Aging Changes in Bladder Hyperpolarization-Activated Cyclic Nucleotide-Gated Channels Are Associated With Increasing Heterogeneity of Adrenergic/Mucosal Influence on Detrusor Control in the Mouse

Abstract

A geroscience-informed approach to the increasing prevalence of bladder control problems in older adults requires understanding the impact of aging on dynamic mechanisms that ensure resilience in response to stressors challenging asymptomatic voluntary control over urine storage and voiding. Bladder control is predicated on sensory neural information about bladder volume. Modulation of volume-induced bladder wall tensions by autonomic and mucosal factors controls neural sensitivity to bladder volume. We hypothesized that hyperpolarization-activated cyclic nucleotide-gated (HCN) channels integrate these factors and thereby mediate adrenergic detrusor tension control. Furthermore, loss of HCN expression compromises that integration and could result in loss of precision of detrusor control. Using a life-span mouse model, reverse transcription quantitative real-time PCR and pharmacologic studies in pretensioned intact and mucosa-denuded bladder strips were made. The dominant hcn1 expression declines with maturation and aging; however, aging is also associated with increased variance around mean values. In strips from Mature animals, isoproterenol had less effect in denuded muscle strips than in intact strips, and HCN blockade diminished isoproterenol responsiveness. With aging, variances about mean response values significantly increased, paralleling hcn1 expression. Our findings support a role for HCN in providing neuroendocrine/paracrine integration and suggest an association of increased heterogeneity of HCN expression in aging with reductions in response precision to neuroendocrine control. The functional implication is an increased risk of dysfunction of brainstem/bladder regulation of neuronal sensitivity to bladder volume. This supports the clinical model of the aging bladder phenotype as an expression of loss of resilience, and not as emerging bladder pathology with aging.

Lower urinary tract symptoms (LUTS) and lower urinary tract dysfunction (LUTD) are increasingly prevalent with advancing age, yet a demonstrably flawed clinical paradigm (1–4) based on a linkage of LUTS to LUTD has not yielded routinely effective and safe therapies. Furthermore, other than clinician intuition, there is little to guide the extent of evaluation or therapeutic attempt in the older patient, with resultant morbidity and costs due to failure to discover correctable causes or unrecognized hopeless intervention. A geroscience-informed approach to address the increasing prevalence of bladder control problems in older adults requires understanding the impact of aging on dynamic mechanisms that ensure resilience in response to stressors challenging asymptomatic voluntary control over urine storage and voiding. Discovery and quantification of the mechanisms underlying resilience are needed to create predictive models and effective individualized approaches to clinical management of LUTS.

Voluntary bladder control is the cognitive response to neural sensory information about bladder volume. The initial mechanotransduction of bladder volume to bladder sensory neuronal activity is not fully understood but has been shown to be best modeled as a linear relationship of bladder sensory neuronal activity to bladder wall stress (5). Stress is the internal distribution of the tension created by resistance of the bladder wall to expansion by increasing volume. During voiding, massive, coordinated detrusor myocyte contractions yield large forces, creating the expulsive pressure that drives outward urine flow. However, even during bladder filling, spontaneous and uncoordinated myocyte contraction augments the acontractile elements (largely extracellular matrix) in resisting volume expansion (6), thereby contributing to bladder wall stress. Since bladder sensory neural activity relates to wall stress, inducers and inhibitors of detrusor myocyte activity during filling, therefore, can modulate the sensory input to the brain about bladder volume (7,8). This constitutes a potential adaptive mechanism aimed at providing an optimal neural data stream about bladder content to higher cortical centers.

The detrusor muscle is subject to neuroendocrine and paracrine influences either of which could be excitatory or inhibitory, such as acetylcholine or nitric oxide (NO). Brainstem-derived autonomic influence drives universal myocyte contraction during voiding via parasympathetics/acetyl choline. During urine storage, sympathetic input via the β-adrenoceptor (β ₃ in humans and β ₂ in mouse) inhibits myocyte activity, thereby relaxing detrusor muscle. Canonically, this is a result of the β-adrenoceptor–induced second messenger cyclic adenosine monophosphate (cAMP), activating a protein kinase A-dependent pathway and inhibiting actin/myosin interaction. In addition to neuroendocrine control, urothelium (UE) and possibly the mucosa as a unit provide a paracrine influence (9). Urothelium/mucosa elaborates factors including adenosine triphosphate (ATP), acetylcholine, NO, prostaglandins, and other peptides, all of which are potentially excitatory or inhibitory to the detrusor myocytes (10–12). Removing the mucosa from the bladder wall reduces tension variations in strips due to spontaneous muscular activity (13,14), suggesting an overall paracrine-driven mechanical excitation of the detrusor by the mucosa. In addition to previously described linkages of urothelial/mucosal ATP release to sensory nerve activation (15), paracrine influence over detrusor tensions could, therefore, also provide a urothelial/mucosal impact on sensory transduction of bladder filling or volumes.

