Cholesteryl ester levels are elevated in the caudate and putamen of Huntington’s disease patients

Huntington’s disease (HD) is an autosomal dominant neurodegenerative illness caused by a mutation in the huntingtin gene (HTT) and subsequent protein (mhtt), to which the brain shows a region-specific vulnerability. Disturbances in neural cholesterol metabolism are established in HD human, murine and cell studies; however, cholesteryl esters (CE), which store and transport cholesterol in the brain, have not been investigated in human studies. This study aimed to identify region-specific alterations in the concentrations of CE in HD. The Victorian Brain Bank provided post-mortem tissue from 13 HD subjects and 13 age and sex-matched controls. Lipids were extracted from the caudate, putamen and cerebellum, and CE were quantified using targeted mass spectrometry. ACAT 1 protein expression was measured by western blot. CE concentrations were elevated in HD caudate and putamen compared to controls, with the elevation more pronounced in the caudate. No differences in the expression of ACAT1 were identified in the striatum. No remarkable differences in CE were detected in HD cerebellum. The striatal region-specific differences in CE profiles indicate functional subareas of lipid disturbance in HD. The increased CE concentration may have been induced as a compensatory mechanism to reduce cholesterol accumulation.


Results
Regional differences in total CE between control and HD tissue. Analysis of total brain CE between the caudate, putamen and cerebellum demonstrated interesting region-specific differences between control and HD tissues (Fig. 2). While no differences were observed in total CE between caudate, putamen and cerebellum within the control tissues, total CE levels were dysregulated between the three regions in the HD patients. For HD tissues, the caudate contained the highest amount of CE compared with both the putamen (+ 27%, p = 0.0360) and cerebellum (+ 60%, p = 0.0001). Within each region of the brain, both the caudate and putamen contained higher amounts of total CE than their corresponding control tissues (+ 27%, p = 0.0045, and + 10%, p = 0.0181, respectively), while no statistically significant differences were observed in total CE between HD and control in the cerebellum. Polyglutamine repeat numbers were not provided in the clinical information. Therefore, a correlation analysis was run to examine if total CE levels were correlated with age of death in HD subjects (Supplementary Table S1), as there is direct relationship between polyglutamine repeat number and age of death 30 . No correlations were observed. Additional correlation analyses were run to determine relationships between biometric data, disease status and total CE (Supplementary Tables S1, S2, and S3). No correlations were observed.
Differences in molecular CE species between control and HD samples. The nomenclature of CE species used in this paper identifies the number of carbons and double bonds of the fatty acyl chain i.e. 16:1 is a fatty acyl chain with 16 carbons and 1 double bond. Seventeen common species of CE were identified across all tissues (Fig. 3). The most dominant CE species in both the control and HD tissues across the three regions of the www.nature.com/scientificreports/ brain were typically 16:0, 18:1, 20:0, 20:4 and 20:5. Little variation in lipid species abundance was observed across the three control brain regions. Six species of CE were primarily responsible for the increase in total CE in HD caudate (Fig. 3a). Species affected included both long and very long-chain fatty acids, and all degrees of saturation (saturated, monounsaturated, and polyunsaturated). These species, with their percentage increase in con-  Total CE concentration in HD vs control in the caudate (n = 11), putamen (HD n = 12, control n = 11) and cerebellum (HD n = 9, control n = 12). Lipid concentrations are expressed as means ± standard error of the mean (nmol/mg brain tissue). Data was assessed for normality using a D' Agostino Pearson Omnibus test. Data was analysed using a two-way ANOVA with a Tukey's test for multiple comparisons between regions. *p < 0.05, ***p < 0.0001. CE cholesteryl ester, HD Huntington's disease.   CE species in the putamen also showed no distinct pattern of species affected (unsaturated, monounsaturated). Four of the same CE species that were increased in HD caudate, were also increased, but to a lesser extent in HD putamen: CE 16:0 (+ 30%, p = 0.0148), CE 18:1 (+ 96%, p = 0.0032), CE 20:3 (+ 105%, p = 0.0026) and CE 22:4 (+ 89%, p = 0.0001) (Fig. 3b). An increase in an additional CE species within the HD putamen, CE 22:5 (+ 46%, p = 0.0073) was also observed. Within the cerebellum, three species of polyunsaturated CE were found to be lower in abundance in HD: CE 18:2 (− 51%, p = 0.0205), CE 18:3 (− 47%, p = 0.0111) and CE 22:6 (− 30%, p = 0.0489) (Fig. 3c). High variability in cerebellar lipid profiles may be responsible for the failure to detect a difference in CE totals between HD and control samples for this region. The cerebellum showed a lower abundance of CE (− 13%, p = 0.1264) in HD compared to controls although this data was not found to be statistically significant. All CEs measured are listed in Supplementary Tables S4, S5 and S6.

