Immature olfactory sensory neurons are intrinsically excitable and show maturation-dependent changes in voltage-gated Na+ and K+ currents
Chiara Ricci, Cesar Adolfo Sánchez Triviño, Uday Rangaswamy, Lorenza Tortella, Remo Sanges, Anna Boccaccio, Anna Menini
TL;DR
This study reveals how immature olfactory neurons differ in electrical activity from mature ones, shedding light on their role in odor processing.
Contribution
The study identifies maturation-dependent changes in voltage-gated ion currents in olfactory sensory neurons.
Findings
Immature OSNs show spontaneous firing at lower frequencies and have distinct electrical properties compared to mature neurons.
Voltage-gated Na⁺ and K⁺ currents in immature OSNs differ significantly from those in mature OSNs.
Transcriptomic data align with functional changes, showing upregulation of ion channel genes during maturation.
Abstract
Olfactory sensory neurons (OSNs) detect odorants and send electrical signals to glomeruli in the olfactory bulb. Unlike most neurons, OSNs are continuously regenerated throughout life and immature neurons contribute to odorant-evoked responses in glomeruli. However, their intrinsic excitability properties are largely unknown. Here, we used acute slices of the olfactory epithelium from neonatal OMP-GFP mice to visually identify mature and immature OSNs and performed patch-clamp recordings to investigate their functional properties. Loose-patch recordings showed that immature OSNs display spontaneous firing at lower frequency than mature neurons. Whole-cell recordings showed that immature OSNs have more depolarized resting potentials, higher input resistance, fire only with phasic patterns, and generate slower action potentials with more depolarized thresholds. Instead, mature OSNs…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —Scuola Internazionale Superiore di Studi Avanzati - SISSA
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsOlfactory and Sensory Function Studies · Neurobiology and Insect Physiology Research · Insect Pheromone Research and Control
Introduction
Olfactory sensory neurons (OSNs) in the olfactory epithelium detect odorant molecules from the environment and transduce their binding to odorant receptors (ORs) into action potentials that are transmitted to the olfactory bulb (OB). OSNs are bipolar neurons that extend an axon to OB glomeruli and a dendrite toward the epithelial surface, with several cilia emerging from the apical part of the dendrite. Airborne odorants interact with ORs on these cilia and initiate a well-characterized transduction cascade. The binding of odorants to ORs activates adenylyl cyclase 3 (AC3) via the G protein G_olf_, leading to a rise in cAMP concentration and the opening of cyclic nucleotide-gated (CNG) channels that allow Ca²⁺ influx. The increase in intracellular Ca²⁺ concentration activates the Ca²⁺-activated Cl^-^ channel TMEM16B (ANO2), generating an inward current that amplifies the CNG current. The resulting depolarization triggers action potentials that are transmitted to the OB [4, 12, 18, 36, 56].
The strategic localization of OSNs at the interface with the external environment enables highly efficient odorant detection, but also makes OSNs particularly vulnerable to environmental stress and damage. However, under physiological conditions, OSNs are continuously replaced throughout life and can also be regenerated after injury. Their lifespan typically ranges from 1 to 3 months, although survival is variable and influenced by both intrinsic and extrinsic factors [17, 28, 35, 40, 42, 46, 49].
Neurogenesis in the olfactory epithelium is sustained by basal stem cells, located near the basal lamina. As newly generated OSNs differentiate, they undergo sequential molecular and morphological changes, with immature OSNs initially expressing the growth-associated protein 43 (Gap43) and the G-protein γ-subunit 8 (Gγ8). Over the course of 7–10 days, these markers are downregulated and mature neurons express the olfactory marker protein (OMP), their well-established marker [24, 34, 40, 51, 61, 64, 70, 73]. Immature OSNs express multiple low-level OR transcripts and components of the cAMP transduction cascade, whereas mature OSNs are well known to express only a single type of OR allele from about 1,000 functional ORs [24, 30, 55, 61]. From a morphological point of view, immature OSNs have short dendrites and their soma are located more basally than those of mature OSNs [20, 21, 49, 50, 68].
At the functional level, mature OSNs have been extensively characterized [33, 48, 67], but little is known about immature OSNs [8, 29]. Cheetham et al. (2016) have shown that immature OSNs can already form functional synapses in the OB and optogenetic photoactivation of their axons elicits robust firing in OB neurons. Moreover, two-photon Ca²⁺ imaging of glomeruli innervated by immature OSNs showed odorant-evoked responses, indicating that these neurons respond to odorants. The response properties of glomeruli innervated by both immature and mature OSNs differed, suggesting that the two neuron populations provide complementary odorant information to individual glomeruli and highlighting the functional relevance of immature OSNs in odorant-evoked responses [29].
Despite their functional relevance, the spontaneous activity and intrinsic excitability properties of immature OSNs remain largely unknown, leaving a gap in our understanding of how these developing neurons contribute to early sensory processing. To address this, we performed patch-clamp recordings in acute slices from neonatal OMP-GFP mice to identify and characterize immature OSNs and compare their properties with those of mature OSNs. We found that immature OSNs are capable of spontaneous firing, though at a lower frequency than mature neurons and have reduced intrinsic excitability. Voltage-gated Na^+^ currents showed maturation-dependent changes in activation and TTX sensitivity, while transient A-type K^+^ currents were observed only in mature neurons. Analysis of publicly available single-cell transcriptomic data further confirmed differential expression of voltage-gated Na^+^ and K^+^ channel genes during OSN maturation.
Materials and methods
Ethical approval
Mice were handled in accordance with the guidelines of the Italian Animal Welfare Act and European Union guidelines on animal research, under a protocol approved by the SISSA Animal Care Committee. All experiments were performed on tissues from OMP-GFP homozygous mice, in which OMP is replaced by GFP and mice are therefore knockout for OMP [57], at postnatal days P0-P4. Every effort was made to minimize the number of animals used.
Acute slices of mouse olfactory epithelium
To obtain acute coronal slices of the olfactory epithelium, P0-P4 mice were decapitated and the heads, after skin removal, were dissected. The method was similar to that previously described [1, 26, 59]. The nose was embedded in 3% Type I-A Agarose prepared in artificial cerebrospinal fluid (ACSF) solution, once the agar had reached a temperature of 37 °C. The ACSF contained (in mM): 120 NaCl, 25 NaHCO_3_, 5 KCl, 1 MgSO_4_, 1 CaCl_2_, 10 HEPES and 10 glucose, pH 7.4. Once solidified, the agar block was fixed in a metal chamber filled with cold, oxygenated ACSF. Coronal slices 300 μm thick were cut using a vibratome (Vibratome 1000 Plus Sectioning System) and kept in cold oxygenated ACSF until use.
