1. Developmental Biology
  2. Neuroscience
Download icon

Lineage does not regulate the sensory synaptic input of projection neurons in the mouse olfactory bulb

  1. Luis Sánchez-Guardado
  2. Carlos Lois  Is a corresponding author
  1. California Institute of Technology, United States
Short Report
  • Cited 1
  • Views 798
  • Annotations
Cite this article as: eLife 2019;8:e46675 doi: 10.7554/eLife.46675

Abstract

Lineage regulates the synaptic connections between neurons in some regions of the invertebrate nervous system. In mammals, recent experiments suggest that cell lineage determines the connectivity of pyramidal neurons in the neocortex, but the functional relevance of this phenomenon and whether it occurs in other neuronal types remains controversial. We investigated whether lineage plays a role in the connectivity of mitral and tufted cells, the projection neurons in the mouse olfactory bulb. We used transgenic mice to sparsely label neuronal progenitors and observed that clonally related neurons receive synaptic input from olfactory sensory neurons expressing different olfactory receptors. These results indicate that lineage does not determine the connectivity between olfactory sensory neurons and olfactory bulb projection neurons.

https://doi.org/10.7554/eLife.46675.001

Introduction

The relationship between cell lineage and neuronal connectivity in the brain is not well understood. Lineage regulates the synaptic connections between neurons in some regions of the invertebrate nervous system. For example, in the Drosophila olfactory system, projection neurons are specified by cell lineage to receive synaptic input from the axons of specific types of olfactory sensory neurons (OSNs) (Jefferis et al., 2001; Li et al., 2018). In mammals, it has been reported that clonally related pyramidal neurons are preferentially connected to each other in the neocortex (Yu et al., 2009; Yu et al., 2012; He et al., 2015). Furthermore, it has been proposed that sister neurons in the visual cortex have a strong correlation to the stimuli to which they respond (Li et al., 2012), while other works suggest that this correlation is much weaker (Ohtsuki et al., 2012). To further investigate the role played by lineage in the assembly of brain circuits we focused on the mammalian olfactory bulb, a brain region with an anatomical organization particularly advantageous to study this question.

The mammalian olfactory system can be divided into three regions: olfactory epithelium, olfactory bulb (OB) and olfactory cortex. The olfactory epithelium harbors the OSNs. Each OSN expresses just one of more than one thousand odorant receptors (Buck and Axel, 1991; Chess et al., 1994). OSN axons expressing the same odorant receptor converge into one or two discrete neuropil structures in each OB called glomeruli, forming a stereotypic map on the OB surface (Ressler et al., 1994; Vassar et al., 1994; Mombaerts et al., 1996; Wang et al., 1998). The projection neurons in the OB are called mitral and tufted cells (M/T cells). In mammals, the majority (>90%) of M/T cells have a single apical dendrite that branches into a single glomerulus (Mori, 1987; Shepherd, 1990; Malun and Brunjes, 1996) where they receive sensory input from OSNs expressing a particular odor receptor (Figure 1A) (Ressler et al., 1994; Vassar et al., 1994; Stewart et al., 1979; Mori, 1987; Malun and Brunjes, 1996; Matsutani and Yamamoto, 2000). Thus, the anatomical organization of the glomerulus in the OB is an ideal system to investigate the possible relationship between lineage and connectivity because the apical dendrite of the M/T cells provides a direct readout of their synaptic input. To address this question, we sparsely labeled M/T cells progenitors and investigated the sensory input that their progeny receives from OSNs. Our results show that sister M/T cells receive synaptic input from different glomeruli, indicating that lineage does not determine the sensory input of the OB projections neurons, and suggest that the connectivity between OB projection neurons and olfactory sensory neurons depends on other mechanisms, including random targeting of dendrites towards glomeruli and activity-dependent mechanisms.

Figure 1 with 3 supplements see all
Clonal analysis of projection neurons using Nestin-CreERT2::Confetti mice to sparsely label neuronal progenitors.