Region-specific comparison of alterations in CE to cholesterol and 24-OHC.
Due to the alterations in cholesterol and 24-OHC found in the previous analysis of the tissue, differences in the total levels of CE in each region were compared. Figure 4 shows the percentage differences in 24-OHC, cholesterol and CE concentrations in HD relative to controls for all three regions. The analyses show that in HD caudate where cholesterol levels were not altered, 24-OHC, used for the removal of cholesterol was reduced (− 66%) along with a similar magnitude increase in CE (+ 72%, p = 0.0045). HD putamen did show an increase in cholesterol concentration (+ 31%), and a decrease in 24-OHC (− 59%) similar to HD caudate, although the increases in CE were not as substantial (+ 25%, p = 0.0181). HD cerebellum showed no differences in 24-OHC, cholesterol or total CE. Correlation analyses are available in Supplementary Table S7. ACAT1 expression was not different in HD striatum. Given the increased abundances of CE in HD caudate and putamen, we hypothesised that these abundances were due to increased expression of ACAT1. To test this, ACAT1 expression was determined in HD and control striatum. ACAT1 was detected as a single band at ~ 45 kDa (Fig. 5a). No difference in the expression of ACAT1 was detected between HD and control caudate or putamen (Fig. 5b).
Additionally, no differences were detected between HD caudate and putamen. Due to the absence of changes to total CE in HD cerebellum, ACAT1 expression was not measured in this region. Full western blot images are available for ACAT1 and GAPDH in the caudate (Supplementary Figs. S1 and S3) and in the putamen (Supplementary Figs. S2 and S4). The graph is showing that in HD caudate, the elevations in CE may have assisted in regulating cholesterol levels when removal via 24-OHC was reduced. HD putamen has a lower elevation in CE; however, it was found to have increases in cholesterol known to be detrimental to neural cells when homeostasis is disturbed. This graph is arguing for the use of CE as a compensatory pathway for cholesterol accumulation when removal is disturbed, and that HD caudate may be better at utilising this pathway than HD putamen. No significant disturbances were identified in HD cerebellum. *p < 0.05, **p < 0.01. HD Huntington's disease.

