In vivo longitudinal 1H MRS comparison of hippocampal and cerebellar changes due to Chronic Hepatic Encephalopathy, a rat model study
Veronika Rackayova1, Olivier Braissant2, Corina Berset3, Jocelyn Grosse4, Rolf Gruetter1,3, Valérie A. McLin5, and Cristina Cudalbu3

1Laboratory of Functional and Metabolic Imaging, Center for Biomedical Imaging, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, Lausanne, Switzerland, 2Service of Biomedicine, University Hospital of Lausanne, Lausanne, Vaud, Switzerland, Lausanne, Switzerland, 3Center for Biomedical Imaging, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, Lausanne, Switzerland, 4Laboratory of behavioral genetics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland, Lausanne, Switzerland, 5Swiss Center for Liver Disease in Children, Department of Pediatrics, University Hospitals Geneva, Geneva, Switzerland, Geneva, Switzerland

Synopsis

Chronic liver disease leads to Hepatic Encephalopathy - spectrum of neuropsychiatrical disorders. We investigated potential neurometabolic differences between two key brain regions (hippocampus and cerebellum). Cerebellum shows similar increase of glutamine but lower tNAA, Tau, Cr, Asc and different osmotic response indicating that these regions are influenced unequaly.

Purpose of the study:

Chronic hepatic encephalopathy (CHE) is a severe liver-disease-induced neuropsychiatric disorder, seriously affecting patients' quality of life1. The pathophysiology of CHE is still incompletely understood but ammonia is believed to be a main toxin playing crucial role in the development and progress of the disease, evoking glutamine (Gln) increase in the brain and subsequent diminution of other osmolytes as an osmoregulatory answer2-4. Previous studies have shown differences in neurometabolic profile between brain regions in a hyperammonemia5 model, and that net glutamine synthesis rates differ between the cortex and hippocampus6. Therefore, the aim of this study was to investigate potential metabolic differences between the hippocampus and cerebellum, the rationale being that they are two key brain regions to investigate in CHE since they are implicated in its symptoms (cognitive and fine motor deficits7).

Methods:

Wistar male adult rats (n=3) underwent bile duct ligation (BDL), recognized animal model to study chronic liver disease (CLD) induced HE8,9. Longitudinal in vivo 1H-MRS and blood sampling (bilirubin, alanine aminotransferase (ALAT), glucose) were performed before BDL-surgery (scan0) and after every two weeks (scan2, 4, 6, 8). Open field tests (to evaluate motor activity and the degree of HE10) took place at scan 4, 6 and 8. All MR experiments were performed on a 9.4T system (Varian/Magnex Scientific) using home-built 14mm diameter quadrature 1H-surface coil as a transceiver and ultra-short-echo time SPECIAL spectroscopy sequence11 (TE=2.8ms, TR=4s, 160 averages). Two volumes-of-interest, hippocampus (2×2.8×2mm3) and cerebellum (2.5×2.5×2mm3) were studied. First and second order shims were adjusted using FASTMAP (linewidth of 9-11Hz in hippocampus, 14-19Hz in cerebellum). Concentrations of metabolites were calculated by LCModel using water as reference.

Results and discussion:

Differences in metabolic profile between the hippocampus and cerebellum are presented in Fig.1. In healthy animals (before BDL-surgery) cerebellum showed significantly higher Cr, PCr, Gln, NAA, tNAA and tCr. 8 weeks after BDL this pattern changed, the only remained significant changes were for Cr, PE and Tau (Fig.1A-B). All BDL rats showed increase in plasma bilirubin, correlated with the increase in ALAT (r=0.82, p=0.0003, Fig.2A) proving the presence of CLD. Bilirubin further correlated with brain Gln both in the hippocampus and cerebellum (r=0.7, p=0.01, Fig.2B). In both regions Gln reached ~200% increase at scan8 (Fig.3A-C). However the osmoregulatory response was different. Both brain regions showed similar decrease in mIns and tCho (Fig.3D-F, Fig.4A-B) with non-significant differences between regions. In contrast, the cerebellum showed significantly stronger decrease in Tau (Fig.3G-I) and Cr (Fig.3J-L) leading to a constant sum of osmolytes (Gln+mIns+Tau+tCho+Cr) in cerebellum compared to a tendency toward an increase in the hippocampus (Fig.3M-O). Tau is considered as neuronal osmolyte12, shown to increase/protect cell viability in cerebellum13. Cr is a metabolite involved in energy metabolism but its involvement in osmoregulation and neuroprotection was recently reported14,15. Additionally, the cerebellum showed a trend toward a decrease in tNAA (-5%, ns, Fig5A) in agreement with cerebellar damage and neuronal cell loss16 in patients with end-stage liver disease. This decrease in Cr and tNAA in the cerebellum had a tendency to correlate with performance in behavioural tests (Fig.5B,C). Among neurotransmitters Glu (-11%, ns) and GABA (-20%, ns) showed a similar decrease in both brain regions with a stronger and negative correlation with Gln in cerebellum (Fig.4E,F). Moreover, there was a different antioxidant response to BDL between the cerebellum and hippocampus with a stronger decrease in Asc in the cerebellum (Fig.4G) while GSH dropped more markedly in the hippocampus (Fig.4H).

