In this study, we explored interactions between comorbid hypertension and OA development in both male and female rats. Endpoint measures of cardiovascular function confirmed that all hypertensive animals had increased blood pressure, with no differences due to surgical group. Additionally, comorbid hypertension appears to accelerate OA joint remodeling differently in male and female rats. Hypertensive animals also exhibited enhanced blood pressure responses to pharmacological activation of the vagus nerve. Taken together, these data suggest influence of hypertension on the development of OA joint pathology. While pain and anxiety-related behavioral results were inconclusive in this study, these data motivate future mechanistic studies on the influence of hypertension on OA pathophysiology and symptoms.
The development of symptomatic and radiographic knee OA is associated with hypertension in humans(22). In our study, male hypertensive rats with OA had higher trabecular bone area ratios compared to male normotensive rats with OA, indicating either a greater extent of bone marrow resorption or the development of subchondral bone sclerosis. This is similar to a study by Chan et al., where naïve 40-week old male SHR animals developed greater bone volume fractions, as shown by µCT analysis, compared to a normotensive control(23). Furthermore, bone area ratios in female normotensive rats displayed no difference due to surgical induction of OA; however, in female hypertensive rats with OA, bone area ratios increased compared to the sham control. Combined, these data indicate that comorbid hypertension and OA leads to more severe subchondral bone pathology following a meniscal injury. Further, this remodeling could be related to vascular etiology in the subchondral bone, which has previously been proposed as a mechanism for OA development(24). As such, these questions and the mechanisms driving these changes may be further dissected with in-depth analysis of subchondral bone pathology in hypertensive and normotensive rats.
While surgical induction of OA in male animals resulted in significant cartilage damage (indicated by smaller minimum cartilage thickness), no differences in cartilage damage were observed between hypertensive and normotensive animals in the OA group. This finding is in contrast to a prior study in hypertensive mice with exercise-induced OA, where strenuous exercise combined with hypertension resulted in larger Mankin scores compared to non-exercise and normotensive controls(25). Further, in this prior work, larger Mankin scores positively correlated with the number of angiotensin II receptor type I positive cells in the joint, suggesting a role for the renin-angiotensin system in the development of joint pathology in male animals. While we did not observe increased cartilage damage in male hypertensive animals, female hypertensive animals with OA did have reduced cartilage thickness compared to both hypertensive animals with a sham surgery and normotensive animals with OA. This finding in female hypertensive animals somewhat parallels the findings in male hypertensive mice by Yamagishi et al.(25) Moreover, this loss of cartilage thickness in hypertensive females may reveal a potential explanation of the sex-differences in OA development in rodents, where female animals tend to develop joint damage at a slower rate than males if at all(26, 27). Combined, this may support a sex-dependent role for autonomic dysfunction in the susceptibility of the joint to post-traumatic OA.
Systemic administration of phenylbiguanide was used as a pharmacological activator of the vagus nerve, thereby causing a drop in the heart rate and blood pressure for all animals. In males, pre-existing hypertension led to exaggerated drops in blood pressure in both OA and non-OA animals. However, for females, these exaggerated drops in blood pressure were only observed in animals with pre-existing hypertension and OA. Again, these findings may help explain potential sex-dependent effects on OA development and progression that arise from systemic physiologic differences.
Anti-hypertensive drugs have been associated with decreased odds of developing knee OA in humans(28), and thus, targeting autonomic function may provide some potential to modify the progression of OA. Here, systolic pressure and pulse pressure, but not diastolic pressure, have been associated with the development of knee OA(28), and persons taking three or more anti-hypertensive medications have decreased odds of knee OA compared to persons taking no drugs (28). Our work in this study, along with work by others in hypertensive rodents, provide mechanistic evidence of shared pathophysiology between OA and hypertension. These mechanistic overlaps provide groundwork for future studies evaluating if and how OA progression may be controlled or slowed via anti-hypertensive drugs.
In contrast to our current findings, we previously conducted a similar study using male Lewis rats(15) where heart rate responses to intravenous phenybiguanide administration did not significantly differ between OA and non-OA. However, the Lewis rat has a dysfunctional hypothalamic-pituitary-adrenal axis (HPA axis), which can confound autonomic measures(29). This prior work was motivation for this study using both hypertensive and normotensive animals. Moreover, unlike literature in mice that has shown slowed progression of OA in females(26, 27), Lewis rats do not develop meaningful sex differences in joint pathology(30). Combined, our work continues to question the general use of the Lewis rat strain to model knee OA pathogenesis, particularly when not also evaluating the systemic effects of a dysfunctional HPA axis.
In association with HPA axis functions, serum cortisol levels increase with exposure to stress. In some chronic pain states, which may be considered a type of chronic stress, basal serum cortisol is sustained at higher levels (31). As such, these elevated systemic cortisol levels have been related to the presence of ongoing pain(32). However, the association between chronic OA pain and cortisol remains unclear(33), particularly in preclinical models. In this study, bi-weekly corticosterone measurements did not reveal meaningful differences with either chronic hypertension or the surgical induction of knee OA. Previously, SHR animals were found to have higher basal corticosterone levels compared to normotensive Wistar rats in samples taken immediately after euthanasia without anesthesia(34). However, we did not observe similar findings between SHR rats and normotensive Sprague Dawley rats. A possible confounder of our data was the use of isoflurane anesthesia during serum collection, as isoflurane has been reported to affect corticosterone levels in female animals(35). Because cortisol is increased with acute stress, future studies may consider evaluating responsiveness of the system through measurement of serum corticosterone levels before and after a stressful event. Moreover, hair-based corticosterone tests may provide better data on the long-term corticosterone levels in rats and reduce the variability caused by diurnal fluctuations in cortisol in the serum.
In this work, we attempted to measure the effects of pre-existing hypertension on knee OA symptoms, and the MCLT + MMT model has previously produced detectable decreases in paw withdrawal thresholds across different strains, sexes, and laboratories(30, 36–38). We did not observe evidence of tactile allodynia in this study. Some confounding factors may have contributed to this finding. First, estrus phase of female animals affects withdrawal thresholds(39), and this study did not control for this factor due to logistical constraints. Additionally, male scents can result in higher stress state in females, and thereby affect female responses (40). Again, due to logistical constraints, male and female animals were not tested separately in this study, and all von Frey testing was conducted by a male experimenter. Moreover, SHR animals have shown modality dependent hyperalgesia or hypoalgesia that is likely related to autonomic control of blood pressure, and not sustained high blood pressure per se(41). Though current data of tactile allodynia in comorbid hypertension and OA are inconclusive, future refined studies are warranted to explore autonomic influence on pain-related behaviors. These studies should more carefully control for confounding experimental variables and should consider other pain assessment modalities, such as thermal pain or non-evoked assays.
In knee OA studies, open-field activity has been more commonly used as a non-evoked pain assay to assess exploratory activity(42). Here, reductions in distance traveled is as a surrogate measure of movement evoked pain (or avoidance of movement-evoked pain). However, anxiety-related behaviors can also be assessed in open-field environments, where time spent in different parts of the arena are used as surrogate measures of anxiety. This interpretation of open field results is less common in studies of OA, despite the common comorbid presentation of anxiety and depression in patients with OA and chronic musculoskeletal pain. Moreover, SHR animals are known to typically display less anxiety-like behaviors and higher levels of locomotion(43, 44). In our data, there were no differences between SHR and normotensive animals at baseline, but at week 2, normotensive animals with MCLT + MMT surgery begin to spend more time in corners, potentially associated with higher levels of anxiety.