Inhibition of nitric oxide and soluble guanylyl cyclase signaling affects olfactory neuron activity in the moth, Manduca sexta

Abstract

Nitric oxide is emerging as an important modulator of many physiological processes including olfaction, yet the function of this gas in the processing of olfactory information remains poorly understood. In the antennal lobe of the moth, Manduca sexta, nitric oxide is produced in response to odor stimulation, and many interneurons express soluble guanylyl cyclase, a well-characterized nitric oxide target. We used intracellular recording and staining coupled with pharmacological manipulation of nitric oxide and soluble guanylyl cyclase to test the hypothesis that nitric oxide modulates odor responsiveness in olfactory interneurons through soluble guanylyl cyclase-dependent pathways. Nitric oxide synthase inhibition resulted in pronounced effects on the resting level of firing and the responses to odor stimulation in most interneurons. Effects ranged from bursting to strong attenuation of activity and were often accompanied by membrane depolarization coupled with a change in input resistance. Blocking nitric oxide activation of soluble guanylyl cyclase signaling mimicked the effects of nitric oxide synthase inhibitors in a subset of olfactory neurons, while other cells were differentially affected by this treatment. Together, these results suggest that nitric oxide is required for proper olfactory function, and likely acts through soluble guanylyl cyclase-dependent and -independent mechanisms in different subsets of neurons.

Fig. 1

Inhibition of NO synthesis modified the resting and odor-evoked activity of PNs.a-c, Representative traces are shown for three PNs before, during, and after treatment with either 500 μM 7NI or 15 mM L-NAME (see Table 1 for specific drug used). NOS inhibition resulted in: rhythmic bursting (a n = 10; PN5 shown), increased firing rates (b n = 3; PN12), and decreased firing rates (c n = 2; PN15) in different subsets of PNs. A rise in the resting membrane potential (RMP; dotted line, b, c n = 4) and a reduction in action potential amplitude (a n = 4) were also observed in some PNs. d The average resting firing activity was plotted as a percentage of baseline levels (± SEM) for each PN subset. e-f PN odor responses, measured by net number of spikes, (see Materials and methods) were also significantly abolished (e, g PN6; n = 6; repeated measures ANOVA df = 2; F = 15.6; P < 0.001; Tukey's post-hoc test: P < 0.05) or decreased (f, h PN1; n = 4; dF = 2; F = 6.86; P < 0.05; Tukey's post-hoc test P < 0.05) after NO synthesis inhibition. f The inhibitory potential (I₁ circle) that precedes the bout of action potentials in the PN odor response was also missing in this PN. Calibration a-c 20 mV and 250 ms; e-h 20 mV and 200 ms (gray bar)

Fig. 2

Inhibition of NO synthesis also modified the resting and odor-evoked activity of LNs. a-c Representative traces are shown for three LNs before, during, and after treatment with either 500 μM 7NI or 15 mM L-NAME (Table 1). Similarly to PNs, NOS inhibition in LNs resulted in: bursting (a n = 2; LN16; LN bursts were arrhythmic; compare to Fig. 1a), increased firing rates (b n = 3; LN19), and decreased firing rates (c n = 2; LN21) in different subsets of LNs. One LN had no change in activity when NO synthesis was blocked (Table 1). Changes in the RMP and action potential amplitude were also observed in bursting LNs (a n = 2). d The average resting firing activity was plotted as a percentage of baseline levels (± SEM) for each LN subset. e-g LN odor responses, like PN odor responses, were also abolished (e, h; LN22; n = 2; significance not measured, n < 4) and decreased (f, i; LN 21; n = 3), but also increased (g, j; LN23; n = 4; repeated measures ANOVA df = 2; F = 164; P < 0.0001; Tukey's post-hoc test: P < 0.001) after NO synthesis inhibition. Calibration a-c 20 mV and 250 ms; e-j 20 mV and 200 ms (gray bar)

Fig. 3

NO inhibition modified resting input resistance in both PNs and LNs. a Current injections during NO inhibition revealed dramatic voltage changes (†V; +45%) in this representative PN (PN1). b The input conductance for PN1, calculated from the slope of the regression line when the peak voltage was plotted against multiple current steps, dropped by 54% (from 5.4 to 3.6 nS) when NO synthesis was inhibited. c The average input conductance for all PNs (all) and by activity class (B bursting; I increased; D decreased) was plotted as a percentage of baseline levels (SEM is only shown for average of all PNs). The average conductance decreased significantly during NOS inhibition (means coded by different lowercase letters differed significantly; n = 12; repeated measures ANOVA df = 2; F = 26.7; P < 0.001; Tukey's post-hoc test: P < 0.001). d The mean conductance decreased significantly for all LNs (n = 8; repeated measures ANOVA df = 2; F = 3.4; P < 0.05; Tukey's post-hoc test: P < 0.01), however the bursting LNs (n = 2) showed an average increase in conductance (+16%). Calibration 20 mV, 500 ms

