Genetic deletion of S6k1 does not rescue the phenotypic deficits observed in the R6/2 mouse model of Huntington’s disease

Huntington’s disease (HD) is a fatal inherited autosomal dominant neurodegenerative disorder caused by an expansion in the number of CAG trinucleotide repeats in the huntingtin gene. The disease is characterized by motor, behavioural and cognitive symptoms for which at present there are no disease altering treatments. It has been shown that manipulating the mTOR (mammalian target of rapamycin) pathway using rapamycin or its analogue CCI-779 can improve the cellular and behavioural phenotypes of HD models. Ribosomal protein S6 kinase 1 (S6K1) is a major downstream signalling molecule of mTOR, and its activity is reduced by rapamycin suggesting that deregulation of S6K1 activity may be beneficial in HD. Furthermore, S6k1 knockout mice have increased lifespan and improvement in age-related phenotypes. To evalute the potential benefit of S6k1 loss on HD-related phenotypes, we crossed the R6/2 HD model with the long-lived S6k1 knockout mouse line. We found that S6k1 knockout does not ameliorate behavioural or physiological phenotypes in the R6/2 mouse model. Additionally, no improvements were seen in brain mass reduction or mutant huntingtin protein aggregate levels. Therefore, these results suggest that while a reduction in S6K1 signalling has beneficial effects on ageing it is unlikely to be a therapeutic strategy for HD patients.

A number of studies have demonstrated that inhibition of the mTOR pathway attenuated the pathological effects induced by mHTT. Rapamycin, an inhibitor of the key nutrient signal integrating protein mTORC, attenuated mHTT accumulation and cell death in cell culture models of HD, and also protected against degeneration of photoreceptor neurons in a fly overexpressing 120-CAG repeat huntingtin in the eye 24 . Furthermore, the rapamycin analogue CCI-779 improved rotarod performance and grip strength of the Ross/Borchelt HD mouse model, which has a late disease onset 24 . Everolimus, which binds with high affinity to FKB12 and thereby inhibits mTOR, decreased the phosphorylation of the mTOR kinase target protein S6 kinase and delayed the decline in motor coordination, as well as reducing the levels of soluble mHTT in the skeletal muscle 25 . Moreover, mHTT enhanced mTORC1 activity, which in turn is thought to contribute to the pathogenesis of HD. These studies therefore indicate that manipulation of the mTOR signalling pathway could be of benefit in the treatment of HD.
S6 protein kinase 1 (S6K1) is a key downstream target of mTORC1 and its activity is reduced by rapamycin via the latter's effects on mTORC1. We have previously shown that genetic knockout of S6k1 in mice leads to an increase in life span and resistance to age-related pathologies 26 . Previous studies in C. elegans have shown that polyglutamine aggregate accumulation and onset of toxicity in muscle is delayed in long-lived insulin/IGF-1-like pathway mutants 27 . Furthermore, a recent study showed that intercrossing heterozygous insulin like-growth factor receptor 1 (Igf-1r) knockout mice, which have been reported to be long-lived, with female N171-82Q HD mice delayed tremor onset in this HD model 28 . This finding suggests that the rate of progression of HD may be linked with the genetic regulation of aging. Based on this, and the studies showing that treatment with rapamycin and its analogues can ameliorate HD pathogenesis, we hypothesized that loss of S6k1 may alleviate the symptoms observed in the R6/2 mouse model of HD.
To address this question, we generated R6/2 mice lacking S6k1 (R6/2 × S6k1 −/− ) by crossing these animals with global S6k1 knockout mice and measured body weight, locomotor activity, rotarod performance, forelimb strength, blood glucose and insulin levels, brain weight and mHTT aggregate load. We showed that genetic knockdown of S6k1 had no beneficial effect on the levels of aggregated mHTT or on any of the behavioural or physiological deficits observed in the R6/2 mouse model, indicating that inhibition of S6K1 is unlikely to be of benefit in the treatment of HD.
