Impact of overactive bladder on the brain: Central sequelae of a visceral pathology

Abstract

Neural circuits that allow for reciprocal communication between the brain and viscera are critical for coordinating behavior with visceral activity. At the same time, these circuits are positioned to convey signals from pathologic events occurring in viscera to the brain, thereby providing a structural basis for comorbid central and peripheral symptoms. In the pons, Barrington's nucleus and the norepinephrine (NE) nucleus, locus coeruleus (LC), are integral to a circuit that links the pelvic viscera with the forebrain and coordinates pelvic visceral activity with arousal and behavior. Here, we demonstrate that a prevalent bladder dysfunction, produced by partial obstruction in rat, has an enduring disruptive impact on cortical activity through this circuit. Within 2 weeks of partial bladder obstruction, the activity of LC neurons was tonically elevated. LC hyperactivity was associated with cortical electroencephalographic activation that was characterized by decreased low-frequency (1–3 Hz) activity and prominent theta oscillations (6–8 Hz) that persisted for 4 weeks. Selective lesion of the LC–NE system significantly attenuated the cortical effects. The findings underscore the potential for significant neurobehavioral consequences of bladder disorders, including hyperarousal, sleep disturbances, and disruption of sensorimotor integration, as a result of central noradrenergic hyperactivity. The results further imply that pharmacological manipulation of central NE function may alleviate central sequelae of these visceral disorders.

Overactive bladder is a prevalent disorder, affecting 17% of the population and negatively impacting quality of life (1). Partial bladder obstruction is a common cause of overactive bladder in males and is used to model overactive bladder in laboratory animals (2). Whereas the structural and functional changes induced in bladder by partial obstruction are well studied (3, 4), the impact on brain function has been neglected. The potential for central consequences of bladder dysfunctions exists because neural circuits are present that communicate pelvic visceral status to the brain so that behavior can be coordinated with visceral functions. Barrington's nucleus (the pontine micturition center) and the locus coeruleus (LC) are integral components of a circuit that performs this task (5). Barrington's nucleus neurons project to spinal preganglionic parasympathetic neurons, where they regulate activity of the bladder and other pelvic viscera (6). These projections form the efferent limb of the micturition reflex, in which Barrington's nucleus neurons are activated by bladder distention and in response, initiate bladder contraction (7). The same Barrington's nucleus neurons project to the LC, a major norepinephrine (NE)-containing nucleus with divergent efferent projections that densely innervate the forebrain (8, 9). Among its functions, the LC regulates arousal, shifts in attention, and involvement in behavioral tasks (10, 11). LC neurons are activated by bladder and colon distention via Barrington's nucleus projections and this is temporally correlated to cortical electroencephalographic activation that is indicative of arousal (12–14). Thus, Barrington's nucleus is poised to coregulate pelvic visceral activity and the LC-NE system in response to pelvic visceral stimuli (5). Through its forebrain projections, the LC can convey visceral information to the cortex and facilitate a central response that is coordinated with the visceral response. This central limb likely involves increased arousal and a shift from ongoing behavior to behavior that is compatible with the visceral response.

By engaging the same circuitry, sensory signals arising from bladder or colonic pathology will be processed by the cortex and can potentially be expressed as central symptoms (e.g., hyperarousal, attention disorders, anxiety) that are comorbid with visceral pathology. This study investigated the potential for central consequences of pelvic visceral dysfunction by quantifying the impact of partial bladder obstruction on brain activity in the rat. Activity of Barrington's nucleus, LC neurons, and cortical EEG activity were quantified during the development of bladder pathology produced by partial bladder obstruction.

Results

Adult male rats underwent partial bladder obstruction or sham surgery, and urodynamic endpoints were recorded 2 weeks later. Rats with partial bladder obstruction had greater bladder capacity and bladder weight by 2 weeks, consistent with hypertrophy that has been reported (15) (Fig. 1A). This duration of partial bladder obstruction also resulted in spontaneous nonvoiding contractions, resembling overactive bladder in many cases (e.g., Fig. 1A, obstructed). Residual volume, measured by subtracting micturition volume from bladder capacity, was 280 ± 113 μl and 67 ± 27 μl in obstructed (n = 11) and sham (n = 15) rats, respectively (P < 0.05).

