Evaluation of the Trypanosoma brucei 6-oxopurine salvage pathway as a potential target for drug discovery

Abstract

Due to toxicity and compliance issues and the emergence of resistance to current medications new drugs for the treatment of Human African Trypanosomiasis are needed. A potential approach to developing novel anti-trypanosomal drugs is by inhibition of the 6-oxopurine salvage pathways which synthesise the nucleoside monophosphates required for DNA/RNA production. This is in view of the fact that trypanosomes lack the machinery for de novo synthesis of the purine ring. To provide validation for this approach as a drug target, we have RNAi silenced the three 6-oxopurine phosphoribosyltransferase (PRTase) isoforms in the infectious stage of Trypanosoma brucei demonstrating that the combined activity of these enzymes is critical for the parasites' viability. Furthermore, we have determined crystal structures of two of these isoforms in complex with several acyclic nucleoside phosphonates (ANPs), a class of compound previously shown to inhibit 6-oxopurine PRTases from several species including Plasmodium falciparum. The most potent of these compounds have Ki values as low as 60 nM, and IC50 values in cell based assays as low as 4 μM. This data provides a solid platform for further investigations into the use of this pathway as a target for anti-trypanosomal drug discovery.

Fig 1. Purine salvage pathway of Trypanosoma brucei [17].

NH, nucleoside hydrolase; AK, adenosine kinase; APRT, adenine phosphoribosyl transferase; AMP, AMP deaminase; HGPRT, hypoxanthine guanine phosphoribosyl transferase; HGXPRT, hypoxanthine guanine xanthine phosphoribosyl transferase; IMPD, inosine-5´-monophopshate dehydrogenase; GMPS, GMP synthase; GDA, guanine deaminase; GMPR, GMP reductase; AMP DA, AMP deaminase; ADSS, adenylosuccinate synthetase; ADSL, adenylosuccinate lyase.

Fig 2

(A) Reaction catalyzed by the 6-oxopurine PRTases. (B–D) General structures of ANPs. Single chain ANPs (B); aza-ANPs (C); and branched ANPs with an attachment at one of the first two carbons from the N⁹ nitrogen in the base (D). When xanthine, guanine and hypoxanthine is the base, Y = OH, Y = NH₂ and Y = H, respectively.

Fig 3. Subcellular localization of HGPRT-I, HGPRT-II and HGXPRT in the bloodstream form of T. brucei.

(A) Immunoblot analysis of BF cells over-expressing v5-tagged HGPRT-I, HGPRT-II and HGXPRT was performed to reveal the subcellullar localization of these proteins. Cytosolic (CYT) and organellar (ORG) fractions were obtained by digitonin fractionation. Purified fractions were analyzed by immunoblot with the following antibodies: anti-V5, anti-enolase (cytosol), anti-hexokinase (organellar fraction, glycosomes), anti-mt hsp70 (organellar fraction, mitochondrion). The relevant sizes of the proteins are indicated on the left. (B) Immunofluorescence microscopy of the same cell lines as in (A) was used to determine subcellular localization of the v5-tagged HGPRT-I, HGPRT-II and HGXPRT proteins within the cell. HGPRT-I, HGPRT-II and HGXPRT were visualized by immunostaining using a monoclonal anti-v5 antibody and anti-mouse secondary antibody conjugated with fluorescein isothiocyanate (FITC). Antibodies against enolase and hexokinase served to mark cytosolic and glycosomal localization, respectively. The DNA content (nucleus and kinetoplast) was visualized using DAPI (4,6-diamidino-2-phenylindole).

Fig 4. Expression of HGPRT-I, HGPRT-II and HGXPRT in the PF and BF cells.

(A) The steady state abundance of HGPRT-I, HGPRT-II and HGXPRT was determined in PF and BF cells by Western blot analysis of whole cell lysates. Densitometric analysis was performed using the Image Lab 4.1 software and the number beneath the blots represents the abundance of immunodetected proteins expressed as a percentage of the PF sample. (B) Cellular fractionation of PF and BF cells was used to distinguish between the localization of cytosolic HGPRT-I and glycosomal HGPRT-II. Antibodies against hexokinase and enolase were used to mark the cytosol and glycosome, respectively. The protein marker sizes are indicated on the left.

Fig 5. Effects of RNAi silencing of HGPRT-I, HGXPRT and simultaneous RNAi silencing of HGPRT-I/HGXPRT on T. brucei BF cell growth.