(A) Schematic representation of the olfactory bulb (OB). Axons from olfactory sensory neurons (OSNs) expressing the same receptor project to a single glomerulus, forming synaptic contacts with the apical dendrites of mitral and tufted cells. Two possible scenarios of the relationship between lineage and connectivity are presented. (left) The apical dendrites of clonally related M/T cells innervate the same glomerulus, indicating that lineage regulates their connectivity. (right) The apical dendrites of sister M/T cells innervate different glomeruli, indicating that connectivity of M/T cells is independent of their lineage (B) Experimental design to label neuronal progenitors with tamoxifen (TMX) at embryonic day 10 (E10.5), and their posterior analyses at postnatal day 21 (P21). (C) The Confetti cassette encodes four different fluorescent proteins (nuclear GFP (nGFP), membrane CFP (mCFP), and cytoplasmic YFP (cYFP) and RFP (cRFP)). Upon Cre recombination, the STOP sequence is excised and randomly expressed one out four possible fluorescent proteins.

https://doi.org/10.7554/eLife.46675.002

Results and discussion

Labeling of progenitors of OB projection neurons

The projection neurons in the OB are called mitral and tufted cells (M/T cells). M/T cells originate from progenitors located in the OB primordium, which is derived from the rostral part of the dorsal telencephalon (Hinds, 1968a; Hinds, 1968b). To investigate the lineage of M/T cells, we crossed two transgenic mouse: Nestin-CreERT2 (Kuo et al., 2006), which can be used to activate Cre in neuronal progenitors in a sparse manner, and Confetti (Snippert et al., 2010), which can label individual cells with one out four possible fluorescent proteins upon Cre-mediated recombination (Figure 1B,C and Figure 1—figure supplement 1) (Kuo et al., 2006; Snippert et al., 2010).

To investigate whether Nestin promoter drives Cre recombinase activity into M/T cell progenitors, we crossed the driver Nestin-Cre mouse (Tronche et al., 1999) with the reporter Ai9 mouse (Madisen et al., 2010) and confirmed the labeling both of OB progenitors in the embryo, and M/T cells in the adult (Figure 1—figure supplement 2 and Figure 1—figure supplement 3). To be able to perform clonal analysis, we optimized the conditions to label just a handful of progenitors, ideally a single progenitor per OB. First, we confirmed that our transgenic mice Nestin-CreERT2::Confetti did not label any neurons in the brain without tamoxifen (TMX) administration (n = 3; data not shown). Second, we found that with an injection of 1 mg of TMX per 40 g of body weight into a 10-day pregnant female (E10.5) we observed a handful of labeled pyramidal neuron clones in the neocortex, and around 20 M/T cells labeled in the OB when the brains were examined at postnatal day 21 (P21) (Figure 1B and Figure 1—figure supplement 1). Third, we confirmed that this TMX concentration labeled a few progenitors per brain when animals were analyzed 2 days after TMX administration (E12.5) (Figure 2). With these conditions, we observed between none to a single progenitor labeled per fluorescent protein in the OB primordium (n = 6 embryos), the presumptive location of the M/T progenitors. Although we observed a very low number of progenitors labeled, we cannot unambiguously conclude whether a group of cells labeled at P21 with the same fluorescent protein in the OB originated from a single progenitor or two independent progenitors. However, because of the low number of clones labeled with these conditions we will work under the assumption that any group of M/T cells labeled with the same fluorescent protein in the OB are part of a single clone.

Figure 2 with 1 supplement see all
Sparse labeling of progenitor cells in the embryonic mouse brain.

