Abstract
Neuronal pattern separation is thought to enable the brain to disambiguate sensory stimuli with overlapping features, thereby extracting valuable information. In the olfactory system, it remains unknown whether pattern separation acts as a driving force for sensory discrimination and the learning thereof. We found that overlapping odor-evoked input patterns to the mouse olfactory bulb (OB) were dynamically reformatted in the network on the timescale of a single breath, giving rise to separated patterns of activity in an ensemble of output neurons, mitral/tufted (M/T) cells. Notably, the extent of pattern separation in M/T assemblies predicted behavioral discrimination performance during the learning phase. Furthermore, exciting or inhibiting GABAergic OB interneurons, using optogenetics or pharmacogenetics, altered pattern separation and thereby odor discrimination learning in a bidirectional way. In conclusion, we propose that the OB network can act as a pattern separator facilitating olfactory stimulus distinction, a process that is sculpted by synaptic inhibition.
This is a preview of subscription content, access via your institution
Access options
Subscribe to this journal
Receive 12 print issues and online access
$209.00 per year
only $17.42 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
References
Friedrich, R.W. & Laurent, G. Dynamic optimization of odor representations by slow temporal patterning of mitral cell activity. Science 291, 889–894 (2001).
Friedrich, R.W. & Laurent, G. Dynamics of olfactory bulb input and output activity during odor stimulation in zebrafish. J. Neurophysiol. 91, 2658–2669 (2004).
Niessing, J. & Friedrich, R.W. Olfactory pattern classification by discrete neuronal network states. Nature 465, 47–52 (2010).
Barnes, D.C., Hofacer, R.D., Zaman, A.R., Rennaker, R.L. & Wilson, D.A. Olfactory perceptual stability and discrimination. Nat. Neurosci. 11, 1378–1380 (2008).
Chapuis, J. & Wilson, D.A. Bidirectional plasticity of cortical pattern recognition and behavioral sensory acuity. Nat. Neurosci. 15, 155–161 (2012).
Leutgeb, J.K., Leutgeb, S., Moser, M.B. & Moser, E.I. Pattern separation in the dentate gyrus and CA3 of the hippocampus. Science 315, 961–966 (2007).
McHugh, T.J. et al. Dentate gyrus NMDA receptors mediate rapid pattern separation in the hippocampal network. Science 317, 94–99 (2007).
Nakashiba, T. et al. Young dentate granule cells mediate pattern separation, whereas old granule cells facilitate pattern completion. Cell 149, 188–201 (2012).
Friedrich, R.W. Neuronal computations in the olfactory system of zebrafish. Annu. Rev. Neurosci. 36, 383–402 (2013).
Friedrich, R.W. & Wiechert, M.T. Neuronal circuits and computations: pattern decorrelation in the olfactory bulb. FEBS Lett. 588, 2504–2513 (2014).
Rolls, E.T. The mechanisms for pattern completion and pattern separation in the hippocampus. Front. Syst. Neurosci. 7, 74 (2013).
Vincis, R., Gschwend, O., Bhaukaurally, K., Beroud, J. & Carleton, A. Dense representation of natural odorants in the mouse olfactory bulb. Nat. Neurosci. 15, 537–539 (2012).
Patterson, M.A., Lagier, S. & Carleton, A. Odor representations in the olfactory bulb evolve after the first breath and persist as an odor afterimage. Proc. Natl. Acad. Sci. USA 110, E3340–E3349 (2013).
Abraham, N.M. et al. Maintaining accuracy at the expense of speed; stimulus similarity defines odor discrimination time in mice. Neuron 44, 865–876 (2004).
Linster, C., Johnson, B.A., Morse, A., Yue, E. & Leon, M. Spontaneous versus reinforced olfactory discriminations. J. Neurosci. 22, 6842–6845 (2002).
Linster, C. et al. Perceptual correlates of neural representations evoked by odorant enantiomers. J. Neurosci. 21, 9837–9843 (2001).
Rubin, B.D. & Katz, L.C. Spatial coding of enantiomers in the rat olfactory bulb. Nat. Neurosci. 4, 355–356 (2001).
Abraham, N.M. et al. Synaptic inhibition in the olfactory bulb accelerates odor discrimination in mice. Neuron 65, 399–411 (2010).
