Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Aug 3;8(1):11678.
doi: 10.1038/s41598-018-30149-7.

Metabolic Network-Based Predictions of Toxicant-Induced Metabolite Changes in the Laboratory Rat

Affiliations
Free PMC article

Metabolic Network-Based Predictions of Toxicant-Induced Metabolite Changes in the Laboratory Rat

Venkat R Pannala et al. Sci Rep. .
Free PMC article

Abstract

In order to provide timely treatment for organ damage initiated by therapeutic drugs or exposure to environmental toxicants, we first need to identify markers that provide an early diagnosis of potential adverse effects before permanent damage occurs. Specifically, the liver, as a primary organ prone to toxicants-induced injuries, lacks diagnostic markers that are specific and sensitive to the early onset of injury. Here, to identify plasma metabolites as markers of early toxicant-induced injury, we used a constraint-based modeling approach with a genome-scale network reconstruction of rat liver metabolism to incorporate perturbations of gene expression induced by acetaminophen, a known hepatotoxicant. A comparison of the model results against the global metabolic profiling data revealed that our approach satisfactorily predicted altered plasma metabolite levels as early as 5 h after exposure to 2 g/kg of acetaminophen, and that 10 h after treatment the predictions significantly improved when we integrated measured central carbon fluxes. Our approach is solely driven by gene expression and physiological boundary conditions, and does not rely on any toxicant-specific model component. As such, it provides a mechanistic model that serves as a first step in identifying a list of putative plasma metabolites that could change due to toxicant-induced perturbations.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
Experimental design to measure early perturbations in rat liver metabolism. Preliminary studies using clinical chemistry markers to determine acetaminophen (APAP) dose and duration. Alterations in ALT (a) and AST activity levels (b) for control (dashed line with circles, n = 6), 1 g/kg APAP (solid line with triangles, n = 6), and 2 g/kg APAP (dotted line with squares, n = 7, *p < 0.05). (c) Schematic showing the design of the study, using Sprague Dawley rats exposed to a single dose of 2 g/kg of APAP under fasting conditions. In Study 1, rats were administered APAP and observed for 10 h (n = 8) together with control animals (n = 8), after which they were infused with 2H/13C labeling to obtain flux measurements using metabolic flux analysis. In Studies 2 and 3, rats were given APAP and observed for 5 h (n = 8) and 10 h (n = 8), respectively, with the corresponding control groups treated with vehicle (n = 8 each). Samples of blood and liver tissue were collected at 5 h or 10 h after APAP exposure and subjected to global metabolic profiling analysis or RNA sequencing, respectively (See Materials and Methods for further details).
Figure 2
Figure 2
Volcano plots of differentially expressed genes (DEGs) in the liver, induced by acetaminophen (APAP). False discovery rates (FDRs) plotted against APAP-induced log2 fold changes in DEGs for one group of rats collected at 5 h (a) and a second group at 10 h (b). Genes from the RNA-seq data mapped onto the iRno model at 5 h (c) and 10 h (d). Circles in red/green show genes/transcripts that were significantly up-/down-regulated (FDR < 0.10), whereas black circles show those that were unchanged. Cyp8b1: Cytochrome P450 family 8 subfamily B member 1; Daglb: Diacylglycerol lipase, beta; Fads1: Fatty acid desaturase 1; Fdft1: Farnesyl diphosphate farnesyl transferase 1; Hmox1: Heme oxygenase 1; Hsd17b2: Hydroxysteroid (17-beta) dehydrogenase 2; Ldlr: Low density lipoprotein receptor; Slco2a1: Solute carrier organic anion transporter family, member 2a1; Slc1a4: Solute carrier family 1 member 4; Slc25a15: Solute carrier family 25 member 15; Slc34a2: Solute carrier family 34 member 2; Tdo2: Tryptophan 2,3-dioxygenase; Tnxrd1: Thioredoxin reductase1; Upp2: Uridine phosphorylase 2.
Figure 3
Figure 3
The minimal network used for estimating flux measurements in the tracer dilution study. Abbreviations shown are names of enzymes for which absolute flux was calculated based on metabolic flux analysis. ALDO: Aldolase; CS: Citrate synthase; ENO: Enolase; GADPH: Glyceraldehyde-3-phosphate dehydrogenase; GK: Glycerol kinase; GPI: D-glucose-6-phosphate isomerase; G6PC: D-Glucose-6-phosphatase; IDH: Isocitrate dehydrogenase; LDH: Lactate dehydrogenase; OGDH: Oxoglutarate dehydrogenase; PC: Pyruvate:carbon-dioxide ligase (ADP-forming); PCC: Propionyl-CoA carboxylase; PCK: Phosphoenolpyruvate carboxykinase; PK: GTP:pyruvate 2-O-phosphotransferase; PYGL: glycogen phosphorylase; SDH: Succinate dehydrogenase.
