The study was designed to screen for early sex- and genotype-dependent differences in the congenic F344tgHD rat model for Huntington’s disease. Three experimental cohorts (A-C) of male and female, wt, het and hom animals of the F344tgHD rat model were investigated using a test battery covering motor and emotional as well as ethological behaviors and metabolism monitored within an automated home cage-like test system (PhenoMaster System, TSE) (Figure 1A). Three Cohorts were investigated: Cohort A was tested at young (1–3 Months), middle (3–5 Months), and adult age (5–8 Months). Cohort B was tested at young and middle age and Cohort C only at young age. Each Cohort was investigated with the home cage-like Phenomaster system after the last experimental round and then sacrificed for brain extraction. The order of behavioral experiments per test phase was strictly defined and the exact age of the animals at each experiment follows the order within each test phase. The exact number of animals used per test varies due to technical reasons and is summarized in the supplements (Supplementary Table 1) or can be found in the respective graph in the results section. Brains were collected after the last experimental trial at an age of 3, 5, and 8 months and investigated with immunohistochemistry as well as stereological and molecular biological approaches (n = 5 per group, unless stated otherwise).

All behavioral experiments were conducted during the light phase of the light: dark cycle (7 to 2 h prior to the dark phase). The automated home cage phenotyping lasted 72 h comprising three full light: dark cycles.

Male and female animals were tested on the same day using the same equipment to minimize variance among animals. To prevent olfactory disturbances, equipment was cleaned with ethanol after every animal.

Data was analyzed with a 3-way analysis of variance (ANOVA) (factor “genotype,” “sex,” “age”) to verify an age effect. Afterward, data was split and differences in the mean values between groups were compared within each age-point using a 2-way ANOVA with two between-subject factors (“genotype” and “sex”). Either Tukey’s (factor: genotype) or Šídák (factor: sex) correction was used for post hoc multiple comparison.

Statistical analysis of body weight was performed using a 3-way ANOVA for repeated measures (factor “time”) with two between-subject factors “genotype” and “sex.” Statistical analysis was conducted separately for time periods 6–8, 9–18, 20–30 weeks of age to correct for different group sizes due to sample collection. A total of 2-way ANOVAs were performed with Prism9.3 (GraphPad Software, San Diego, CA, USA). A total of 3-way ANOVAs were performed with Jamovi open software (v2.3.18).

For all analysis, a statistically significant difference between groups was defined as *p < 0.05, **p < 0.02, ***p < 0.001, and ****p < 0.0001. Genotype-differences are indicated by asterisks, sex-differences by hashtags and interaction by paragraphs. Further statistical information comprises F_{(DFn, DFd)}. All data are presented as mean ± SEM.

Longitudinal observation of F344tgHD rats did not show apparent differences in general health between sexes, nor genotypes. No anomalies, like tremor, staggering gait, seizures (neither spontaneous nor handling-induced), ataxia or obvious tumor developments were observed within the experimental period of 8 months. Adult transgenic females demonstrated opisthotonos-like movements of the head observed during the open field test. Detailed evaluation is described in the corresponding results section. Determination of steroid hormones (17ß-estradiol, progesterone and testosterone) in plasma revealed physiological level for both sexes without any apparent difference between genotypes (Supplementary Figure 1). All animals continuously gained weight over time reaching a stable weight at adulthood independent of their genetic condition (Figure 1B) [5–8 weeks: F_(3,580) = 929.867, p < 0.001; 9–18 weeks: F_(7,624) = 2182.255, p < 0.001; 20–30 weeks: F_(5,367) = 251.43, p < 0.001]. No difference in weight was detected at any age point due to the animals’ genotype. A strong but physiological difference between males and females was present from 5 weeks on [5–8 weeks: F_(1,582) = 136.522, p < 0.001; 9–18 weeks: F_(1,630) = 1664.423, p < 0.001; 20–30 weeks: F_(1,372) = 1002.193, p < 0.001].

Breeding via het × het mating of F344tgHD animals produced litter sizes of 6.98 ± 2.48 pups. A total of 50 breeding events resulted in 349 living descendants of which 27.51 % were wt, 53.30 % het and 19.19 % hom animals (Figure 1C). The genotype ratio in all offspring was 1:1.94:0.68 (wt:het:hom) with a significant difference between sex [p = 0.001, F_{(2, 182)} = 7.201]. This discrepancy between theoretical (1:2:1) and actual Mendelian ratio manifested especially in female animals, where the ratio was 1:2.02:0.51, while 1:1.84:0.91 in the male litter. A total of 49 pups died before genotyping due to cannibalism or other unknown factors.

