Noradrenergic and cholinergic modulation of olfactory bulb sensory processing

Abstract

Neuromodulation in sensory perception serves important functions such as regulation of signal to noise ratio, attention, and modulation of learning and memory. Neuromodulators in specific sensory areas often have highly similar cellular, but distinct behavioral effects. To address this issue, we here review the function and role of two neuromodulators, acetylcholine (Ach) and noradrenaline (NE) for olfactory sensory processing in the adult main olfactory bulb. We first describe specific bulbar sensory computations, review cellular effects of each modulator and then address their specific roles in bulbar sensory processing. We finally put these data in a behavioral and computational perspective.

Keywords: acetylcholine; neuromodulation; noradrenaline; olfactory bulb.

Figure 1

Olfactory bulb processing. (A) Schematic of olfactory bulb organization. Olfactory sensory neurons (OSN) expressing a common receptor and therefore exhibiting similar odor receptive fields project to common glomeruli in the olfactory bulb (MOB). Within this glomerulus, OSNs make excitatory synapses onto mitral and tufted cells, the primary output neurons of the MOB, as well as glomerular layer interneurons comprising periglomerular (PG) and external tufted (ET) cells. Most PG cells (~70%), and a third type of interneuron, short axon cells (SA), are not directly activated by OSNs. PG, ET, and SA cells form intricate networks within the glomerular layer that have been proposed to perform operations such as normalization, contrast enhancement and synchronization. In deeper processing layers mitral cells (Mi) interact with at least one other class of interneurons, granule cells (Gr). These provide extensive feedback and lateral interactions between mitral cells by interacting with their elongated secondary dendrites. This layer of processing is thought to be involved in creating olfactory bulb oscillatory rhythms and generating synchronized spike patterns. Noradrenergic inputs from the locus coeruleus activate three classes of noradrenergic receptors distributed across the MOB. NE α1 receptors are thought to be predominantly located on Mi and Gr cell bodies as well as secondary dendrites with a sparser distribution in the glomerular layers; NE α 2 receptors are mainly located on granule cells with a sparse distribution on Mi cell bodies and the glomerular layer; NE α receptors have been reported on Mi cell bodies and in the glomerular layer. Muscarinic ACh receptors are thought to be located on granule cells (mACh) whereas nicotinic ACh receptors are located on mitral and periglomerular cells (nACh). (B) Schematic depiction of glomerular layer functions and Mi cell activity patterns in response to odor stimulation. (B_i–B_iii) show simulated distributed odor responses at two concentrations with lower concentration in the left column and higher concentration in the right column. The two-dimensional simulations are color coded with red indicating high and dark blue low levels of activity. (B_ii) shows the same patterns after amplitude-invariance processing has been performed by the network of local interneurons and (B_iii) shows the same pattern after contrast enhancement. The patterns in (B_iii) represent the end result of the glomerular computations transmitted to deeper layers by Mi cells. The details of these computations are given in Cleland and Sethupathy (2006); Cleland et al. (2007). (B_iv) shows how spikes generated in Mi cells in response to activation patterns (left side) are transformed into sparser, oscillatory and highly synchronized spike patterns by the interactions with deeper interneuron networks (Mandairon et al., ; Escanilla et al., 2010).

Figure 2

Schematic depiction of cholinergic modulation of glomerular contrast enhancement. (A) Within OB glomeruli, mitral and PG cells receive direct inputs from OSNs. PG cells have a higher input resistance and hence respond more quickly and saturate earlier than mitral cells. As a consequence, PG cells inhibit mitral cells in response to low affinity odorants but mitral cells override this inhibition in response to high affinity odorants. This microcircuit leads to contrast enhancement within each glomerulus, independent of location and receptive field (Cleland and Sethupathy, ; Cleland and Linster, 2012). The graph shows mitral (black) and PG (gray) cell activation levels as a function of varying the odor affinity between the receptor expressed in OSNs projecting to that glomerulus. PG cells inhibit mitral cells, resulting in a final activation (red) that is sharpened with respect to the incoming OSN signal. The schematic network below depicts the types of spike trains that might result from these interactions for odors located outside, at the lower end and in the middle of the OSNs receptive field. (B) Because ACh activates PG cells via nicotinic receptors, PG cell receptive fields (gray) are enlarged and as a consequence mitral cell receptive fields are sharpened (red). Nicotinic depolarization of mitral cells increases the amplitude of response within the positive part of the receptive field (dotted red line). As a consequence, mitral cell receptive fields are sharpened by ACh and contrast between odorants activating overlapping OSN populations is increased.