Confocal imaging
Imaging was performed on acute coronal slices cut with the vibratome. Slices were fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline solution (PBS) for 1 h at 4 °C. After washing in PBS, slices were stained with DAPI (2.5 µg/ml) in 0.2% Tween-20 in PBS for 2 h at room temperature to label cell nuclei. After additional wash in PBS, slices were mounted with Fluoromount-G (Thermofisher). Images were acquired using a Nikon A1R confocal microscope equipped with a 40X objective. Z-stack images were acquired at a resolution of 1024 × 1024 pixels with a 2X zoom using NIS Elements software (Nikon) and further analyzed with ImageJ 1.54f software (NIH) to merge the different channels. Projections of 18 μm optical sections at maximal intensity were displayed.
Electrophysiological recordings
Acute coronal slices were anchored to the recording chamber using a custom-made U-shaped silver wire and continuously perfused with oxygenated ACSF. All the experiments were performed at room temperature (20–25 °C). The tissue was visualized with an upright microscope (BX51WI Olympus) equipped with a 40x water-immersion objective. Fluorescent neurons were identified using a mercury lamp (U-LH100HG Olympus) and a GFP filter set for GFP-expressing neurons or a red fluorescence filter for Alexa Fluor 594-labelled cells.
Extracellular solutions were delivered through an eight-in-one multibarrel perfusion pencil connected to a ValveLink 8.2 pinch-valve perfusion system (Automate Scientific). The perfusion pencil, with a tip diameter of 360 μm, was positioned approximately 0.5 mm from the slice.
Loose-patch recordings were obtained with patch pipettes of 2–4 MΩ filled with filtered ACSF. Seal resistances of 15–50 MΩ were established on the soma of OSNs to obtain the loose-patch configuration. Spontaneous activity was recorded in voltage-clamp mode with a holding potential of 0 mV.
Whole-cell recordings were performed from the soma of OSNs. For current- and voltage-clamp recordings, patch pipettes of 4–6 MΩ were filled with an intracellular solution containing (in mM): 80 KGluconate, 60 KCl, 2 MgATP, 1 EGTA and 10 HEPES, adjusted to pH 7.2 with KOH. The same solution was used to isolate K^+^ currents in voltage-clamp experiments. For Na^+^ current isolation, a Cs^+^-based intracellular solution was used, containing: 135 CsCl, 5 NaCl, 5 EGTA and 10 HEPES, adjusted to pH 7.2 with NaOH. All intracellular solutions included Alexa Fluor-594 (1 µg/ml) to allow fluorescent visualization of cells (Fig. 1).Fig. 1. Imaging of immature and mature OSNs. (A) Confocal images of a 300 μm coronal slice of the olfactory epithelium from a P2 OMP-GFP mouse, cut with a vibratome. Left: brightfield image. Right: merged brightfield and fluorescence images showing DAPI-stained nuclei (blue) and GFP-positive mature OSNs (green). (B) Epifluorescence image of an immature, GFP-negative OSN, filled with Alexa Fluor-594 (shown in white) via the patch pipette in a coronal slice of the olfactory epithelium. Left: overlay of Alexa Fluor-594 fluorescence with the brightfield image, showing the bipolar morphology of the immature neuron. Right: fluorescence image of the same slice showing GFP-positive mature OSNs (green), with the image of the immature neuron extracted from the left panel and superimposed to indicate its relative location
Resting membrane potential was measured in current-clamp mode without current injection and input resistance was evaluated in voltage-clamp mode by holding the cell at −80 mV and applying a 10 mV hyperpolarizing pulse.
Voltage-clamp recordings were repeated at least twice and only stable recordings were considered.
Electrophysiological recordings were performed using a Multiclamp 700B amplifier controlled by a Digidata 1550B via the Clampex 10.7 software (Molecular Devices). Data were low-pass filtered at 2 kHz and sampled at 10–20 kHz. Patch pipettes were pulled from borosilicate capillaries (WPI) using a PC-10 puller (Narishige).
The following chemicals were prepared as stock solutions as indicated and diluted in ACSF to the final concentrations on the day of the experiment: 1 mM tetrodotoxin citrate (TTX) (Latoxan) and 0.5 M 4-aminopyridine (4-AP) stored at −20 °C; 1 M CdCl_2_ stored at 4 °C.
All chemicals were obtained from Sigma-Aldrich, unless otherwise specified.
Data analysis and statistics
Data analysis and figure preparation were performed using IgorPro software (Wavemetrics) or ImageJ 1.54f (NIH).
Epifluorescence images taken during electrophysiological experiments were pseudocolored in ImageJ to enhance the visualization of OSNs. The area corresponding to the immature OSN (Fig. 1) was extracted and relocalized using the Intelligent Scissors and Interactive Boundaries tools in GIMP 2.10.28.
All data from different OSNs (n) are presented as mean ± standard deviation (SD). In voltage-clamp experiments, capacitive transients were cut off from figures for clarity.
Loose-patch recordings were filtered with a bandwith from 5 to 1500 Hz. Spikes were detected using a custom procedure based on an arbitrary threshold and verified by visual inspection of individual spike shapes. Mean firing frequency was calculated as the total number of spikes divided by 180 s recording duration. The interspike interval (ISI) was defined as the time interval between consecutive spikes, and the instantaneous frequency as the reciprocal of the ISI.
In current-clamp experiments, action potential parameters were determined as follows: half-width values were obtained by measuring the action potential width at 50% of its maximal amplitude. The action potential threshold was determined using phase-plot analysis by identifying the membrane potential at which the second derivative of voltage (d²V/dt²) reaches its first prominent peak. Maximal depolarization and repolarization were evaluated from phase-plot analysis of the first action potential, in which the rate of change of membrane potential over time (dV/dt) is plotted as a function of voltage (V).