Abraham, N.M., Guerin, D., Bhaukaurally, K. & Carleton, A. Similar odor discrimination behavior in head-restrained and freely moving mice. PLoS ONE 7, e51789 (2012).
Abraham, N.M., Vincis, R., Lagier, S., Rodriguez, I. & Carleton, A. Long term functional plasticity of sensory inputs mediated by olfactory learning. eLife 3, e02109 (2014).
Stettler, D.D. & Axel, R. Representations of odor in the piriform cortex. Neuron 63, 854–864 (2009).
Miura, K., Mainen, Z.F. & Uchida, N. Odor representations in olfactory cortex: distributed rate coding and decorrelated population activity. Neuron 74, 1087–1098 (2012).
Wiechert, M.T., Judkewitz, B., Riecke, H. & Friedrich, R.W. Mechanisms of pattern decorrelation by recurrent neuronal circuits. Nat. Neurosci. 13, 1003–1010 (2010).
Schaefer, A.T. & Margrie, T.W. Spatiotemporal representations in the olfactory system. Trends Neurosci. 30, 92–100 (2007).
Cury, K.M. & Uchida, N. Robust odor coding via inhalation-coupled transient activity in the mammalian olfactory bulb. Neuron 68, 570–585 (2010).
Rinberg, D., Koulakov, A. & Gelperin, A. Speed-accuracy tradeoff in olfaction. Neuron 51, 351–358 (2006).
Uchida, N. & Mainen, Z.F. Speed and accuracy of olfactory discrimination in the rat. Nat. Neurosci. 6, 1224–1229 (2003).
Gschwend, O., Beroud, J. & Carleton, A. Encoding odorant identity by spiking packets of rate-invariant neurons in awake mice. PLoS ONE 7, e30155 (2012).
Tatti, R. et al. A population of glomerular glutamatergic neurons controls sensory information transfer in the mouse olfactory bulb. Nat. Commun. 5, 3791 (2014).
Arevian, A.C., Kapoor, V. & Urban, N.N. Activity-dependent gating of lateral inhibition in the mouse olfactory bulb. Nat. Neurosci. 11, 80–87 (2008).
Giridhar, S., Doiron, B. & Urban, N.N. Timescale-dependent shaping of correlation by olfactory bulb lateral inhibition. Proc. Natl. Acad. Sci. USA 108, 5843–5848 (2011).
Koulakov, A.A. & Rinberg, D. Sparse incomplete representations: a potential role of olfactory granule cells. Neuron 72, 124–136 (2011).
Bathellier, B., Buhl, D.L., Accolla, R. & Carleton, A. Dynamic ensemble odor coding in the mammalian olfactory bulb: sensory information at different timescales. Neuron 57, 586–598 (2008).
Guerrieri, F., Schubert, M., Sandoz, J.C. & Giurfa, M. Perceptual and neural olfactory similarity in honeybees. PLoS Biol. 3, e60 (2005).
Youngentob, S.L. et al. Experience-induced fetal plasticity: the effect of gestational ethanol exposure on the behavioral and neurophysiologic olfactory response to ethanol odor in early postnatal and adult rats. Behav. Neurosci. 121, 1293–1305 (2007).
Friedrich, R.W., Habermann, C.J. & Laurent, G. Multiplexing using synchrony in the zebrafish olfactory bulb. Nat. Neurosci. 7, 862–871 (2004).
Yaksi, E., Judkewitz, B. & Friedrich, R.W. Topological reorganization of odor representations in the olfactory bulb. PLoS Biol. 5, e178 (2007).
Kato, H.K., Gillet, S.N., Peters, A.J., Isaacson, J.S. & Komiyama, T. Parvalbumin-expressing interneurons linearly control olfactory bulb output. Neuron 80, 1218–1231 (2013).
Miyamichi, K. et al. Dissecting local circuits: parvalbumin interneurons underlie broad feedback control of olfactory bulb output. Neuron 80, 1232–1245 (2013).
Cleland, T.A. & Sethupathy, P. Non-topographical contrast enhancement in the olfactory bulb. BMC Neurosci. 7, 7 (2006).
