The Nitric Oxide/cGMP Pathway Tunes the Thermosensitivity of Swimming Motor Patterns in Xenopus laevis Tadpoles

Discussion

We examined the contribution of the NO/cGMP pathway to determine the thermosensitivity of the swimming CPG in post-hatching Xenopus laevis larvae. Our main conclusion is that the NO/cGMP pathway plays a role in tuning the thermosensitivity of the swimming circuit by setting the thresholds for failure and recovery, as well as by modulating the duration of swim episodes. We show that the presence of an NO donor (SNAP) greatly hindered the recovery from hyperthermic failure of the swim circuitry, regardless of whether the motor pattern was activated by electrical stimulation of the tail skin or by superfusion with the glutamate receptor agonist NMDA. This demonstrates that the effects of NO on recovery from hyperthermia are mediated via actions at the level of the central circuits and not the sensory pathways from the skin to the spinal cord, though the latter may also be affected by NO (Alpert et al., 2007). Using tail-stimulated swimming, we show that activation of the NO/cGMP pathway renders most measures of circuit function more sensitive to hyperthermia, whereas inhibition of the pathway has the opposite effect and reduces their thermosensitivity. The measure that was not noticeably affected was the temperature sensitivity of swim rhythm frequency. We suggest that the functional role of this effect of the NO/cGMP pathway is to modulate excitability and energy consumption, dependent on previous environmental conditions. Thus, increases in abiotic stressors activating the NO/cGMP pathway, such as hyperthermia or hypoxia, would reduce locomotion and hence conserve energy, enabling the organism to cope with these conditions more effectively.

The effects of extreme high temperature on neural processes can be divided into two main types: those that have immediate consequences as a result of the Q₁₀ relationships of the rates of neural phenomena such as the opening and closing of ion channels; and those that have a temporal component reflecting an accumulating disturbance such as gradual protein denaturation or the loss of ionic equilibria (Robertson, 2004). The increase in swim cycle frequency with increasing temperature is not surprising, because this is a well established property of CPGs (Walker, 1975; Tryba and Ramirez, 2003) that is due to the reduction of conduction and synaptic delays between elements in the circuit (Robertson, 2004). The profound decrease in swim episode duration with increasing temperature is not as easy to explain. At room temperature, the duration of swim episodes in stage 42 larvae is determined partly by the timing of activation of GABAergic inhibition (Reith and Sillar, 1999), which can prematurely terminate swimming. One possibility is that increased temperature simply speeds up the processes controlling this timing via Q₁₀ relationships, i.e., in the same way that the durations of action potentials and synaptic potentials decrease with increasing temperature (Janssen, 1992). Indeed, one of the prominent actions of NO is to decrease swim episode duration by potentiating GABA transmission (McLean and Sillar, 2002), and thus an associated hypothesis is that increased temperature increases the production of NO by increasing the enzymatic activity of NOS and that it is the action of NO at GABAergic synapses that decreases episode duration. In this scenario, it would have to be proposed that the action of NO to decrease cycle frequency (via potentiation of glycinergic transmission) (McLean and Sillar, 2002) is more than counterbalanced by the direct actions of temperature to increase cycle frequency. Moreover, one would need to propose that any release of NO during the temperature ramp is sufficient to affect episode duration but insufficient to have the dramatic effect on recovery temperature and recovery time that pharmacological activation of the NO/cGMP has. These difficulties, and the observation that manipulation of the NO/cGMP pathway had no effect on temperature sensitivity of the rhythm frequency, lead us to believe that most of the temperature-induced shortening of swim episodes is not mediated via NO.

Our experiments did not address the mechanisms underlying hyperthermic failure and subsequent recovery of motor pattern generation in the spinal cord. Many different processes are likely to be implicated, including the failure to generate action potentials, the increased likelihood of conduction failure at axonal regions of low safety factor, and the failure of synaptic transmission (Janssen, 1992; Robertson, 2004). An important common feature of all of these possibilities is a dysregulation of ion homeostasis or conductance. Action potential failure is predicted by the Hodgkin–Huxley equations as K⁺ currents overwhelm Na⁺ currents at high temperature (Chapman, 1967). Synaptic failure at the Drosophila neuromuscular junction results from a loss of the ability to regulate intracellular Ca²⁺ concentrations (Klose et al., 2008). Given these precedents and the time dependence of recovery of the tadpole swim CPG, we believe it to be a reasonable supposition that hyperthermic failure and recovery in this system are closely related to loss and restoration of appropriate ion gradients.

