Age-related neuroendocrine, cognitive, and behavioral co-morbidities are promoted by HIV-1 Tat expression in male mice

In the U.S. about half of the HIV-infected individuals are aged 50 and older. In men living with HIV, secondary hypogonadism is common and occurs earlier than in seronegative men, and its prevalence increases with age. While the mechanisms(s) are unknown, the HIV-1 trans-activator of transcription (Tat) protein disrupts neuroendocrine function in mice partly by dysregulating mitochondria and neurosteroidogenesis. We hypothesized that conditional Tat expression in middle-aged male transgenic mice [Tat(+)] would promote age-related comorbidities compared to age-matched controls [Tat(−)]. We expected Tat to alter steroid hormone milieu consistent with behavioral deficits. Middle-aged Tat(+) mice had lower circulating testosterone and progesterone than age-matched controls and greater circulating corticosterone and central allopregnanolone than other groups. Young Tat(+) mice had greater circulating progesterone and estradiol-to-testosterone ratios. Older age or Tat exposure increased anxiety-like behavior (open field; elevated plus-maze), increased cognitive errors (radial arm water maze), and reduced grip strength. Young Tat(+), or middle-aged Tat(−), males had higher mechanical nociceptive thresholds than age-matched counterparts. Steroid levels correlated with behaviors. Thus, Tat may contribute to HIV-accelerated aging.

Herein, we investigate the combined effects of aging and HIV-1 Tat expression on the development of neuroHIV-like sequelae in young adult (6-8 months) and middle-aged (11-13 months) male mice to determine whether Tat precipitates age-related dysfunction. We hypothesized that conditional Tat expression in transgenic male mice [Tat(+)] would accelerate the development of (i.e., catalyze earlier presentation) or accentuate (i.e., increase magnitude) age-related affective, cognitive, neuromuscular, and neuropathic pain symptomatology compared to agematched controls [Tat(-)] ( Figure 1). We further expected that Tat would alter the production of circulating and central steroids (T, E2, P4, corticosterone, or alloP) concurrent with behavioral impairment.
Older age and Tat expression also interacted to influence brain alloP content. We previously observed that Tat expression increased alloP protein content in the whole brain of young male mice, without changes in serum [36]. However, regional analysis of Tat-provoked alloP in the brain has not been conducted heretofore. In hippocampus, older age [F(1,22) = 6.21, p < 0.05; Figure 3A, see †] or Tat expression [F(1,22) = 14.12, p < 0.05; Figure 3A, see *] increased alloP content. In midbrain, age and Tat expression interacted to increase alloP content in middle-aged Tat(+) mice to a greater extent than that seen in all other groups [F(1,24) = 13.71, p < 0.05, Figure 3B, see ‡]. No significant differences were observed in alloP content within the frontal cortex ( Figure 3C).

Aging or Tat increased anxiety-like behavior among male mice
Aging and Tat expression promoted anxiety-like behavior. In the open field, there were main effects for age [F(1,35) = 9.14, p < 0.05; Figure   respectively. As well, there was an interaction wherein age or Tat genotype increased anxiety-like behavior [F(1,36) = 5.86, p < 0.05; Figure 4B]. Both middle-aged mice and Tat(+) young adult mice made fewer entries into the brightly-lit center of the open field compared to young adult Tat(−) controls (p = 0.001-0.027; Figure  4B, see ^). In the elevated plus-maze, we observed a main effect for Tat [F(1,38) = 13.13, p < 0.05; Figure  4C, see *] to increase the latency to enter the open arms compared to Tat(−) controls. As well, Tat(+) mice demonstrated a notable decrease in the proportion of open arm time compared to Tat(−) mice; albeit, this did not reach statistical significance (p = 0.09; Figure 4D). Notably, motor differences were detected for the distance (cm)  did Tat(+) mice (p = 0.03). No differences were observed in the total arm entries made in the elevated plus-maze (Table 1).
Affective-like behavior correlated with steroid fluctuations. In the open field, circulating corticosterone ( Figure 4A'; Table 2) and frontal cortex alloP (Table 2) were negatively correlated with the time spent in the center of the open field. Similarly, circulating P4 (Table  2) and hippocampal alloP ( Figure 4B'; Table 2) were negatively correlated with open field central entries. P4 also negatively correlated with the velocity travelled   (Table 2). In the elevated plus-maze, circulating corticosterone ( Figure 4C'; Table 2) and alloP, within all brain regions examined (i.e., hippocampus, midbrain, frontal cortex; Table 2 Table 2).