Assessment of anxiety levels were conducted with the social interaction test. The average time per active interaction between each rat pair was evaluated (Figure 3D). A clear trend toward an age effect was measured [F_(2,89) = 2.763, p = 0.069]. No significant difference between sexes could be measured at any investigated age point, although a trend for increased female interaction was present at middle age [F_(1,40) = 3.518, p = 0.068]. A significant genotype effect was detected at young and middle age [young: F_(2,28) = 4,720, p = 0.0171; middle age: F_(2,40) = 4,04, p = 0.0252] with an indicative gene-dosage dependency at young age. Post hoc comparison revealed increased socio-positive interaction in tg animals (young: wt vs. hom p = 0.0193; middle age: het vs. hom p = 0.0316).

The open field test takes advantage of the animal’s subjective conflict between the drive to explore and aversion to the exposure toward a brightly lit open arena, hence the assay provides information on spontaneous locomotion and anxiety-related parameters. All animals spent the majority of the test period close to the peripheral walls of the arena (thigmotaxis) and displayed sufficient exploratory behavior (Supplementary Figure 5A).

The time spent in the center of the arena (Figure 3E) showed a trend for an increase over time [F_(2,259) = 2.609, p = 0.076]. At young and middle age, animals spent equal time in the center of the arena regardless of their sex or genotype. At adulthood, females stayed more often in the arena’s center than males [F_(1,42) = 5.767, p = 0.0208] with hom females spending the most time in the center and significantly differing from hom males (p = 0.0078). Time spent in the center of the arena was additionally calculated as percentage of the total time spent in locomotion (Supplementary Figure 5B), which supports the findings described above, though not reaching statistical significancy.

Analysis of grooming behavior (Figure 3F) confirmed a significant age effect with a decrease over time [F_(2,236) = 25.618, p < 0.001]. Significant interaction between factors “sex” and “genotype” was present at young age [F_{(2, 101)} = 5.063, p = 0.008] with significant less grooming in het compared to wt males (p = 0.0164). Animals of adult age displayed significant sex- and genotype-dependent overall effects [sex: F_(1,47) = 5.470, p = 0.0237; genotype: F_(2,47) = 3.439, p = 0.0404] with males grooming more than females and hom animals more than het and wt littermates.

Rearing showed significant increase over time [F_(2,238) = 18.512, p < 0.0001] (Figure 3G). Beginning with middle age, females spent significantly more time rearing than males [middle age: p = 0.0013, F_(1,78) = 11.21; adult: p = 0.0145; F_(1,46) = 6.458] with significant post hoc between homozygous animals (middle age: p = 0.0429; adult: p = 0.0420).

Hom females showed ophistotonus-like head movements at the oldest time point investigated, as described for adult SPRDtgHD rats (Cao et al., 2006). These choreiform movements were detected in 6 out of 10 hom females with 5.5 ± 0.5627 events per 15 min observation time (Supplementary Figure 6). No such anomalies were seen in het females, nor male F344tgHD animals.

Detection of ethological behaviors, their circadian patterns, as well as corresponding energy expenditure (derived from indirect calorimetry) were performed in a home cage-like environment within the PhenoMaster System (TSE Systems GmbH, Germany). All animals’ behaviors displayed typical circadian rhythmicity (data not shown).

Energy expenditure (EE) and respiratory exchange rate (RER) were measured only at middle and adult age due to technical issues. Both parameters showed an age-dependent decrease [EE: F_(2,85) = 23.934, p < 0.001, RER: F_(2,88) = 12.570, p < 0.001]. The RER gives information about the main energy source with values around 0.8 for a fat-based and >0.8 for carbohydrate-based metabolism. All groups displayed fat-based calorimetry with values around 0.8. RER was significantly lower within the female group at both age points [middle age: F_(1,41) = 9.489, p = 0.0037; adult: F_(1,38) = 9.770, p = 0.0034] without significant post hoc comparison (Figure 4A).

In contrast, male animals demonstrated significantly lower EE than females at middle age and adulthood [middle age: F_(1,42) = 8.405, p = 0.0059; adult: F_(1,36) = 136.8, p < 0.0001] with no significant post hoc at middle age but within all genotypes at adult age (wt: p < 0.0001, het: p < 0.0001, hom: p < 0.0001) (Figure 4B). No genotype effect could be shown for EE or RER.