Conclusions:

We analysed for the first time in vivo and longitudinally the metabolite changes in cerebellum and compared them to those in hippocampus during CHE. Based on our measurement we can conclude that hippocampus and cerebellum show similar increase in glutamine but differences in some other metabolites (tNAA, Tau, Cr, Asc, GSH) demonstrating that these regions are influenced differently by CLD-induced HE. These changes in cerebellum could stand behind the observed deterioration in locomotor activity.

Acknowledgements

Supported by CIBM of the UNIL, UNIGE, HUG, CHUV, EPFL, the Leenaards and Jeantet Foundations. EU: FP7-PEOPLE-2012-ITN project 316679 TRANSACT; The authors thank Prof Carmen Sandi (Laboratory of behavioral genetics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland,) for her support in behavioural part.

References

1Morgan, Metab Brain Disease 2007; 2Rackayova et al, Metab Brain Dis 2015; 3Cudalbu, Metab Brain Dis 2013; 4Brusilow et al, Neurotherapeutics 2010; 5Dejong et al, J Neurochem 1992; 6Cudalbu et al, Proc Intl Soc Magn Reson Med 18 2010; 7Norenberg et al, Metab Brain Dis 2009; 8Biecker et al, J Pharmacol Exp Ther 2005; 9Butterworth et al, Liver Int 2009; 10Leke et al, Plos one 2012; 11Mlynárik et al, Magn Reson Med 2006; 12Lei et al, J Cereb Blood Flow Metab 2009; 13Boldyrev et al, Neurosci Lett. 1999; 14Rae, Neurochem Res 2014; 15Braissant O, Mol Genet Metab 2010; 16Butterworth, Metab Brain Dis 2007.

Figures

Fig.1

A,B-Graphs represent significantly different metabolites between brain regions at corresponding time-points.

C-Representative spectra from one animal at scan0 and scan8 from hippocampus and cerebellum, changes highlighted between scans 0 and 8.

We admit a difference in linewidths in the hippocampus and cerebellum but it do not influence integral comparison.


Fig.2

A-Correlation between plasma bilirubin and plasma levels of liver enzyme alanine aminotransferase (ALAT) during the evolution of chronic liver disease in BDL rats.

B-Correlation between plasma bilirubin and brain glutamine (Gln) in cerebellum and hippocampus.

Bilirubin in [mg/dl], ALAT in [U/l]; brain Gln in [mmol/kg ww]


Fig.3

Longitudinal evolution of glutamine and other brain osmolytes in hippocampus (hippo) and cerebellum.

Glutamine (A-C), myoInositol (D-F), Taurine (G-I), Creatine (J-L), sum of main brain osmolytes (Gln+mIns+Tau+tCho+Cr; M-O) in [mmol/kg ww].

Repeated measures ANOVA were used for statistical comparison, significance level: ****p<0.0001; *** p<0.001; ** p<0.01; * p<0.05.


Fig.4

Correlations and dependencies in cerebellum (in green) and hippocampus (in black) between:

A-D glutamine and other brain osmolytes (mIns, tCho, Tau, Cr),

E-F glutamine and neurotransmitters (Glu, GABA),

G-H glutamine and antioxidants (Asc, GSH)

Metabolites in [mmol/kg ww].


Fig.5

A-Negative correlation between glutamine and cerebellar tNAA.

B-Relation between distance moved in Open field behavioural test and cerebellar tNAA.

C-Relation between distance moved in Open field behavioural test and cerebellar Cr.

Brain metabolites in [mmol/kg ww].




Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
1460