Distinct representation of cue-outcome association by D1 and D2 neurons in the ventral striatum’s olfactory tubercle

Article activity feed

1. 

   Version published to 10.7554/elife.75463 on eLife

   Jun 16, 2022
2. 

   eLife

   Apr 26, 2022

   Author Response

   Reviewer #3 (Public Review):

   Martiros et al. investigated whether medium spiny neurons in the striatal component of the olfactory tubercle (OT) acquire conditioned responses to odors that have been paired with unconditioned stimuli (water and airpuff). The authors found that both cells containing D1-type receptors (D1R) and D2-type receptors (D2R) acquire conditioned responses to odors. D1R cells appear to have responded to the valence more readily than the identity of unconditioned stimuli, and vice versa for D2R cells. The authors also found that tones can be used to condition D1R and D2R cells; however, conditioned responses with tones were not as much correlated with the valence as those of odors. The conclusions of the paper should be written with a nuanced manner.

   Strength:

   The authors used state-of-the-art …

   Author Response

   Reviewer #3 (Public Review):

   Martiros et al. investigated whether medium spiny neurons in the striatal component of the olfactory tubercle (OT) acquire conditioned responses to odors that have been paired with unconditioned stimuli (water and airpuff). The authors found that both cells containing D1-type receptors (D1R) and D2-type receptors (D2R) acquire conditioned responses to odors. D1R cells appear to have responded to the valence more readily than the identity of unconditioned stimuli, and vice versa for D2R cells. The authors also found that tones can be used to condition D1R and D2R cells; however, conditioned responses with tones were not as much correlated with the valence as those of odors. The conclusions of the paper should be written with a nuanced manner.

   Strength:

   The authors used state-of-the-art techniques to monitor changes in cellular activity over multiple days in awake animals engaging in conditioning tasks.

   Weaknesses:

   1. The grin lens, used to detect cellular activities, was large relative to the mouse brain, causing an extensive brain damage. A 1.0-mm diameter lens was unilaterally placed from the top of the brain to the bottom where the olfactory tubercle is situated. The lateral width of one hemisphere at the level of the olfactory tubercle is approximately 3.5 mm, indicating a large portion of the brain was damaged by the placement of the grin lens. This may be estimated that approximately 20-25% of the brain hemisphere anterior to the thalamus was damaged. The implication of this issue needs to be discussed.

   We agree that the damage caused by the GRIN lens is an unfortunate outcome of the experimental procedure. We will attempt to address the issue in two ways. First, to address the degree to which the principal findings related to the OT neuronal activity could be related to this damage, and second to address the degree to which the behavior of the mice may have been affected by it.

   The olfactory bulb projects to the olfactory tubercle via the lateral olfactory tract which runs along the ventral portion of the brain and thus would not be damaged by the cannula/GRIN lens insertion. The dopaminergic projections from the VTA to the OT run along a similarly ventral track and do not intersect with the inserted lens. The areas that were primarily damaged by the lens were motor cortex and striatum, neither of which are known to project strongly to the OT. We agree that it is likely that the damage caused to the striatum may have altered some of the indirect inputs to the OT and had this damage not occurred we may have observed some differences in the neuronal activity in the OT. However, the main comparisons we make in our study are likely not a function of the damage caused by the lens. First, the differences in the valence coding of D1 and D2 neurons are unlikely to be a result of lens damage, since there is no reason to suspect that damage caused by the lens will differentially affect D1 and D2 neuronal activity in the OT. D1 and D2 type neurons in the OT and the striatum typically receive inputs from similar upstream structures and their inputs were not generally altered by the lens damage. Second, there is no reason to suspect that the robust valence coding in the absence of the instrumental response and outcomes by D1 neurons could be a result of the lens damage. Third, there is no particular reason to suggest that the distinction between the odor and sound responsive neurons in the last set of experiments would be a result of the damage caused by the lens as the possible auditory cortical projections to the OT arrive from the posterior direction. Finally, the nature of the valence-related responses we observe in the OT are similar to those observed by others using tetrode recordings (Gadziola et al., 2015, Millman and Murthy, 2020) in which there was presumably less damage to the striatum.