Deletion of S6k1 has no beneficial effect on the body weight loss observed in the R6/2 mouse model. R6/2 mice maintain normal body weight during the first part of their lives, but gradually lose weight after the onset of the disease, a phenotype that resembles the low BMI reported for HD patients 5 . In this study, male and female mice were weighed weekly from 4 to 20 weeks of age ( Fig. 1a,b; Table 1). Consistent with previous findings, both male and female R6/2 mice weighed significantly less overall than their control littermates (males: F 1,33 = 27.356, P < 0.001; females: F 1,31 = 19.877, P < 0.001), and had significant weight loss over time (males: F 16,528 = 52.465, P < 0.001; females: F 16,496 = 9.954, P < 0.001). As anticipated, S6k1 −/− mice were dwarves and so were also significantly smaller than their WT littermates (males: F 1,33 = 119.521, P < 0.001; females: F 1,31 = 113.564, P < 0.001). Genetic knockout of S6k1 in R6/2 mice did not attenuate the rate of weight loss as the R6/2 × S6k1 −/− mice began to progressively lose weight at a similar age, and rate, as the R6/2 mice (males: F 16,528 = 3.185, P = 0.036; females: F 16,496 = 2.976, P = 0.015). Therefore, S6k1 deletion in the R6/2 mouse model did not improve the weight loss phenotype.

Discussion
Previous studies, in both cells and model organisms, have suggested that attenuation of mTORC1 signalling may provide therapeutic benefit in the treatment of HD. For example, rapamycin treatment of Drosophila expressing mHTT in their photoreceptors reduced neurodegeneration in the eye 24 . Likewise, in the Ross/Borchelt mouse model of HD, treatment with the rapamycin ester CCI-779 improved a range of neurological parameters including rotarod performance and grip strength, and reduced HTT aggregates in the brain 24 . The underlying mechanisms for this beneficial effect have been suggested to involve the induction of autophagy upon inhibition of mTOR and subsequent clearance of polyglutamine aggregates. Consistent with this hypothesis it has also been shown that HTT itself promotes mTORC1 signalling and increased mTOR activity in the striatum of N171-82QHD mice exacerbates neurological phenotypes 31 . Thus, increased activity of the mTORC1 may in turn contribute to HD pathogenesis, providing further rationale for therapeutically targeting this pathway. www.nature.com/scientificreports www.nature.com/scientificreports/ Figure 3. Deletion of S6k1 does not reverse the brain weight reduction or the mHTT aggregation observed in R6/2 mice. (a) Male and (b) female mouse brain weight was measured at 14 and 20 weeks. As the S6k1 −/− mice have smaller brains compared to WT mice data are represented as % of appropriate control, R6/2 to WT and R6/2 × S6k1 −/− to S6k1 −/− . Deletion of S6k1 did not modify the brain weight loss in R6/2 mice as the % reduction in brain weight was equivalent for R6/2 and R6/2 × S6k1 −/− mice. For 14 weeks, males n = 4-9/genotype and females n = 4-5/genotype and for 20 weeks, males n = 12-14/genotype and females n = 14/genotype. Data were analysed by GLM univariate with Bonferroni correction. ns = non-significant and P > 0.999 for all. (c-f) The Seprion ligand based ELISA assay was used to quantify mHTT aggregation in the striatum, cortex and brain stem of (a, c and e) male and (b, d and f) female R6/2 and R6/2 × S6k1 −/− mice. Aggregate levels increase with age but are not modified by S6k1 deletion. For 14 weeks, males n = 4/genotype and females n = 4-6/genotype and for 20 weeks, males n = 4-7/genotype and females n = 4-8/genotype. Data were analysed with GLM univariate with Bonferroni correction and are presented as mean ± SEM. *P <0.05, **P < 0.01, ***P < 0.005 and ****P < 0.001. www.nature.com/scientificreports www.nature.com/scientificreports/ S6K1 is a major downstream effector of mTORC1 signalling. A broad range of studies have implicated S6K1 in a number of cellular processes including protein translation, ribosomal biogenesis, mRNA splicing, transcription, cytoskeletal organization, regulation of cell signalling intermediates and cell proliferation [32][33][34][35] . At an organismal level S6K1 signalling has been shown to control growth and metabolism including glucose homeostasis and insulin sensitivity. Importantly, S6K1 has been shown to control the ageing process as mice lacking S6k1 globally are long-lived and resistant to the diseases of ageing 26 . Moreover, tissue-specific loss of S6k1 reduces the accelerated pathology and physiological decline associated with loss of lamin A 36 . Furthermore, and of relevance to HD, genetic reduction of S6k1 has been shown to reduce neurological defects and neuropathology in the triple transgenic mouse model of Alzheimer's disease 37 . Taken together these findings suggest a connection between S6K1, neurodegeneration and ageing, and motivated our studies to investigate whether loss of S6k1 may ameliorate the disease course in the R6/2 mouse model of HD.