(A–D) Sagittal sections through the brain of an E12.5 mouse treated with TMX at E10.5. (A–B) Confocal images of individual clones labeled in the OB expressing cRFP (A–A’) and cYFP (B–B’). (A’–B’) High-magnification images of the clones shown in A and B. (C–D) Single clones labeled in the neocortex expressing cRFP (C–C’) and cYFP (D–D’). (C’–D’) High-magnification images of the clones shown in C–D). Cell nuclei are labeled with DAPI (blue). Scale bar in D is 200 µm and applies to A-D, scale bar in D’ is 50 µm applies to A’-D’. Orientation of brains: D, dorsal; A, anterior.

https://doi.org/10.7554/eLife.46675.006

To study the lineage of the M/T cells we induced Cre activity at E10.5, the peak time for mitral cell generation (Hinds, 1968a; Hinds, 1968b; Blanchart et al., 2006; Kim et al., 2011; Imamura et al., 2011). Brains were analyzed at P21, once M/T cells have completed the refinement of their dendrites and they have a mature morphology with a single apical dendrite projecting into a single glomerulus (Figure 1A) (Malun and Brunjes, 1996; Lin et al., 2000; Matsutani and Yamamoto, 2000; Blanchart et al., 2006). Confetti mice can produce four different fluorescent proteins with distinct subcellular locations (cytosolic (cRFP and cYFP), membrane (mCFP), and nuclear (nGFP)) (Figure 1C, Figure 1—figure supplement 1 and Figure 2—figure supplement 1) (Snippert et al., 2010). Consistent with previous works, we observed that many clones in the OB were labeled by RFP (n = 9), whereas YFP (n = 4) and CFP (n = 1) clones appeared less frequently (Reeves et al., 2018). We did not analyze any of the nGFP+ cells for two reasons. First, the most reliable way to unambiguously identify M/T cells is by their distinctive morphology. Nevertheless, if a cell is only labeled in the nucleus (as in nGFP+ cells), we cannot tell apart M/T cells from other OB cell types (e.g. short axon cells, granule cells, juxtaperiglomerular). Second, to identify the connectivity between M/T cells and glomeruli, it is necessary to follow the projection of their apical dendrites (Figure 1—figure supplement 1), and we cannot observe any dendrites in the nGFP+ cells.

In total, we analyzed 28 OBs. 15 of them did not have any labeled M/T cells. 11 OBs had both M and T cells labeled (n = 14 clones). Of these 11 OBs, eight had putative clones of a single color, and the remaining three OBs had two clones labeled with different fluorescent proteins. Two OBs had clones that contained only M cells (n = 2 clones). We do not know the reason why these two OBs had only M cells labeled, and several reasons may account for this observation, including progenitors committed to produced only M cells. We did not find any OB with only T cells labeled when TMX was administered at E10.5.

Size of clones and distribution of neurons in the OB and neocortex

We measured the putative clone size in the OB and compared them with neocortex clones. We found 310 labeled M/T cells in 14 putative clones in the OB, such that the average OB clone contained 22.14 ± 6.61 M/T cells (average ± standard deviation). We found 556 labeled cells in six neocortex clones, such that the average cortical clone contained 92.67 ± 23.18 pyramidal neurons (average ± standard deviation), consistent with previous results (Franco et al., 2012; Gao et al., 2014) (Figure 3A). These observations suggest that the clone size in the neocortex is approximately four times larger than in the OB, consistent with the previous results (Cárdenas et al., 2018).

Figure 3 with 4 supplements see all
Clone size and distribution of cells labeled in the olfactory bulb and neocortex.