   Second, we found that after the extensive recovery period of 1 month after the implantation surgery, the mice learned the association tasks extremely rapidly. Most mice began to learn the odor associations within the first day of training, and clearly exhibited anticipatory licking responses to only the rewarded odors by day 2 of training (Fig. 1C and D). This suggests that their behavior relevant to the task was not impaired by the damage caused by the lens. In addition, the observation that OT neurons gain valence coding over learning also suggests that circuitry allowing plasticity is at least partially preserved. We are also able to compare the anticipatory licking rates of unimplanted WT mice trained in the odor-sound task, to the anticipatory licking rates of the mice implanted with the GRIN lens. These mice underwent surgery for the attachment of the head fixation plate to the skull, but did not undergo a craniotomy or GRIN lens implantation. The anticipatory licking of the mice with no GRIN lens implant is similar to that of the implanted mice (Fig. 5B), suggesting that the implanted mice are not impaired in their ability to learn the stimulus-outcome associations. This data is presented in Figure 1 figure supplement 5.

   Finally, we made concerted efforts to minimize the damage caused by the implanted cannula. First, the cannula was constructed from highly biocompatible and thin-walled polyamide tubing and quartz floor which have previously been shown to minimize glial scar tissue (Bocarsly et al., 2015). Second, 2mm of the cortex were removed by suction prior to the virus injection and cannula implantation in order to minimize pressure in the brain. Finally, the mice were allowed to recover for at least one month prior to the onset of behavioral training. We have added text in the Discussion section (paragraphs 8 and 9) of the manuscript to summarize these points.

   1. The recording of cells may have included non-striatal cells. The olfactory tubercle consists of three major components: striatal, pallidal, and islands of Calleja units. These units are interwoven within the OT. Although the stratal unit is filled with medium spiny GABAergic neurons that the authors was interested in, there are other cells. The pallidal region contains GABAergic, cholinergic, and glutamatergic neurons, and the island Calleja contains granule cells. The authors need to inform readers whether the cells of the pallidal and islands of Calleja units contain D1R or D2R. For example, granule cells of the islands of Calleja have been shown to express D1R (Ridray et al 1998). This fact affects the interpretation of the present study. The implication of this issue needs to be discussed.

   We agree that this is a notable concern and have addressed this more explicitly in the paper in the Discussion (paragraph 3) and Results section (Figure 1 figure supplement 4). We will address possible imaging of the two structures separately.

   With regards to the ventral pallidum – we believe it to be unlikely that our dataset includes ventral pallidal neurons for two reasons. First, ventral pallidal neurons are dorsal to the tubercle and we assessed the position of each GRIN lens to include only those positioned ventral enough to image the tubercle. In fact, we imaged two Drd1-Cre mice in which we suspect the GRIN lens was positioned at the level of VP (based on histological estimates), and others where the GRIN lens was positioned at the level of NA, which were not included in the data analysis. Second, the Allen Brain atlas demonstrates very low expression of Drd1 and A2A in the portion of VP adjacent to the OT as compared to the OT and striatum (see figure below), which has been also previously observed in anatomical studies (Mansour et al., 1990). Since we focused the imaging on the focal plane with the strongest GCaMP fluorescence, this is unlikely to have been the VP. Certainly, we cannot rule out the possible inclusion of some VP neurons in the dataset, but these are likely to be very rare and unlikely to change the main conclusions of the study.