Fig. 1.

Effect of bladder obstruction on urodynamics and neuronal activity in the pontine micturition circuit 2 weeks after surgery. (A) Cystometry records of bladder pressure and micturition volume from a sham rat (Left) or rat with partial bladder obstruction (Right). Recordings were obtained in the unanesthetized state. Note the spontaneous nonvoiding contractions in the obstructed rat. The bar graph compares bladder weight and capacity in sham (solid bars) and obstructed (open bars) rats. The ordinate is unlabeled because the units are different for the two sets of bars. Vertical lines indicate standard error (**, P < 0.005, n = 11–16 subjects). (B) Records of bladder pressure and discharge rate of Barrington's nucleus neurons from a sham (Left) and obstructed (Right) rat. The bar graph compares mean discharge rate of Barrington's nucleus neurons of sham or obstructed rats before (solid bars) and after (open bars) bladder distention. Distention activated Barrington's nucleus neurons only in sham rats (*, P < 0.05, paired t test, n = 10 cells from 8 sham rats and 26 cells from 12 obstructed rats). (C) Records of bladder pressure and discharge rate of LC neurons from a sham (Left) and obstructed (Right) rat. The bar graph compares mean LC discharge rate of sham or obstructed rats before (solid bars) and after (open bars) bladder distention. Distention activated LC neurons of sham rats only (**, P < 0.005, paired t test, n = 26 cells from 10 sham rats and 34 cells from 12 obstructed rats). Example LC cells chosen for comparison in C were specifically matched for basal discharge rates to make the point that the lack of response in obstructed rats is not due to the higher basal discharge rate.

Single unit activity was recorded from Barrington's nucleus neurons or LC neurons 24 h after cystometry sessions. As previously reported (16), Barrington's nucleus neurons had a wide range of baseline discharge rates, and although the mean rate was somewhat higher in rats with partial obstruction than in sham rats (3.7 ± 0.9 vs. 1.9 ± 1.2), this was not statistically significant (P = 0.2). As in surgically naïve rats (16), Barrington's nucleus neurons of sham rats were consistently activated by acute increases in bladder pressure (Fig. 1B). In contrast, Barrington's nucleus neurons of rats with partial bladder obstruction were relatively insensitive to increases in pressure up to the micturition threshold, with only 6 of 26 neurons increasing discharge rate by 0.5 Hz or more (Fig. 1B).

Given that the LC receives pelvic visceral information from Barrington's nucleus (14), it was not surprising that LC neurons of rats with partial bladder obstruction were also unaffected by acute increases in bladder pressure equivalent to the micturition threshold, whereas LC neurons of sham rats were activated by bladder pressure increases, similar to those of surgically naïve subjects (13) (Fig. 1C). The lack of response of LC neurons to bladder pressure increases did not generalize to a lack of response to other sensory modalities. Thus, LC neurons of obstructed rats were robustly activated by repeated sciatic nerve stimulation. Although the duration of the evoked response was somewhat shorter in obstructed rats than in sham rats (29 ± 6 msec vs. 47 ± 5 msec, respectively; P < 0.05), the mean evoked discharge rate was similar between groups (18.8 ± 4 Hz, n = 11 and 20 ± 4 Hz, n = 10 in obstructed and sham, respectively), suggesting that the loss of LC responses to bladder pressure increases was relatively selective. Response latency was not different between groups (36 ± 3 ms vs. 31 ± 1 ms, obstructed vs. sham, respectively; P = 0.2).