While the precise mechanism responsible for coordinating the neuroendocrine (autonomic) with paracrine (local urothelial) influences on detrusor activity during bladder filling remains unknown, hyperpolarization-activated cyclic nucleotide-gated (HCN) channels have properties which could provide neuroendocrine regulation of detrusor muscle responses to mucosa-derived paracrine excitation. Hyperpolarization-activated cyclic nucleotide-gated channels are a family of 4-protein paralogs providing an inward depolarizing current (I_h) in excitable cells in response to hyperpolarization (16) and have been described in mouse and human bladder (17–20). The sensitivity to hyperpolarization and the opening kinetics are regulated by intracellular cAMP. This sensitivity to intracellular cAMP, therefore, links HCN channel dynamics to the β-adrenergic metabotropic receptor. By providing a relatively slow and adjustable depolarizing current, I_h, HCN could provide several possible mechanistic contributions. These include a feedback current in oscillatory/pacing cellular excitation (eg, as found in neural oscillatory circuits or the cardiac sinoatrial node), and regulation of cellular susceptibility to external excitatory potentials. Although HCN has a pacemaking role in the upper urinary tract (21), reduction in the detrusor smooth muscle (DSM) relaxation response to the adrenergic agonist isoproterenol by blocking or knocking out HCN without a change in baseline tensions suggests HCN provides the latter function in detrusor (17). Hyperpolarization-activated cyclic nucleotide-gated blockade fails to reduce spontaneous detrusor micromotional activity in rat bladder strips, which indicates that HCN does not primarily provide a pacing current in the bladder (19). By enhancing the opening probability of HCN channels over time, cAMP could also diminish excitability of the myocyte membrane to paracrine influence. Loss of HCN channel function with aging would not necessarily prevent other intracellular actions induced by β-adrenoceptor activation; however, it would render the myocyte more susceptible to paracrine influence. We therefore generally hypothesize that activation of HCN via sympathomimetic-induced cAMP renders detrusor less sensitive to excitatory mucosal paracrine factors.

We have previously shown that in Young mouse transverse bladder strips, isoproterenol-induced relaxation is impaired by HCN blockade (17). Old mice showed diminished HCN expression, and in contrast to Young (2–4 months) mouse strips, Old mouse strips (18–22 months) showed an enhancement rather than reduction to the same β-adrenergic stimulation following HCN blockade. While demonstrating HCN’s functionality in adrenergic detrusor relaxation and an age dependence, that study did not separate maturation, healthy aging, and advanced age, nor did it consider the impact of mucosal/paracrine signaling on neuroendocrine/sympathomimetic responses. The primary objective of this follow-up study was to test the more specific hypothesis that the sympathomimetic-induced relaxation response is dependent upon both the mucosa (as a paracrine influence) and on HCN function. Second, if, in Mature healthy systems, HCN provides an optimal integration of neuroendocrine and paracrine influence and thus contributes to a precise response to sympathetic control over detrusor relaxation, we hypothesized that age-associated declines in HCN expression would be associated with loss of precision of detrusor relaxation responses to sympathomimetics.