Growth curves of the noninduced (NON) and RNAi induced (IND) cells in which the expression of HGPRT-I (A) HGXPRT (B) and HGPRT-I/HGXPRT (C) was RNAi silenced. Cells were cultured in HMI-9^full media and the cumulative cell number was calculated from cell densities adjusted by the dilution factor needed to see the cultures at 10⁵ cells/ml each day. The figure is representative of at least three independent RNAi inductions. The steady-state abundance of HGPRT-I and HGXPRT in noninduced (NON) RNAi cells and in cells induced with tetracycline for 2, 4 and 6 days (A, B and C, bottom panels) was determined by immunoblotting using specific anti-HGPRT-I and anti-HGXPRT serums. Densitometric analysis was performed using the Image Lab 4.1 software and the determined values were normalized to an enolase loading control signal. The protein marker sizes are indicated on the left.

Fig 6. Effects of RNAi silencing of HGPRT-I and HGXPRT on T. brucei BF cell growth in the presence of indicated purine base.

The RNAi cell lines were grown in HMI-9 medium containing hypoxanthine (A) or xanthine (B) as the sole purine source. The cumulative cell number was calculated from cell densities adjusted by the dilution factor needed to have the cultures at an appropriate concentration on each day. The steady-state abundance of HGPRT-I and HGXPRT in noninduced (NON) cells and in cells induced for given time points (bottom panels) was determined by immunoblotting using specific anti-HGPRT-I and anti-HGXPRT serums. Mt hsp70 served as a loading control. The protein marker sizes are indicated on the left.

Fig 7. DMS crosslinking of recombinant HGPRT-I, HGPRT-II and HGXPRT.

Proteins were crosslinked by DMS for 0, 10, 60 and 180 minutes and separated on the 15% SDS-PAGE followed by Coomassie R-250 staining. The HGPRT-I, HGPRT-II and HGXPRT homodimers were identified based on their sizes. The protein marker sizes are indicated on the left. (A) test for homodimerization, (B) test for heterodimerization.

Fig 8. Immunoprecipitation of v5-tagged HGPRT-I, HGPRT-II and HGXPRT.

Whole cell lysates of non-induced (NON) cells and cells induced for expression of v5-tagged HGPRT-I, HGPRT-II and HGXPRT (IND) were subjected to immunoprecipitation using anti-v5 monoclonal antibody. Immunoblots containing 5% input and 100% eluate were probed for the presence of v5-tagged HGPRT-I, HGPRT-II and HGXPRT, and for native untagged HGPRT-I and HGXPRT.

Fig 9. General structures of ANPs 1–7 and their phosphoramidate prodrugs 1p-7p.

Fig 10. K_i values of the selected ANPs for HGPRT-I, HGPRT-II and HGXPRT and a comparison of the in vitro antitrypanosomal activity of their prodrugs against T. brucei BF427 cell line with their cytotoxicity in human cell lines.

*Estimated selectivity index (SI) = average CC₅₀ for A549 (Human lung carcinoma cells) divided by the average EC₅₀ for T. brucei BF427 cell line. Synthesis of ANP inhibitors and their prodrugs is described in [, , and 29].^aData from [29]; ^bData from [28]; ^cData from [20]; ^dData from [22].

Fig 11. The active site of HGPRT-I and HGXPRT.

Each enzyme is represented by its Connolly surface and the inhibitor is shown as a stick model. The F_o-F_c electron density for each inhibitor is overlaid. (A) HGPRT-I.6 complex, (B) HGPRT-I.1 complex, (C) HGPRT-I.7 complex, (D) HGPRT-I.2 complex, (E) HGXPRT.7 complex and (F) HGXPRT.2 complex.

Fig 12. Stereoimages of the active sites of HGPRT-I and HGXPRT in complex with the four ANPs.

(A) HGPRT-I. 6 complex (cyan carbon atoms), (B) HGPRT-I.1 complex (orange carbon atoms), (C) HGPRT-I.7 complex, (D) HGPRT-I.2, (E) HGXPRT.7 and (F) HGXPRT.2.

Fig 13. Active sites of HGPRT-I, HGXPRT and human HGPRT after superimposition.

(A) compound 6 complexes. HGPRT-I. 6 in cyan, human HGPRT.6 in gold. (B) compound 1 complexes. HGPRT-I.1 in orange and human HGPRT.1 in purple. (C) compound 2 complexes. HGPRT-I.2 in green, human HGPRT.2 in grey (D) compound 2 complexes. HGXPRT.2 in pink, human HGPRT.2 in grey. Black labels are for HGPRT-I, green labels are for the HGXPRT and red labels are for human HGPRT.