Despite the loss in sensitivity to acute bladder distention, the basal activity of LC neurons was elevated by nearly 40% in bladder-obstructed rats (2.2 ± 0.2, n = 31) compared with sham rats (1.6 ± 0.1, n = 25; P = 0.02) by 2 weeks after surgery. This was apparent in recordings of baseline discharge rates as well as in the mean prestimulus discharge rate determined during trials of repeated sciatic nerve stimulation, where LC discharge rate was ≈50% higher in bladder-obstructed rats (2.0 ± 0.3 Hz, n = 10) than in sham rats (1.3 ± 0.2 Hz, n = 11) (P < 0.05).

The ability of partial bladder obstruction to elevate the basal level of LC activity is significant because relatively small selective increases in LC activity are sufficient to produce forebrain EEG indices of arousal, i.e., cortical desynchronization and hippocampal theta rhythm (17, 18). To determine whether the elevated basal tone of LC neurons in subjects with partial bladder obstruction was expressed as cortical activation, we recorded EEGs in weekly sessions for up to 4 weeks after partial bladder obstruction or sham surgery. Power spectrum analysis of EEGs recorded in sham rats revealed a shift to the left, toward lower frequencies, as time after surgery progressed (Figs. 2 A and C1). This shift, which likely reflected decreased arousal associated with habituation to the recording environment, did not occur in obstructed rats (Fig. 2 B and C2). Importantly, cortical EEGs showed substantial activation as the duration of bladder obstruction increased. This was evident by 2 weeks after surgery as decreased power in low frequencies (1–3 Hz), and this difference was maintained up to 4 weeks after surgery (Fig. 3A). Particularly striking was the development of prominent theta oscillations (6–8 Hz) that were maintained up to 4 weeks after surgery (Fig. 3A).

Fig. 2.

Bladder obstruction produces enduring alterations in cortical EEGs. (A and B) Representative examples of cortical EEGs from a sham rat (A) and a rat with partial bladder obstruction (B) at 1 and 3 weeks after surgery. Upper traces are raw EEG signals, and the lower graphs are PSD plots showing relative power in frequencies from 1–20 Hz. (C1) Mean PSD plots of sham rats at 1 (black symbols, n = 7) and 4 (gray symbols, n = 6) weeks after surgery indicate a significant shift to the left, toward lower frequencies, as time after surgery increased. The power in each frequency bin across weeks 1–4 was statistically different [F(3, 93) = 3.6, P < 0.001; factorial ANOVA]. *, P < 0.05; **, P < 0.005, Fisher LSD post hoc comparisons. (C2) The mean PSD plots of obstructed rats at 1 (black symbols, n = 11) and 4 (gray symbols, n = 4) weeks after surgery did not exhibit this shift [F(3, 93) = 0.37].

Fig. 3.

Partial bladder obstruction results in a decrease in low frequency activity and prominent theta activity. (A) Mean PSD plots of sham (black symbols) and obstructed (gray symbols) subjects at 1, 2, 3, and 4 weeks after surgery. The power in frequency bins across weeks statistically differed between obstructed and sham rats (F (4, 124) = 5.2, P < 0.001; ***, P < 0.005; **, P < 0.01; *, P < 0.05; Fisher LSD post hoc test). For obstructed rats, n = 11 for week 1, n = 7 for week 2, n = 7 for week 3, and n = 4 for week 4. For sham rats, n = 7 for week 1, n = 7 for week 2, n = 7 for week 3, and n = 6 for week 4. (B) Mean PSD plots of sham rats (black circles, n = 6), obstructed rats having a pattern of spontaneous nonvoiding contractions (gray circles, n = 6), and obstructed rats having a pattern of frequent voiding (black squares, n = 5). Some of the obstructed rats in this analysis included vehicle-treated rats. There was no statistically significant difference between the obstructed groups [F(1, 9) = 32, P = 0.19]. Standard errors were not plotted to better visualize the points and because there was no statistically significant difference between the obstructed groups. (C) Mean PSD plots for obstructed rats pretreated with vehicle (gray symbols, n = 8) or DSP-4 (black symbols, n = 5) at 4 weeks after surgery. Statistical significance [F = 1.92, P < 0.001] was demonstrated with the omnibus F-statistic, which assessed surgery and treatment effects on the power in each frequency across weeks 1–4 (repeated measure). ***, P < 0.005; **, P < 0.01; *, P < 0.05; Fisher LSD post hoc comparisons.

