Optimization of tissue extraction for GCMS metabolomic analysis reveals altered metabolism of barbiturates in a polycystic kidney disease model

Hayley Abbiss1,2,3, Garth L. Maker1,2,4, Gabrielle C. Musk1,5, Catherine Rawlinson2,4,6, Joel P.A. Gummer1,2,4, Patricia A. Fleming1, Jacqueline K. Phillips7, Mary Boyce8, John Moncur3, Robert D. Trengove2,4

1School of Veterinary and Life Sciences, Murdoch University, WA, Australia; 2Separation Science and Metabolomics Laboratory, Murdoch University, WA, Australia; 3SpectralWorks, Cheshire, United Kingdom; 4Metabolomics Australia, Western Australian Node, WA, Australia; 5Animal Care Services, The University of Western Australia, WA, Australia; 6Faculty of Science and Engineering, Curtin University, WA, Australia; 7Department of Biomedical Science Macquarie University, NSW, Australia; 8School of Science, Edith Cowan University, WA, Australia;

First Published: ASMS, 2016

Overview

During extraction optimization for an untargeted GCMS metabolomic analysis of the Lewis Polycystic Kidney rat, it was found that barbiturates (euthanasia drug) are altered in both kidney and liver tissue.

Introduction

Polycystic kidney diseases (PKDs) are inherited and lead to end-stage kidney failure. There are many forms of PKD, including the nephronophthisis (NPHP) group of ciliopathies, for which the Lewis Polycystic Kidney (LPK) rat has been classified as a model. LPK tissues were investigated using GCMS metabolomic analysis to determine biochemical changes in the diseased tissues. Prior to this, the amount of tissue extracted was first optimised. During optimisation, we identified four barbiturate metabolites in the profiles of kidney and liver from LPK rats which were significantly different to controls. The study was repeated without the use of barbiturate drugs, as the impact on metabolism may lead to misinterpretation of data.

Methods

Four 16-week old LPK and 4 Lewis age- and sex-matched control animals were used. Animals were anaesthetised and euthanized by intra cardiac injection of pentobarbitone (Jurox, 325 mg.mL-1). Organs were harvested, frozen on dry ice, and stored (-80°C). Organs were freeze-dried whole, ground using a mortar and pestle and weighed (40 mg) into lysis tubes. Metabolites were extracted with methanol and water containing 13C6-sorbitol (IS). Extracts (2, 5 and 10 mg equivalents) were frozen, freeze-dried and stored (-80°C). Metabolite extracts were derivatised with methoxyamine-HCl and MSTFA. For the analysis, a Shimadzu QP2010 Ultra GC-MS was used and for data analysis, SpectralWorks’ AnalyzerPro® and SPSS were used. Barbiturate compounds were matched to the NIST mass spectral database.

Results

Kidney tissue extraction optimization

For the optimisation of kidney tissue extraction, there was a sample weight effect on the number of resolved compounds (P = 0.007) and total peak area (TPA) of the total ion chromatogram (TIC; P < 0.001), where the 5 mg extract had a greater number of compounds than 2 mg (P = 0.006), but no difference existed between 2 and 10 mg (P = 0.501) or 5 and 10 mg (P = 0.137). Differences were seen in the TPA of the TIC between all weights where 5 > 10 > 2 mg (P < 0.001).

Figure 1. Mean (± SD) total peak area of the total ion chromatogram for 2, 5 and 10 mg tissue extractions using healthy, diseased and mixed (50:50 healthy:diseased) pooled kidney organ tissue samples.

Figure 2. Mean (± SD) number of components for 2, 5 and 10 mg tissue extractions using healthy, diseased and mixed (50:50 healthy:diseased) pooled kidney organ tissue samples.

Liver tissue extraction optimization

Similar results were seen for the optimisation of liver tissue extraction. There was a sample weight effect on the number of compounds (P < 0.001) where 5 and 10 mg extracts had a greater number of compounds than 2 mg (P < 0.001) but 5 and 10 mg were not different (P = 1.000). There was a sample weight effect on the TPA of the TIC (P < 0.001) where 5 and 10 mg showed greater TPA of the TICs than 2 mg (P < 0.001) but 5 and 10 mg were not different (P = 0.190).

Figure 3. Mean (± SD) total peak area total ion chromatogram (TIC) for 2, 5 and 10 mg tissue extractions using healthy, diseased and mixed (50:50 healthy:diseased) pooled liver organ tissue samples.

Figure 4. Mean (± SD) number of components for 2, 5 and 10 mg tissue extractions using healthy, diseased and mixed (50:50 healthy:diseased) pooled liver organ tissue samples.

Identification of barbiturates in tissue extracts

In kidney tissue, B1-4 (Figure 5; 6) were significantly increased (P ≤ 0.036) in control 2 mg extracts; B1, 3 and 4 were increased (P < 0.001) in control 5 mg and B2-4 were increased (P ≤ 0.012) in control 10 mg. In liver tissue, B1, 3 and 4 were significantly increased (P ≤ 0.004) in LPK 2 mg extracts; B3 and 4 were increased (P ≤ 0.046) in LPK 5 mg and B2-4 were increased (P ≤ 0.024) in LPK 10 mg.

Figure 5. Example extracted ion chromatogram from a 2 mg extraction from healthy control kidney tissue showing the four barbiturate peaks detected in Lewis and LPK kidney and liver tissue after euthanasia with pentobarbitone. The mass spectra of barbituric acids 1-4 are shown in Figure 6.

Figure 6. Mass spectra and structures of the four barbituric acid peaks detected in Lewis and LPK kidney and liver tissue: barbituric acid 1 = a (dimethylpentobarbital); 2 = b (methylbarbital); 3 = c (methylbarbital) and; 4 = d (pentobarbital)

Figure 7. Comparison of mean (± SD) barbiturate peak area relative to IS 13C6-sorbitol (a; dimethylpentobarbital), (b; methylbarbital), (c; methylbarbital) and (d; pentobarbital) in each extract (2, 5 and 10 mg; n=3) for pooled kidney and liver tissue from Lewis (□) and LPK (■) rats. ***P < 0.001; **P < 0.01; *P < 0.05; one-way ANOVA

Conclusion

Since for both kidney and liver tissues, the 5 mg extract had a significantly greater number of compounds and peak area of the TIC than 2 mg, and no significant increase in compounds or TIC was seen for the 10 mg extractions, 5 mg of tissue were used for kidney and liver extractions for future metabolomic analyses of organ tissue. In this optimization, it was also found that Lewis polycystic kidney rats appear to metabolize barbiturate compounds differently to age- and sex-matched controls, which may alter broader metabolite profiles. To avoid interpretation of the euthanasia method rather than the disease biochemistry of the polycystic kidney rat, future studies adopted an exsanguination euthanasia method. The benefit of an exsanguination method is two-fold: 1) no euthanasia drug is introduced and; 2) a near total blood volume is removed from the animal allowing organs free from circulating blood to be harvested.

Acknowledgements

Metabolomics Australia and Shimadzu are acknowledged with thanks for supporting this work.