Scientific RepoRtS
In our studies, we found that global deletion of S6k1 did not attenuate the features of HD in the R6/2 mouse model. Abrogation of S6k1 had no effects on aggregated mHTT load in the striatum, cortex or brainstem. This was reflected in the failure to see an improvement in a range of behavioural parameters such as locomotor activity, rotarod performance and grip strength. This lack of benefit was seen in both male and female mice. S6k1 −/− mice display growth retardation which in part complicated the interpretation of body weight profiles in the R6/2 × S6k1 −/− intercross but again for this parameter there was no apparent benefit in male animals.
R6/2 mice display abnormalities of glucose homeostasis with elevated fasting blood glucose levels, reduced insulin levels and impaired glucose tolerance 19,20 . S6k1 −/− mice have increased insulin sensitivity and improved glucose homeostasis under conditions of nutrient excess and in old age 26,38 . It therefore might have been anticipated that reduction of S6k1 might show improvement in glucose handling. However, fed and fasting blood glucose levels and glucose tolerance showed no improvements in both R6/2 × S6k1 +/− and R6/2 × S6k1 −/− mice. Therefore, taken together it is clear that global genetic reduction of S6k1 does not improve the neurological features of R6/2 mice or result in any metabolic benefits.
There are a number of potential explanations for the failure to observe improvements in R6/2 HD phenotypes. The beneficial effects of rapamycin-induced mTORC1 inhibition may be due to the more general effects upon the pathway than the specific deletion of just one of its downstream effectors. It appears that the induction of autophagy by rapamycin is key to its effects upon HTT aggregate clearance 39 . While S6K1 has many molecular targets and cellular effects the induction of autophagy is not a prominent and consistent finding with deletion of S6k1. Furthermore, rapamycin inhibits the activation both S6K1 and S6K2 and the beneficial effects may require suppression of both these kinases. Therefore, it is conceivable that simply deleting S6k1 alone does not capture the underlying signalling and cellular events required to see a therapeutic benefit. Consistent with this concept, deletion of Igf1r, an upstream activator of S6K1 signalling, had some benefits in N171-82Q HD mice, but this manipulation was not associated with a reduction in the phosphorylation p70S6k, suggesting that suppression of S6K1 activity was not mediating these effects 28 . It should also be noted that this study demonstrates both beneficial and detrimental effects on HD pathology indicating complexity in the signalling pathways. The beneficial effects of S6k1 deletion upon AD pathology probably reflects the different aetiopathophysiology of HD and AD although both are aggregation-associated neurological diseases.
An additional reason for the lack of a beneficial effect could be the choice of HD model used. In the original description of the beneficial effects of rapamycin on HD, the Ross/Borchelt mouse model was used 24,40 , which has a slightly longer life expectancy and slight delay in some behavioural features compared to the R6/2 model. Although our manipulation of S6k1 occurred in the germ-line the more progressive R6/2 phenotype could have overcome any potential beneficial effects of the loss of S6k1. Indeed, treatment of R6/2 mice with everolimus, which acts with a very similar mechanism to rapamycin but with more mTORC1 selectivity, has subsequently been reported to not reduce brain HTT aggregates or have beneficial neuroprotective effects but did alter skeletal muscle HTT load. Interestingly, in this study everolimus suppressed brain rpS6 phosphorylation suggesting inhibition of S6K1 activity but did not induce autophagy 25 .
In summary while abrogation of S6K1 signalling has beneficial effects on ageing and age-related disease including attenuation of the pathology seen in AD, this therapeutic effect does not extend to the amelioration of the HD phenotype in the R6/2 mouse model of HD.
Body weight. Body weight was measured at the same time of the day each week from 4 weeks of age until 20 weeks using a Sartorius BP610 balance. The same cohort of mice was used for the body weight curves and all the behavioural tests. At the start of the study the n numbers for the males was: WT, n = 9, R6/2, n = 9, S6k1 −/− , n = 11 and R6/2 × S6k1 −/− . n = 9, and for the females was: WT, n = 9, R6/2, n = 11, S6k1 −/− , n = 8 and R6/2 × S6k1 −/− . n = 9. One male and 2 female R6/2 × S6k1 −/− mice had to be culled before termination of the study due to their weight loss reaching that of the limits of the Home Office PPL.
Open field. Locomotor activity was assessed using the HVS Image tracking software (HVS Image 2100, Hampton, UK). Mice were individually placed in a wooden grey arena (45 × 45 × 30 cm) that had a base covered in sawdust. Each mouse was released into a corner of the box and was allowed to explore for 5 min. The tracking system recorded distance travelled. Testing began at 6 weeks of age and mice were tested every 2 weeks until they reached 20 weeks of age.
