Fig 1.
The traditional (A) and enhanced (B) neural circuits of micturition have been adapted and redrawn from Fig 5 in Fowler et al. [18]. The circuits for both micturition and continence, shown separately by Fowler et al [18], were incorporated into a single circuit in Fig 1A and 1B. The circuit elements controlling urine storage are shown in red; the circuit elements controlling voiding are shown in green; and shared parts of the network are shown in blue. Neuronal populations are represented by individual nodes in the model calculations (and in the original figure drawn by Fowler et al. [18]). The figure emphasizes efferent control of the bladder and urethra, but the nerves innervating the bladder and urethra also carry afferent information, which is not, with the exception of the pelvic nerve, explicitly incorporated into either the traditional or enhanced models. The afferent connection of the pelvic nerve to the bladder is shown in red and the efferent connection is shown separately in green for clarity—even though the afferent and efferent activity run in a single nerve in reality. Only the efferent activity in the pudendal nerve controlling contraction of the external sphincter is shown and incorporated in the model (even though the pudendal nerve carries afferent information that contributes to bladder control). The circuits shown in Fig 1A, are adequate to explain the storage and voiding operations of healthy individuals. When an upper-level injury occurs at T10 or above (green dashed line, Fig 1B), the communication of the spinal circuits with higher centers was blocked. Additions to the traditional model were necessary to allow the spinal, interneuronal circuit controlling micturition to replicate both overactive bladder and detrusor-sphincter dyssynergia, which are seen after SCI. Three additional elements have been added to the circuit shown in B in order to replicate dysfunctional micturition after SCI. We added a spinal voiding reflex loop involving the afferent pelvic nerve (Node 7) that communicates with preganglionic elements in the spine that control detrusor contraction (Node 8) (shown in yellow; note also that an inhibitory Node 6 was added to the model to allow the pontine storage center to suppress the spinal voiding reflex in normal subjects). Afferent information flow through the pelvic nerve to the PSC (to enhance storage) was made explicit (green arrow), and last, an inhibitory connection from the PMC to the PSC was added so that storage was inhibited during the act of micturition (shown in black). In addition, three lumbar interneuronal nodes (excitatory Nodes 1 and 3 –green, and one inhibitory Node 2—grey) were added to receive TMS. PAG—periaqueductal grey; PSC—pontine storage center; PMC—pontine micturition center; HC—higher cortical centers; SCI—spinal cord injury; TMS—transcutaneous magnetic stimulation; Nodes 1 and 3 -excitatory interneurons; Node 2 inhibitory interneuron; Node 4 –efferent hypogastric nerve ganglion; Node 5—afferent pelvic nerve ganglion; Node 6 lumbar inhibitory interneuron; Node 7 a lumbar interneuron; Node 8 preganglionic parasympathetic neuron; Node 9 somatic motor neuron. The unnumbered nodes were maintained for continuity with the original figure, but these nodes were not incorporated in the model since they incorporate no internodal information processing (they simply pass information to the next node without modification as explained in the text). The colors of the nodes differ to enhance visual contrast, but the colors have no mechanistic significance. This figure was created in part with BioRender.com.
Table 1.
Structural additions to the traditional model that were necessary to replicate bladder dysfunction after SCI and replicate the response to lumbar TMS.
Fig 2.
The effect of low and high frequency, biphasic and monophasic TMS on inhibitory and excitatory interneurons.
In each plot, the electromotive forces associated with biphasic or monophasic TMS pulses are plotted (blue lines, first trac: biphasic A and B; monophasic C and D). The inhibitory interneuronal responses in the lumber spine (Node 2) are plotted in red (second trace), and excitatory lumbar neuronal responses (Nodes 1 and 3) are shown in black (third trace). The responses to 1 Hz TMS are shown on the left (top panel biphasic stimulation and bottom panel monophasic TMS), and the neuronal responses to 30 Hz TMS are shown in the right panels. For the same amplitude TMS, biphasic stimulation was more effective (elicited more action potentials) than the equipotent monophasic stimulation. Inhibitory neurons were more sensitive to the TMS stimulation than excitatory neurons so that for any level of TMS, more action potentials were elicited from the inhibitory neuron. High frequency pulses had cumulative effects (see text) and created bursts of activity, which were both at a higher frequency and lasted slightly longer in the inhibitory neurons. These differences in sensitivity to TMS led to important differences in response to 1 Hz or 30 Hz TMS within the enhanced model of control of micturition. This figure was created by MATLAB.