   With regards to the islands of Calleja – the neurons in the IC do express Drd1 but do not express A2A according to the Allen brain atlas and anatomical studies (Barik and de Beaurepaire, 1998, Mengod et al., 1992). Due to this, we don’t believe that IC neurons were included in the D2 neuronal dataset. With regard to the D1 neuronal dataset, it is possible that some IC neurons were included; however, we also believe this to be rare. This is due to the fact that IC neurons are characteristically small in size (~ 8 micron diameter) and densely clustered together likely appearing differently in the imaging field of view than typical OT neurons. In rare cases, we observed such regions in the imaging field of view which may have been IC regions as shown in the figure below (panel B). In these cases, the putative neurons that were clustered in these regions were not included in the data analysis. We further quantified the approximate diameter of the D1 and D2 neurons (see figure below). While this size is a rough estimate based on the number of pixels in the field of view occupied by the footprint of the neuron as selected by the CaImAn analysis tool, and may exceed the size of the neurons’ soma due to scatter of the fluorescence signal, we found that there were very few neurons with diameters of < 10 um, suggesting the absence of the small densely clustered IC neurons in our dataset. Additionally, we chose the imaging focal plane with the brightest GCaMP activity focusing on a layer of OT neurons. Due to the fact that the IC are typically most dense above and below the OT layer, it is also less likely that we focused on the IC neurons. While we cannot rule out the possibility that some IC neurons are included in the D1 neuronal dataset, we don’t believe they are contributing significantly to the main results of the study as the majority of the D1 neuronal population was strongly responsive to odor valence. These results are in strong agreement with previous electrophysiology studies in which authors used criteria such as firing rate, interspike interval distribution, and spike waveforms to classify SPN type neurons and analyze their spiking in response to valenced odors.

   Here are notable authors' claims and this reviewer's responses.

   Claim 1: The authors state "the OT is likely to be involved in learning about both positive and negative odor associations, rather than the alternative possibilities that the OT is only involved in learning rewarded odor associations, or that it encodes odor salience rather than signed odor valence." I believe that this is a reasonable conclusion.

   Claim 2A: The authors state "D1 OT neurons selectively and bidirectionally encode learned odor valence, unlike D2 neurons". This statement should be attenuated. Their data suggest that both D1R and D2R neurons are involved in both valence and identity and that D1R neurons are more likely involved in valence than identity, and vice versa.

   We have modified this sentence to “D1 OT neurons are more likely to encode learned odor valence than D2 neurons, and conversely less likely to encode odor identity”.

   Claim 2B: The authors state "stimulus valence representation by D1 OT neurons is limited to olfactory stimuli, and does not generalize to multimodal stimuli." This statement is premature, and the authors should provide a more nuanced statement. I note two issues: First, the mice had different experimental histories between odors and sounds; the mice were trained with odors first (4 sessions) and then with sounds (3 sessions). Therefore, differential responses between odors and tones can be attributed to their experimental histories rather than olfactory and auditory modalities. Indeed, the data showed that the mice had not fully learned to discriminate between two tones as the mice displayed anticipatory licks upon the tone paired with airpuff. In addition, it is not warranted to have a sweeping generalization because the authors examined only one reward (water) and one type of sounds (tones). Odors may be better conditioned with water while other rewards may work better with tones. Tones may have affective qualities that may have interfered with water conditioning and were needed additional pairing sessions. Remember that the absence of evidence does not provide a proof. It could simply that it was not done well. Moreover, OT neurons clearly responded to the auditory stimuli. Although the OT is strongly linked with the olfactory system compared to other sensory systems, the OT can receive sensory-related information from the limbic cortical structures that provide afferents, including the medial prefrontal cortex, basolateral nucleus of amygdala, and subiculum. Perhaps, the authors should discuss possible roles of these cortical structures in conditioned signals that were detected in the present study and how the large brain damage caused by the grin lens might have compromised afferent inputs to the OT.

   The reviewer is indeed correct, and that statement may not be warranted. Indeed, our own data show that there are D1 neurons that respond positively to rewarded tones. Given the differences in the way we did the tone and odor experiments (which the reviewer highlights), a sweeping generalization is not warranted. We have modified the text to remove such generalizations and discuss the possibility that the OT may represent valence of stimuli from other modalities if those signals reach the OT through polysynaptic pathways.