Rotarod. The rotarod apparatus (Ugo Basile, Italy) was used to measure fore-and hind-limb motor coordination, balance and strength. Mice performed three trials per day with an inter-trial interval of 1 h for three consecutive days. The rod accelerated from 5 to 60 rpm over a period of 570 s and the latency to fall was recorded. For the subsequent analysis the mean latency was taken for the 9 trials and plotted as a value for the week. Testing began at 6 weeks of age and mice were tested every 2 weeks until they reached 20 weeks of age.

Grip strength.
A grip strength meter was used to measure forelimb strength (Linton Instruments, UK). To measure grip strength the mouse was swung gently by the tail so that its forelimbs contacted the grip bar. The mouse instinctively gripped the bar and was then pulled horizontally backwards, exerting a tension until the mouse released the bar. The maximum load was recorded by the grip strength meter, which was then returned to zero before the next test. The mouse was placed back in its home cage for a minute to rest before the next test. Each mouse performed five consecutive tests, and the three highest scores were used for statistical analysis. This protocol is preferred because lower scores are often due to the mouse failing to grip the bar effectively, rather than reflecting muscular strength. Testing began at 6 weeks of age and mice were tested every 2 weeks until they reached 20 weeks of age. All scores were normalized to body weight. Metabolic analysis. Fed and fasted blood samples were collected from mice via tail vein or trunk bleeds using a capillary blood collection system (Sarstedt, Nümbrecht) and blood glucose was measured using a Contour glucometer (Bayer Corp). Fasted blood samples were collected after a 16 h overnight fast. Plasma insulin levels were analysed using a mouse insulin ELISA (Crystalchem Inc.).

Seprion ligand ELISA.
Brains from all genotypes were sub-dissected into specific brain regions, snap-froze in liquid nitrogen and then stored at −80 °C until required. A 2.5% lysate (w/v) was prepared in ice-cold RIPA buffer (50 mM Tris-HC pH 8.0; 120 mM NaCl; 1% Igepal; 3.125% sodium deoxycholate; 0.01% SDS; 1 mM β-mercaptoethanol; 1 mM PMSF; 1 mM DTT; protease inhibitor cocktail; Roche) by ribolysing for 30 s in Lysing matrix tubes (Lysing matrix D; MP Biomedicals). Samples were stored on ice for 5 min and used immediately. Homogenate (15 ml) was mixed with 3 ml 10% SDS, diluted to 80 ml with water, and then made up to 100 ml with 5x capture buffer (Microsens Biotechnologies). For detection of aggregated protein, lysate was transferred to the well of a Seprion ligand-coated ELISA plate, and incubated with shaking for 1 h at room temperature (RT). After removal of the lysate, the well was washed 5 times in PBS-T (PBS; 0.1% Tween) and 100 ml S830 primary anti-body (diluted 1:2000 in conjugate buffer (150 mM NaCl; 4% BSA (98% electrophoretic grade); 1% non-fat dried milk; 0.1% Tween 20 in PBS) was added and incubated with shaking for 1 h at RT. After five washes with PBS-T, 100 ml horseradish peroxidase-conjugated rabbit anti-goat secondary antibody (DAKO) (1:2000 in conjugation buffer) was added and incubated with shaking for 45 min at RT. After washing five times with PBS-T, 100 ml of TMB substrate (SerTec) at RT was added and incubated in the dark (wrapped in foil) at RT for 30 min. Reactions were terminated by the addition of 100 ml 0.5 M HCl and the absorption at 450 nm was measured using a plate reader (BioRad). For analysis, the WT mice absorption values were subtracted from those of the R6/2, R6/2 × S6k1 +/− and R6/2 × S6k1 −/− mice. www.nature.com/scientificreports www.nature.com/scientificreports/ Statistical analysis. Statistical analyses were performed using SPSS (IBM) or Prism 7 (GraphPad). For all longitudinal physiological and behavioural tests to determine the effect of either genotype or treatment over the course of the study General Linear Model repeated measures ANOVA was used with genotype (R6/2), treatment (S6k1 −/− ) and age as within-subject factors, with the Greenhouse-Geisser correction for non-sphericity. For all other studies, General Linear Model univariate and ANOVA analysis with Bonferroni correction, or the Kruskall Wallis test if data were not normally distributed. Normality of distribution was assessed using the Shapiro-Wilk test. Data are expressed as mean ± SEM and statistical significance is defined as p < 0.05.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.