Figure 4
Figure 4
Acetaminophen (APAP)-induced absolute flux measurements in the glucose production pathway obtained from the tracer dilution study under fasting conditions. (a) Bar graphs of flux measurements calculated from metabolic flux analysis 10 h after treatment with APAP (unfilled bars) or vehicle (filled bars). (b) Bar graph of approximate absolute flux values derived from the literature for control rats studied after 5 h of fasting (*p < 0.05, #values in Hexose units and abbreviations as in Fig. 3).
Figure 5
Figure 5
Volcano plots of global plasma metabolite changes induced by acetaminophen (APAP). False discovery rates (FDRs) plotted against APAP-induced log2 fold changes in plasma metabolites for one group of rats collected at 5 h (a) and a second group at 10 h (b). Metabolites mapped onto the iRno model based on KEGG ID annotation at 5 h (c) and 10 h (d). Red/green circles show metabolites significantly elevated/depressed (FDR < 0.10) for 5 and 10 h post APAP treatment, respectively; symbols in black show unchanged metabolites.
Figure 6
Figure 6
Schematic representation of how multi-omics data were integrated into the iRno model. The TIMBR algorithm estimates the network feasibility of producing a metabolite given changes in gene expression to the iRno model. Here we integrated TIMBR with in vivo multi-omics data. We used in vivo differential gene expression data to determine the reaction weights (W) in the iRno model, and then used flux measurements (MFA) from a tracer labeling study (vmfa) as well as the physiological boundary conditions of exchange metabolites (vex) to constrain the model. TIMBR then calculates the global network demand required for production of a metabolite (Xmet) by minimizing the weighted sum of flux across all reactions, under a control (Xcontrol) or treatment (Xtreatment) condition. We next z-transformed each raw metabolite production score (Xraw) to calculate the TIMBR production score (Xs) for that metabolite, which we compared with the global metabolic profiling data to assess whether its level had increased or decreased under the treatment condition relative to the control condition.
Figure 7
Figure 7
Heat map of TIMBR production scores compared to metabolic profiling data obtained 10 h after APAP treatment under fasting conditions. iRno model predictions calculated under two integration conditions were compared against log2 fold changes of metabolites that significantly (FDR < 0.10) changed in the global metabolic profiling data (Data). In one condition, only gene expression changes were used (No MFA), whereas in the other both gene expression changes and MFA data were used as constraints (MFA). The numbers in the heat map show the log2 fold changes of the metabolic profiling data (left column) and TIMBR production scores under the no MFA (center column) and MFA (right column) conditions. The color scheme on the far right shows the degree of change in the level of a plasma metabolite, from highly increased (dark red) to highly decreased (dark blue).

Similar articles

See all similar articles

Cited by 5 articles

References

    1. Larson AM, et al. Acetaminophen-induced acute liver failure: results of a United States multicenter, prospective study. Hepatology. 2005;42:1364–72. doi: 10.1002/hep.20948. - DOI - PubMed
    1. Taylor, L. G., Xie, S., Meyer, T. E. & Coster, T. S. Acetaminophen overdose in the Military Health System. Pharmacoepidemiology and Drug Safety21, 375–83, 10.1002/pds.3206 (2012). - PubMed
    1. Kaplowitz N. Idiosyncratic drug hepatotoxicity. Nature reviews. Drug discovery. 2005;4:489–99. doi: 10.1038/nrd1750. - DOI - PubMed
    1. Giesen PL, et al. Greater than expected alanine aminotransferase activities in plasma and in hearts of patients with acute myocardial infarction. Clin Chem. 1989;35:279–83. - PubMed
    1. Halkes S, van den Berg A, Hoekstra M, du Pont J, Kreis R. Transaminase and alkaline phosphatase activity in the serum of burn patients treated with highly purified tannic acid. Burns. 2002;28:449–53. doi: 10.1016/S0305-4179(02)00041-4. - DOI - PubMed

Publication types

Feedback