Methods

Animals

Mixed sex cohorts of C57BL/6 (wild type) mice (41 M/39 F) were used for this study in accordance with Institutional Animal Care protocol 101873. Mice were obtained through the National Institute on Aging’s aged animal colony and acclimated for at least 1 week prior to use. Animals for use at ages older than 18 months were procured at 18 months and further aged in the institutional animal care facility, following approved animal care policies. Four age groups were used, Young (2–4 months, n = 6 F/4 M), Mature (10–14 months, n = 5 F/5 M), Old (18–22 months, n = 5 F/5 M), and Oldest Old (24–28 months, n = 5 F/5 M). The choice of these age groups is derived from published data regarding food consumption, weight, and survival curves (22) and our prior cystometric findings (23,24). The weight versus age curve for ad libitum-fed female C57BL/6 mice maximizes at about 22 months and achieves 80% of this value at about 12 months. As the maximum rate of increase in weight and food consumption occurs between 2 and 12 months, we assigned Maturation to the transition from Young to Mature. Between Mature and Old, growth slows/stops, and food consumption does not increase; survival remains common; therefore, this interval is considered Aging. After 22 months, a progressive loss of resilience is evidenced by diminished survival and an increasing prevalence of voiding failure consistent with a detrusor underactivity phenotype in response to cystometric preparations and testing (23,25). Therefore, the transition from Old to Oldest Old group is representative of functional loss of resilience, or Senescence. These categories are intended as representative of measures of overall life-span processes in this organ system–specific study, parallel to aging categorization within the geriatric clinical literature (26–28). Animals were euthanized with CO₂ immediately prior to experiments, and bladders were excised just above the urethra and immediately transferred to bubbled buffer for further preparation.

Reverse Transcription Quantitative Real-Time PCR, hcn1-4

Bladders were harvested into RNAlater (Qiagen), kept overnight at 4 °C before freezing at −20 °C. All bladders were processed simultaneously when all age groups were collected. Bladders were placed in lysis buffer and homogenized (four 15 second pulses with 30 second intervals on ice). The Nucleospin RNA isolation kit (Clonetech) was used to purify RNA following manufacturer’s protocol. RNA quality and concentration were determined by Nanodrop 2000c (Thermo Scientific, Waltham, MA). Equal amounts of RNA were reverse transcribed to cDNA using the iScript Reverse Transcription Supermix (Bio-Rad Laboratories, Inc.) following manufacturer protocol. One hundred nanograms cDNA was used per qRT-PCR reaction, using the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories, Inc.) and predesigned Biorad PrimePCR SYBR Green Assays for mouse HCN1-4 (Unique Assay IDs: qMmuCID0015896 (Hcn1), qMmuCID0007901 (Hcn2), qMmuCID0022774 (Hcn3), and qMmuCID0022862 (Hcn4)). PCR, gDNA, RT and RNA quality controls were all run simultaneously. No RT controls were also run, and no Hcn1-4 was detected. Gene expression was calculated via a modified Pflaffl method utilizing multiple reference genes. Rps18, Yhwaz and B2m, showed the least variability between young and old bladders and thus were determined to be suitable reference genes in a screen of 14 commonly used reference genes (data not shown). Data were normalized to gene expression of young mice to give comparable fold changes by geometric averaging of multiple internal control genes.

Bladder Tissue Pharmacomyography (Strip Studies)

Bladder strips were prepared in room temperature bubbled buffer (NaCl 135 mM, KCl 5 mM, CaCl₂ 2 mM, MgCl₂ 1.2 mM, HEPES 10 mM, glucose 8.9 mM, pH 7.2; aqueous solution) immediately following harvest: an incision was made longitudinally along the ventral mid-line and pinned out in a Sylgard-lined petri dish. Two 1-mm-wide longitudinal strips were cut from either side of the midline and pinned out under minimal tension. One strip remained intact (INT), the other strip had the mucosa (UE + lamina propria [LP]) teased from the DSM using fine forceps (Figure 1A); previous histologic examination of sample preparations confirmed the cleavage plane at the LP–detrusor junction. Suture loops were secured on each end of the tissue strip, and the preparation was mounted on a horizontal myography chamber kept at 37 °C (Mayflower, Harvard Apparatus) (Figure 1B). The strip was slowly tensioned and stabilized over 30–60 minutes at ca. 1 g (~10 mN) in 37 °C bubbled buffer. Preliminary experiments demonstrated that strips remain viable and consistently responsive over at least 3 hours under these conditions (data not shown).

Figure 1.

Bladder muscle strip preparations model. (A) Bladder strips were prepared from longitudinally harvested tissue. For INT (intact) strips, whole thickness tissue was used. For DSM (detrusor smooth muscle only) strips, mucosa was gently removed. (B) Experimental setup: Bladder strip studies were performed in horizontal myobaths. (C) Dose–response protocol: Strips were stabilized in buffer, and then exposed to sequential log-fold differences in isoproterenol concentration, with and without block. Viability was confirmed using electrical field stimulation (EFS; lightning bolt).