Discussion
This study has identified an increased abundance of CE in HD striatum (caudate and putamen), the brain region most severely atrophied in HD 2,7 . Previous research in R6/1 and R6/2 HD murine models identified increases in key metabolites from both the Kandutsch-Russell and Bloch cholesterol synthesis pathways with little effect on steady-state levels of cholesterol, suggesting additional possible compensation mechanisms in the human brain.
The current data suggest that the increased abundance of CE in HD striatum may represent the induction of a pathway that supports a capacity for cholesterol sequestration in the HD brain. The higher concentration of CE in HD caudate in the current study suggests an alternative compensatory pathway for sequestration of excess cholesterol through esterification if the 24-OHC pathway is disturbed. Although exhibiting similar decreases in the cholesterol removal product 24-OHC, the caudate and putamen show differences in CE and cholesterol. We suggest that HD caudate compensated cholesterol accumulation by sequestering it into cholesteryl esters. The putamen, which also had 24-OHC disturbed in a similar fashion to the caudate, had approximately half the increase of CE alongside cholesterol increases, indicating that this region may not have been able to compensate in the same way. ACAT1 activity is stimulated by a surplus of cholesterol in the cell, allowing for sequestration of cholesterol into esters which exist as lipid droplets in the cytoplasm 29,31 . We were unable to measure ACAT1 activity in this study therefore the possibility that ACAT1 activity is increased in HD cannot be ruled out. An alternative explanation may be an accumulation of CE over time. Synthesis and metabolism of CE is allosterically regulated 22,32 . Therefore when cholesterol excess subsides, CE are hydrolysed by cholesterol ester hydrolyase for conversion back to cholesterol 31 . However, this enzyme may have failed to either degrade excess CE in time or to override the continual formation of CE in response to high cholesterol within the cell. Recently, a study identified two additional configurations of ACAT1 as a dimer and a tetramer 32 . The study proposed the dimer as the active configuration, however, our protein analysis detected only the monomer. Although the accumulation of cholesterol in the cell is detrimental, the consequences of elevations in CE is still unknown. Increases in CE have been found in the CNS of other neurodegenerative diseases such as multiple sclerosis (MS), amyotrophic lateral sclerosis (ALS), and brain injury, which may indicate CE accumulation as a general  www.nature.com/scientificreports/ neuropathological consequence and not a feature of HD alone [23][24][25]33,34 . Specifically, in MS, CE are increased primarily in and surrounding white matter plaques. Analysis of human MS tissue identified no changes in the synthesis of CE, but instead a substantial reduction in the breakdown of CE, indicating that the high levels of CE were due to insufficient hydrolysis in MS 33 . Inhibition of ACAT1 activity can decrease the formation of amyloidbeta plaques in a mouse model of AD 35,36 . Therefore, CE is suspected to contribute to AD pathology. Cholesterol esterification in neurodegeneration may also prevent the synthesis of toxic cholesterol oxidation products, and therefore protect cells 23 . A study examining excitotoxic injuries in rat hippocampus suggested that raised CE concentrations are a direct result of increased cholesterol in response to neural injury 23 , which may be occurring in HD. Increases in CE have been observed in brain lesions that involved the breakdown of myelin 23 . Magnetic resonance imaging of HD patients has identified microstructural abnormalities in the white matter tracts which surround the striatum, indicating myelin breakdown 37 . These results support the hypothesis that CE may also be increased in HD striatum due to increased demand in cholesterol for the repair of myelin and/or increased available cholesterol from degenerating myelin. Esterification may, therefore, be stimulated to avoid cytotoxic damage. The authors acknowledge that CE represent a significantly small portion of the total cholesterol pool in the brain (globally < 0.1%) 38 , which seemingly precludes their ability to have a profound effect on cholesterol metabolism. However, a previous study in AD transgenic mice demonstrated that depletion of ACAT1 activity significantly increases levels of 24-OHC and cholesterol synthesis, demonstrating a potentially noteworthy impact on other metabolites 26 . Comparison of CE levels in this study with 24-OHC in the same human tissue 7 reveals that control CE tissue concentrations are approximately 12% and 15% of their respective 24-OHC concentrations in the putamen and caudate. In cerebellum CE levels are equivalent to 24-OHC, which means that alterations in its abundance could affect the concentrations of other metabolites in the cholesterol pathway.
Our study also revealed specific alterations in CE between striatal subregions, with HD caudate showing a larger increase in CE than HD putamen (Fig. 2). This was unexpected since previous analysis showed no increase in cholesterol despite reduced removal via 24-OHC in HD caudate 7 . The striatum atrophies earliest of the brain regions in HD, however the caudate and putamen atrophy at different rates [39][40][41] , and correlate with different disease indices i.e. caudate with age of onset, putamen with disease severity 42 This may contribute to the observed differences in the magnitude of lipid disturbance and in this instance, CE accumulation ( Fig. 2) 2,7,18,42 . It is also possible that these regions represent different stages of cholesterol disturbance in HD, and that the caudate may have a greater capacity to maintain cholesterol balance than the putamen, possibly due to its earlier and slower degeneration in the disease 7,42 . Although a difference could not be detected between the total CE concentrations in HD and control cerebellum, the decreased abundance of several CE species is in opposition to the differences in HD striatum. Our subjects showed high variability in cerebellar lipid profiles, which has been identified in other HD studies along with variances in cerebellar degeneration leading to difficulty in statistical analysis 43 . Previous research has identified the cerebellum as an area of little cholesterol disturbance in HD, perhaps attributable to a later disturbance for this region in adult onset HD 44 . Post mortem analysis identified that reductions in cerebellar Purkinje cells were only detected in HD patients with a predominant motor phenotype, and not in mood phenotypes, indicating that the clinical expression of the patient may indicate disturbance of the cerebellum 45 . Additionally, the white matter of the cerebellum degenerates to a greater extent than the grey. Our analysis of the lipid profile of the cerebellum is consistent with that of the other grey matter regions (caudate, putamen and grey cortex) and less aligned with that of white matter regions (white cortex) indicating that the cerebellar tissue in this study was grey matter (unpublished data). We cannot therefore comment on disturbances to cholesteryl esters in the white matter of the cerebellum.
Due to the advanced stage of HD within our subjects, it is unknown if esterification of cholesterol was able to maintain homeostasis earlier on in the disease within the striatum. Previous studies have identified the early reduction of 24-OHC in HD 18,21 . This may be relevant to the maintained steady-state levels of cholesterol identified in previous HD studies despite disturbances to cholesterol metabolism. It has been hypothesized that cholesterol accumulation alters membrane organisation and cell signalling in HD striatal neurons 12,14,46 . Consideration should be made for the possible influence of medications on the brain chemistry of the subjects in this study. Some antidepressants and antipsychotics can influence the cholesterol metabolic pathway by upregulating genes that are modulated by sterol regulatory element binding proteins 47,48 . However, studies examining the effects of these drugs observed different effects on lipid metabolism dependant on the drug itself and the brain region. For example, a study looking at the effect on neural lipids after antidepressant treatment in rats found changes specific to the frontal cortex and no reported changes to lipids in the striatum or cerebellum 49 . Additionally, due to the variations in drugs prescribed among patients and the absence of clinical information regarding the duration and dosage of these drugs, analysis of their influence on our results is challenging.
This study was the first to analyse CE in human HD post-mortem tissue. Our analyses have uncovered striatal subregion specific accumulations of CE in HD. The striatal subregion specific alterations of CE in HD patients suggests an alternative mechanism of sequestration of excess cholesterol into CE, however, since no difference in the expression of ACAT1 was identified this may be a consequence of increased ACAT1 activity. This discovery highlights not only the unique disturbances of CE in the caudate and putamen but also provides further insight into the alteration of specific lipid pathways in HD pathology, which may help us to uncover the nature and specific mechanism of mhtt's interaction with neural lipids.