After 4 weeks of obstruction, two urodynamic profiles could be distinguished. One pattern was characterized by the presence of nonvoiding contractions (e.g., Fig. 1 A and Fig. 4A, obstruction 1). Another pattern was characterized by frequent micturition events and a general absence of nonvoiding contractions (e.g., Fig. 4A, obstruction 2). Despite the different urodynamic patterns, power spectral density (PSD) plots of the cortical EEG were remarkably similar in both cases, showing a shift toward higher frequencies, a decrease in power in the delta frequency range, and an increase in theta oscillations (Fig. 3B). Subtle trends suggested by separating the two groups included a more distinct peak at 7–8 Hz and increased amplitude at 14–15 Hz for subjects with nonvoiding contractions, although these differences were not statistically significant.

Fig. 4.

Effect of partial bladder obstruction on the relationship between bladder pressure and cortical EEG. (A) Simultaneous EEG and cystometry recordings from sham and bladder-obstructed rats. From top to bottom, the traces show the raw EEG, bladder pressure (BP, mmHg), bladder capacity (BC, μl), micturition volume (MV, ml), and PSD plots (PSD). The urodynamic pattern of an obstructed rat (obstruction 1) is similar to that of the sham rat, with the exception of spontaneous nonvoiding contractions, but has a distinct EEG pattern characterized by a peak at 7.5 Hz and a smaller peak at 14 Hz. Obstructed rat 2 (obstruction 2) does not exhibit nonvoiding contractions but shows increased micturition frequency. In this case, the EEG is characterized by lower amplitude and a shift toward higher frequencies. (B) Perievent spectrograms generated from the same subjects as in A, indicating how power in different EEG frequency bands covaries with bladder pressure during individual micturition cycles. (B1) The traces represent mean bladder pressure over four to five micturition cycles and are centered at the micturition threshold (time = 0). The heat map above each trace represents bladder pressure for each micturition cycle. For the sham rat, a uniform increase in bladder pressure up to micturition threshold can be seen. In contrast, for obstruction 1, nonvoiding contractions are indicated in the heat map as sporadic episodes (lighter blue blocks interspersed within darker blue) that occur up to the point at which the micturition threshold is reached. Obstruction 2 does not exhibit nonvoiding contractions. (B2) Heat maps that indicate the mean relative power in different EEG frequency bands (0–20 Hz, ordinate), and these are time-locked with B1, such that time = 0 indicates the point of the micturition threshold. Note how the different patterns of bladder activity produced by obstruction impact the relationship between bladder pressure and cortical EEG activity. In sham rats, a decrease in power in all frequencies (i.e., desynchronization) precedes the micturition threshold and is maintained. Obstruction 1, which has nonvoiding contractions, exhibits greater power in higher frequencies (7–10 Hz and 14–15 Hz) that fluctuate like the contractions. In obstruction 2, which has frequent micturition cycles and no nonvoiding contractions, the cortical EEG is desynchronized throughout the session, and increases in bladder pressure up to the micturition threshold are without effect.

Simultaneous cystometry and EEG recordings at 4 weeks after surgery illustrate how cortical activity covaries with bladder pressure and the impact of partial bladder obstruction on this temporal relationship (Fig. 4). As in surgically naïve rats (19), in sham rats the EEG desynchronized and amplitude in all frequencies decreased as bladder pressure rose but before the micturition threshold (Fig. 4B, sham, time = 0), consistent with an arousal response occurring before micturition. In rats with obstruction, the temporal relationship between micturition and desynchronization was diminished (Fig. 4B, obstruction 1) or absent (Fig. 4B, obstruction 2), consistent with the loss of LC responses to acute bladder distention. In a subject with nonvoiding contractions, oscillations in theta (7–8 Hz) and higher frequencies (14–15 Hz) were prominent and fluctuated with the contractions (Fig. 4B, obstruction 1). However, an increase in pressure up to the micturition threshold had little impact on this pattern. In a case characterized by frequent micturition events, the EEG showed desynchronization, and the power spectrum shifted toward higher frequencies throughout the entire session (Fig. 4A, obstruction 2) irrespective of increases in bladder pressure up to the micturition threshold (Fig. 4B, obstruction 2).

