Differential protein expression and post-translational modifications in metronidazole-resistant Giardia duodenalis

Abstract

Background: Metronidazole (Mtz) is the frontline drug treatment for multiple anaerobic pathogens, including the gastrointestinal protist, Giardia duodenalis. However, treatment failure is common and linked to in vivo drug resistance. In Giardia, in vitro drug-resistant lines allow controlled experimental interrogation of resistance mechanisms in isogenic cultures. However, resistance-associated changes are inconsistent between lines, phenotypic data are incomplete, and resistance is rarely genetically fixed, highlighted by reversion to sensitivity after drug selection ceases or via passage through the life cycle. Comprehensive quantitative approaches are required to resolve isolate variability, fully define Mtz resistance phenotypes, and explore the role of post-translational modifications therein.

Findings: We performed quantitative proteomics to describe differentially expressed proteins in 3 seminal Mtz-resistant lines compared to their isogenic, Mtz-susceptible, parental line. We also probed changes in post-translational modifications including protein acetylation, methylation, ubiquitination, and phosphorylation via immunoblotting. We quantified more than 1,000 proteins in each genotype, recording substantial genotypic variation in differentially expressed proteins between isotypes. Our data confirm substantial changes in the antioxidant network, glycolysis, and electron transport and indicate links between protein acetylation and Mtz resistance, including cross-resistance to deacetylase inhibitor trichostatin A in Mtz-resistant lines. Finally, we performed the first controlled, longitudinal study of Mtz resistance stability, monitoring lines after cessation of drug selection, revealing isolate-dependent phenotypic plasticity.

Conclusions: Our data demonstrate understanding that Mtz resistance must be broadened to post-transcriptional and post-translational responses and that Mtz resistance is polygenic, driven by isolate-dependent variation, and is correlated with changes in protein acetylation networks.

Figure 1:

Protein identification, differential expression, and functional enrichment in Mtz lines. Proportional Venn diagrams showing unique and overlapping protein identifications in the 3 TMT 10-plexes (left) and for differentially expressed proteins in each MtzR line compared to MtzS parents (right). (B) The 6 functional clusters identified as enriched from differentially expressed proteins identified in MtzR isolates and their total protein number.

Figure 2:

EGF-like differentially expressed proteins and VSP subpopulations. (A) Distribution of differentially expressed EGF-like proteins within Giardia protein families by MtzR line. (B) Proportional Venn diagrams showing overlapping identities of all differentially expressed EGF-like proteins in MtzR lines (above) and all differentially expressed VSP proteins (below). (C) Heat map showing fold change in expression of VSPs in MtzR lines compared to the MtzS parents; replicate details are shown on the bottom axis to represent biological variation within lines. Lower fold changes are represented by blue, while higher fold changes are represented in red.

Figure 3:

Protein expression in antioxidant and electron transport networks. Gene annotations including accession numbers (prefix “GL50803_”) and average protein expression fold change between MtzR from MtzS parents in (A) antioxidant proteins and (B) electron transport proteins.

Figure 4:

Differential protein expression in glycolysis and pyruvate catabolism. Pyruvate catabolism, with emphasis on enzymes with electron transport in upstream glycolysis. Enzymes with a white background were not identified in the protein dataset but have been included for completeness of pathway. Metabolites are shown in black boxes. Direction of differential expression in proteins is indicated using arrows and colors as designated in the top right corner. Fd, ferredoxin; ADH LTC, alcohol dehydrogenase lateral transfer candidate; KB, ketobutyrate; KG, ketoglutarate.

Figure 5:

Western blots of post-translational protein modifications in MtzS and MtzR lines. Total protein lysate from trophozoites (15 μg) from WB, 713, and 106 MtzS and MtzR lines was probed with antibodies against acetylated lysine (KAc), mono-methylated lysine (K-MMe), ubiquitin (Ubi), phosphorylated tyrosine (pY), and the 14–3-3 binding motif (including phosphorylated serine) (14–3-3). Protein loading was verified after transfer using Ponceau S staining (first row, left). MtzS and MtzR lanes are designated by “S” and “R,” respectively. Altered protein features detected in 3/3 MtzR lines are designated with a solid red arrow, while protein features changed in 2/3 lines are designated with a broken red arrow.

Figure 6:

Results of trichostatin A exposure in MtzS and MtzR lines. (A) Dose–response curves for MtzR and MtzS lines of WB, 106, and 713 to deacetylase inhibitor TSA. Error bars represent ±1 standard deviation; experiments were performed in triplicate. (B) Western blots for total protein lysate from trophozoites (10 μg) exposed to TSA from WB, 713, and 106 MtzS and MtzR lines probed with antibodies against acetylated lysine (KAc). Trophozoites were exposed for 18 hours to 2 μM TSA, with control flasks exposed to the same volume DMSO as used as a vehicle for TSA exposure. Red arrows designate protein features with significant changes between TSA exposed and control trophozoites in the 6 lines.

Figure 7:

IC₅₀ profiles and post-translational modifications upon discontinuation of Mtz selection. (A) Dose–response curves for MtzR lines of 106 and 713 upon discontinuation of drug selection at 4 (P8), 8 (P16), and 12 (P24) weeks compared to MtzS lines. MtzS lines were used and are designated 106_WT and 713_WT. Error bars represent ±1 standard deviation; experiments were performed in triplicate. The table records the calculated IC₅₀ for each time point, with resistance factors calculated against the IC₅₀ of the MtzS parent isolate. (B) Western blots against lysate from trophozoites (15 μg) from 713 and 106-MtzR (P0) and 4 (P8), 8 (P16) after discontinued drug selection was probed with antibodies against acetylated lysine (KAc), mono-methylated lysine (K-Mme), and phosphorylated tyrosine (pY). Altered features within 713 and 106 lines are designated on the right of the blot using a solid red arrow. Exposure times have been reduced by 20% to prevent overexposure of major bands (e.g., histone variants).