Once a stable baseline was obtained, the strips were exposed to sequential isoproterenol in buffer in log-fold increasing concentrations (0.1, 1.0, and 10 µM). Strips were held for 10–15 minutes in each concentration and the final 3 minute window was recorded for analysis. The isoproterenol was thoroughly washed out of the preparation, and electrical field stimulation (EFS) was applied to confirm tissue viability. The HCN blocker, CsCl (5 mM), was added to the buffer and the tension was recorded. Isoproterenol dose response was repeated in the presence of CsCl, and after washing out isoproterenol thoroughly, viability was confirmed again via EFS.

The length of each strip was measured with a calibrated video camera. If needed, fine adjustments were made to length (± <3% of baseline length) following the initial isoproterenol washout and prior to adding CsCl to the buffer, to ensure initial tension for the CsCl phase was the same as for the plain buffer phase. Strip weights were recorded following myograph after the removal of the suture loops and blotting excess fluid from the tissue. Tissue stress in the baseline state was calculated by dividing the baseline tension by the estimated cross-sectional area, determined by dividing tissue mass (weight) by length. Tension data versus time were recorded using WinDAQ Pro+ (DataQ, Dayton, OH). Tension data were later transferred to LabChart (AD Instruments, Sydney, Australia) for initial review and analysis. The final 3-minute window of tension data under each drug condition was analyzed. Spontaneous smooth muscle activity causes low amplitude and low-frequency tension variations in bladder strips (29). A spike filter was applied to remove any momentary, artifactual spikes (not consistently observed), and the tension over the 3-minute collection window was then averaged for each drug condition. The mean responses to isoproterenol under intact and DSM conditions, without and with the presence of HCN blockade, were calculated by dividing drug responses by mean baseline tension.

Statistical Methods

Reverse transcription quantitative real-time PCR results were compared across age groups using analysis of variance (ANOVA). Baseline tissue stresses in intact and DSM across age groups were calculated and log transformed to normalize distributions. Strip weights, lengths, and baseline stresses were compared by age and strip type using repeated measures 2-way ANOVA, matching on strip type. The impact of CsCl on baseline tension was tested by comparing tension in buffer with the immediately following CsCl-containing buffer using t test. In the interest of strip viability, dose responses were limited to only 3 concentrations; therefore, a typical sigmoid curve with associated measures (eg, ED50) could not be generated. Tension was modeled by age group, using linear mixed-effects models with a random mouse intercept and fixed effects of dose, treatment, and dose–treatment interaction. Least square means for the 4 treatments by age group and dose level were presented in forest plots. Pairwise comparisons between treatments were conducted using the Tukey method to adjust for multiple testing. Two confidence intervals may be overlapped in the forest plot, but the mean difference was statistically significant as the statistical test gained power for modeling the dependency between treatments due to shared mice. Adjusted p values less than .05 were considered statistically significant. All the statistical analyses were conducted in R version 3.6.1.

Results

Hyperpolarization-activated cyclic nucleotide-gated channel mRNA expression declines principally with maturation, however displays increased variance about the mean with aging in the dominant paralog hcn1. All 4 paralogs of hcn mRNA are found in the mouse bladder, with relative expression hcn1 >> hcn 2 > hcn3, hcn4 in both sexes (Figure 2A). Expression of the dominant paralog hcn1 significantly decreased with maturation and aging, as shown in Figure 2B (top left). Interanimal variance about mean values appeared to narrow with maturation and considerably widen with early aging, and with advanced aging show overlap into values typical of all age groups. Sex was not a significant factor in mean hcn1 expression. The expression of hcn2 and hcn3 did not significantly change across life span in either sex (Figure 2B, top right/lower left). Maturation, aging, and sex were associated with changes in hcn4 (Figure 2B, lower right). In females, expression of hcn4 was significantly decreased between Young and Old mice; however, Mature and Oldest-Old females were not significant. In males, hcn4 expression decreased with maturation and recovered to Young expression levels with aging, and showed less interanimal variance with aging.

Figure 2.

Hyperpolarization-activated cyclic nucleotide-gated channel mRNA expression changes in the mouse bladder across life span. Data are presented from 10 bladders per age group obtained from Young (2–4), Mature (10–14), Old (18–22), and Oldest Old (24–28 mo of age). (A) Relative expression by Cq, hcn1 >> hcn 2 > hcn3, hcn4. (B) Log-fold change of hyperpolarization-activated cyclic nucleotide-gated (HCN) paralog mRNA for hcn1 (top, L), hcn2 (top, R), hcn3 (lower, L), and hcn4 (lower, R). Female animals represented by filled dots, and male animals represented by open circles. Bars represent means, while vertical lines indicate SD. Brackets and asterisks represent statistical significance.