Aging and Tat altered nociceptive tolerance in male mice
Older age and Tat expression interacted to influence the threshold for paw withdrawal in response to a mechanical stimulus [F(1,37) = 5.87, p < 0.05; Figure  6C]. Unexpectedly, middle-aged Tat(−) and young adult Tat(+) mice exhibited a greater mechanical threshold than did young adult Tat(−) controls (p = 0.02-0.03; Figure 6C, see ^). No differences were observed in thermal hyperalgesia among middle-aged or Tatexposed mice ( Figure 6D); however, hippocampal alloP was positively correlated with the latency to paw withdrawal from a thermal stimulus ( Figure 6D'; Table  2).

DISCUSSION
HIV-infected individuals contend with an accelerated onset of age-related diseases and disorders [49][50][51]; however, the pathophysiology underlying accelerated aging is poorly understood. Findings from this study support the hypothesis that aging and HIV-1 Tat protein exert independent and interactive effects to induce neuroendocrine dysfunction which coincide with neuroHIV-like symptomatology in male mice. Consistent with observations in HIV + patients [16][17]20], Tat expression in mice reduced circulating T among middle-aged males. Moreover, HIV + men (ages 18-70 y/o) experience increased E2: T ratios [18], and we find that Tat expression recapitulates this endophenotype in male mice. Tat induction also elevated circulating P4 in young adult mice, an endocrine profile observed in our middle-aged controls suggesting that Tat expression may accelerate the onset of endocrine dysregulation. These preclinical data support the notion that exposure to Tat is sufficient to induce changes in the HPG axis that are consistent with clinical phenotypes. We observed further changes in the HPA axis as the animals aged. Tat expression elevated circulating corticosterone and central alloP in the hippocampus and midbrain of middle-aged mice. These neuroendocrine responses correlated with Tat-induced behavioral deficits, confirming our prior findings of Tat-disrupted alloP neurosteroidogenesis [36] and HPA axis dysregulation [52] and revealing their long-term consequences in the development of age-related neurological comorbidities. In addition, Tat altered peripheral organs, inducing an increase in the proportion of heart wet-weight. In particular, middleaged Tat(+) mice had the greatest heart wet-weight of any group. These data are consistent with premature heart remodeling to promote hypertrophy; heart disease is notably a leading cause of mortality among PLWH [53] and its incidence increases with age [54,55]. Middle-aged Tat(+) mice also had greater spleen weights than other groups which could be associated with several age-related comorbidities including lymphoma. Future pathological studies may expand on these findings.
Mood and cognitive disorders are common among HIV + patients [49,[56][57][58][59] and their prevalence increases with age [60] and with endocrine dysfunction [61,62]. Irrespective of HIV status, age-related androgen decline is associated with a greater risk of developing dementia, Alzheimer's disease [63,64], and cognitive impairment in men [65]. In the pre-cART era, hypogonadal men had higher depression scores than eugonadal men [61]. Among individuals having progressed to AIDS, risk for depression was greater than in eugonadal HIV + men [66]. While, endocrine disorders have declined in the post-cART era, secondary hypogonadism is still prevalent among HIV + patients (16-25%) [18,21,67,68]. Testosterone insufficiency is associated with depression/apathy, cognitive impairment, sexual dysfunction, fatigue, reduced muscle strength, and increased risk for HIV-associated lipodystrophy [19,20]. Herein, we find that older age and/or Tat expression increase anxiety-like behavior in an open field and in an elevated plus-maze. In both tasks, Tat expression engendered an anxiety-like phenotype in young adults comparable to that observed in middleaged mice, consistent with accelerated aging. Testosterone positively correlated with anti-anxiety-like behavior on the elevated plus maze supporting a role for androgens to improve mood. On the RAWM, Tat expression increased the frequency of errors made throughout the learning phase of the task, consistent with prior demonstrations of Tat-impaired spatial memory [41,[69][70][71][72][73][74]. These data extend prior findings to demonstrate that older age also increases errors, an effect that was positively correlated with greater circulating E2. The distance travelled to escape the RAWM was negatively correlated with circulating T, further supporting the benefits of a eugonadal state. Contrary to expectation, Tat-exposed middle-aged mice made fewer errors than their age-matched Tat(−) counterparts. This may reflect greater activation of the HPA axis in older Tat(+) mice. In support, middle-aged Tat(+) mice were faster than controls on day 1 of the RAWM and had greater corticosterone levels that also positively correlated with swim speed. This group also had the lowest circulating T and lower body weight than their non-Tat expressing middle-aged counterparts supportive of a reduced anabolic state. Consistent with the possibility of a generalized HPG/HPA activation in response to stress, circulating steroids and hippocampal alloP levels were positively correlated with RAWM swim speed. Thus, age and Tat effects on HPG/HPA axis function influenced 'performative measures' (i.e., open field and swim velocity) and these were associated with affective and cognitive dependent measures.