Fukunaga, I., Herb, J.T., Kollo, M., Boyden, E.S. & Schaefer, A.T. Independent control of gamma and theta activity by distinct interneuron networks in the olfactory bulb. Nat. Neurosci. 17, 1208–1216 (2014).
Chow, S.F., Wick, S.D. & Riecke, H. Neurogenesis drives stimulus decorrelation in a model of the olfactory bulb. PLOS Comput. Biol. 8, e1002398 (2012).
Linster, C. & Cleland, T.A. Decorrelation of odor representations via spike timing-dependent plasticity. Front. Comput. Neurosci. 4, 157 (2010).
Alonso, M. et al. Activation of adult-born neurons facilitates learning and memory. Nat. Neurosci. 15, 897–904 (2012).
Carleton, A. et al. Making scents of olfactory neurogenesis. J. Physiol. Paris 96, 115–122 (2002).
Padmanabhan, K. & Urban, N.N. Intrinsic biophysical diversity decorrelates neuronal firing while increasing information content. Nat. Neurosci. 13, 1276–1282 (2010).
Boyd, A.M., Sturgill, J.F., Poo, C. & Isaacson, J.S. Cortical feedback control of olfactory bulb circuits. Neuron 76, 1161–1174 (2012).
Markopoulos, F., Rokni, D., Gire, D.H. & Murthy, V.N. Functional properties of cortical feedback projections to the olfactory bulb. Neuron 76, 1175–1188 (2012).
Haddad, R. et al. Olfactory cortical neurons read out a relative time code in the olfactory bulb. Nat. Neurosci. 16, 949–957 (2013).
Wilson, D.A. & Sullivan, R.M. Cortical processing of odor objects. Neuron 72, 506–519 (2011).
Li, J., Ishii, T., Feinstein, P. & Mombaerts, P. Odorant receptor gene choice is reset by nuclear transfer from mouse olfactory sensory neurons. Nature 428, 393–399 (2004).
Zariwala, H.A. et al. A Cre-dependent GCaMP3 reporter mouse for neuronal imaging in vivo. J. Neurosci. 32, 3131–3141 (2012).
Bathellier, B., Van De Ville, D., Blu, T., Unser, M. & Carleton, A. Wavelet-based multi-resolution statistics for optical imaging signals: Application to automated detection of odour activated glomeruli in the mouse olfactory bulb. Neuroimage 34, 1020–1035 (2007).
de Saint Laumer, J.-Y., Cicchetti, E., Merle, P., Egger, J. & Chaintreau, A. Quantification in gas chromatography: prediction of flame ionization detector response factors from combustion enthalpies and molecular structures. Anal. Chem. 82, 6457–6462 (2010).
Lagier, S., Carleton, A. & Lledo, P.M. Interplay between local GABAergic interneurons and relay neurons generates gamma oscillations in the rat olfactory bulb. J. Neurosci. 24, 4382–4392 (2004).
Lagier, S. et al. GABAergic inhibition at dendrodendritic synapses tunes gamma oscillations in the olfactory bulb. Proc. Natl. Acad. Sci. USA 104, 7259–7264 (2007).
Acknowledgements
We thank S. Barthelme, B. Bathellier, R. Friedrich, A. Holtmaat, F. Markopoulos, A. Pouget, and members of the Carleton and Rodriguez laboratories for helpful discussions and/or comments on the manuscript. We thank J. Bourquin and S. Pellat for technical assistance. This research was supported by the University of Geneva, the Geneva neuroscience center (common grant to A.C. and I.R.), the Swiss National Science Foundation (grant numbers: 31003A_153410 to A.C., CR33I13_143723 to A.C. and I.R., and 310030E_135910 to I.R.), the National Center of Competence in Research (NCCR) 'SYNAPSY - The Synaptic Bases of Mental Diseases' financed by the Swiss National Science Foundation (51AU40_125759, A.C.), the European Research Council (contract number ERC-2009-StG-243344-NEUROCHEMS to A.C.) and the European Molecular Biology Organization (young investigator program to A.C.; long-term postdoctoral fellowship to N.M.A.).