Voltage-gated activation curves were fitted with the Boltzmann equation: G/G_max_ = 1/{1 + exp[(V_1/2_ -V)/k]} where G is the conductance, G_max_ is the maximal conductance, V_1/2_ is the membrane potential at which G is half of G_max_, V is the membrane potential and k is the slope factor. Statistical analysis was performed using Prism 8 (GraphPad) or IgorPro (Wavemetrics) software. Normal distribution of the data was assessed with the Shapiro-Wilk test and equality of variances with the F test. For normally distributed data with equal variances, an unpaired t-test was used to determine statistical significance. When variances were different, Welch’s correction was applied. Not normally distributed data were analyzed with the Mann-Whitney test. For multiple comparisons, one-way ANOVA followed by Tukey’s post-hoc test was performed. In the case of not normally distributed data, a Kruskal-Wallis followed by Dunn’s post-hoc test was used. For the I-V curves, statistical significance was assessed using a mixed model two-way ANOVA followed by Bonferroni correction.
Cumulative distributions were compared with the Kolmogorov-Smirnov test. P values < 0.05 were considered statistically significant. In the Figures, * indicates P < 0.05, ** P < 0.001, *** P < 0.0001.
Single-cell RNA seq analysis
Single-cell RNA seq analysis was performed on a published dataset of whole olfactory mucosa from five adult male C57BL/6 mice (8–12 weeks old) [7]. Analysis was performed in R (version 4.4.3) using the Seurat (version 5.3.0) [25] package. The raw count matrix and metadata were downloaded from the Gene Expression Omnibus (GEO) under accession number GSE151346. A seurat object was created using the CreateSeuratObject() function from this data. The dataset had been pre-processed for quality control metrics including total RNA counts (nCount_RNA), number of detected genes (nFeature_RNA), mitochondrial gene percentage (percent_mito), and doublet scores. Data were normalized using NormalizeData(), and the top 2000 variable features were identified with FindVariableFeatures(). These features were then scaled using ScaleData(). Principal component analysis was run on the scaled data with RunPCA(), and the first 30 principal components were used to generate a UMAP embedding with the RunUMAP() function.
Cell type refinement and differential expression analysis
To investigate the molecular profiles of mature and immature OSNs in a manner consistent with OSN identification in our electrophysiological experiments, we refined the original annotations of the OSN population. Cells were classified as immature OSNs (iOSNs) if they were negative for Omp detection while being positive for either Gap43 or Gng8. All cells positive for Omp were classified as mature OSNs (mOSNs). Differential gene expression analysis between these two populations was performed using the FindMarkers() function with default parameters. Genes were considered significantly differentially expressed if they had an adjusted p-value < 0.05, an absolute log₂ fold change > 0.25, and were expressed in at least 5% of cells in either population. Results were visualized using a volcano plot generated with the IgorPro (Wavemetrics) software.
Results
To visually identify immature and mature OSNs, we used homozygous OMP-GFP mice, in which OMP is replaced by GFP [57]. In olfactory epithelium slices, mature OSNs were easily identified by their GFP signal, whereas immature OSNs were non-fluorescent (Fig. 1A). Among the non-fluorescent cells, immature OSNs were identified based on their bipolar morphology and the location of their soma below the layer of GFP-positive neurons, toward the basal region of the olfactory epithelium. Figure 1B shows the typical bipolar morphology and soma location of an immature OSN. The image was obtained in the whole-cell configuration, in which the Alexa Fluor dye contained in the intracellular solution diffused into the neuron through the patch pipette, showing its morphology. In GFP-positive cells, the same intracellular solution enabled confirmation of successful recordings from mature OSNs through co-localization of GFP and Alexa Fluor signals.
Immature OSNs display spontaneous activity
Although the basal firing activity of mature OSNs in the absence of odorant stimulation is well established, the spontaneous firing activity of immature OSNs has remained uncharacterized. Using loose-patch recordings from the soma of OSNs in acute slices of the olfactory epithelium, we found that immature OSNs do have spontaneous firing in the absence of external stimuli, showing a wide range of firing patterns with frequencies ranging from 0.20 to 1.81 Hz (Fig. 2A, B, E). In comparison, mature OSNs had frequencies varying from 0.19 to 3.87 Hz (Fig. 2C, D, E). On average, immature OSNs fired at 0.78 ± 0.49 Hz (n = 16), whereas mature OSNs had a significantly higher mean firing frequency of 1.69 ± 0.98 Hz (n = 21; Fig. 2F).Fig. 2. Spontaneous firing activity differs between immature and mature OSNs. (A, C) Representative loose-patch recordings showing 180 s of spontaneous spiking activity from an immature (A) and a mature (C) OSN. Insets below each trace show regions indicated by the dashed boxes at an expanded time scale. (B, D) Raster plots of spontaneous activity from 10 representative immature (B) and mature (D) OSNs. Each row corresponds to the activity of a single neuron. (E) Scatter plots with averages ± SD of mean spontaneous firing frequencies from immature (n = 16) and mature (n = 21) OSNs (p = 0.0009, one-tailed t-test). (F) Normalized interspike interval (ISI) distributions (bin = 10 ms) for all recorded immature (orange) and mature (blue) OSNs. (G) Cumulative distribution of instantaneous firing frequencies (bin = 0.5 Hz) showing a rightward shift in mature compared to immature OSNs (p = 0.02, Kolmogorov–Smirnov test)
To further analyze these firing patterns, we calculated the interspike interval (ISI) distributions, which describe the probability of occurrence of time intervals between consecutive spikes. Immature OSNs displayed a broad ISI distribution that lacked the prominent peak at short intervals observed in mature neurons (Fig. 2F). This difference is further supported by the instantaneous firing frequency analysis, calculated as the inverse of ISI: the cumulative fraction of events for mature OSNs was significantly shifted to the right compared to that of immature OSNs (Fig. 2G).
Taken together, these results demonstrate that immature OSNs exhibit spontaneous firing activity, characterized by lower firing rates and distinct firing patterns compared to mature OSNs. These findings suggest that neuronal excitability undergoes significant functional changes during OSN maturation.
Maturation-dependent changes in evoked firing and membrane properties of OSNs
To further investigate maturation-dependent changes in functional properties, we recorded OSN responses to depolarizing current injections in the whole-cell current-clamp configuration. The membrane potential was held near − 65 mV by applying a small holding current when needed, and depolarizing current steps, ranging from 2 to 12 pA, were delivered in 2 pA increments from this baseline.