In the present work, an attenuated HPA axis was associated with improved neuromuscular function and hyperalgesia. Moreover, young adult Tat(+) mice demonstrated grip strength that was commensurate to that of middle-aged Tat(−) or Tat(+) mice suggestive of Tat's capacity to accelerate age-related neuromuscular dysfunction. These observations correlated negatively with circulating corticosterone and P4. In HIV + patients, neuromuscular deficits occur concurrent with disease progression and the onset of painful neuropathies [75]. As such, we also assessed mechanical antinociception and thermal hyperalgesia. We observed that aging or Tat expression increased pain-like thresholds among mice. While, this is contrary to the expectation of peripheral neuropathy, we previously observed similar effects in young adult male Tat(+) mice, which had greater antinociceptive thresholds compared to their Tat(−) counterparts [76]. This effect may involve peripheral sensory damage given evidence for intraepidermal nerve fiber regression [77], which may impact sensation. In support, ~50% of HIV + patients experience neuropathic pain and/or paresthesias, the latter of which are difficult to assess in mouse models [78]. We did not observe any thermal hyperalgesia, consistent with prior reports [41,77], but hippocampal alloP was positively correlated with greater thermal thresholds. Several factors may contribute to neuropathic pain in the HIV + population, including neurotoxic antiretroviral treatment (nucleoside reverse transcriptase inhibitors; albeit, used more sparingly in developed nations [79]), or metabolic syndrome [80], and age [81,82]. As well, HIV-1 proteins beyond Tat, including the envelope glycoprotein 120 (gp120), likely contribute to neuropathic pain. For example, in cultured dorsal root ganglion neurons, gp120 exerts chemokinelike effects via increased neural excitation and substance P release [83]. Intradermal injection of gp120 induces allodynia [83] and reduces epidermal nerve fiber density [84]. Discrepancies in findings for Tatinduced mechanical allodynia may also involve sex differences. We previously found that middle-aged Tat(+) females have reduced mechanical pain thresholds compared to controls [41], in contrast to current findings in middle-aged Tat(+) males. Bagdas et al.
(2020) observed similar effects in a cohort of agematched young adult Tat-transgenic mice.
Beyond typical hormone replacement therapies, neurosteroid-based therapeutics may hold promise for numerous age-related diseases and disorders. In rodents, androgen replacement therapy including testosterone and its neuroactive metabolite, 5αandrostan-3α, 17β-diol (a.k.a. 3α−androstanediol), reduces anxiety-and depression-like behavior and improves neurocognitive performance [85][86][87]. Similarly, exogenous alloP improved hippocampaldependent learning and memory among aged male mice including the triple-transgenic Alzheimer's model [87,88]. We have found that exogenous alloP attenuated affective dysfunction, mitotoxicity, and neurotoxicity associated with HIV-1 Tat [36,45]. Together, these data support the notion that maintaining the endocrine milieu of the CNS promotes resilience to neurodegenerative insults, including those associated with older age. In support, we recently found that T declines with age in the medial prefrontal cortex and striatum of Tat-expressing transgenic mice, concurrent with decrements in excitatory and inhibitory neurotransmitters as well as endogenous antioxidants detected with magnetic resonance spectroscopy [89]. These findings are consistent with clinical observations of cognitive impairment [90] that occurred in parallel with cerebral metabolic disturbances including a reduction of N-acetyl aspartate and an increase of choline and myoinositol, markers of inflammation, in HIV + patients [29,90,91]. Further, older age and HIV tended to exert additive effects at reducing subcortical gray matter in regions that included amygdala, caudate, and corpus callosum in one study [92] and altered βamyloid deposition in another [93]. These findings are consistent with observations of accumulated hyperphosphorylated Tau in the hippocampus of HIV + patients [94]. Others have found synergistic effects of older age and HIV on the exacerbation of verbal memory deficits [95]. Given the potential protective benefits of neurosteroids for neurodegenerative disease states ranging from Alzheimer's [96] to neuroHIV, neurosteroids may serve as scaffolds for a new generation of age-based therapeutics.
The present study had some limitations. We measured total circulating steroid content but did not assess the free, bioavailable fraction. Given increases in sex hormone-binding globulin among HIV + patients [16,17,67,97,98], future studies will discern the amount of bioavailable androgen. As well, we have recently found that middle-aged female mice also display HPG dysregulation that correlated with some behavioral deficits [41]. Together, these data support the assessment of exogenous steroid hormone treatment for potential benefits using aged models. While the present study did not include mice greater than 13 months of age, the current work highlights Tat's capacity to precipitate several characteristic phenotypes of advanced age on its own; thus, suggesting that Tat accumulation can decrease health span within mice.
The current work provides potentially actionable information that could help to limit premature or accentuated aging in the context of HIV. The hypogonadism observed in clinical populations may reflect substantial neuroendocrine dysfunction that can be readily assessed in preclinical models. In support, the formation of neurosteroids are dysregulated in postmortem HIV + brains [99] and in the circulation of HIV + patients [100]. The latter included the dehydroepiandrosterone sulfate-to-cortisol ratio, a biomarker for HPA dysregulation, which was associated with depressive symptomatology [100]. As well, T replacement improves mood disorders and lean body weight in HIV-infected men [66,101,102]. In conclusion, our data suggest that older age and Tat expression exert independent and interactive effects to worsen neuroendocrine, affective, cognitive, and neuromuscular comorbidities. Novel steroid replacement therapies may be useful adjunctive therapeutics to cART in the aging HIV + population.