Age, Mucosa, and HCN Blockade Status Had Minimal Impact on Stress in Strips Adjusted to a Baseline Tension

Confirming the experimental design of a uniform baseline tension, the set baseline tensions were consistent among strip types and age (mean: 9.8, SD 0.098 mN; Figure 3). The adjustment of strip length after the initial buffer-only isoproterenol testing (to ensure same starting tensions) did not result in different mean strip length within each age group (paired t-test comparisons, p = .13 overall). The calculated stress in the baseline predrug condition, incorporating strip weight and length, did not significantly differ by strip type. The appearance of a decline in baseline stress with maturation and aging was statistically significant only for the comparison of Oldest-Old intact versus intact Young strips (strip type p = .269, age p = .015, interaction p = .501; Figure 3A). Tensions before and after the addition of HCN blocker CsCl in each group are presented in Figure 3B; while no formal statistical testing was done (in order to retain statistical power for the other comparisons), no consistent tension change after the addition of CsCl was observed at baseline. Our previous report found no statistically significant impact of HCN blockade on baseline tension (17); the absence of an effect on baseline tension after administration of CsCl blockade provides evidence in support of HCN channels as an integrator during storage phase: the effect of blocking HCN is not apparent until adrenergic stimulation is provided.

Figure 3.

Baseline strip characteristics. (A) Calculated tissue stress (tension/estimated cross-sectional area) in Intact vs detrusor smooth muscle (DSM) strips at baseline passive tension. Stress diminishes with aging but is not dependent on mucosa. (B) The lack of impact of CsCl on passive tension.

Isoproterenol-Induced Bladder Relaxation Is Enhanced by HCN and Demonstrates a Maturation-Associated Dependence on Mucosa

Isoproterenol-induced (adrenergic) relaxations appeared to increase with isoproterenol concentration (Figure 4). Consistent with our prior report, HCN blockade inhibited isoproterenol-induced relaxation in intact strips from Young but not Old animals. Inhibition was more robustly observed in strips from Mature animals, in both intact and DSM (at 0.1 µM, DSM vs DSM + Block was not significant). A statistically significant difference in means between Intact and DSM strip responses to isoproterenol was found only in the Mature group (p = .018). A consistently similar pattern was seen in the older groups; however, this did not prove statistically significant in Old and Oldest Old, the exception being in Oldest Old at 0.1 µM isoproterenol in which isoproterenol had a stimulatory effect in some mice albeit reduced by HCN blockade. However, standard errors of the means were notably larger in Old and Oldest Old than in Young and Mature, and extreme values were found in Oldest Old under non-HCN–blocked conditions. While our testing cohorts were evenly split between male and females, with the multiple comparisons being made regarding isoproterenol, strip type, and HCN status, we did not formally evaluate the impact of sex.

Figure 4.

Least square mean analyses on the impact of strip type and hyperpolarization-activated cyclic nucleotide-gated (HCN) channel status (unblocked or CsCl-blocked), at 3 isoproterenol concentrations, n = 10 per age group. Ratios of drug to baseline tensions are shown. Isoproterenol induces relaxation of passively pretensioned bladder strips, except at 0.1 μM isoproterenol in Oldest-Old strips; ratio of pre- and post-isoproterenol is shown for Intact and detrusor smooth muscle (DSM) strips in the absence or presence of CsCl (HCN block). CsCl blockade diminishes isoproterenol effect, statistically significant as shown. Response to isoproterenol varies by strip type, statistically significant in Mature/isoproterenol 10 μM. While the trend of mean values looks similar in Old and Oldest Old, variance around the means was larger, as shown by an expanding 95% confidence interval.

Response to isoproterenol increases in variance around mean values with aging and is associated with HCN function and presence of mucosa. In analyzing the tension data, we observed similar mean impact of mucosa in Mature, Old and Oldest-Old strips but increasing standard errors of mean responses to isoproterenol with aging. The Young and Mature strips showed similar variances. However, with aging and senescence, variances progressively increase, with the appearance of extreme values in Oldest Old in the non-blocked condition, and more marked in Intact strips versus DSM strips (Figure 5). Inclusive of the extreme values, in Old and Oldest Old, HCN blockade reduced variances in each strip type.