Cortical EEG desynchronization and diminished responses to bladder distention reflect the generalized activation of LC neuronal activity and loss of responses to acute increases in bladder pressure. To test causality between the effects of bladder obstruction on LC neuronal activity and cortical EEG activity, we selectively lesioned the LC–NE system by using DSP-4 before partial bladder obstruction. Cortical EEGs of DSP-4-pretreated rats resembled those of sham rats, rather than those of obstructed rats (Fig. 3C). Compared with cortical EEGs from vehicle-pretreated rats, those from DSP-4-treated rats had increased power in low frequency bands (1–3 Hz) and did not exhibit the typical theta oscillations associated with bladder obstruction (Fig. 3C).

Discussion

This study demonstrates that bladder pathology has consequences on activity of neurons in the pontine micturition circuit (i.e., Barrington's nucleus and the LC) that can contribute to both the visceral and central pathology of this disorder. At the level of the pons, partial bladder obstruction decreased responses of Barrington's nucleus neurons and LC neurons to acute bladder distention and elevated tonic LC activity. LC hyperactivity translated to cortical activation, expressed as a decrease in low frequency activity, and prominent theta oscillations that gradually increased over 4 weeks. Together, the results underscore the potential for hyperarousal and neurobehavioral impairments in individuals with overactive bladder and implicate the LC–NE system as a therapeutic target for alleviating these central symptoms.

The finding that Barrington's nucleus neurons were less responsive to acute bladder distention in subjects with partial obstruction is relevant to ongoing visceral pathology, because this represents a loss of central regulation of bladder function. Analogous to spinal cord injury, which disrupts supraspinal regulation of micturition, this could contribute to the development of local spinal regulation of the micturition reflex, which is thought to underlie the hyperreflexia that characterizes overactive bladder (20). Thus, consequences of bladder obstruction on Barrington's nucleus neuronal activity may feed-forward to further contribute to the bladder dysfunction.

Given the proposed functions of the LC, the effects of bladder obstruction on LC activity reported here have potentially important neurobehavioral consequences. The LC has been implicated in the modulation of arousal states because selective excitation or inhibition of LC neurons is sufficient to activate or inhibit forebrain EEG activity, respectively (17, 21). Early studies describing LC activation by diverse sensory stimuli also implicated this system in vigilance (22). More recent theories of LC function in relation to behavior suggest that the system facilitates decisions related to task-directed behavior, i.e., whether to maintain behavior in an ongoing task or to disengage and seek alternative strategies in a dynamic environment (11). Specifically, tonic LC activation is thought to favor disengagement from ongoing behavior and tasks involving focused attention and to promote scanning of the environment for alternate strategies. Evidence suggests that this can occur with relatively small magnitudes of LC activation (e.g., 40–50%) (23). With this perspective, the transient tonic excitation of LC neurons elicited by bladder pressure as it rises toward the micturition threshold in control rats is speculated to be a central limb of the micturition reflex that serves to increase arousal as well as to facilitate disengagement from ongoing behavior and a shift to elimination-related behaviors. In subjects with partial bladder obstruction, LC activity is persistently elevated and is not further activated by increases in bladder pressure up to the micturition threshold. The persistently elevated LC tone would be associated with hyperarousal, disordered attention, and inability to stay on task, effects that would disrupt normal behavioral function. LC activation has also been linked to anxiety, which along with sleep disorders, is reported in men with benign prostatic hypertrophy (24). The inability of bladder distention to further increase LC discharge may play a role in the loss of sensation of bladder filling that occurs in many subjects with overactive bladder (25).