MATERIALS AND METHODS
All experimental procedures were approved by the Institutional Animal Care and Use Committee at the University of Mississippi and conduced in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals (NIH Publication No. 85-23) ethical guidelines.

Behavioral assessment
Mice were tested over the course of four-weeks using several behavioral assessments in the following order: open field, elevated plus maze, radial arm water maze, grip strength, electronic Von Frey (eVF), and thermal probe. A 48-h break occurred between each behavioral test except for neuropathic-like pain assessments, eVF and thermal probe, which were conducted on the same day ( Figure 1). Mice were handled daily before behavioral assessments, which were conducted ~1 h into their dark phase. Thirty min prior to testing, mice were transferred to a testing room and habituated to 70dB of white noise. Behavioral apparatus was cleaned with 70% ethanol between trials to avoid olfactory bias. All behavioral assessments were tracked and digitally encoded by EthoVision animal tracking software (Noldus, Leesburg, VA, USA). AGING

Open field
The open field test was used to assess anxiety-like behavior and ataxia in rodents [46,110]. Mice were placed in the center of the brightly-lit area of the open field apparatus (40 × 40 × 35 cm) and allowed to freely explore the apparatus for 5 min. Greater time spent, and more entries made into, the brightly-lit center area were used as indices of anti-anxiety-like behavior. Total distance moved (cm) and velocity (cm/s) were considered indices of locomotor activity [46,110].

Elevated plus maze
The elevated plus maze was used to assess anxiety-like behavior [41,111]. Briefly, the maze consisted of two cross-arms (two open and two enclosed arms; 61 × 5 cm ea.) that were elevated from the floor (37.5 cm). Mice were placed in the plus-maze center area facing the open arm and allowed to freely explore for 5 min. A shorter latency to enter the open arm and a greater proportional time spent on the open arms were considered indices of greater anti-anxiety-like behavior. The total number of entries made into arms was used as an index of locomotor activity [41].

Radial arm water maze
The radial arm water maze (RAWM) was used to assess learning and spatial memory [112,113]. The maze consisted of eight arms (24 × 8 cm) attached to a central area (18 cm). Briefly, the maze was filled with roomtemperature tap water (equilibrated overnight) and a hidden platform (6 × 8 cm) was located at the end of a predetermined arm (goal arm). Mice completed 6 trials daily for 3 days. In each trial, mice were placed in a randomized, predetermined (starting arm) and allowed to swim freely for 90 s to reach a hidden platform. When mice failed to locate the hidden platform, they were gently guided by the investigator to the platform and allowed to remain for 15 s. The starting arm was randomized for each trial. A shorter time to locate the platform and fewer total errors made were considered indices of greater spatial cognitive performance [112,113]. To assess the proportional improvement from day 1, the following calculation was performed: 1 -(latency to escape on days 2 or 3/latency to escape on day 1) × 100 [41]. Entries into an incorrect arm were counted as errors. Total distance (cm) and velocity (cm/s) were used as indices of locomotor activity.

