Olfactory tuning does not correlate with the recording depth

We found no significant correlation between olfactory tuning under voltage clamp and the neuronal recording depth; this was the case for both EPSCs and IPSCs (see Figure 7-4A for r and p values). We have shown in a previous chapter that excitatory tuning under current clamp conditions is graded across the PC cortical axis; however, this correlation was only revealed when the AHP was used as a proxy for neuronal identity (see Figure 5-3). Note that it was not possible to measure the AHP in voltage clamp experiments due to the use of a Cs-based internal solution.

An additional difficulty is that it appears that most of the recordings were made from locations within layer 2b, where SP cells are found (Figure 7-4A, Recording depth). This biased sampling may have resulted from difficulties associated with the ‘blind’ patching process, as we may have inadvertently favoured deeper neurons when none were encountered at more superficial locations. If most of the neurons in the dataset were indeed SP cells, tuning differences across the cortical superficial-to- deep axis will likely be concealed (see Chapter 8. General discussion). Nevertheless, we found that excitatory and inhibitory tuning of principal neurons correlated with each other (see Figure 7-4B for r and p values), suggesting that principal neurons (or at least SP cells) appear to receive excitatory and inhibitory inputs with matched or correlated odour tuning.

Figure 7-4 Olfactory tuning measured under voltage clamp does not correlate with the recording depth

A. Plots showing the tuning index and the recording depth of principal neurons for EPSCs (red, left) and IPSCs (blue, right). r and p values as shown; EPSC, n = 27 cells; IPSC, n = 18 cells, Pearson’s correlation. B. Excitatory and inhibitory tuning indices (voltage clamp) are plotted against each other. Each point represents a single neuron in the dataset. r and p

7.4

Discussion

In this chapter, we show that odour-evoked EPSCs and IPSCs can be quantified based on their magnitude (z-score) and respiratory entrainment (cross correlation) relative to baseline levels. Due to the abundant spontaneous activity in vivo, particularly with respect to synaptic inhibition, we found that the quantification of odour responses (and the subsequent estimation of olfactory tuning indices) benefits from the additional cross correlation analysis. It was found that the amalgamation of the two analyses significantly increased the measured excitatory and inhibitory tuning indices (Figure 7-3 A & B); the increase in inhibitory tuning was particularly substantial. These results are consistent with other findings of this study: 1) we show that tonic inhibitory activity is high under in vivo conditions, which frequently prevented odour-evoked IPSCs from being detected using the z-score method; 2) we show that odour stimulation selectively enhances the respiratory patterning of IPSCs more than EPSCs (Figure 6-4 and Table 6-2). For these reasons, it is unsurprising that the addition of cross correlation analysis had a stronger effect on the measured inhibitory tuning.

Our current clamp data indicate that principal neurons are more broadly inhibited than excited by odours (see Section 5.3.2 Excitatory tuning is graded across the PC cortical axis). However, even though neuronal inhibitory tuning measured under voltage clamp conditions appeared broader than its excitatory counterpart, this difference was not statistically significant (p = 0.19; Figure 7-3C). Similar to the current clamp results, we found no direct correlation between the somatic depth and olfactory tuning (excitatory and inhibitory). However, it was not possible to examine the relationship between the AHP and neuronal tuning under voltage clamp due to the use of a Cs-based internal solution. We reasoned that these results could be explained by the fact that we may have inadvertently measured mostly from SP and DP cells in voltage clamp experiments, the implications of which are explored in Chapter 8. General Discussion.

Chapter 8. General Discussion

The aim of the current study was to systematically investigate the olfactory tuning of PC principal neurons using whole-cell patch clamp electrophysiology in vivo. In the first part of the study we conducted current clamp experiments to examine odour responses of principal neurons under physiological conditions (Chapters 3~5); in the second part of the study, we conducted voltage clamp experiments in order to separately study odour-evoked excitatory and inhibitory responses (Chapters 6 & 7). The main findings of the study and their interpretations are summarised in the body of this general discussion.

Although the core focus of this study was to investigate olfactory tuning, we were also struck by the diversity of neuronal odour responses. We came to the conclusion that odour responses are variable and richly nuanced, exhibiting diverse properties including temporal synchrony (respiration & LFP oscillations), response adaptation and post-stimulus dynamics. Despite the complexity of odour responses, we simplified matters by focussing only on the supra- and sub-threshold magnitudes of responses during odour presentation relative to spontaneous activity (effectively discarding information that may be encoded in the temporal pattern of responses or post-stimulus dynamics). We chose to do this in part because of recent reports that odorant identity is encoded in the sniff-locked spike rate and not in the phase of the spikes relative to respiration (Miura et al., 2012; Gire et al., 2013). Interestingly, another study showed that odour-evoked spike trains are time-locked to distinct phases of the LFP beta oscillation (Poo and Isaacson, 2009), hinting at the possibility that temporal synchrony on a finer timescale may be informative. Such findings do not necessarily contradict the rate code, as the two are not mutually exclusive. Indeed, instances in which a common neural substrate streams or multiplexes different stimulus information through rate and temporal codes have been documented in different sensory systems (Friedrich et al., 2004; Harvey et al., 2013). Relevantly, Friedrich and colleagues showed that in the OB of zebra fish, information regarding odorant identity and odour category are separately encoded in the rate response and the spike-LFP synchrony, respectively (Friedrich et al., 2004). Whether the spike-LFP phase relationships convey sensory information in the mammalian system, however, remains to be elucidated. Nevertheless, recent data indicate that information regarding odour value is transmitted in the non-sniff-locked spike rate,

in parallel to information regarding odour identity, suggesting that sensory information is indeed multiplexed in the PC of the mouse (Gire et al., 2013). In

summary, the PC neural network performs remarkable transformation of afferent sensory input, which ultimately enriches odour perception and supports odour- guided behaviours.