Immature OSNs responded to current injections with phasic firing patterns, generating one or a few action potentials followed by a sustained depolarized plateau (Fig. 3A). Mature OSNs, instead, had greater heterogeneity in their firing behavior, showing either phasic or tonic firing, with tonic firing characterized by repetitive action potentials throughout the depolarizing step (Fig. 3B, C). No spontaneous activity was observed in immature neurons, whereas some mature neurons (19%, 6 out of 32) fired spontaneously.Fig. 3. Evoked firing activity in immature and mature OSNs shows differences in excitability. (A–C) Representative whole-cell current-clamp recordings from immature (A) and mature OSNs (B, C) in response to 2 s current injections of 0, 2, 4, or 6 pA. The immature OSN in (A) displayed a phasic firing pattern, whereas mature OSNs exhibited either phasic (B) or tonic (C) firing patterns. The black dashed lines indicate –65 mV. (D-F) Expanded view of the first action potential evoked by 4 pA current injection from recordings in (A-C). (G–L) Scatter plots with averages ± SD for the indicated properties in immature (orange circles, n = 8), phasic firing (blue filled squares, n = 12) and tonic firing mature OSNs (open squares, n = 14). (G-I) Action potential threshold, maximum depolarization rate (Max Vdep) and maximum repolarization rate (Max Vrep) calculated at 4 pA from phase-plot analysis. (J) Action potential half-width duration at 4 pA. (K) Resting membrane potential. (L) Input resistance. Statistical analysis: data in G and K were analyzed using one-way ANOVA followed by Tukey’s post-hoc test, and data in panels H, I, J, and L were analyzed using Kruskal–Wallis test followed by Dunn’s post-hoc test. Comparisons showed that immature neurons differed significantly from phasic firing mature neurons (G, p= 0.002; K, p = 2.53 × 10⁻⁴; H, p = 0.06; I, p = 0.07; J, p = 0.048; L, p = 0.06) and from tonic firing mature neurons (G, p = 2.24 × 10⁻⁶; K, p = 1.35 × 10⁻⁵; H, p = 7.06 × 10⁻⁵; I, p = 2.20 × 10⁻⁴; J, p = 5.90 × 10⁻⁵; L, p = 3.34 × 10⁻⁵). Phasic and tonic firing mature neurons did not differ significantly in most panels: G, p = 0.05; K, p = 0.58; H, p = 0.11; I, p = 0.19; J, p = 0.13;L, p = 0.08
Current injections of 2 or 4 pA were sufficient to elicit firing in several immature and mature neurons. Since most neurons fired in response to a 4 pA current step, we selected this amplitude for quantitative analysis. At 4 pA, 57% of immature OSNs (8 out of 14) fired action potentials, while the remaining neurons required a higher current injection of 6 pA. At each current step that elicited firing, immature OSNs responded exclusively with a phasic firing pattern.
In mature OSNs, a 4 pA step evoked firing in 81% of neurons (26 out of 32), with 37% (12 out of 32) showing phasic firing and 44% (14 out of 32) tonic firing, whereas 19% (6 out of 32), did not fire at 4 pA but responded only to higher current injections (6–12 pA).
To compare excitability parameters, we examined the first action potential evoked by the 4 pA step (Fig. 3D-F) and analyzed some of its properties with phase-plot analysis. Immature OSNs had a more depolarized action potential threshold (−40.4 ± 4.8, n = 8) than both phasic (−50.1 ± 6.5 mV, n = 12) and tonic (−55.5 ± 5 mV, n = 14) mature neurons (Fig. 3G). The maximal depolarization and repolarization rates were smaller in immature compared with tonic mature neurons but did not differ significantly from phasic mature neurons (Fig. 3H, I). These slower kinetics are consistent with the broader action potential half-width calculated in immature (21.2 ± 9 ms, n = 8) compared with tonic (3.9 ± 2.5 ms, n = 14) and phasic mature neurons (8.5 ± 7.3 ms, n = 12; Fig. 3J).
Passive membrane properties also differed between groups. The resting membrane potential, measured at 0 pA holding current, was more depolarized in immature (–40.7 ± 12 mV, n = 8) than in both phasic (–62 ± 12.8 mV, n = 12) and tonic (–66.4 ± 6.7 mV, n = 14) mature OSNs (Fig. 3K). The input resistance was higher in immature (6.6 ± 2.9 GΩ, n = 8) than in tonic (1.2 ± 0.9 GΩ, n = 14) mature OSNs. A difference was also seen between immature neurons and phasic firing mature neurons (2.7 ± 1.6 GΩ, n = 12), which however did not reach statistical significance (Fig. 3L). Taken together, these results show that OSN maturation is associated with changes in both passive and active membrane properties. Immature neurons had a more depolarized resting potential and higher input resistance compared with mature OSNs. Immature neurons fired only with phasic patterns and had slower action potentials with a more depolarized threshold, whereas mature OSNs showed a diverse excitability profile and the emergence of tonic repetitive firing.
Maturation-dependent changes in voltage-gated currents of OSNs
To investigate the ionic mechanisms underlying differences in excitability, we recorded voltage-gated currents from the same immature and mature OSNs characterized in current-clamp experiments. Currents were measured from a holding potential of −80 mV using 200 ms voltage steps ranging from − 90 mV to + 50 mV in 10 mV increments. Depolarizing voltage steps elicited transient inward currents followed by outward currents in immature and mature neurons (Fig. 4A-C).Fig. 4. Current-voltage relationships of immature and mature OSNs. (A-C) Representative whole-cell voltage-clamp recordings from immature (A), phasic firing mature (B), and tonic firing mature (C) OSNs. Voltage steps of 200 ms duration were given from a holding potential of −80 mV to voltages between − 90 and + 50 mV in 10 mV steps. (D-F) Average ± SD current-voltage relationship for (D) peak inward currents and for outward currents measured at the beginning (E) or at the end (F) of the voltage steps. Recordings were obtained from the same OSNs analyzed in Fig. 3G-L: immature, n = 8; phasic firing mature, n = 12; tonic firing mature, n = 14. Statistical comparisons performed using two-way ANOVA. Immature vs. tonic firing mature: D, p = 0.002; E, p = 0.002; F, p = 5.72 × 10^− 4^. Immature vs. phasic firing mature: D, p = 0.44; E, p = 0.02; F, p = 0.004. Phasic firing mature vs. tonic firing mature: D, p = 0.01; E, p = 0.48; F, p = 0.29
The average current-voltage relationship indicates some differences between immature and mature neurons. For example, the onset of inward currents in immature neurons occurred at about − 50 mV, a more depolarized potential compared with the range between − 70 mV and − 60 mV measured in phasic and tonic firing mature OSNs. Peak inward currents occurred between − 20 and − 10 mV in immature neurons and at about − 30 mV in both phasic and tonic firing mature OSNs (Fig. 4D). Furthermore, outward currents in immature neurons were mainly sustained, whereas mature neurons showed an additional transient peak current at the beginning of the voltage step (Fig. 4A-C, E, F). The average current-voltage relationship indicates that outward current activation occurred at about − 10 mV in immature OSNs and between − 30 mV and − 20 mV in mature neurons (Fig. 4E). Immature neurons had smaller inward and outward current amplitudes compared with tonic firing mature neurons, whereas only outward currents were reduced relative to phasic mature neurons. These observations indicate maturation-dependent changes in voltage-gated currents in OSNs.