   We however, would like to note that the reviewer is not quite correct in stating that “ the mice had not fully learned to discriminate between two tones”. While it is true that some anticipatory licks remain even for the airpuff predicting tone, the difference in lick rates for the two tones is large and significant. Mice tend to lick more for both tones compared to odors, possibly due to the startling nature of the sound onset. In day 3, the mice clearly and strongly discriminated between sounds 1 and 2. They perform an average of 3.7 anticipatory licks in response to sound 2 and 1.1 licks in response to sound 1 (p = 8.0914e-28, Wilcoxon rank-sum test). The difference between the number of licks between sound 1 and sound 2 is ~2.6 licks. In comparison, the mean number of anticipatory licks for small rewarded odor 4 in the odor-odor task was 2.2 and the number of licks for aversive odor 1 was 0.08, for a smaller difference of 2.1 licks between the two stimuli. Therefore, we are confident that mice are indeed discriminating quite well between sounds 1 and 2. We have now added this text to the Results section when describing the odor-sound experiments.

   Claim 3: The authors state "valence representation is not correlational in nature, but likely serves to inform downstream brain regions of the value of odor stimuli". The present study showed that OT responded to conditioned odors in the absence of behavior. Although this result makes it difficult to hypothesize the functional role of those signals, it is reasonable to infer that the acquired conditioned signals influence downstream systems for some unknown function. However, the method used to obtain the data was correlational. The authors should avoid misleading phrases. Additional experiments are needed to understand functional role of the acquires signals.

   We have tempered the sentence to “… valence representation could inform downstream brain regions of the …”

   Read the original source
3. 

   eLife

   Apr 26, 2022

   Evaluation Summary:

   In this manuscript the authors carefully describe the activity of individual neurons within the mouse olfactory tubercle, comprised of the two principal cell types, in the context of odor and tone associative learning. The use of 2-photon microscopy to monitor activity of the neurons is a major step forward and unveiled new insights into the dynamics of these neurons. This manuscript will be of interest to a wide range of readers, including those interested in affective circuits, learning, and sensory processing.

   (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

   Read the original source
4. 

   eLife

   Apr 26, 2022

   Reviewer #1 (Public Review):

   In this impressive manuscript the authors sought to monitor the activity of the two principal cell types within the ventral striatum's olfactory tubercle -- those expressing either the dopamine D1 or D2 receptors. Work by several groups, including from this specific group, have provided evidence that olfactory tubercle neurons flexibly encode the learned meaning of odors. Reports using immediate early gene expression (Murata et al 2015) or fiber photometry (Gadziola et al 2020) have begun to uncover the possibly unique activities of tubercle D1 or D2 neurons during or following associative learning of odors. Here, Martiros and colleagues take the important step of real time monitoring of these neurons with 2-photon microscopy, as mice learned to associate odors [and tones] with wither positive or aversive …

   Reviewer #1 (Public Review):

   In this impressive manuscript the authors sought to monitor the activity of the two principal cell types within the ventral striatum's olfactory tubercle -- those expressing either the dopamine D1 or D2 receptors. Work by several groups, including from this specific group, have provided evidence that olfactory tubercle neurons flexibly encode the learned meaning of odors. Reports using immediate early gene expression (Murata et al 2015) or fiber photometry (Gadziola et al 2020) have begun to uncover the possibly unique activities of tubercle D1 or D2 neurons during or following associative learning of odors. Here, Martiros and colleagues take the important step of real time monitoring of these neurons with 2-photon microscopy, as mice learned to associate odors [and tones] with wither positive or aversive outcomes.

   There are many strengths of this study. These include a carefully designed behavioral task which manipulated assigned valence to the stimuli. Ideal cellular-level resolution monitoring of D1 or D2 (via adora2a) neural dynamics via GRIN lenses coupled with multiphoton imaging. And concise analyses of the dynamics in relationship to the behavior. An additional important strength is the forthcoming and fair discussion of their results in context of prior literature which allows the reader with an appropriate appreciation for the advances of the present work and what future directions it may foster.