(A) Clone size quantification in the OB and neocortex. Data are shown as average showing all data points. (B–D) 3D reconstruction of a NestinCreERT2::Confetti P21 mice OB (B–C) and neocortex (D) treated with TMX at E10.5. Gray lines indicate the contours of the brain and red dots represent the cell bodies of labeled neurons. (B) Frontal and (C) lateral views of the 3D reconstruction of one OB. (D) Frontal view of the neocortex 3D reconstruction. (E) Cumulative percentage of the NNDs of sister neurons labeled in the OB (red) and neocortex (dark blue). Data are shown as average ± standard deviation (OB, n = 310 neurons in 14 clones; neocortex, n = 556 neurons in six clones). Pink and light blue lines represent 100 datasets of random simulations of OB and neocortex NND, respectively (see also Figure 3—source data 1). No significant differences were observed when real OB clones were compared to different real OB clones, or when real neocortex clones were compared to different real neocortex clones (OB, p=0.96; neocortex, p=0.95; two-way ANOVA). However, significant differences were observed when real clones were compared with their respective simulated clones (for both OB and neocortex, p<0.01: two-way ANOVA). Scale bar in C is 0.5 mm and applies to B-C. Scale bar in D is 1 mm. Orientation of diagrams in B-D: D, dorsal; A, anterior; M, medial.

https://doi.org/10.7554/eLife.46675.008

We analyzed the distribution of the cell bodies of the labeled M/T cells in the 14 clones containing M and T labeled cells in the OB (n = 310 neurons) and labeled pyramidal neurons in the six neocortex clones (n = 556 neurons) by performing 3D reconstructions using the Neurolucida software (Figure 3B–D, Figure 3—figure supplement 1 and Figure 3—figure supplement 2). The 3D reconstructions revealed that sister M/T cells were distributed in a broader area than the tight columns of sister pyramidal neurons in the neocortex (Figure 3—figure supplement 1 and Figure 3—figure supplement 2). To analyze the distribution of cells from each clone, we calculated the nearest neighbor distance (NND) based on the distances of neurons in our 3D reconstructions (Figure 3E and Figure 3—figure supplement 3). We found that sister M/T cells were more separated from each other (287.47 µm ± 61.23; average ± standard deviation) than sister pyramidal neurons (59.56 µm ± 9.86) (Figure 3E). The dispersion of sister M/T cells that we observed is consistent with the tangential migration of immature M/T cells reported in the embryonic OB (Blanchart et al., 2006; Imamura et al., 2011).

To investigate whether the distribution of sister M/T cells observed was random, we compared the NNDs of the labeled M/T cells observed (n = 310) with a simulated random dataset. The same strategy was followed for neocortex clones. We found that the NNDs between clonally related neurons were shorter than the simulated random datasets both for the OB and neocortex (Figure 3E, p<0.01; two-way ANOVA). Similar results were reported for pyramidal neurons in the neocortex (Gao et al., 2014). This indicates that although sister M/T cells are not obviously clustered, their distribution in the OB is not random. Interestingly, previous works have observed that the tangential migration of immature M/T cells in the embryonic OB may be regulated by gradients of secreted (Inokuchi et al., 2017) or cell adhesion molecules (Bastakis et al., 2015), biasing their distribution to specific regions within the OB.

Previous experiments have demonstrated that migration of M/T cells is biased toward the dorsal or ventral regions of the OB at different developmental times (Imamura et al., 2011). In addition, it has been hypothesized that the dorsal and ventral domains of the OB may have a preference to process innate and learned odorants, respectively (Kobayakawa et al., 2007). To investigate whether the cell distribution in a clone was biased toward a specific OB domain, we divided the OB into two domains based on the expression of the OSN markers NQO1 and OCAM, that label the dorsal and ventral regions of the OB, respectively (Figure 3—figure supplement 1K; Gussing and Bohm, 2004; Yoshihara et al., 1997). Then, we analyzed the distribution of clonally related M/T cells throughout these two domains (Figure 3—figure supplement 1). Of the 14 OB clones we analyzed, four clones had a bias toward the ventral OB domain, three clones for the dorsal domain, and the remaining seven clones had similar number of cells in the dorsal and ventral domains. Overall, when all the clones were analyzed together, there were no preferences in the distribution of M/T cells towards the dorsal or ventral domains (Figure 3—figure supplement 1L; p=0.67, t-test). Similarly, we did not detect any bias for the distribution of M/T cells OB in the lateral or medial domains (data not shown).