To better characterize inward and outward current components, we performed experiments under conditions that selectively isolated voltage-gated Na^+^ or K^+^ currents (Figs. 5 and 6).Fig. 5. Changes in Na^+^ channel activation and TTX sensitivity across OSN maturation. Representative whole-cell voltage-clamp recordings obtained with a Cs^+^-based intracellular solution from immature (A, B) and mature (G, H) OSNs. Voltage steps of 200 ms duration were given from a holding potential of −80 mV to voltages between − 90 and + 20 mV in 5 mV steps, or from a 100 ms prepulse to − 110 mV in control conditions or after application of 200 nM TTX. For clarity, only recordings every 10 mV are shown and capacitive transients were cut off. (C, I) Normalized conductance–voltage relationships for the representative immature (A, B) and mature (G, H) OSNs. Lines are fits to the Boltzmann equation. (D, J) V₁/₂ and k values from Boltzmann fits. Immature OSNs (D): p = 0.34 for V₁/₂; p = 0.83 for k, (paired t-test, n = 11); mature OSNs (J): p = 0.03 for V₁/₂; p = 0.03 for k (Wilcoxon test, n = 6). (E, K) Maximal peak inward currents from prepulses of −80 or −110 mV. Immature OSNs (E): p = 0.001 (paired t-test, n = 11); mature OSNs (K), p = 0.002 (Wilcoxon test, n = 10). (F, L) Percentage of maximal current inhibited by 200 nM TTX from prepulse of −80 or −110 mV. Immature OSNs (F), p = 0.35, n = 4 at −80 mV and n = 6 at −110 mV; mature OSNs (L), p = 0.002, n = 6 at −80 mV and − 110 mV (Mann-Whitney test). (D-F, J-L) Data are presented as scatter plots with averages ± SDFig. 6Voltage-gated K^+^ current components differ between immature and mature OSNs. (A) Representative whole-cell voltage-clamp recordings obtained from mature (A) or immature (B) OSNs using a K^+^-based intracellular solution. The extracellular solution contained 1 µM TTX and 100 µM CdCl_2_. Voltage-gated K⁺ currents were elicited by 800 ms voltage steps from a holding potential of −80 mV to voltages between − 50 and + 50 mV in 10 mV steps. (C) Average ± SD current-voltage relationship measured at the beginning (squares) or end (circles) of the voltage steps in immature (orange, n = 13) and mature (blue, n = 28). Statistical comparisons by two-way ANOVA: mature beginning vs. end, p = 1.33 × 10^− 6^; immature beginning vs. end, p = 0.28; mature vs. immature at beginning: p = 2.72 × 10^− 8^, mature vs. immature at end, p = 5.14 × 10^− 9^. (D) Representative currents from mature OSN elicited from an 800 ms prepulse potential of + 20 mV to inactivate transient currents. (E) Difference current obtained by subtracting the prepulse-inactivated current in (D) from the total current in (A). (F) Block of the transient current component following application of 5 mM 4-AP. (G) Difference current obtained by subtracting the 4-AP blocked current in (F) from the total current in (A). (H, I) Average ± SD current-voltage relationships for the subtracted currents measured at the beginning (diamonds) or at the end (triangles) of the voltage steps in mature OSNs (H: p = 0.1; I: p = 0.1, two-way ANOVA, n = 6)
To examine how voltage-gated Na^+^ currents change during maturation, we used a Cs^+^-based intracellular solution to block voltage-gated K^+^ channels and, in some experiments, added 100 µM CdCl_2,_ a blocker of voltage-gated Ca^2+^ channels. To investigate whether different Na^+^ channel subtypes are functionally expressed in immature and mature OSNs, we took advantage of biophysical and pharmacological differences, such as inactivation properties and sensitivity to TTX. Therefore, we applied prepulses to different potentials (−80 mV vs. −110 mV) to modify the availability of Na^+^ channels and tested whether Na^+^ currents were TTX-sensitive or TTX-resistant (Fig. 5).
In immature OSNs, hyperpolarization of the prepulse from − 80 mV to −110 mV produced a small increase in peak current amplitude (−0.63 ± 0.27 nA vs. −0.89 ± 0.34 nA, n = 11), indicating that a small fraction of channels were inactivated at −80 mV and hyperpolarization to −110 mV allowed recruitment of additional channels (Fig. 5A, B, E). Activation curves obtained by plotting the normalized conductance versus the step voltages were well fitted by a Boltzmann equation. Hyperpolarizing the prepulse from − 80 to −110 mV did not modify either the voltage dependence of activation or the slope factor (V_1/2_: −36.2 ± 6 mV vs. −37.2 ± 8.3 mV; k: 6.0 ± 2.3 mV vs. 6.0 ± 2.2 mV, n = 11; Fig. 5C, D). Application of 200 nM TTX almost completely blocked the current at both prepulse potentials, with an average current block of 89 ± 15% (n = 4) at – 80 mV and 87 ± 14% (n = 6) at −110 mV prepulse, indicating that all or the majority of Na^+^ currents in immature OSNs are TTX-sensitive (Fig. 5A, B, F).