Figure 5.

Variances about the means. Same data as in Figure 4; however, here plotted as individual data points to clarify the impact of mucosa and hyperpolarization-activated cyclic nucleotide-gated (HCN) channel blockade on data distributions about the mean values. Bars show means. Variances are impacted by age and HCN blockade status. Data presented by sex to show potential sex dependence; however, analyses were conducted with sexes grouped together.

Discussion

Our results confirm previous findings regarding interactions between age, HCN, and adrenergic detrusor relaxation, and provide new insights regarding the impact of aging on detrusor control.

Role of HCH in Increasing Heterogeneity of Lower Urinary Tract Resilience With Aging

While hcn1 expression declines across the life span, early and late aging are associated with increased variance around mean fold-change values, clarifying our early report comparing Old and Young (17). In Mature and therefore optimally resilient animals, HCN involvement in adrenergic relaxation is impacted by the presence of mucosa, supporting our hypothesis that HCN mediates an interactive paracrine/mucosal and neuroendocrine/adrenergic control over detrusor myocyte activity. A similar pattern of changes in mean values was observed in Old and Oldest-Old animals; however, paralleling the changes in hcn1 expression, standard errors of the mean responses to isoproterenol increased. The mean differences did not achieve the specified level of statistical significance; therefore, we cannot reject the possibility that mucosa has no impact in Old and Oldest-Old animals.

However, our findings of increased variances with aging suggest a more complicated significance, and in the context of HCN’s role in an adaptive system, this variance may be a more biologically important finding. Overall, our findings suggest that HCN is associated with range, consistency, and reproducibility—precision—of response to the interaction of neuroendocrine sympathomimetic control over mucosal paracrine influence on detrusor tonus. Loss of HCN reduces this precision, which could have the functional implication of an increased risk of dysfunction of brainstem/bladder regulation of neuronal sensitivity to bladder volume. These new data are consistent with this functional loss being primarily a feature of early aging (Mature to Old), with increasing heterogeneity of response with more advanced early and late aging. These changes in hcn expression and response heterogeneity parallel our previous demonstration of diminishing resilience to cystometric challenge with aging and senescence, as well as common experience of asymptomatic bladder control in older adults despite objective function changes (1,2) and the increasing prevalence of LUTS.

Role of Different HCN Subunits

In our previous study, we discovered hcn1 to be associated with the detrusor muscle layer, and hcn2 to be more diffusely expressed by RNAScope (17). The HCN channel is a heterotetramer of subunits encoded by these transcripts. It is possible that changes in gene/mRNA expression observed in this study could be indicative of altered subunit composition and different channel dynamics. Furthermore, changes in HCN mRNA levels may not directly correlate with parallel changes in protein expression. For example, HCN4 has the lowest mRNA expression but the most abundant HCN protein in pituitary and kidney (30). As our work suggests a changing role of HCN with maturation in aging, translational regulation, regulation of channel assembly, and senescence associated pathways need to be considered. The dominant HCN proteins and channel compositions in murine bladder remain undefined.

Cellular Localization of HCN Expression and Its Functional Implications

The cellular expression of HCN within the mouse and rat bladder wall has been reported as an “interstitial cell” (20,31); however, this has not been confirmed, and the role of interstitial cells in myocyte control is not certain (32). The impact of HCN blockade on isoproterenol relaxation in DSM-only strips (lacking LP + UE) from Mature and older animals suggests that HCN is a feature of detrusor myocytes. Finally, it is likely that the HCN-associated slowly activating inward current I_h likely has multiple roles in detrusor and sensory control. For instance, hcn2 and the I_h have been described in rat bladder dorsal root ganglia neurons (33). Overall, we interpret our findings to be consistent with age-related changes in I_h current dynamics influencing an age-evolving role of HCN in detrusor adrenergic responses, and more generally the interactions of paracrine and neuroendocrine control of detrusor tensions.

