Discussion

Administration of either an NMDA (MK801) or an AMPA (CP465022) glutamatergic receptor antagonist suppresses PNS inhibition of bladder overactivity induced by AA irritation in chloralose anesthetized cats, indicating that both types of receptors play an important role in the inhibition. In addition, both antagonists increase bladder capacity and decrease the amplitude of the reflex bladder contractions, indicating that tonic activation of NMDA and AMPA receptors contributes to the AA-induced bladder overactivity. These results extend previous reports regarding the important role of glutamatergic excitatory transmission in the normal bladder function (Yoshiyama et al., 1993a; Matsumoto et al., 1995a, 1995b; Kakizaki et al., 1998) and demonstrate that glutamatergic excitatory mechanisms are also important in the reflex pathways mediating bladder overactivity.

Because glutamic acid is considered to be a major excitatory transmitter released by primary afferents in the spinal cord, it is likely that pudendal inhibition as well as bladder overactivity is modulated by actions of MK801 and CP465022 at primary afferent-interneuronal synapses in the spinal dorsal horn. Under basal conditions, AMPA are more important than NMDA receptors in generation of primary afferent evoked monosynaptic EPSPs in the dorsal horn (Larsson, 2009). However, NMDA activation contributes to more delayed excitatory synaptic responses under basal conditions and to the enhancement of excitatory transmission that occurs after nociceptive stimulation (Larsson, 2009). Although large doses of the AMPA and NMDA antagonists suppress both pudendal inhibition and bladder overactivity, a subtle difference was noted in the dose response studies of the two agents. CP465022 suppresses pudendal inhibition at doses that do not significantly alter bladder capacity or the amplitude of bladder contractions (Figs. 5 and 6), whereas MK-801 suppresses pudendal inhibition only in doses that also suppress bladder activity (Figs. 2 and 3). This suggests that transmission mediated by NMDA receptors is more important and/or more sensitive to block by MK801 in the bladder reflex pathway than in the pudendal inhibitory pathway. This difference may reflect an enhanced role of NMDA receptors in bladder reflexes after AA infusion into the bladder, which activates central nociceptive mechanisms and thereby may unmask an effect of MK801 on bladder reflex pathways. This possibility could be tested in future experiments by comparing the effects of the antagonists during saline CMGs in the absence of a nociceptive stimulus.

In addition to the contribution of ionotropic glutamate receptors to pudendal neuromodulation, previous studies in cats have shown that metabotropic glutamate receptor 5 (mGluR5) also plays an important role in pudendal neuromodulation of bladder overactivity induced by AA irritation (Larson et al., 2011). However, these receptors are not involved in the reflex mechanisms mediating bladder overactivity (Larson et al., 2011). Based on the known functions of mGluR5, it is likely that activation of these receptors during pudendal neuromodulation facilitates synaptic transmission mediated by NMDA/AMPA receptors.

Pudendal neuromodulation of bladder overactivity also depends on multiple inhibitory transmitter mechanisms involving GABA_A, β-adrenergic, and 5-HT receptors (Matsuta et al., 2013b; Schwen et al., 2013; Xiao et al., 2014; Kadow et al., 2016). GABA_A-receptor mediated inhibition occurs in the spinal cord (Xiao et al., 2014), and therefore very likely involves glutamatergic activation of local GABAergic inhibitory interneurons. β-Adrenergic receptor-mediated inhibition occurs via reflex activation of sympathetic inhibitory pathways arising in the rostral lumbar spinal cord (Kadow et al., 2016), and therefore must involve a lumbosacral intersegmental excitatory pathway that also depends on glutamatergic excitatory synapses. The 5-HT receptor component of pudendal neuromodulation depends on serotonergic inputs to the sacral spinal cord from the raphe nuclei in the brain stem (Matsuta et al., 2013b; Schwen et al., 2013; Reese et al., 2014) and therefore must involve activation of raphe 5-HT neurons by ascending glutamatergic projections from the sacral spinal cord to supraspinal sites. Thus it is likely that the glutamatergic antagonists administered intravenously in the present experiments suppress pudendal neuromodulation by acting at various sites within the central nervous system.

Our study showing that AMPA and NMDA receptors are involved in pudendal neuromodulation differs from a previous study of sacral neuromodulation of bladder overactivity in rats, which showed that spinal NMDA receptors but not spinal non-NMDA receptors play a key role in sacral neuromodulation (Riazimand and Mense, 2005). This may indicate a species difference or a difference in the mechanisms involved in the two types of neuromodulation. In cats, sacral neuromodulation of bladder overactivity depends in part on activation of opioid receptors and supraspinal GABA_A receptors (Jiang et al., 2016; Bandari et al., 2017), whereas pudendal neuromodulation does not involve opioid receptors but depends in part, as mentioned above, on activation of spinal GABA_A receptors (Mally et al., 2013; Xiao et al., 2014).