Importantly, the effects of bladder obstruction on LC activity and responses to bladder distention are reflected at the level of the cortex. Decreased power in the delta frequency range is indicative of elevated arousal and consistent with reported sleep disturbances in subjects with partial bladder obstruction (26). Theta oscillations were a particularly salient effect of bladder obstruction. Theta oscillations have been implicated in sensorimotor integration, among other functions, and they are hypothesized to coordinate activity in different brain regions in preparation for motor responses to sensory input (27). Their persistence in overactive bladder may reflect constant or disordered processing of bladder signals. Their presence, even when selective responses to the major micturition event are attenuated, suggests a loss of ability to grade or discriminate between different magnitudes of bladder pressure changes, rather than a complete loss of sensation. Importantly, an effect on sensorimotor processing could affect cortical processing of nonbladder related stimuli and adversely impact functions requiring focused attention.

Although bladder obstruction resulted in two distinct urodynamic profiles that were apparent by 4 weeks, the effects on cortical activity were comparable, indicating that the mechanisms underlying these central changes were engaged regardless of urinary pattern and that the central sequelae described here would be applicable in subjects exhibiting either type of dysfunction. This would occur if mechanisms underlying the central changes occur at a relatively early stage and are maintained, whereas the urinary patterns become distinct at a later time as a result of adaptations at the level of the bladder. It is noteworthy that a study of cognitive function of human subjects with lower urinary tract symptoms suggests that two profiles of urge incontinence, which may be analogous to those described here, were both associated with cognitive impairments (28).

The finding that a toxin specific to the LC–NE system alleviates the cortical changes produced by bladder obstruction implies a causal role for the LC in these effects and suggests that targeting this system may be useful in attenuating the central sequelae of overactive bladder. Still, the mechanisms by which bladder obstruction produces persistent LC activation are unknown. Several findings suggest that this is not likely a result of chronic stress. First, body weight gain was comparable in sham and obstructed rats (see Methods). Reduced weight gain over time is a cardinal feature of chronic stress that endures long after stress termination (29, 30). This is seen with both physical (29, 30) and social stressors (31) and is sufficiently sensitive to serve as an endpoint of chronic mild stress (32). This is in contrast to hyperthermia, which does not consistently occur with certain chronic physical stressors (33) and habituates with repeated stress, while reduced weight loss is maintained (30). It is noteworthy that the elevation in tonic LC discharge produced by bladder obstruction in this study is not seen with chronic physical stressors (i.e., repeated shock or chronic cold), which affect stimulated but not tonic LC activity (34, 35). Finally, corticotropin-releasing factor, the neuropeptide that initiates the stress response (36) and mediates LC activation during stress (37), has effects on cortical EEG that are substantially different from those produced by bladder obstruction, decreasing power in all frequencies from 1–32 Hz (38).

Overactive bladder as a result of partial obstruction is particularly prevalent in the elderly, a population that is also vulnerable to neurobehavioral deficits and sleep disturbances (1). The present findings suggest that this visceral dysfunction may contribute to both neurobehavioral and sleep deficits in this population. However, given that overactive bladder affects a diverse population, the results have relevance to a broad group and emphasize the previously unrecognized potential for this visceral disorder to affect cognitive and behavioral functions. Importantly, evidence for a causal role of the LC–NE system suggests that fine-tuning the activity of this system may be a useful therapeutic approach for central sequelae of overactive bladder.

Finally, although this study focused on bladder dysfunction, it is noteworthy that the results likely can be generalized to pathologies that arise from other pelvic viscera and engage the same circuitry. For example, the LC is activated by colonic distention, and this is mediated by afferents from Barrington's nucleus and correlates to cortical electroencephalographic activation (12, 14). This circuit may underlie the well documented comorbidity of psychiatric and colonic symptoms that characterize irritable bowel disorder. The results of this study underscore the need to consider the impact of visceral dysfunctions on ongoing cognitive and behavioral status.

Methods

Subjects and Methods.