Fig 3.
Enhanced model simulation of normal bladder control during storage (A) and voiding (B). The elements of the network controlling storage and voiding are highlighted (thick lines) in A (storage) and B (voiding). The activities of Nodes 4, 8 and 9 are shown sequentially in separate panels (1–3). The activity shown represents the simulated changes in membrane potential at each node, which resembles somatic neuronal activity. The activities of Nodes 4, 8 and 9 during urine storage (A and C1-3), Nodes 4 simulates volume-dependent sympathetic inhibition of detrusor contraction (C1; note that the bladder volume was increased in three large steps—each step indicated by an arrowhead below the tracing of nodal activity and each step associated with a commensurate increased in nodal activity); parasympathetic activity stimulating detrusor contraction, which is inhibited during urine storage is shown in C2; and sphincter contraction originating from the Nucleus of Onuf as represented by Node 9 as shown in C3. Activity of the PSC and inactivity of the PAG during urine storage mean that detrusor muscle relaxes and sphincter muscle contracts. During voiding, the PAG is activated, which in turn activates the PMC, the detrusor contracts (note high activity in the parasympathetic preganglionic Node 8 (D2), and the sphincter relaxes so that activity in Node 9 is reduced. In order to permit the detrusor to contract, sympathetic inhibition of the detrusor is withdrawn, and Node 4 is quiet, thereby enabling urine to flow out of the bladder during detrusor contraction. This figure was created in part with BioRender.com.
Fig 4.
Simulation based on enhanced model replicates detrusor-sphincter dyssynergia after upper-level SCI (B) and hyperactive bladder activity after lower-level SCI (C). The schematic of the enhanced neural circuit of micturition with both an upper-level SCI (green dashed line) and a lower-level SCI (red dashed line) shown. B: After an upper-level SCI, supraspinal control of micturition is lost, and both voiding and storage reflexes are active, causing the detrusor-sphincter dyssynergia. The activity of Nodes 4, 8 and 9 are shown in B1. 2 and 3, respectively. Note that as the bladder is full, afferent activity from the bladder drives sympathetic (Node 4) to keep the detrusor muscle relaxed. However, there is no input from the PSC to inhibit activity of the preganglionic parasympathetic neurons (Node 8) and the detrusor receives active parasympathetic drive to contract the muscle. Simultaneously, the motor neurons driving sphincter contraction, represented by Node 9, are active. As a consequence, the bladder and external sphincter contractions occur simultaneously and the coordination of bladder sphincter activity necessary to empty the bladder is lost. C: After a lower-level SCI, sympathetic activity increases (Node 4, C1) as the bladder is filled (the time scale of filling has been accelerated), but there is no activation of detrusor contraction (Node 8, C2), and no external sphincter contraction (Node 9). As a result, the bladder leaks continuously as it is filled. This figure was created in part with BioRender.com.
Fig 5.
Modeling the effect of TMS in the enhanced model of control of micturition.
The response to 1 Hz TMS is shown in C1-3. The timing of TMS is shown by arrows, and after each arrow, the bladder briefly contracts, and sympathetic inhibition of detrusor contraction ceases briefly (Node 4, C1), and the external sphincter is relaxed briefly (Node 9, C3). As a consequence, the parasympathetic drive to the detrusor contracts the bladder (Node 10, C2). In contrast, 30 Hz TMS suppresses sympathetic drive to the bladder (Node 2 inhibits Node 4, D1); excitation of Node 1 inhibits preganglionic parasympathetic activity, and the detrusor lacks activation to contract (Node 10, D2); while excitatory Node 3 drives the Nucleus of Onuf and Node 9 activity (D3) results in contraction of the external sphincter. This figure was created in part with BioRender.com.