Tetrahydroisoquinolines affect the whole-cell phenotype of Mycobacterium tuberculosis by inhibiting the ATP-dependent MurE ligase

Abstract

Objectives: (S)-Leucoxine, isolated from the Colombian Lauraceae tree Rhodostemonodaphne crenaticupula Madriñan, was found to inhibit the growth of Mycobacterium tuberculosis H37Rv. A biomimetic approach for the chemical synthesis of a wide array of 1-substituted tetrahydroisoquinolines was undertaken with the aim of elucidating a common pharmacophore for these compounds with novel mode(s) of anti-TB action.

Methods: Biomimetic Pictet-Spengler or Bischler-Napieralski synthetic routes were employed followed by an evaluation of the biological activity of the synthesized compounds.

Results: In this work, the synthesized tetrahydroisoquinolines were found to inhibit the growth of M. tuberculosis H37Rv and affect its whole-cell phenotype as well as the activity of the ATP-dependent MurE ligase, a key enzyme involved in the early stage of cell wall peptidoglycan biosynthesis.

Conclusions: As the correlation between the MIC and the half-inhibitory enzymatic concentration was not particularly strong, there is a credible possibility that these compounds have pleiotropic mechanism(s) of action in M. tuberculosis.

Keywords: Mycobacterium tuberculosis; SPOTi; TB; aporphine alkaloids; cell wall peptidoglycan; drug resistance; tuberculosis.

Figure 1.

Bischler–Napieralski-mediated synthesis to ortho-cyclized THIs 36–40. (a) NCS or NIS, CH₂Cl₂; 85% and 81% yield, respectively; (b) For X = Cl and I, B₂H₆.SMe₂, 42% and 55%, respectively. (c) Br₂, CH₃CO₂H, 67%. (d) 3,4-(Methylenedioxy)phenylethanoic acid, N,N-dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP), CH₂Cl₂; X = Cl and X = Br, 99%, X = I, 74%. (e) POCl₃, toluene, reflux. (f) NaBH₃CN, MeOH; 36 19%, 37 19%, 38 2% over two steps. (g) 36, NBS, CH₂Cl₂, 35%. (h) 37, LiAlH₄, –78°C, 89%.

Figure 2.

Chemical structures of the natural aporphines 3-methoxynordomesticine and (S)-leucoxine (1) and the synthetic THIs 2-40. Compounds 1-40 were evaluated in this study.

Figure 3.

Comparison of the three-dimensional surface of four active alkaloids using FieldTemplater (Cresset software). Alignment of 3-methoxynordomesticine, leucoxine (1), 28 and 35 showing surfaces of different chemical environments. Turquoise surfaces represent negative patches, red surfaces correspond to positive sections, brick yellow surfaces denote lipophilic regions and light yellow volumes relate to dominant van der Waals areas. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 4.

Acid fast staining of M. bovis BCG after 48 h of treatment with d-cycloserine (DCS) and 39, and comparison of the cell length. (a) Control with DMSO. (b) DCS at 100 mg/L. (c) Compound 39 at 160 mg/L. Magnification ×1000. (d) The cell length of M. bovis BCG after treatment with vehicle (DMSO), DCS (100 mg/L) and 39 (160 mg/L). The dots represent the raw data and the diamonds represent the mean value with SD (error bars). The cell length of 100 cells was manually measured for each treatment from microscopic images. The mean cell sizes of the three treatments were statistically significant at P < 0.01. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.