cyclization of the β-aminoalcohol moiety and of (S)-N-benzoyl 1-[1-(2-pyrrolyl)alkyl]amines and their N-substituted derivatives by oxidative cleavage of the chiral auxiliary. 1-Allyl-2-pyrroleimines obtained from (S)-phenylglycinol and (S)-valinol underwent highly diastereoselective addition of allylmetal reagents, used in excess amounts, to give the corresponding secondary amines with concomitant allyl to 1-propenyl isomerisation of the 1-pyrrole substituent. Protection of the 2-aminoalcohol moiety as oxazolidinone, amide or Boc derivate followed by ring closing metathesis of the alkene groups gave the unsaturated bicyclic compound, whose hydrogenation afforded the indolizidine derivative as a mixture of separable diastereomers. The absolute configuration of the main diastereomer was assessed by X-ray crystallographic analysis.

Stereoselective synthesis of heterocycles as intermediates of biologically active compounds

The aim of this thesis was to investigate the synthesis of enantiomerically enriched heterocycles and dehydro-β-amino acid derivatives which can be used as scaffolds or intermediates of biologically active compounds, in particular as novel αvβ3 and α5β1 integrin ligands. The starting materials of all the compounds here synthesized are alkylideneacetoacetates. Alkylidene derivates are very usefull compounds, they are usually used as unsaturated electrophiles and they have the advantage of introducing different kind of functionality that may be further elaborated. In chapter 1, regio- and stereoselective allylic amination of pure carbonates is presented. The reaction proceeds via uncatalyzed or palladium-catalyzed conditions and affords enantiopure dehydro-β-amino esters that are useful precursor of biologically active compounds. Chapter 2 illustrates the synthesis of substituted isoxazolidines and isoxazolines via Michael addition followed by intramolecular hemiketalisation. The investigation on the effect of the Lewis acid catalysis on the regioselectivity of the addition it also reported. Isoxazolidines and isoxazolines are interesting heterocyclic compounds that may be regarded as unusual constrained -amino acids or as furanose mimetics. The synthesis of unusual cyclic amino acids precursors, that may be envisaged as proline analogues, as scaffolds for the design of bioactive peptidomimetics is presented in chapter 3. The synthesis of 2-substituted-3,4-dehydropyrrole derivatives starting from allylic carbonates via a two step allylic amination/ring closing metathesis (RCM) protocol is carried out. The reaction was optimized by testing different Grubbs’ catalysts and carbamate nitrogen protecting groups. Moreover, in view of a future application of these dehydro-β-amino acids as central core of peptidomimetics , the malonate chain was also used to protect nitrogen prior to RCM. Finally, chapter 4 presents the synthesis of two novel different classes of integrin antagonists, one derived from dehydro-β-amino acid prepared as described in chapter 1 and the other one has isoxazolidines synthesized in chapter 2 as rigid constrained core. Since that these compounds are promising RGD mimetics for αvβ3 and α5β1 integrins, they have been submitted to biological assay. and to interpret on a molecular basis their different affinities for the αvβ3 receptor, docking studies were performed using Glide program.