Author information
Authors and Affiliations
Contributions
O.G., N.M.A., S.L., I.R. and A.C. carried out the study conceptualization and experimental design. O.G. acquired and analyzed electrophysiological data. S.L. acquired and analyzed imaging data. N.M.A. performed and analyzed behavioral experiments. O.G. and N.M.A. performed virus injection, window implantation, opto/pharmacogenetics and related behavior, immunohistochemistry, and confocal imaging and quantification. N.M.A. and F.B. acquired and analyzed GC-FID data. O.G. and A.C. wrote the manuscript with comments from N.M.A., S.L., F.B. and I.R.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Integrated supplementary information
Supplementary Figure 1 Correlation matrices for the glomeruli inputs computed with different methods.
(a) Matrix of inputs similarity computed in each mouse for all possible mixtures tested. The analyses are performed on manually selected regions of interest (ROIs, see methods). (b) Mean correlation matrix averaged across the single animal matrices plotted in A (same matrix as in Fig. 1c). (c) Matrix of inputs similarity computed in each mouse for all possible mixtures tested. The analyses are performed on all pixels of the OB surface (see methods), giving similar results to the ROI analysis. (d) Mean correlation matrix averaged across the single animal matrices plotted in c. (e) Matrix of inputs similarity computed for all possible mixtures tested. All the ROIs values from different animals are concatenated in a single population vector before calculating correlation coefficients (giving similar results to b,d). (f) Bar graph showing the average correlation for different subgroups of mixtures as represented by the color code on the right schematic matrix (data presented as mean ± sem; averaged across odors from data presented in e; Mann-Whitney test). (g) Correlation matrices computed over 40ms time window across the first breath after odor onset (1 out of 2 consecutive matrices are ommitted for clarity). Analysis performed on all ROIs from different mice concatenated (as in e).
Supplementary Figure 2 Absence of relationship between odorant-evoked glomerular response amplitude and pairwise correlation.
Each circle corresponds to a pair of mixtures (same data as in Figure 1; the change in fluorescence is averaged across all glomeruli and the two odorants). Note that the regression does not capture at all the variance of the data (i.e. R2 close to 0).
Supplementary Figure 3 Odorant-evoked glomerular responses monitored by GCaMP3 does not saturate at the odorant concentration used in the study.
(a) Average number of glomeruli activated by different odorants (white numbers indicate number of OB imaged, at least n = 2 mice for each). In the study, odorants are diluted by mixing an odorized air stream to clean air stream at a ratio of 1/20 (i.e. 0.05; total flow 400 ml/min). Increasing and decreasing odorant concentration, by changing relative flows, recruited more or less glomeruli, respectively. It means that the odorant concentration used in the study does not saturate OSN response. Data presented as mean ± sem. (b) Cumulative distributions of response amplitude averaged on the first breath for all glomeruli at different concentrations (red curves). Note that the entire distribution can be shifted at higher concentration than the one used in the study (thin vs thick red lines, Kolmogorov-Smirnov test P < 0.05), indicating that the dye is not saturated. The glomeruli responses shown in Fig. 1 have been split in two groups based on response amplitude (median + 1 SD, see Supplementary Fig. 4). The cumulative distribution for these “weakly” and “strongly” responding glomeruli are shown in black for comparison.
Supplementary Figure 4 Binary mixtures evoke correlated input patterns to the mouse olfactory bulb irrespective of glomeruli response strength.
(a) We tested whether glomeruli displaying different response strengths may differentially contribute to inputs correlation. Using the data presented in Fig. 1, we plotted for each mouse the distribution of glomeruli response amplitude evoked by all mixtures (top panels). For each mouse, a median plus one standard deviation is defined (vertical dashed lines) as a cut-off to distinguish “weakly” from “strongly” activated glomeruli. Bottom panels: cumulative distributions of glomerular response amplitude for all mixtures plotted in each mouse for all glomeruli (thin line), weakly responding glomeruli (dashed lines; n = 11, 18, 18, 13 glomeruli for mouse 1 to 4, respectively) and strongly responding glomeruli (thick lines, n = 19, 21, 23, 28 glomeruli for mouse 1 to 4, respectively). (b) Cumulative distributions of glomerular response amplitude for all mixtures plotted in each mouse for all glomeruli (same color code as in A). (c) Cumulative distributions of glomerular response amplitude for all mixtures plotted across all animals for all glomeruli (thin line), weakly responding glomeruli (dashed line) and strongly responding glomeruli (thick line). (d) We computed the input correlation matrices for the different subgroups of glomeruli and compared them to the entire population: matrix of inputs correlation computed for all glomeruli (same as in Fig. 1d), weakly or stongly activated glomeruli. We observed similar results for all the different groups of glomeruli. (e) Bar graph showing the average correlation for different subgroups of mixtures as represented by the color code on the right schematic matrix (data presented as mean ± sem, n = 4 mice). No significant difference between the different groups of glomeruli could be found (repeated measures MANOVA F = 3.125, P = 0.055). Since the test is close to significance, we analyzed the posthoc tests which only revealed a significant difference between all glomeruli and strongly activated glomeruli groups (LSD test P = 0.28, 0.016, 0.17 for All vs. weakly, All vs. strongly and weakly vs. strongly, respectively). In summary, these data indicate that weakly and strongly activated glomeruli do not contribute differently to the correlation.