   There is not a single objective weakness in this study.

   Conclusions are well supported by the results.

   The authors results uncovering differing roles of tubercle D1 and D2 neurons are important and help steer the field by recognizing that these cell populations may both a)have differing inputs but also b)may differentially influence downstream circuits which may afford affective learning and responding.

   Read the original source
5. 

   eLife

   Apr 26, 2022

   Reviewer #2 (Public Review):

   In this manuscript, Martiros and colleagues characterize the functional properties of neurons in the mouse olfactory tubercle. They record, using 2-photon microscopy, neuronal responses to odors associated with appetitive or aversive unconditioned stimuli and show that the activity of D1 and D2 type dopamine receptor expressing neurons is modulated by odor outcome associations. D1 neurons robustly encode learned odor valence, and D1 neurons maintain an odor valence representation even when appetitive (water drop) or aversive (air puff) unconditioned stimuli are removed. In contrast, D2 neurons are more selective to odor identity than D1 neurons, and odor valence coding in D2 neurons is dependent on odor outcome associations. Finally, the authors show that in D1 neurons, odor and tone outcome associations …

   Reviewer #2 (Public Review):

   In this manuscript, Martiros and colleagues characterize the functional properties of neurons in the mouse olfactory tubercle. They record, using 2-photon microscopy, neuronal responses to odors associated with appetitive or aversive unconditioned stimuli and show that the activity of D1 and D2 type dopamine receptor expressing neurons is modulated by odor outcome associations. D1 neurons robustly encode learned odor valence, and D1 neurons maintain an odor valence representation even when appetitive (water drop) or aversive (air puff) unconditioned stimuli are removed. In contrast, D2 neurons are more selective to odor identity than D1 neurons, and odor valence coding in D2 neurons is dependent on odor outcome associations. Finally, the authors show that in D1 neurons, odor and tone outcome associations recruit largely non-overlapping neuronal ensembles.

   Together, the experiments and analyses described in this manuscript address important yet poorly understood questions about the cellular and circuit mechanisms underlying odor valence coding. However, several concerns about experimental design and data analysis and interpretation need to be addressed.

   Read the original source
6. 

   eLife

   Apr 26, 2022

   Reviewer #3 (Public Review):

   Martiros et al. investigated whether medium spiny neurons in the striatal component of the olfactory tubercle (OT) acquire conditioned responses to odors that have been paired with unconditioned stimuli (water and airpuff). The authors found that both cells containing D1-type receptors (D1R) and D2-type receptors (D2R) acquire conditioned responses to odors. D1R cells appear to have responded to the valence more readily than the identity of unconditioned stimuli, and vice versa for D2R cells. The authors also found that tones can be used to condition D1R and D2R cells; however, conditioned responses with tones were not as much correlated with the valence as those of odors. The conclusions of the paper should be written with a nuanced manner.

   Strength:
   The authors used state-of-the-art techniques to monitor …

   Reviewer #3 (Public Review):

   Martiros et al. investigated whether medium spiny neurons in the striatal component of the olfactory tubercle (OT) acquire conditioned responses to odors that have been paired with unconditioned stimuli (water and airpuff). The authors found that both cells containing D1-type receptors (D1R) and D2-type receptors (D2R) acquire conditioned responses to odors. D1R cells appear to have responded to the valence more readily than the identity of unconditioned stimuli, and vice versa for D2R cells. The authors also found that tones can be used to condition D1R and D2R cells; however, conditioned responses with tones were not as much correlated with the valence as those of odors. The conclusions of the paper should be written with a nuanced manner.

   Strength:
   The authors used state-of-the-art techniques to monitor changes in cellular activity over multiple days in awake animals engaging in conditioning tasks.

   Weaknesses:

   1. The grin lens, used to detect cellular activities, was large relative to the mouse brain, causing an extensive brain damage. A 1.0-mm diameter lens was unilaterally placed from the top of the brain to the bottom where the olfactory tubercle is situated. The lateral width of one hemisphere at the level of the olfactory tubercle is approximately 3.5 mm, indicating a large portion of the brain was damaged by the placement of the grin lens. This may be estimated that approximately 20-25% of the brain hemisphere anterior to the thalamus was damaged. The implication of this issue needs to be discussed.