Figure 3

Schematic depiction of cholinergic modulation of deeper layer synchronization and dynamics. (A_i) OSN activation by odorants is processed in the glomerular layer (Figure 2) and creates an odor-specific response in mitral cells. In the example shown here, mitral cells respond with a spike train and in turn activate granule cells though excitatory synapses. (A_ii) Nicotinic ACh modulation depolarizes mitral cells and, at the same time, muscarinic receptor activation renders granule cells more responsive to input (Castillo et al., ; Pressler et al., 2007). Together, these enhance the excitatory-inhibitory feedback loop between these two groups of cells, increasing oscillatory dynamics and mitral cell spike synchronization. (B) Simulation of cholinergic effects in a computational model of olfactory bulb. (B_i) The graph shows the membrane potential and action potentials of a subset of granule cells in the model. When ACh is “ON,” granule cells are depolarized and fire more easily in response to odor stimulation. (B_ii) shows the corresponding membrane potential and action potentials of modeled mitral cells. When ACh is “ON,” mitral cells are more excitable due to activation of nicotinic receptors. The additional excitation is balanced by the higher inhibitory inputs from granule cells shown in (B_i). This balance does not significantly change mitral cell synchronization but because the loop is more strongly activated, both mitral and granule cells become more oscillatory and their action potentials are more synchronized (synchronization index calculates the number of synchronized spikes divided by the total number of spikes). The graph also shows the receptive field sharpening in mitral cells which is due to the nicotinic activation of PG cells described in Figure 2. (C) Summary of cholinergic modulation. ACh in the glomerular layer sharpens mitral cell receptive fields (Figure 2), whereas ACh modulation in deeper layers increases synchronization among odor responsive mitral cells.

Figure 4

Schematic depiction of noradrenergic modulation of odor sensitivity. (A_i) OSN inputs are processed in the glomerular layer and result in activation of a subpopulation of mitral cells. The spiking response of these active cells is further influenced by inhibitory inputs from granule cells. In the example shown here the applied stimulus was below threshold in amplitude, resulting in no spiking response in mitral cells. (A_ii) NE α2 activation (at very low NE concentrations) decreases granule cell spontaneous activity, and disinhibits mitral cells (Nai et al., 2009, 2010). Mitral cells are more responsive to odorants and to low amplitude electrical stimulation (Jiang et al., ; Ciombor et al., 1999). At higher NE concentrations, α1 receptors are activated, granule cell spontaneous activity increases and mitral cells are inhibited (Nai et al., 2009, 2010); at the same time activation of mitral cell α1 receptors depolarizes mitral cells, rendering them more responsive to stimulation. As a consequence, α1 activation leads to increased sensitivity while preserving discrimination (Linster et al., 2010). (B) Behavioral experiments show that odor detection at near-threshold concentration initially decreases, then increases with increasing concentrations of NE, following the non-linearities observed in brain slice recordings (Escanilla et al., ; Linster et al., 2010). The graph shows the degree of odor detection measured in a spontaneous odor detection task as a function of NE concentration infused directly into the OB of adult rats.

Figure 5

Schematic depiction of noradrenergic modulation of signal-to-noise ratio in the EPL. (A_i) OSN inputs are processed in the glomerular layer and result in activation of a subpopulation of mitral cells. The spiking response of these active cells is further influenced by inhibitory inputs from granule cells. (A_ii) NE α1 activation increases granule cell spontaneous activation, resulting in less spontaneous in activity in the connected mitral cells. At the same time, a1 activation lowers mitral cell response thresholds, rendering them more sensitive to odor inputs. As a consequence, α1 activation leads to increased signal-to-noise ratio. (B_i–B_ii) Simulations [from Linster et al. (2010)] of a1 effects in the OB network. The graphs show the membrane potential and action potentials of modeled mitral cells during spontaneous and odor triggered activity. Odor stimulation is shown in grey below the traces. In the unmodulated state (B_i), spontaneous activity is high and odor triggered activity comparatively low. In the modulated state (B_ii), spontaneous activity is low (low noise), and odor triggered response comparatively strong (high signal).