Supplementary Figure 5 Analysis of output pattern correlations in two other datasets.
(a–d) Summary of M/T cell responses to all tested odors recorded in different datasets. Percentage of responsive odor-breath pairs for recorded neurons (a,c) and percentage of responsive cell-breath pairs for each odor (b,d) of different datasets. Neurons displaying an increase in firing rate (averaged on the entire breath), a decrease in firing rate or only a change of spike distribution (temporal change) between baseline and odor epochs are presented in red, blue and black, respectively. (e) Temporal evolution of the matrices of output patterns similarity computed for all possible mixtures tested in dataset 2 (subgroups of mixtures are indicated on the left). For clarity, only few matrices of the first breath after odor onset are shown. Note that pure monomolecular odorants have also been used for comparison with the mixtures. (f) Average correlation for different subgroups of mixtures plotted over time for dataset 2 (grey boxes indicate odor application; thick line: mean, colored surface: sem). The color codes correspond to the different groups shown on the schematic matrix. For clarity, few subgroups of mixtures are not plotted. (g) Same as in e, but for a third dataset using other mixtures. (h) Same as in c, but for dataset 3. Note that the results obtained for datasets 2 and 3 are similar to the ones presented for dataset 1 in Fig. 3a,b.
Supplementary Figure 6 Temporal evolution of output pattern correlation in different parts of the first breath after odor onset.
(a,b) Bar graphs showing the average correlation computed in different parts of the first breath after odor onset for different subgroups of mixtures (same colour code as in Fig. 1F; data are presented as mean ± sem). The correlation levels differ for different groups of mixtures (Kruskal-Wallis ANOVA H = 28.2 and 32.3, P < 0.0001 for the first and second half of the breath, respectively). Values of the posthoc tests for multiple comparisons are indicated on each graph. A decrease of correlation over the breath duration was visible for most mixtures (Wilcoxon paired test P = 0.028, 0.028, 0.17, 0.028, 0.0004 for AA/EBliq, AA/EBair, HX/EBliq, HX/EBair, AA/EB-EB/HX, respectively.
Supplementary Figure 7 Temporal evolution of output pattern correlation in the first odor breath and relationships to behavioral performances.
(a) Temporal evolution of the correlations for all possible pairs of mixtures during the first breathing cycle after odor onset. Correlation matrices are computed using vectors of firing rate averaged over consecutive 21ms time windows. (b–e) The discrimination performance averaged over 300 trials is plotted as a function of the output pattern correlation computed on different parts of the breath for 42ms (b,c) and 21ms (d,e) time window analysis for the raw data (b,d) and data corrected for baseline firing (c,e). Data presented as mean ± sem.
Supplementary Figure 8 Immunohistochemical analysis of AAV-mediated infection pattern.