   2. The recording of cells may have included non-striatal cells. The olfactory tubercle consists of three major components: striatal, pallidal, and islands of Calleja units. These units are interwoven within the OT. Although the stratal unit is filled with medium spiny GABAergic neurons that the authors was interested in, there are other cells. The pallidal region contains GABAergic, cholinergic, and glutamatergic neurons, and the island Calleja contains granule cells. The authors need to inform readers whether the cells of the pallidal and islands of Calleja units contain D1R or D2R. For example, granule cells of the islands of Calleja have been shown to express D1R (Ridray et al 1998). This fact affects the interpretation of the present study. The implication of this issue needs to be discussed.

   Ridray S, Griffon N, Mignon V, Souil E, Carboni S, et al. 1998. Coexpression of dopamine D1 and D3 receptors in islands of Calleja and shell of nucleus accumbens of the rat: opposite and synergistic functional interactions. The European journal of neuroscience 10: 1676-86

   Here are notable authors' claims and this reviewer's responses.

   Claim 1:
   The authors state "the OT is likely to be involved in learning about both positive and negative odor associations, rather than the alternative possibilities that the OT is only involved in learning rewarded odor associations, or that it encodes odor salience rather than signed odor valence." I believe that this is a reasonable conclusion.

   Claim 2A:
   The authors state "D1 OT neurons selectively and bidirectionally encode learned odor valence, unlike D2 neurons". This statement should be attenuated. Their data suggest that both D1R and D2R neurons are involved in both valence and identity and that D1R neurons are more likely involved in valence than identity, and vice versa.

   Claim 2B:
   The authors state "stimulus valence representation by D1 OT neurons is limited to olfactory stimuli, and does not generalize to multimodal stimuli." This statement is premature, and the authors should provide a more nuanced statement. I note two issues: First, the mice had different experimental histories between odors and sounds; the mice were trained with odors first (4 sessions) and then with sounds (3 sessions). Therefore, differential responses between odors and tones can be attributed to their experimental histories rather than olfactory and auditory modalities. Indeed, the data showed that the mice had not fully learned to discriminate between two tones as the mice displayed anticipatory licks upon the tone paired with airpuff. In addition, it is not warranted to have a sweeping generalization because the authors examined only one reward (water) and one type of sounds (tones). Odors may be better conditioned with water while other rewards may work better with tones. Tones may have affective qualities that may have interfered with water conditioning and were needed additional pairing sessions. Remember that the absence of evidence does not provide a proof. It could simply that it was not done well. Moreover, OT neurons clearly responded to the auditory stimuli. Although the OT is strongly linked with the olfactory system compared to other sensory systems, the OT can receive sensory-related information from the limbic cortical structures that provide afferents, including the medial prefrontal cortex, basolateral nucleus of amygdala, and subiculum. Perhaps, the authors should discuss possible roles of these cortical structures in conditioned signals that were detected in the present study and how the large brain damage caused by the grin lens might have compromised afferent inputs to the OT.

   Claim 3:
   The authors state "valence representation is not correlational in nature, but likely serves to inform downstream brain regions of the value of odor stimuli". The present study showed that OT responded to conditioned odors in the absence of behavior. Although this result makes it difficult to hypothesize the functional role of those signals, it is reasonable to infer that the acquired conditioned signals influence downstream systems for some unknown function. However, the method used to obtain the data was correlational. The authors should avoid misleading phrases. Additional experiments are needed to understand functional role of the acquires signals.

   Read the original source
7. 

   Version published to 10.1101/2021.11.01.466363v2 on bioRxiv

   Apr 20, 2022
8. 

   Version published to 10.1101/2021.11.01.466363v1 on bioRxiv

   Nov 2, 2021