Materials and methods
Human Post-Mortem Brain Tissue. The Victorian Brain Bank (VBB) provided human post-mortem tissue from the left hemisphere of 13 advanced stage (IV) HD subjects and 13 age and sex-matched controls. Clinical disease stage was measured using the Unified Huntington's Disease Rating Scale. All HD cases had a pathological Vonsattel grading of IV as deemed by the Victorian Brain Bank. Tissue was selected from the caudate,

Scientific Reports
| (2020) 10:20314 | https://doi.org/10.1038/s41598-020-76973-8 www.nature.com/scientificreports/ putamen and cerebellum of each subject and stored at − 80 °C until use. The cerebellum was included as a comparison tissue as it is one of the least disturbed brain regions in adult-onset HD, both in terms of degeneration and its low degree of lipid disturbance 7,43 . The mean age, post mortem interval and brain pH between controls and HD subjects were not different ( Mass spectrometry. Nanoelectrospray ionization mass spectrometry of lipid extracts was performed using a hybrid triple quadrupole linear ion trap mass spectrometer (QTRAP 5500, Sciex, Concord, Canada), equipped with an automated chip-based nanoelectrospray source (TriVersa Nanomate, Advion Biosciences, Ithaca, USA) as described previously 52 . Samples were loaded onto a 96-well plate (Eppendorf Twin-Tec 96) and sealed before direct infusion. All lipids were analysed in positive ion mode. Spray parameters were set at a gas pressure of 0.4 psi and a voltage of 1.2 kV 51,53 . Declustering potential was set to 100 V, collision cell exit potential 8 V, entrance potential 10 V and scan rate at 200 Da/s. Lipid data was acquired by targeted precursor ion scans as shown in Supplementary Table S8. Target lists for molecular lipid species within each class were generated after the manual review of spectra in Analyst (v1.6, Sciex, Ontario, Canada). Mass spectrometry data were analysed and Statistical analysis. Outliers were identified using a 2.2 interquartile range outlier rule and excluded from further data analysis. Statistical analysis was conducted using GraphPad Prism (V5, GraphPad Software, California USA). Normality of data was assessed using a D' Agostino Pearson Omnibus test. To determine differences in CE molecular species between HD and control ( Fig. 3), data fitting normality assumptions were analysed by an unpaired t-test with Welch's correction, while data that did not fit normality assumptions were transformed or analysed by Mann-Whitney U test if the transformation was unsuccessful. A Two-Way ANOVA with a Tukey's test was used to determine regional differences between HD and control, in total CE and ACAT1. p values less than 0.05 were considered statistically significant. Data with p < 0.01 are also indicated. Detailed statistical information is supplied in Supplementary Tables S2, S3 and S4. Data is presented as mean ± standard error of the mean (SEM). Spearman's correlation analysis was used to examine relationships between the biometric and total CE concentrations in SPSS (v25, IBM, USA).

Ethics approval. This research was granted ethics approval by the University of Wollongong Human
Research Ethics Committee (HE10/327) and was carried out in accordance with the Declaration of Helsinki (2008).