The subjects were male Sprague–Dawley rats (300–350 g) [housing details can be found in supporting information (SI) Methods]. Care and use of rats was approved by the Children's Hospital of Philadelphia Institutional Animal Care and Use Committee.

Barrington's nucleus and LC neuronal recordings were obtained 2 weeks after bladder obstruction surgery. Surgical details are described in SI Methods. Two weeks after surgery, a bladder catheter was implanted, and cystometry recordings were obtained in the unanesthetized state 24–48 h later for 1 h (19).

Single-unit extracellular activity was recorded from Barrington's nucleus or LC neurons in the halothane-anesthetized state 24 h after cystometry, as described elsewhere (16). Once a single neuron was isolated, basal activity was recorded for at least 3 min. The bladder was then distended for 1 min by injection of saline through the implanted catheter with a volume that matched the capacity for that subject, determined by previous cystometry measurements. Bladder pressure was recorded simultaneously with neuronal activity. To determine whether bladder obstruction altered LC responses to sensory stimuli, we recorded neuronal responses to repeated sciatic nerve stimulation (1 mA, 0.5 ms, 0.5 Hz, 60 trials) as peristimulus time histograms (PSTHs) (39). When more than one cell was recorded in one LC, these were all in the same dorsal–ventral track, and the depth of each recording was noted for later verification of cellular location. Pontamine sky blue (PSB) dye was iontophoresed at the most ventral recording site for histological localization of the recording site. Barrington's nucleus was localized after identifying the LC, and the recording protocol was similar to that described for the LC. It was not always possible to obtain stable recordings from both the LC and Barrington's nucleus in the same subject. Data presented are only from neurons that were histologically identified as being within the LC or Barrington's nucleus.

For EEG recordings, electrodes were implanted during the obstruction or sham surgery (see SI Methods for details on EEG electrodes and recordings). Rats were habituated to EEG recording in the cystometry chamber 48 h after surgery. At 1, 2, and 3 weeks after surgery, rats were placed into the cystometry chamber for EEG recordings only. At ≈4 weeks after surgery, a bladder catheter was inserted. Rats were placed into the cystometry chamber 48 h later for simultaneous cystometry and EEG recordings in a 60-min session.

Behavior was observed throughout the cystometry session, and there was no evidence of stress in the obstructed rats. Indeed, these subjects appeared more active in later cystometry sessions compared with their sham counterparts. The mean presurgery weights were 378 ± 27 g (sham) and 345 ± 19 g (obstructed). For rats that survived 2–4 weeks, weights were 422 ± 19 g (sham) and 390 ± 19 g (obstructed), indicating a gain of ≈45 g for both groups.

To determine the role of LC activation in EEG changes induced by bladder obstruction, we pretreated rats with the serotonin reuptake inhibitor, citalopram (10 mg/kg, i.p.), and then treated these animals, 30 min later, with saline or the selective neurotoxin, 50 mg/kg i.p. DSP-4 [N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine], which preferentially lesions NE terminals originating from the LC (40). Citalopram is used to prevent uptake into serotonergic terminals. Agents were administered 2–3 weeks before bladder obstruction.

Data Analysis.

Between-group comparisons of discharge rate and urodynamic parameters were performed by using the Student's t test for unpaired samples. Discharge rates before and during bladder distention were compared by using the Student's paired t test. PSTHs were analyzed as described elsewhere (39). All t tests were two-tailed, with P < 0.05 being deemed statistically significant.

Raw EEG data were converted to PSD plots indicating the relative power in 32 frequency bins from 0–20 Hz using Neuroexplorer (Nex Technologies). Differences were assessed by repeated measures ANOVA, with the frequency bins and subject group as independent variables and power in each frequency as the dependent variable, which was measured repeatedly across weeks 1–4. The Fisher LSD was used as a post hoc test to identify frequency values that differed between the obstructed and sham groups for each individual week. Significance was determined at P < 0.05. For DSP-4 experiments, power in each frequency bin was compared between vehicle- and DSP-4-treated obstructed rats.