NO has many and various important roles to play in neural function and dysfunction (Guix et al., 2005; Garthwaite, 2008). The modulatory effects of NO on a variety of neuronal circuits across different phyla have been described. NO affects rhythmic motor activity in jellyfish swimming (Moroz et al., 2004), snail feeding (Kobayashi et al., 2000), and cardiac activity (Taylor et al., 2003), crab stomatogastric activity (Scholz et al., 2001; Stein et al., 2005), locust oviposition (Newland and Yates, 2007), and mouthpart movements (Rast, 2001), frog respiration (Hedrick and Morales, 1999), tadpole swimming (McLean and Sillar, 2000), lamprey swimming (Kyriakatos and El Manira, 2007), and mammalian respiration (Pierrefiche et al., 2007; Reeves et al., 2008). The mechanisms NO uses to exert its effects are less well known, with a notable exception being in the tadpole spinal cord (McLean and Sillar, 2002, 2004). Nevertheless, it remains difficult to explain the effects of NO on hyperthermic failure and recovery through its described facilitatory effects on glycinergic and GABAergic neurotransmission, and better understanding of the mechanisms will require further investigation.

Behavioral flexibility in the face of highly variable environmental conditions is considered to be a primary role for neuromodulation of CPGs (Dickinson, 2006). Despite this awareness, and the often very detailed knowledge of the cellular and synaptic mechanisms underlying neuromodulator action, the environmental contexts requiring motor pattern changes are less well known. This may be simply because the appropriate contexts have not been presented to the animal or preparation. Thus, we know that, at these early stages of Xenopus development, NO acts as a brake on swimming, but we do not know the conditions appropriate for this. NO is known to be released by, and mediate responses to, cellular stressors such as hypoxia (Wingrove and O'Farrell, 1999; Reeves et al., 2008) and hyperthermia (Armstrong et al., 2009), and we can interpret our findings as NO reducing the probability of swimming to conserve energy during exposure to more stressful environments. NO makes the swim CPG output less intense and causes it to cease functioning at a lower temperature. These are modifications that would conserve energy. Moreover, the modulation of spinal circuitry by NO is long lasting (Kyriakatos and El Manira, 2007), befitting a response to a changing environment.

The control animals in our study were variable, and their responses to hyperthermia fell along a continuum between those of preparations treated with activators of NO/cGMP and those treated with inhibitors. For example, we found that the tadpoles from the breeding colony had markedly variable swim durations (several seconds to several tens of minutes) before any manipulations (see supplemental information and supplemental Fig. S1, available at www.jneurosci.org as supplemental material). Interestingly, the duration of swim episodes also correlates with rearing temperature (see supplemental information and supplemental Fig. S2, available at www.jneurosci.org as supplemental material). Cool-reared individuals swim for longer on average than those reared under warm conditions. Swim duration is determined at least partly by levels of NO (McLean and Sillar, 2000). We interpret the evident variability in swim duration as reflecting variable conditions in the rearing chambers (e.g., temperature, oxygenation, crowding, lighting, etc. that were not precisely controlled) and that short swim durations indicate previous experience of a more stressful nature causing increased levels of NO. Another source of variability, which may affect the NO systems, is the dissection before recording which could, at least in principle, lead to variable levels of iNOS activation and hence alter the production of NO and episode duration. It is clear that NO has a continuous role to play, because inhibition of the NO/cGMP pathway has effects without previous pharmacological activation (this paper) (McLean and Sillar, 2000).

We conclude that NO modifies the thermosensitivity of the swim circuit in the spinal cord of Xenopus tadpoles, making it more sensitive to hyperthermia. Thus, we describe here a possible functional role for the well characterized effect of NO on the swimming CPG in tadpoles and propose that the NO/cGMP pathway may be involved more generally in tuning the CNS and organismal responses to the abiotic environment.