Drug Development Today

With the recent development of molecular biology and genetics, the bottleneck in drug discovery has shifted from target identification to the development of candidate drugs^[1]. The techniques of combinatorial chemistry have evolved to meet this demand of rapid and diverse substance production, by focusing on automated handling of a large number of samples rather than on experimental methodologies.^[2] For those combinatorial synthesizers available where gases can be distributed via a common and open link, contamination between the different reaction wells constitutes an immediate risk. The construction of fast, reliable and convenient gaseous protocols for chemistry is therefore important for the chemical development of combinatorial chemistry.

Mo(CO)₆ as Carbonylating Agent - Development ^[3]

The finding that Mo(CO)₆ under certain conditions of fast microwave heating demonstrated a controlled liberation of carbon monoxide encouraged us to investigate the construction of a new convenient synthetic strategy for combinatorial carbonylations where carbon monoxide was generated in-situ.

By heating a mixture of an aryl bromide or aryl iodide (1 equiv.), an amine (1.2 equiv.), Mo(CO)₆ (0.33 equiv.), a palladium catalyst, K₂CO₃ (aq) as base and diglyme as solvent in a closed vessel with microwaves, amides were formed in medium to high yields (65 - 85 % isolated, Table 1, Figure 1).

Figure 1: Conditions for the palladium-catalyzed amidation utilizing in-situ generated carbon monoxide from Mo(CO)₆
 

Iodides could be coupled with solid Pd/C as catalyst, while bromides required a homogeneous catalyst. Thus, for aryl iodides an extremely simple workup by extraction with 2M HCl(aq) and diethyl ether, followed by filtration, afforded the pure amides. Among the homogeneous catalysts tested for aryl bromides, the most suitable to use was the deep red 2:1 mixture of BINAP and Herrmann’s catalyst (1) dissolved in toluene.
 

Table 1: Carbonylations according to figure 1 employing the in-situ CO generation protocol.
 

Entry	R’’	Reagent (RR’NH or other)	Yield
1 2 3 4 5 6 7 8 9	MeO- Me- F3C- Ac- MeO- Me- F3C- Ac- Me	n-BuNH2 n-BuNH2 n-BuNH2 n-BuNH2 Piperidine Piperidine Piperidine Piperidine Water	68 [a] 72 [a] 78 [a] 79 [a] 65 [a] 68 [a] 75 [a] 76 [a] 87 [a,b]
[a]: Average isolated yields from 2-6 runs. Aryl iodides and bromides produced essentially the same yields, although Pd/C was used for the iodides while 1/BINAP was used for the bromides. [b]: Carboxylic acid. Ethylene glycol was added as co-solvent, making up ca. 30 % of the solvent.

Interestingly, the carbonylations could also be performed in open vessels without any loss in yield, suggesting that enough carbon monoxide is dissolved in the reaction mixture to furnish complete conversion.

The addition of ethylene glycol as co-solvent to the diglyme/K₂CO₃(aq) mixture resulted in competitive formation of the corresponding benzoic acid derivative instead of the amide (Table 1, entry 9). The omission of amine made this procedure a powerful tool for making the corresponding benzoic acid from an aryl halide.^[4]

Library Synthesis of HIV-Protease Inhibitors - Application
 
 

An efficient HIV-drug must ideally have a therapeutic activity not only on the native forms of HIV but also on its mutants. Inhibition of native HIV-1 protease has been demonstrated by the C₂-symmetric structure 2.^[5] A recent discovery demonstrates that infected patients prescribed with the HIV-drug Ritonavir™ evolve specific mutations in the protease active site. These mutations are situated very close to the phenyl rings of 2, within its enzyme-inhibitor complex. Therefore, the synthesis of aryl analogues to 2 was attractive in our search for new HIV-drugs.

Figure 2: The application of the carbonylation protocol for rapid production of HIV-protease inhibitors.

By using halogenated analogues of 2 (meta, meta dibromo (3) and ortho, ortho diiodo (4)) in combination with the carbonylation protocol described above, a focused department stock-room library of amide-derivatives was synthesized (Figure 2). Generation of the library was undertaken employing the microwave heating Smith Synthesizer™. Rapid synthesis and repeated mass-triggered LC-MS purification resulted in a 16 % library yield of 36 isolated compounds. The crude reaction mixtures contained mono- and di-amidated inhibitors as well as unreacted starting material, where the contribution of the mono-amidated product varied from 0 – 30 % of the total yield. Amine 1 –10 (Figure 3) are some of the amines utilized.

Figure 3: Different amines employed in the synthesis of a small sized department stock-room library of HIV-1 protease inhibitors.

Test results on native HIV-1 protease indicate that compounds of the ortho- and meta-substituted structure type might be potent and mutant resistant inhibitors.

& References:
[*] [1] Candidate drugs are new substances that have a beneficial therapeutic activity towards a disease and a commercial potential.
[2] a) G. Jung, Combinatorial Chemistry, Whiley-VCH, Weinheim, Germany, 1999; b) R. E. Dolle, K. H. Jr. Nelson, J. Comb. Chem. 1999, 1, 235.
[3] N-F. K. Kaiser, A. Hallberg, M. Larhed, In-Situ Generation of CO from M(CO)_x, A Convenient & Fast Route to Pd-catalyzed Amidations. Submitted.
[4] For other similar approaches to the synthesis of benzoic acids see: a) Y. Masuyama, R. Hayashi, K. Otake, Y. Kurusu, Chem. Commun. 1988, 44; b) H. Kumobayashi, S. Mitsuhashi, S. Akutagawa, S. Otsuka, Chem. Lett.1986, 157.
[5] M. Alterman, H. O. Andersson, N. Garg, G. Ahlsén, S. Lövgren, B. Classon, U. H. Danielsson, I. Kvarnström, L. Vrang, T. Unge, B. Samuelsson, A. Hallberg, J. Med. Chem. 1999, 42, 3835.