Investigation of locomotion (Figure 4C) in x- and y-axis revealed a significant age effect [F_(2,141) = 3.198, p = 0.044] and higher exploratory activity in females throughout testing [young: F_(1,31) = 4.383, p < 0.0446; middle age: F_(1,58) = 52,18, p < 0.0001; adult: F_(1,37) = 33.96, p < 0.0001]. Significant post hoc comparison was present between all genotypes at middle age (wt p = 0.0004; het p < 0.0001; hom p = 0.0110) and between transgenic animals at adult age (het p = 0.0446; hom p < 0.0001). Genotype differences were additionally observable at middle age [p < 0.0253, F_(2,58) = 3.919] with increased activity in het females compared to controls (p = 0.0196). An interaction between factors genotype and sex could be shown in the adult group [F_(2,37) = 5.855, p = 0.0062] with significant higher interaction of hom females compared to wt (p = 0.0107) and het (p = 0.00111) animals. Additionally, a trend for increased locomotion of transgenic animals was detected at adult age [F_(2,37) = 2.918, p = 0.0666].

Neurohistological examination of F344tgHD revealed first polyglutamine-containing aggregates at the age of 8 months (Figures 5C, D). Aggregates were detectable in medial forebrain nuclei of hom animals. No difference in the appearance of HTT aggregates was observed between sexes. Brains derived from wt and het animals did not show 1C2-antibody-specific immunoreactivity in the investigated coronal section (Figures 5A, B).

Western blot analyses of total HTT were performed on striatum enriched tissue to ensure a gene-dosage effect in protein levels (heterozygous vs. homozygous) while excluding sexual dimorphism (Figure 5E and Supplementary Figure 7). Full length HTT was increased in young females compared to males [F_(1,24) = 5.972, p = 0.0223], while both sexes showed equal expression at adult age. As anticipated, transgenic (fragmented) HTT was expressed with a significant gene-dose effect at both ages [young: F_(1,16) = 33.90, p < 0.0001; adult: F_(1,16) = 5.718, p = 0.0294] with significant post hoc between het and hom animals in both sexes at young age (young het vs. hom: male p = 0.0017; female p = 0.0015). Additionally, females expressed less transgene at adult age compared to males [F_(1,16) = 0.0046, p = 0.0046] with significant post hoc comparison in hom animals (p = 0.0118).

Stereological quantification of the number of DARPP32-positive cells in the striatum was performed in a subset of 8 months old F344tgHD rats (n = 3/group) (representative staining in Figures 6A–C) and revealed no significant difference in the absolute number of neurons (Figure 6D). Yet, looking at the relative numbers, a reduction of around 26 % in heterozygous and nearly 40% in homozygous males and 30% in homozygous females was evident compared to the respective wt controls.

At the same time, no pathological change in striatal volume was detected yet (Figure 6E), and males exhibited significantly larger portions of striatum than female rats [F_(1,12) = 7.915, p = 0.0156].

The F344tgHD rat line is of general health and does not develop any impairments that could affect its behavioral characterization at early age. In line with published data on older, male F344tgHD rats and the previous SPRDtgHD model (Kirch et al., 2013; Urbach et al., 2014; Plank et al., 2018) neither female nor male transgenic animals showed significant differences in weight compared to control animals. As expected, females had physiologically lower body weights beginning already at 4 weeks of age. Intra-home cage ethophysiological analysis using an automated PhenoMaster system revealed no difference in food ingestion when corrected for body size. However, the increased energy expenditure seen in females might contribute to the observed weight difference. A significant connection between these parameters has been reported in different mouse strains (König et al., 2020).

Analysis of the genotype ratio among F344tgHD descendants revealed slight deviation from the Mendelian distribution known for a monogenetic, autosomal dominant disease, like HD. The discrepancy manifested especially in female animals with decreased numbers of hom descendants. Although these deviations are not in a magnitude calling for major impairment of embryonic development in the F344tgHD rat model, our data might indicate pre-natal decease of preferentially hom animals due to yet unknown factors and we consider them as important information for future biostatistical planning of studies.

Investigation of HD neuropathological hallmarks was performed in the striatum of F344tgHD rats to examine their occurrence in relation to behavioral abnormalities over time. Aggregates containing mHTT were detectable at 8 but not 5 months of age in hom F344tgHD rats. In the SPRDtgHD model, first aggregates were already seen at 6 months (Nguyen et al., 2006), indicating a slower disease progression in the congenic F344tgHD model.

Male F344tgHD rats showed pronounced striatal cell loss with a gene-dosage effect, i.e., smaller cell numbers in hom compared to het animals. While hom females also showed a decrease in the number of DARPP32^±cells compared to wt controls, het females demonstrated an unexpected increase (+10 %). Besides the underlying zygosity effect, compensatory mechanisms, e.g., via 17ß-estradiol (Bode et al., 2008) might be causal for this deviation.