It is likely that MK801 and CP465022 also act at multiple sites in the central nervous system to increase bladder capacity and reduce the amplitude of bladder contractions. The micturition reflex is mediated by a spinobulbospinal pathway consisting of 1) an ascending limb projecting from the sacral spinal cord to the brain stem, 2) supraspinal circuitry in the periaqueductal gray and the pontine micturition center, and 3) a descending limb from the pontine micturition center back to the sacral spinal cord (Fowler et al., 2008). In rats, glutamate has been identified as an excitatory transmitter in each segment of this pathway (Yoshiyama et al., 1993a; Matsumoto et al., 1995a,b; Kakizaki et al., 1998). AMPA and NMDA glutamatergic antagonists administered systemically, intrathecally, or intracerebroventricularly suppress bladder reflexes during saline or AA CMGs in urethane anesthetized rats (Yoshiyama et al., 1993b, 1997). The antagonists also suppress c-fos expression induced in spinal dorsal horn neurons by AA irritation of the bladder (Kakizaki et al., 1996), indicating AMPA and NMDA receptors have an important role in the afferent limb of the overactive bladder reflex triggered by a nociceptive stimulus. During saline CMGs in unanesthetized decerebrate rats (Yoshiyama et al., 1994, 1997), AMPA antagonists are also effective in suppressing reflex bladder activity, whereas NMDA antagonists are ineffective or facilitate bladder activity. On the basis of these experiments it was proposed that excitatory glutamatergic transmission in the normal micturition reflex pathway in rats is mediated primarily by AMPA receptors and that the contribution of NMDA receptors is only unmasked when transmission mediated by AMPA receptors is depressed by an anesthetic. On the other hand, patch clamp recording in spinal cord slices of neonatal rats revealed that both AMPA and NMDA receptors are involved in excitatory transmission in spinal reflex pathways to the parasympathetic preganglionic neurons (Araki and de Groat, 1996, 1997; Miura et al., 2003) as well as in axonal projections from the lateral funiculus to these neurons (Miura et al., 2001). These axonal projections could be part of propriospinal or supraspinal pathways.

Relatively little information is available about the role of glutamatergic transmission in bladder reflexes in cats. It has been reported that glutamatergic receptor agonists injected into the pontine micturition center in decerebrate cats induce voiding or facilitate reflex bladder contractions (Mallory et al., 1991), whereas intrathecal injections of MK-801 or kynurenic acid, a broad spectrum glutamatergic receptor antagonist, blocks bladder contractions elicited by electrical stimulation of the pontine micturition center (Iwabuchi, 1997). However the latter contractions are not blocked by CNQX, a non-NMDA receptor antagonist. Iontophoretic application of glutamatergic receptor agonists have an excitatory effect on bladder parasympathetic preganglionic neurons in the cat spinal cord (de Groat and Ryall, 1968; De Groat, 1971) and iontophoretic application of kynurenic acid blocks the firing of these neurons elicited by electrical stimulation of the pontine micturition center (Iwabuchi, 1997). These studies indicate that in cats NMDA receptors in the spinal cord are involved in the descending limb of the micturition reflex pathway, whereas spinal AMPA receptors are not involved in this pathway. Therefore, in the present study the reduction in contraction amplitude by MK801 (Fig. 3), which occurs by the same dose (0.1–0.3 mg/kg i.v.) in cats (Fig. 3) and rats (Yoshiyama et al., 1993a), can be explained very well by the important role of spinal NMDA receptors in the descending limb of the micturition reflex pathway. On the other hand, the reduction in contraction amplitude as well as the increase in bladder capacity by CP465022 (Fig. 6) is probably caused by targeting AMPA receptors in the ascending limb of the micturition reflex or in the periaqueductal gray-pontine circuitry, thereby reducing the input from the pontine micturition center to the sacral spinal cord.

In MK801-treated cats, pudendal inhibition was eliminated only when the prestimulation bladder capacity was more than double the capacity measured in the untreated condition (see Fig. 2). This raises a concern that the loss of pudendal inhibition might not be due to the blockade of NMDA receptors in the pudendal-to-bladder inhibitory pathway, but instead could be due to the large bladder volume that produces a strong micturition reflex that overcomes the pudendal inhibition or due to the bladder reaching a maximal volume that cannot be further increased by PNS. However, the data in this study support the opposite conclusion. At the 3 mg/kg dose of MK801, the prestimulation capacity is only 134.5 ± 30.3% of saline control capacity (Fig. 3), which is not likely to be the maximal bladder volume during an inhibition. Our previous studies in cats (Matsuta et al., 2013b; Schwen et al., 2013) showed that during the inhibition produced by PNS and/or drugs, the micturition reflex can occur at a bladder volume about 300–400% of saline control capacity, indicating that the maximal bladder volume is much larger. In addition, the micturition reflex after 0.3 mg/kg of MK801 seems to be weaker because the amplitude of micturition contraction is only 50% of the control (see the prestimulation CMG in Figs. 1 and 3). This may be due to MK801 suppressing glutamatergic transmission in the central descending limb of the spinobulbospinal micturition reflex pathway, thereby producing a weaker micturition reflex rather than a stronger one. Because the micturition reflex is weak and the maximal bladder volume for inducing a reflex is not reached, the 3 mg/kg dose of MK801 must have suppressed the micturition reflex and blocked pudendal inhibition at the same time. Otherwise, a significant increase in bladder capacity should have occurred during PNS.

In summary, this study in cats revealed that both NMDA and AMPA glutamatergic mechanisms are involved in the generation of reflex bladder overactivity as well as in pudendal neuromodulation of this activity. These results together with previous reports showing the involvement of inhibitory neurotransmitters (Kadow et al., 2016; Matsuta et al., 2013b; Schwen et al., 2013; Xiao et al., 2014) suggest that the pudendal neuromodulation depends on the activation of NMDA and AMPA glutamatergic excitatory transmission, which in turn stimulates the inhibitory pathways that suppress reflex bladder activity.