Fig. 4

Inhibitors of sGC signaling caused similar modifications as NO synthesis inhibitors in the resting and odor-evoked activity in a subset of PNs. a-c Representative traces are shown for three PNs before, during, and after treatment with 500 μM ODQ. As found with NOS inhibition, sGC inhibition resulted in: rhythmic bursting (a n = 3; PN5 shown, compare to Fig. 1a), increased firing rates (b n = 2; PN12, compare to Fig. 1b), and decreased firing rates (c n = 2; PN27), coupled with changes in the RMP and action potential amplitude in a subset of PNs (a n = 2). d The average resting activity plot revealed the bursting PN firing rate decreased during sGC inhibition, which deviates from the increased firing rate observed in the PNs during NOS inhibition (Fig. 1d). e The average conductance of all PNs decreased significantly during sGC inhibition (n = 6; repeated measures ANOVA df = 2; F = 9.41; P < 0.001; Tukey's post-hoc test: P <0.01). f-k PN odor responses during sGC inhibition were also abolished (f, i; PN 26; n = 3; significance not measured, n < 4), decreased (g, j; PN27; n = 4; df = 2; F = 48.4; P < 0.0001; Tukey's post-hoc test: P < 0.001), and, unlike during NOS inhibition, increased (h, k PN24; n = 4; dF = 2; F = 36.5; P < 0.0001; Tukey's post-hoc test: P < 0.01). g The I₁ (inset, circle) and the I₂ (see Results) were also modified in this PN during sGC inhibition. Calibration a-c 20 mV and 250 ms; f-k 20 mV and 200 ms (gray bar)

Fig. 5

Inhibitors of sGC signaling caused similar modifications in the resting and odor-evoked activity as NO synthesis inhibitors in a subset of LNs. a-c Representative traces are shown for three LNs before, during, and after treatment with 500 μM ODQ. As found with NOS inhibition, sGC inhibition resulted in: arrhythmic bursting (a n = 1; LN28 shown), increased firing rates (b n = 2; LN29), and decreased firing rates (c n = 1; LN31), coupled with changes in the RMP and action potential amplitude in the bursting LN (a). e The average conductance of all LNs decreased significantly during sGC inhibition (n = 4; repeated measures ANOVA df = 2; F = 16.5; P < 0.001; Tukey's post-hoc test: P < 0.001). f-k LN odor responses during sGC inhibition were also abolished (f, i LN30; n = 1; significance not measured, n < 4), decreased (g, j; LN28; n = 1), and increased (h, k; LN32; n = 2). Calibration a-c 20 mV and 250 ms; f-k 20 mV and 200 ms (gray bar)

Fig. 6

sGC inhibition caused differential effects in the resting and odor-evoked in a subset of neurons. a-c A subset of LNs (n = 3) showed no change (X) in resting activity (a, b; LN33 shown) and conductance (c) when sGC was inhibited (changes were similar to those observed during control saline application, see Materials and methods). d, e The effects of NOS and sGC inhibition on the resting (d) and odor-evoked (e) activity were directly compared in PN4. d PN4's resting activity became burst-like after NOS inhibition, but no bursting activity was observed after sGC inhibition. e Odor-evoked activity also differed during NOS and sGC inhibition. NOS inhibition caused a decrease in net spikes and the appearance of bursting, intermittent firing, while sGC inhibition caused a more moderate decrease in net spikes and an increased delay to onset. Calibration a, d 20 mV and 250 ms; e 20 mV and 200 ms (gray bar)

Fig. 7

Immunohistochemistry revealed many olfactory neurons do not contain high levels of sGC-immunoreactivity. A subset of neurons in this study (Tables 1, 2) was labeled with lucifer yellow (LY) dye and subsequently tested for sGC-immunoreactivity (sGCir). a A whole-mount view of a LY filled Type Ib LN (LN28, see Fig. 5a) with a cell body in the lateral cell body cluster (dashed oval c) and no ramifications in the MGC (dashed oval in a). b The AL was frontally sectioned (100 μM sections) and c, labeled with the MsGCα1/Cy3 antibody (images show single optical sections 3 μM). A magnified view of the cell body is shown in the inset. d LN28 contained strong levels of sGCir. e-h, Another LN, Type Ia (LN17), was found to contain little or no sGCir. i-l sGC-immunohistochemistry in PN4 (Fig. 6d, e) revealed that while neighboring PNs appeared to contain strong labeling, this PN and its projection (arrow) did not contain high levels of sGCir. Calibration 100 μM; inset is 20 μM. Orientation D: dorsal; L: lateral