Figure 6

Summary of behavioral results. (A) Spontaneous odor discrimination task. (A_i) Rats are first habituated to an odor during four habituation trials (H1-H4). A significant decrease in odor investigation time over the course of these trials indicates the formation of an odor memory. Rats are then presented with a test odor (T) of variable similarity to the habituated odor; a significant increase in investigation indicates that rats discriminate between the habituated and the test odor (Cleland et al., 2002). (A_ii) Manipulations of cholinergic or noradrenergic modulation in the olfactory bulb in these experiments did not affect rats' formation of the odor memory; in our hands, all rats habituated to the odorants over the course of several experiments manipulating Ach or NE activity in the OB (Mandairon et al., , ; Escanilla et al., 2010). (A_iii) Manipulation of ACh and NE activity in the OB modulated odor discrimination in this task. In summary, test odorants NOT discriminated by control rats (C1, usually a straight chain aldehyde with a 1 carbon difference to the habituated odor), could be discriminated when increasing cholinergic (+ACh) or noradrenergic (+NE) modulation in the OB (Mandairon et al., , ; Escanilla et al., 2010). On the other hand, test odorants discriminated by control rats (two or more carbons removed from the habituated odorant) are less well-discriminated when nicotinic ACh receptors (−nACh) or α1 NE receptors (−α1 NE) are blocked (Mandairon et al., , ; Escanilla et al., 2010). (B) Spontaneous odor detection task. (B_i) Rats are first habituated to the odor carrier, mineral oil during three or four habituation trials (MO). Rats are then presented with a test odor (T) of variable concentrations; a significant increase in investigation indicates that rats detect the odorant at that concentration. (B_ii) Manipulations of cholinergic or noradrenergic modulation in the olfactory bulb in these experiments did not affect rats' response to mineral oil. (B_iii) Manipulation of NE but not ACh activity in the OB modulated odor discrimination in this task. In summary, test odorant concentrations NOT detected by control rats (10⁻⁴ Pa) were detected by rats infused with additional NE into their bulbs (+NE), in contrast, infusion of the non-specific cholinergic agonist CCh (+ACh) had no effect on odor detection. Similarly, odor concentrations easily detected by control rats (10-2 Pa) were detected by rats with blocked ACh receptors (−ACh) but not rats with blocked NE receptors (−NE). The blockade of NE receptors was shown to be specific to α1 receptors (Mandairon et al., , ; Escanilla et al., 2010, 2011). (C) Rewarded discrimination task. (C_i) In this task, rats are presented with two odorized cups and learn to dig for a reward in the rewarded odor (O1 + R) and to ignore the non-rewarded odor (O2–R) over the course of 20 trials (Cleland et al., 2002). (C_ii) Rats with cholinergic receptors blocked (−ACh) learn a discrimination between highly similar odorants at similar rates to saline infused control rats whereas rats with all NE receptors blocked (−NE) made significantly more mistakes over the course of a session (Mandairon et al., , ; Escanilla et al., 2010, 2011). The graph shows the percentage of correct choices made by the rats over the course of 20 trials as a function of drug treatment and test odor concentration. (D) Rewarded detection task. (D_i) In this task rats are trained to retrieve a reward from an odorized cup (O + R) presented at the same time as an unodorized cup (MO–R). Detection thresholds are measured by using variable near-threshold concentrations for the odorized cup and recording the percentage of correct choices made. (D_ii) In this task, rats with NE receptors blocked (−NE) were significantly impaired at detecting low concentration odorants as compared to saline infused control rats and rats with cholinergic receptors blocked (−ACh). The graph shows the average number of correct trials during a session as a function of drug treatments (Escanilla et al., 2011).

Figure 7

Cholinergic and noradrenergic modulation of granule cells. (A) Simplified integrate and fire model of a granule cell. (A_i) Baseline condition. The cell is driven with 100 ms pulses of current injection separated by 100 ms. (A_ii) Activation of muscarinic receptors changes afterhyperpolarization into afterdepolarization; as a consequence the cell is more active during the interstimulus intervals. (A_iii) Activation of α1 NE receptors depolarizes the cell leading to an increase in the evoked response followed by a strong hyperpolarization during the interstimulus intervals. (A_iv) When driven by a short current pulse (1 ms), the baseline granule cells emits a single spike followed by a small after hyperpolzarization (upper trace) whereas the granule cell under muscarinic activation responds with a spike followed by a afterdepolarization capable of triggering additional spikes. (B) In this simulation, a single mitral cell (B_ii) receives 100 ms current injections, drives an attached granule cell and is inhibited by this same granule cell (B_i). (C) In this simulation the granule cell undergoes modulation of its after hyperpolarization in response to muscarinic receptor activation similar to that shown in A. Note the increased firing of the granule cell (C_i) accompanied by a decrease in firing of the connected mitral cell (C_ii). (D) Effect of a1 NE activation on a single mitral (D_ii) and granule (D_i) cell loop. Both cells are more excitable and fire more during the current injections is observed. (E) Effect of cholinergic and noradrenergic modulation of granule cells on mitral cell spike latency. The graph shows the average latency to first spike during respiration cycles 1, 2, and 3 as well as the standard deviation of the latency. Under control conditions, the latency does not vary across respiration cycles (Control). When muscarinic receptors on granule cells are activated, the average latency as well as the standard deviation is decreased over successive respiration cycles (ACh). In contrast, activation or noradrenergic a1 receptors on granule cells does not affect spike latency nor its standard deviation (NE).