In mature OSNs, currents showed a stronger dependence on the prepulse potential. Hyperpolarizing the prepulse from − 80 mV to −110 mV doubled the average peak current amplitude from − 0.70 ± 0.34 nA to −1.40 ± 0.31 nA (n = 10), shifted the voltage dependence of activation to more negative potentials, from V_1/2_ of −40.8 ± 8.6 mV to −54.0 ± 11.4 mV, and increased the slope factor (3.9 ± 2.0 mV vs. 4.9 ± 1.9 mV; n = 6) (Fig. 5G-K). Application of 200 nM TTX blocked most of the current at a prepulse of −80 mV (84 ± 12%, n = 6) but only partially at −110 mV (44 ± 11%, n = 6), revealing a substantial TTX-resistant component (Fig. 5G, H, L).
In a set of experiments designed to evaluate the possible contribution of Ca^2+^ channels, we added 100 µM CdCl_2_ to the extracellular solution to block voltage-gated Ca^2+^ channels. The activation curves from a prepulse of −110 mV in the presence of Cd^2+^ were similar to those measured in its absence, showing that voltage-gated Ca^2+^ channels did not significantly contribute to inward currents. In immature OSNs, V_1/2_ and the slope factor were − 35.9 ± 5.1 mV and 7.9 ± 2.9 mV (n = 7) in the presence of Cd^2+^, not significantly different from − 37.2 ± 8.3 mV and 6.0 ± 2.2 mV (n = 11) measured in the absence of Cd^2+^ (p = 0.72 for V_1/2_; p = 0.14 for k, unpaired t-test). In mature OSNs, the activation parameters were also unaffected by Cd^2+^: V_1/2_ and the slope factor were − 47.8 ± 10.2 mV and 7.2 ± 3.0 mV (n = 15) with Cd^2+^, compared with − 54.0 ± 11.4 mV and 4.9 ± 1.9 mV (n = 6) without Cd^2+^ (p = 0.24 for V_1/2_, unpaired t-test; p = 0.13 for k, Mann-Whitney test).
The quantitative comparison of voltage-gated Na^+^ currents shows that currents in immature OSNs are mainly carried by TTX-sensitive channels, whereas in mature OSNs both TTX-sensitive and TTX-resistant channels contribute to the total current, with channel availability strongly dependent on prepulse potential, indicating that different Na^+^ channel subtypes are functionally expressed during maturation.
To isolate voltage-gated K^+^ currents, we recorded from a holding potential of −80 mV and blocked voltage-gated Na^+^ and Ca^2+^ channels with 1 µM TTX and 100 µM CdCl_2_. Currents were recorded in response to 600 ms depolarizing steps ranging from − 50 mV to + 50 mV in 10 mV increments (Fig. 6A). Consistent with data shown in Fig. 4, outward currents in immature neurons were mainly sustained, whereas mature neurons showed both transient and sustained components. These differences are more evident in Fig. 6A and B, where isolation of K^+^ currents and the longer step duration (600 ms instead of 200 ms) allow the distinct kinetic components to be more clearly resolved. In immature neurons, the current-voltage relationship measured at the onset and at the end of the voltage steps were nearly identical, indicating the absence of a transient component (Fig. 6 B, C). In mature neurons, instead, currents were larger at the beginning than at the end of the voltage step over the − 20 and + 50 mV range, reflecting the presence of a transient component (Fig. 6A, C). Furthermore, the amplitude of outward currents was higher in mature compared to immature OSNs (Fig. 6C).
To further characterize the outward current components in mature OSNs, we combined biophysical and pharmacological approaches. First, we applied the same series of depolarizing steps either from a holding potential of −80 mV or after a depolarizing prepulse to + 20 mV designed to inactivate the transient current (Fig. 6A, D). Subtracting the prepulse-inactivated current from the total current revealed a clear transient current followed by a plateau (Fig. 6E). We next examined the pharmacological sensitivity of the outward currents using 4-aminopyridine (4-AP), a well-established blocker of transient K^+^ currents [22, 62, 63]. Both application of 5 mM 4-AP abolished the transient component and subtraction of the 4-AP-resistant current from the total current gave current kinetics similar to those obtained with the prepulse protocol (Fig. 6F, G). On average, both experimental procedures produced similar current-voltage relationships for the transient current, as well as comparable amplitudes of the remaining plateau currents, indicating that the two approaches isolated the same current components (Fig. 6H, I).
The comparison of voltage-gated K⁺ currents shows the absence of a transient component in immature OSNs, whereas mature OSNs have a prominent transient, 4-AP–sensitive K⁺ current, indicating a maturation-dependent change in voltage-gated K⁺ channels.
Public scRNA-seq data analysis suggests a maturation-linked transcriptional shift related to voltage-gated Na+ and K+ channels
To investigate the molecular basis underlying the electrophysiological differences in voltage-gated currents between immature and mature OSNs, we analyzed their transcriptional profiles using a published single-cell RNA sequencing dataset of the whole olfactory mucosa from C57BL/6 mice [7]. The original study assigned cell types to 29,585 cells based on canonical marker gene enrichment within unsupervised clusters. We further refined the annotation of OSNs by classifying as immature cells lacking Omp expression but expressing Gap43 or Gng8, whereas cells expressing Omp were classified as mature OSNs (Fig. 7A) [24]. We then performed differential gene expression analysis between mature and immature OSNs (Supplementary Table 1) and specifically examined genes encoding voltage-gated Ca^2+^ α subunits, Na^+^ α and β subunits, and K^+^ channels of the Shaker-,* Shab-, Shaw-*, and *Shal-*related families (Fig. 7B) [22]. However, while no genes related to voltage-gated Ca^2+^ channels resulted upregulated, several Na⁺ channel genes were upregulated in mature OSNs, including Scn5a, Scn8a and Scn9a, which encode the Na_v_1.5, Na_v_1.6 and Na_v_1.7 α-subunits, respectively, as well as Scn3b and Scn4b, which encode the β3 and β4 auxiliary subunits (Fig. 7B). Na_v_1.6 and Na_v_1.7 are TTX-sensitive channels, whereas Na_v_1.5 is TTX-resistant [19].Fig. 7scRNA-seq analysis reveals differential expression of voltage-gated channel genes in mature and immature OSNs. (A) Uniform Manifold Approximation and Projection (UMAP) plot from the whole olfactory mucosa dataset from Brann et al., 2020. Mature OSNs (mOSNs, blue) are Omp+ and immature OSNs (iOSNs, orange) are Omp^−^ and Gap43^+^ or Gng8^+^. (B) Volcano plot of differentially expressed genes between mOSNs and iOSNs. Genes encoding voltage-gated Na^+^ and K^+^ channel subunits are highlighted. Dashed vertical lines indicate log_2_ fold-change thresholds at ±0.25, and the dashed horizontal line marks the adjusted P-value cutoff at 0.05. Upregulated Na^+^ channel genes in mOSNs include Scn5a, Scna8 and Scn9a (α subunits) and Scn3b and Scn4b (β subunits); upregulated K^+^ channel genes in mOSNs include Kcna2, Kcnc4, Kcna5, Kcna6 and Kcnb1
Voltage-gated K⁺ channel genes were also upregulated in mature OSNs, including Kcna2, Kcnc4, Kcna5, Kcna6 and Kcnb1, with Kcnc4 showing the highest statistical significance (Fig. 7B). These genes encode the K_v_1.2, K_v_3.4, K_v_1.5, K_v_1.6 and K_v_2.1 channels, respectively. K_v_3.4 channels primarily mediate fast-inactivating A-type K^+^ currents, whereas K_v_1.2, K_v_1.5, K_v_1.6 and K_v_2.1 channels mainly contribute to non-inactivating or slowly inactivating delayed rectifier K⁺ currents [22].