(a-c) Confocal images of different animals transfected with AAV-GiDREADD (different mice for each panel). Infection is restricted to the GCL with no sign of M/T cell labelling (a,c). By staining M/T cells using a reelin antibody (b), the absence of M/T cell infection is confirmed at higher magnification even when few infected granule cells were found just below M/T cell somata (yellow arrows in b, right image is a higher magnification of the yellow dashed box shown on the left image). Note that a short axon cell is also infected (green cell in the IPL). Scale bars: 65 μm (a,c), 25μm (b). (d–e) Confocal images of different animals transfected with AAV-ChR2 (different mouse for each panel). As observed for AAV-GiDREADD infection, AAV-ChR2 exclusively infected neurons in the granule cell layer but not M/T cells even when labelled cells with typical granule cell morphology are found just below M/T cell (yellow arrowhead in e points to GC dendritic spine). Scale bars: 65 μm. (f) Fraction of neurons infected by viral vectors in different layers of the OB (each circle represents the quantification for a single mouse, n = 6 for each condition). Quantifications represent the fraction of NeuN+ and Reelin+ neurons in GCL and MCL, respectively. EPL: external plexiform layer, IPL: internal plexiform layer, MCL: mitral cell layer. Data are presented as mean ± sem.
Supplementary Figure 9 Adjustment of LED power for optogenetic manipulation of odor discrimination behavior.
(a,b) Schemas of the experimental procedure (a) and the behavioral protocol (b). At the beginning of a trial, a masking blue light is applied. Head-fixed mice are initially trained to lick when an odor is presented. A water reward is delivered at the end of the trial if a mouse sufficiently licked during odor presentation (see methods). A LED mounted on the cranial window is turned on during odor presentation and its power is adjusted for each mouse to prevent the stop of licking due to strong stimulation of GABAergic neurons (c) Examples of licking patterns evoked by odor application (onset at time 0) for two mice. Increasing LED power progressively suppressed licking behavior. Note that each mouse exhibited a unique pattern of licking blockage (note that power 7.3 is not used for mouse 2 since behavior already shut down at 3 mW.mm−2). For the discrimination behavior, power is set in such a way that licking remains similar to the no light condition (i.e. 3 and 1 mW.mm−2 for mouse 1 and 2, respectively). Thick lines and colored areas represent mean and s.e.m., respectively. (d) Licking patterns evoked by odor application and light modulations for all mice (n = 5-7 mice for different powers). Thick lines and colored areas represent mean and s.e.m., respectively.
Supplementary Figure 10 Optogenetic stimulation of granule cell layer neurons improves discrimination learning for mixtures but not for monomolecular odorants.
(a) Effect of GCL photostimulation on the mean population correlation for monomolecular odorants (Wilcoxon paired test *P = 0.08). (b) Simple odor discrimination performances was not improved by photostimulation in ChR2- when compared to Gi-DREADD- expressing mice (n = 7 and 8, respectively; repeated measures ANOVA F = 0.11 P = 0.74). Data are presented as mean ± sem. (c) Photostimulation improved the learning performances of ChR2- expressing mice when compared to Gi-DREADD expressing animals during mixture discrimination task (repeated measures ANOVA F = 12.3 P = 0.0038, post-hoc LSD test at least *P < 0.05). The odor pairs used were the following mixtures (all gas mixes): AA/EB 60/40 vs. AA/EB 40/60, EB/HX 60/40 vs. EB/HX 40/60. Data are presented as mean ± sem. (d) Reaction times for different discrimination tasks (for which we could extract enough trials at high performance) measured on individual animal basis from 200-300 trials with the highest performances (on average superior to 65–70%, white numbers indicate number of animals; ctrl: control mice either GFP infected or DREADD infected mice from panels b,c and Fig. 6k). Mann-Whitney test *P < 0.05. Data are presented as mean ± sem.
Supplementary information
Supplementary Text and Figures
Supplementary Figures 1–10 and Supplementary Tables 1 and 2 (PDF 2780 kb)
Rights and permissions
About this article
Cite this article
Gschwend, O., Abraham, N., Lagier, S. et al. Neuronal pattern separation in the olfactory bulb improves odor discrimination learning. Nat Neurosci 18, 1474–1482 (2015). https://doi.org/10.1038/nn.4089
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/nn.4089
This article is cited by
-
Robust odor identification in novel olfactory environments in mice
Nature Communications (2023)
-
Artificial olfactory sensor technology that mimics the olfactory mechanism: a comprehensive review
Biomaterials Research (2022)
-
Biologically plausible single-layer networks for nonnegative independent component analysis
Biological Cybernetics (2022)
-
Rapid Bayesian learning in the mammalian olfactory system
Nature Communications (2020)
-
Whitening of odor representations by the wiring diagram of the olfactory bulb
Nature Neuroscience (2020)