Modulatory Role of Bladder Mucosa

In strips from Mature animals, relaxation and the impact of HCN blockade were diminished by the absence of mucosa. The Mature group, having completed maturation but not yet experienced age-associated losses, best represents the optimal model of the control physiology. By showing an interdependency of neuroendocrine (isoproterenol) responses, mucosal presence (paracrine influence), and HCN status (unblocked vs blocked) in this age group, our data support our hypothesis that HCN serves a neuroendocrine/paracrine integrative role in detrusor control and therefore neuronal sensitivity to bladder volume. By reducing the precision of this integration, declines in HCN could diminish resilience to physiologic challenge, analogous to loss of sinoatrial responsiveness and subsequent cardiac reserve observed to accompany declines in sinoatrial nodal HCN expression (34,35).

In line with our findings, a positive contribution of mucosa to norepinephrine-induced relaxation was recently reported (36). In contrast to these findings, another study reported mucosal inhibition of isoproterenol response in bladder strips from male mice (37). Discrepancies between these works may be attributed to differences in strip preparation: longitudinal strips in our study include a heterogenous composition of dome, bladder wall, and trigone, whereas the dome was excised in the prior study. Additionally, differences in tension normalization should be considered, as our work normalized to baseline tension not maximum response to forskolin, an adenyl cyclase activator.

Study Limitations and Interpretation of Literature

The increasing variance in tension responses in the post-Mature groups versus the Mature group is largely driven by extremes in the data distributions. However, such extreme values were not observed in the DSM/blocked condition despite the 10 observations in the 4 conditions being from the same strips, showing that the atypical responses were associated with mucosa and with HCN. Therefore, these findings are both plausible and consistent with increased heterogeneity with aging and senescence. As our primary focus was on impact of mucosa and HCN status (buffer alone vs CsCl block of HCN), formal statistical comparison among age groups and between sexes was not made. The observed qualitative differences may serve as hypothesis generators for follow-up investigations. Variances were compared among age groups and may in fact be the more important metric. Finally, the bladder wall is anisotropic (38) despite the seemingly random directional distribution of myocytes and muscle fibers. While our results agree with the prior report of diminishing influence of HCN on isoproterenol-induced relaxation, we did not observe the previously reported enhanced relaxation of mouse bladder strips in the presence of blockade (17). That study, however, was made using full-thickness transverse bladder strips. An apical pacemaker function has been previously described (14,39), and its potential presence in our longitudinal strips may also contribute to this difference in findings. Finally, single-dimension tension studies do not apply the same slow 3-dimensional forces of natural bladder filling. With these complex stresses, the mucosa may be more relevant to dynamic tension changes.

Summary

The geroscience hypothesis implies that discovery of the mechanisms by which humans successfully adapt to physiologic stressors is a key to discovery of more effective and safer prevention and therapeutics (40). Our findings provide evidence that the increasing prevalence of bladder control symptoms and dysfunction with advancing age is reflective of an increased risk of adaptive failure rather than an inherent “aging pathology.” Since sympathetic regulation is a bounded continuous function in other systems, it can be postulated that brainstem control over the degree of detrusor relaxation during storage provides adaptive regulation of bladder volume sensitivity (7), aimed at satisfying allostatic/homeostatic requirements to integrate bladder volume information within a panoply of other internal and external physiologic demands. The narrower distributions of responses to fixed stimuli in Young and Mature animals suggest control inputs are met with precise responses, optimizing resilience to challenge. The increasing variances in Old and Oldest Old can be interpreted as loss of response precision. Loss of response precision with aging does not exclude the possibility of normal, adaptive responses, but implies that an increasing proportion within a population fall outside of the optimal response “window,” thereby diminishing the systemic capacity for normal system responses. This is characteristic of loss of resilience and contributes to the allostatic load that increases the risks of bladder control deficits with aging. Overall, our findings regarding the adrenergic/HCN/mucosal system suggest that increased response heterogeneity to control inputs is a principal characteristic of the aging bladder phenotype. Response heterogeneity may in fact be a more critical determinant of urinary performance than changes in mean values of variables, suggesting degenerative change within the bladder wall.

Funding

This work was supported by two grants received through the National Institute on Aging (NIA) at the National Institutes of Health (NIH), grant numbers K76AG054777 and R01AG058814 awarded to P.P.S.

Conflict of Interest

None declared.

Author Contributions

C.C.H.: conduct of experiments, manuscript preparation, and review; I.M.A.: conduct of experiments, manuscript preparation, and review; C.-L.K.: statistical analysis and review; G.A.K.: manuscript preparation and review; P.P.S.: study design, conduct of experiments, manuscript preparation, and review.