Grip strength
Grip Strength was used to assess neuromuscular function in rodents as previously described [41,114]. Briefly, mice were suspended by the tail and allowed to grab a metal grid bar connected to a transducer that measured peak force (g). Mice were gently pulled by the base of the tail horizontally away from the bar until grip was lost. The strength of forelimbs, alone or in combination with hindlimbs, was measured. Each mouse was tested in five trials with a 1 min break between trials to prevent fatigue. The mean of five trials was used as the grip strength score. Grip strength values were normalized to body weight and considered as an index of neuromuscular function [41,114].

Electronic Von Frey
The electronic Von Frey (eVF; Top Cat Metrology) test was used to measure mechanical allodynia as previously described [41,115]. Mice were allowed to habituate for 15 min on an elevated wire mesh. A series of eVF probes was applied to the hind paw (at the middle plantar surface) with force gradually increasing until paw withdrawal occurred. Each mouse was assessed in eight trials alternating between right and left paws (4 for each side) with a 3-min inter-trial interval. The mean of 4 trials for each paw was calculated. A lower mechanical threshold indicated an increase in the allodynic response [41,115,116].

Thermal probe test
A thermal probe test was used to evaluate thermal hyperalgesia in rodents as previously described [41,117]. A radiant heat stimulus was applied to the midplantar surface of the hind paw. The heat source was increased from room temperature to 60°C at a rate of 2.5°C/s. The test was stopped once a paw withdrawal occurred. Each mouse was assessed in 4 trials (2 for the right and 2 for the left paw) by alternating between right and left paws. Mice had 3-5 min breaks between trials. Paw withdrawal latencies were averaged and used as an index of hyperalgesia [41,117].

Ultra performance liquid chromatography (UPLC)mass spectrometry (MS)
Brain regions (hippocampus, midbrain, and frontal cortex) were grossly-dissected, wet weighed, and stored at −80°C until UPLC-MS/MS was conducted. Charcoalstripped brain tissues derived from Tat(−) and Tat(+) mice were used for calibration and quality control. Samples were homogenized in 100 μl of PBS (pH 7.4) and steroid extraction was achieved via protein precipitation. Samples were precipitated with 100 µl of acetonitrile followed by vortexing for two min and centrifugation for ten min (14,000 rpm). Following centrifugation, supernatants were mixed with 50 μl of derivatizing solution (20 mg/mL 2-hydrazinopyridine prepared in 0.5% trifluoroacetic acid ethanol solution) and incubated for 1 h at 60°C. 20 μl of the internal standard solution (LLOQ: 0.25 ng/ml) was added and vortexed. For analysis, 2 μl of sample was injected onto the UPLC-MS/MS instrument [118].

Statistical analyses
Behavioral endpoints (open field, elevated plus-maze, grip strength, eVF and thermal probe) were analyzed via two-way analyses of variance (ANOVA) using agegroup and Tat genotype as factors. Data obtained from radial arm water maze were analyzed via repeated measures ANOVA with testing day as the withinsubjects factor and age-group and Tat genotype as the between-subjects factors. Main effects were delineated using Fisher's Protected Least Significant Difference post hoc tests to determine group differences. Interactions were delineated via the assessment of simple main effects and main effect contrasts with alpha (0.05) controlled for family-wise error. Outliers were determined by Dixon's test and were excluded. All data were considered significant when p ≤ 0.05.

Data and materials availability
Data are available upon request.

AUTHOR CONTRIBUTIONS
JJP, NMA, and MJK conceptualized the project. ANQ, JJP, NMA, and MJK garnered funding for the project. NMA oversaw behavioral work conducted in the UM Neuropharmacology Core. ANQ, JJP, and NMA analyzed the data. ANQ and FM collected data. ANQ first drafted and revised the manuscript. All authors revised the manuscript and approved the final version.

FUNDING
This work was supported by funds from the University of Mississippi (Graduate Student Council award to ANQ) and the National Institutes of Health: R00 DA039791 (JJP), R01 DA052851 (JJP and NMA), R01 DA039044 (MJK), and an administrative supplement from award P30 GM122733 (pilot project to JJP and MJK).