No genes related to the examined voltage-gated channel families were significantly downregulated.
These transcriptional differences indicate an upregulation of some Na^+^ and K^+^ channel genes during OSN maturation, consistent with the functional diversification of voltage-gated currents measured in immature and mature OSNs.
Discussion
In this study, we investigated maturation-dependent functional properties of OSNs from neonatal OMP-GFP mice. While mature OSNs have been extensively characterized electrophysiologically, no information is available on immature neurons. Our results show maturation-dependent changes in spontaneous activity, intrinsic excitability and the expression of voltage-gated channels.
Using loose-patch recordings, we found that immature OSNs fire spontaneously, showing that they are functionally active before completing maturation. Their firing activity is characterized by lower firing rates and distinct firing patterns compared with mature OSNs, indicating that neuronal excitability undergoes significant functional changes during OSN maturation. Since spontaneous activity is crucial for the establishment and refinement of synaptic connections within the OB [43, 75], intrinsic firing in immature OSNs may play an essential role to initiate and stabilize appropriate synaptic contacts in the OB [8].
Mature OSNs express only a single type of OR and their spontaneous activity primarily is triggered by the stochastic activation of that OR and the downstream cAMP-dependent transduction cascade. It has been shown that firing frequency depends on the specific OR expressed by each neuron [10, 52, 53, 58]. Instead, immature OSNs can express multiple OR transcripts and some components of the transduction cascade [24, 66, 69], suggesting that more than one OR may contribute to their spontaneous activity. Differences in spontaneous firing between immature and mature OSNs are therefore likely to arise from multiple factors, including OR expression, variation in the maturation of the transduction cascade, as well as from intrinsic excitability properties. In this context, our data showed that immature OSNs are less intrinsically excitable than mature ones, as we will discuss later.
The use of OMP-GFP mice, in which GFP replaces the OMP coding sequence and OMP is therefore absent in both mature and immature OSNs, not only allows the visualization of mature OSNs but also ensures that spontaneous activity is independent of OMP. OMP has long been established as a marker for mature OSNs and plays a role in their responsiveness to odorants [11]. Recently, it has been shown that OMP regulates odorant response kinetics by modulating cAMP sequestration [13, 39, 54, 60]. Since basal OR activity contributes to regulate basal cAMP levels [58], modulation of cAMP dynamics by OMP can influence OSN spontaneous firing patterns [13, 54]. Although OMP knockout and heterozygous mice have similar mean firing rates, their firing pattern distributions differ [54]. Furthermore, OMP deletion particularly reduces spontaneous activity in OSNs expressing high basal activity ORs, such as M71, with weaker effects on low basal activity ORs, such as mOR-EG [13]. Thus, using the OMP-GFP line ensures that the observed differences in spontaneous firing between immature and mature OSNs cannot be attributed to OMP, which is absent in both groups. However, expression of OMP might modulate, at least in part, functional properties of mature OSNs, further enhancing or attenuating the existing differences between the two groups.
To address the lack of information on the intrinsic membrane properties of immature OSNs, we performed whole-cell current-clamp and voltage-clamp recordings and found that OSN maturation is accompanied by pronounced changes in both passive and active membrane properties. Immature neurons had a more depolarized resting potential, higher input resistance, a more depolarized action potential threshold, and slower action potential kinetics compared with mature OSNs. Immature OSNs fired only with phasic patterns, whereas mature OSNs showed a broader range of excitability and were either capable of tonic repetitive firing or phasic firing. The phasic firing mature OSNs exhibit intermediate electrophysiological properties between immature OSNs and tonic firing mature OSNs, suggesting that these cells may represent a transitional population during neuronal maturation. Such neurons might express OMP while still retaining immature markers such as GAP43 or Gγ8 [8, 24, 29]. Alternatively, they may reflect functional diversity among mature OSNs, as previously proposed [10, 47], with some of them optimized for higher sensitivity (tonic firing) and others for rapid detection of changes in the environment (phasic firing), potentially contributing to a broader dynamic range in olfactory coding. Thus, whole-cell patch-clamp recordings provide direct evidence that immature OSNs are intrinsically less excitable than mature neurons, and this reduced intrinsic excitability likely contributes to the lower spontaneous firing activity measured in loose-patch recordings.
Regarding passive properties, the combination of a depolarized resting potential and high input resistance likely contribute to the reduced excitability of immature OSNs: although higher input resistance would amplify depolarizing inputs, the more depolarized resting potential cause inactivation of a fraction of voltage-gated Na^+^ channels, reducing action potential generation.
Comparison with previous studies in adult animals shows that the resting membrane potential, input resistance and the presence of both phasic and tonic firing patterns in mature OSNs measured in our study are in general agreement with previous reports across multiple species, including humans [15, 27, 32, 38, 41, 45, 47, 59, 67, 71], whereas recordings from immature OSNs have not previously been reported.