In general, females showed less distinct striatal cell degeneration compared to the respective male group. This sex-effect might be attributed to the physiological higher levels of 17ß-estradiol and their neuroprotective effect, studied intensely in neurodegenerative diseases (reviewed: Dhandapani and Brann, 2002). In the SPRDtgHD rat model, Bode et al. (2008) showed a direct correlation between 17ß-estradiol level and number of DARPP32^±cells supporting the described findings in F344tgHD rats.

The possible reasons for the measurable cell death in het males without the detection of aggregates in the rostral part of the striatum, and the relatively low number of aggregates in hom animals, comprise a variety of hypothesis: First, aggregates were shown to be unequally distributed throughout the brain, with more aggregates in the neocortex compared to the striatum and a gradient within the later toward more aggregates in caudal direction in human patients (Kuemmerle et al., 1999). Second, current literature suggests that it is rather the structure of mHTT aggregates that predicts cell death than their abundance. A study in R6/2 mice, e.g., showed that loop-structured aggregates in the striatum have highly neurotoxic properties, while ß-sheet containing aggregates in the cerebellum are only mildly toxic (Nekooki-Machida et al., 2009). Third, various studies report a neuroprotective effect by the formation of aggregates, sequestering mHTT in the cell (reviewed: Bäuerlein et al., 2020).

Socio-positive behavior in pairs of rats with the same genotype and unfamiliar to each other, were measured with the SI test as the time spent in active social interaction in a novel environment (File and Seth, 2003). Transgenic males demonstrated an early anxiety-reduced phenotype in the form of increased social interaction at young and middle age, confirming previous results (Plank et al., 2018) and matching reports in the SPRDtgHD model (von Hörsten et al., 2003; Nguyen et al., 2006; Bode et al., 2008; Kirch et al., 2013).

Female F344tgHD did not show such clear anxiolytic-like phenotype in the SI test. In the OF test, though, females showed more rearing beginning with middle age, a behavior comprising both an emotional (exploration) and a motoric component (ability to balance on hind paws). Hom females reared more than het littermates, indicative of a gene-dosage effect. They also spent more time in the center of the arena than other animals in this study at adult age, adding to the picture of a less anxious phenotype. It appears that females experience emotional disturbances at older age points and that this domain is affected later than in males.

The impact of intersessional habituation on the corresponding test construct validity is of particular relevance for the interpretation of data derived from the OF test, as “exposure to novelty” as the underlying behavioral construct is gradually lost upon repeated exposure. Thus, loss of novelty of the testing procedure and environment is unavoidable in such experimental designs, if the study is designed longitudinally. In this study, the same animals were repeatedly tested from 1 to 8 months of age, which includes certain habituation effects over time. However, the data presented here do not directly indicate major habituation effects on the OF performance, e.g., no observable decrease in exploratory activity (Supplementary Figure 5A).

A mild impairment in sensorimotor function in form of a decreased ASR was detected in hom F344tgHD males beginning with middle age. Since no hearing deficits were observed and response to acoustic stimuli was normal at young age, basal deficit in sensorimotor gating can be ruled out. Our results correspond with data on other mouse and rat HD models, e.g., BacHD rats (Abada et al., 2013) and R6/2 mice (Carter et al., 1999) who also measured decreased startle reaction in transgenic animals. The additionally observed sexual dimorphism toward higher startle amplitudes in male animals matches data published on other strains, e.g., Wistar (Lehmann, 1999), Long Evans and SPRD rats (Faraday, 2002).

In contrast to published literature, no longitudinal habituation in form of a decreasing startle response over the inspected time course could be observed, stating no influence of the F344tgHD background nor a possible age related hyporeactivity (Gómez-Nieto et al., 2020).

Presence of an acoustic pre-stimulus (PPI) showed transient genotype-associated effects in the ASR feedback-regulation at middle age when normalized to the initial baseline PPI. While wt and het animals showed up to 50% higher PPI at middle age compared to their reaction at young age, hom animals of both sexes remained on the same level. This matches the early anxiolytic-like phenotype of hom F344tgHD animals observed in the SI and OF test. The PPI is exerted by a sensorimotor circuitry via the reticular formation and with participation of several different neurotransmitters. Deficits are often observed in symptomatic HD patients (Swerdlow et al., 1995), while the previous SPRDtgHD rat model showed absence of these (Höhn et al., 2011; Urbach et al., 2014).