To investigate the ion channels underlying maturation-dependence differences in excitability, we performed voltage-clamp recordings. Our experiments showed clear changes in voltage-gated Na^+^ and K^+^ currents. In immature OSNs, Na^+^ currents were almost entirely TTX-sensitive, whereas in mature OSNs we observed both TTX-sensitive and TTX-resistant components. Moreover, the strong dependence of channel availability on prepulse potential further indicates that the expression of Na^+^ channel subtypes changes during maturation. These physiological differences are consistent with our analysis of single-cell RNA sequencing data [7], which shows that mature OSNs upregulate genes encoding Na_v_1.5, Na_v_1.6 and Na_v_1.7 α-subunits, along with β3 and β4 auxiliary subunits of Na_v_ channels. Among these channels, Na_v_1.5 is known to be TTX-resistant, whereas Na_v_1.6 and Na_v_1.7 are TTX-sensitive [19], strongly suggesting that Na_v_1.5 is responsible for the TTX-resistant Na^+^ current measured in mature OSNs.
Our results on mature OSNs are in agreement with previous studies: Frenz et al. (2014) recorded both TTX-sensitive and TTX-resistant Na^+^ currents in dissociated OSNs from adult mice and showed that they contribute to spontaneous firing activity [14, 16]. Their RT-PCR and immunohistochemistry data identified Na_v_1.5 as the only TTX-resistant channel, localized specifically to dendrites. In addition, another finding consistent with our results comes from microarray mRNA profiling in OMP-GFP mice, which demonstrated that Na_v_1.5 is expressed in mature OSNs but not in immature neurons [65]. Together, these findings suggest that the TTX-resistant Na_v_1.5 is a characteristic feature of OSN maturation.
Regarding TTX-sensitive channels, RT-PCR in previous studies detected mRNA for multiple Na_v_ α-subunits, including Na_v_1.2, Na_v_1.3, Na_v_1.6 and Na_v_1.7, but immunohistochemistry showed that only Na_v_1.3, and Na_v_1.7 are expressed at the protein level in OSNs [2, 5, 16, 74]. Thus, although our analysis of single-cell RNA sequencing data [7] indicates upregulation of Na_v_1.6 and Na_v_1.7 in mature OSNs, protein expression appears limited to Na_v_1.7. Together, these results indicate that the TTX-sensitive component is primarily mediated by Na_v_1.3 and Na_v_1.7, with Na_v_ 1.7 expression increasing in mature OSNs.
Among the β-subunits, we observed upregulation β3 and β4, which modulate neuronal excitability by interacting with α-subunits and modifying channel gating [6]. For example, the β4 subunit has been linked to negative shifts in the activation curve of several Na_v_ subtypes [9, 76]. Therefore, differences in Na^+^ current properties during OSN maturation may arise not only from the differential expression of expression of α-subunits but also from modulatory β subunits. In addition, post-translational modifications, such as phosphorylation or glycosylation, may further influence the kinetics and voltage-dependence of Na_v_ channels [37, 78].
Also voltage-gated K⁺ currents showed pronounced maturation-dependent differences. Mature OSNs had a transient, 4-AP–sensitive K⁺ current, whereas immature OSNs lacked a transient component. Both immature and mature neurons showed a sustained K⁺ current, likely mediated by delayed rectifiers, but the amplitude of this sustained current was much larger in mature OSNs. From analysis of single-cell RNA sequencing data [7], we found that mature OSNs upregulate multiple voltage-gated K⁺ channel genes, including K_v_1.2, K_v_1.5, K_v_1.6, K_v_2.1 and K_v_3.4. Among these, K_v_3.4 is known to mediate fast-inactivating A-type K^+^ currents [22], suggesting it is the primary contributor to the transient component observed in mature OSNs. The other upregulated K_v_ channels mainly generate non-inactivating or slowly inactivating delayed rectifier K⁺ currents [22], consistent with the larger amplitude of the sustained K⁺ current in mature compared with immature OSNs.
Our results, showing both transient and sustained K⁺ current components in mature OSNs, are in agreement with previous reports in several species [3, 15, 23, 38, 44, 45, 72, 77]. In addition, our transcriptional analysis identified K_v_3.4 as a likely mediator of the transient A-type K⁺ current in mature OSNs, in agreement with microarray mRNA profiling in OMP-GFP mice showing K_v_3.4 expression in mature but not in immature OSNs [65]. Although other genes associated with A-type K⁺ currents, K_V_1.4, K_v_4.2 and K_v_4.3, have been reported in adult mouse OSNs (Han & Lucero, 2006), our analysis did not detect their differential expression, further indicating the specific role of K_v_3.4 in OSN maturation.
Limitations related to developmental stage and genetic background should be considered when interpreting the transcriptomic analysis. Our electrophysiological recordings were performed in neonatal OMP-GFP mice, whereas the single-cell RNA sequencing dataset was derived from adult C57BL/6 animals. Although OSN maturation follows a conserved sequential molecular program across developmental stages [40], quantitative differences in gene expression between immature and mature OSNs in neonatal animals cannot be excluded. Accordingly, the transcriptional differences identified in adult tissue should be interpreted as supportive and consistent with the physiological findings, but not as direct mechanistic evidence for the neonatal condition studied here. Furthermore, although OMP is replaced by GFP in the OMP-GFP line, previous transcriptomic analyses have demonstrated highly correlated gene expression profiles between OMP-GFP and wild-type OSNs [31], indicating that major alterations in voltage-gated channel gene expression due to OMP replacement are unlikely. Nevertheless, subtle strain- or genotype-specific differences cannot be entirely ruled out.
In summary, our data provide a previously uncharacterized functional profile for immature OSNs, defined by a reduced excitability, the absence of TTX-resistant Na⁺ currents, reduced sustained K⁺ currents, and the lack of A-type K⁺ currents, compared with mature OSNs. The progressive changes during maturation produce increased excitability, broader firing patterns, and faster action potential kinetics of mature OSNs. Transcriptomic analysis identified upregulation of some Na^+^ and K^+^ channel genes during OSN maturation, further supporting the functional diversification of voltage-gated currents during the transition from immature to mature OSNs.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1 (Excel file including the differentially expressed genes. Any gels or blots were performed CSV 860 KB)