The clear sexual dimorphism found in F344tgHD rats regarding the PPI is in line with findings in healthy human subjects, where women show less PPI than men (Swerdlow et al., 1993; Kumari et al., 2003). However, not only sex in general affects PPI measures in humans, but intraindividual changes dependent on the female estrous cycle were detected as well (Swerdlow et al., 1997). To our knowledge, this has not been intensely studied in rodents so far, but our results do not indicate such effect, since no alterations of the PPI were observed between pre-pubescent and pubescent females nor does the magnitude of inter-individual variation suggest such phenomenon.

Although ASR and PPI function via involuntary sensorimotor gating processes, the individual’s emotional status is a known side factor influencing this reflex (Koch and Schnitzler, 1997; Grillon and Baas, 2003). In patients with stress-induced psychological diseases the startle reaction is a commonly examined parameter (Swerdlow, 1994; Kaviani et al., 2004; Ray et al., 2009) and the ASR amplitude can be increased by various stressors in the SPRD rat strain (Acri, 1994; Garrick et al., 1997; Servatius et al., 2000; Faraday, 2002). Hence, the anxiety-reduced phenotype in F344tgHD males observed in the SI approach is further supported by the decrease in ASR amplitude in this study.

Motor coordination, balance, and gait were analyzed using the classical RotaRod test as well as the CatWalk system. Complementary data were obtained with the intra-home cage-like Phenomaster analysis.

On the RotaRod, significantly better performance of all females was detected already at young age. This observation has also been described for Sprague Dawley rats (Bagg et al., 2017) and might be, at least to some extent, explainable by the existing sex dimorphism in body weight. Boakes et al. (1999) showed that activity on a running wheel can be increased in male rats through weight loss. In consistency with the RotaRod data, females were also more active in the OF paradigm as well as in the Phenomaster home cage system, where potential influence through the experimental setup (presence of experimenter, preceded handling, etc.) can be ruled out. The observed activity profile of female individuals also matches the comparatively higher energy expenditure measured in the automated Phenomaster system. The later seems to be physiologically increased in female rodents compared to male littermates (König et al., 2020).

No obvious gene-dosage effects could be detected in RotaRod performance of neither sex, but tg males demonstrated increased coordination skills at young age corresponding with previous data from 2 months old male F344tgHD (Plank et al., 2018) and 1 month old SPRDtgHD rats (Nguyen et al., 2006). “Hypercompensation” through increased neuronal excitability has been discussed as causal, as seen in other motor diseases like ALS, but still needs to be further characterized (Nguyen et al., 2006). The expanded CAG repeat stretch in the Htt gene supposedly constitutes an evolutionary benefit in form of increased fitness to the carrier (Eskenazi et al., 2007; Mühlau et al., 2012; Schultz et al., 2021) which might additionally take effect here.

In addition to the classical RotaRod approach, we addressed the demand made in Plank et al. (2018) to apply an additional, more sensitive method to investigate motor function and gait. The CatWalk assay can measure over 250 parameters and allows a more holistic investigation of locomotion compared to the classical RotaRod approach (Vandeputte et al., 2010; Timotius et al., 2023). Clear differences were detected between both sexes of the F344tgHD model, as already observed with the RotaRod approach.

Increased values for stride length and swing speed measured in females are in line with a general higher activity as described above. Male animals tended to display a higher BOS throughout the experimental phase, an observation also made in Sprague Dawley rats (Bagg et al., 2017) and most probably accountable their higher body weight compared to females.

In regard to genotype-related differences, subtle changes could be detected already at the earliest investigated time point. Hom females displayed increased velocity, swing speed and stride length with larger BOS in hind paws. They also moved significantly more in the PM paradigm and showed opisthotonos head movements in the OF arena. The later has also been described for older SPRDtgHD females (Cao et al., 2006) and is a unique observation in HD rodent models, reminiscent of the choreatic movements in human HD.

Our observations on a hyperkinetic phenotype in female F344tgHD beginning with young age matches data of other HD rat (Nguyen et al., 2006; Vandeputte et al., 2010) and mouse models (Menalled and Chesselet, 2002). In patients, hyperkinetic movements are observable in the early phase of the disease (Delval et al., 2006) with increased stride length in one out of six HD patients (Reynolds et al., 1999).

A higher BOS has also been described for symptomatic HD patients (Rao et al., 2008; Pyo et al., 2017) and might be a compensatory mechanism for the increased instability through choreiform movements (Koller and Trimble, 1985).

Male tg F344tgHD showed less impairment in the motor domain as seen in the Catwalk analysis, besides an increased alteration in the step pattern (regularity index) at 8 months, also matching reports on HD patients (Pyo et al., 2017; Gaßner et al., 2020; Radovanović et al., 2020).