To analyze whether labeling of M/T progenitors at different developmental times could influence the distribution of M/T cells to a specific OB domain, we performed additional experiments to label M/T progenitors at a later time point by injecting TMX into 12 day pregnant females (E12.5), and brains were examined at P21, as in the E10.5 experiment. Previous works have demonstrated that in the neocortex the number of neurons per clone is reduced as progenitors are labeled at later embryonic stages (Angevine and Sidman, 1961; Walsh and Cepko, 1988; Luskin et al., 1988; Price and Thurlow, 1988; Rakic, 1988; Gao et al., 2014). Consistent with this observation, the clones that were labeled at E12.5 in the OBs contained fewer cells than at E10.5: 8.44 ± 6.37 M/T cells per clone (average ± standard deviation, n = 76 cells) when labeled at E12.5, compared with 22.14 ± 6.61 M/T cells per clone when labeled at E10.5 (Figure 3—figure supplement 4B, C-I'). Similarly, we observed a reduction in the number of cells per clone in the neocortex when labeling progenitors at later developmental stages (22.5 ± 6.47 pyramidal neurons (n = 135 cells) at E12.5 versus 92.67 ± 23.18 pyramidal neurons at E10.5)), consistent with previous results (Franco et al., 2012; Gao et al., 2014) (Figure 3—figure supplement 4B,K–P). In total we analyzed 18 OBs with progenitors labeled at E12.5. Eleven OBs did not have any M/T labeled cells. Seven OBs had nine clones with labeled M/T cells. Of these seven OBs, five s had a single putative clone, each clone labeled with a single fluorescent protein. Each of the other two OBs had two clones labeled with different fluorescent proteins. As in our E10.5 experiment, we observed that sometimes the cells in a clone were preferentially located in a specific domain (dorsal or ventral, or medial or lateral), although overall we did not find any significant differences in their distribution (Figure 3—figure supplement 4J).

Our experiments were designed to investigate the relationship between lineage and connectivity in the main olfactory bulb (MOB). Although not the primary goal of our work, these experiments gave us the opportunity to investigate whether M/T cells in the accessory olfactory bulb (AOB) were clonally related to the M/T cells in the MOB. When TMX was administered at E12.5, we did not find any M/T cells labeled in the AOB, consistent with the observation that AOB M/T cells are born earlier than MOB M/T cells (Hinds, 1968a). When we injected TMX at E10.5 we observed a small number of labeled M/T cells in the AOB. We inspected 28 OBs labeled at E10.5, and found that 10 OBs contained 18 M/T cells labeled in the AOB, with only 1–3 labeled M/T cells per AOB. Four OBs had 1–2 labeled M/T cells in the AOB and none in the MOB. Four OBs had M/T cells labeled with the same fluorescent proteins in both MOB and AOB, with only 1–3 cells in each AOB. The remaining two OBs had one cell in each AOB labeled with a fluorescent protein different from the M/T cells labeled in the MOB (see table in Supplementary file 1). Although these small numbers do not allow for a definitive conclusion, our results suggest that there are separate progenitors for the M/T cells in the MOB and AOB. This hypothesis is consistent with recent works indicating that some M/T cells in the AOB are born from progenitors located in the diencephalic-telencephalic boundary, which then migrate rostrally to the posterior AOB (Huilgol et al., 2013; Ruiz-Reig et al., 2017). Further experiments will be required to clarify these questions.

Synaptic input of sister M/T cells

It has been proposed that the anatomical organization of the OB may be analogous to the neocortex columnar organization. In the neocortex it is thought that the pyramidal neurons forming part of a column perform a similar task (Rakic, 1988; Mountcastle, 1997). Similarly, M/T cells receiving synaptic input from the same glomerulus respond to the same odorant (Kauer and Cinelli, 1993; Mori et al., 1999; Bozza et al., 2002). Our results indicate that sister M/T cells are widely distributed throughout the OB (Figure 3, Figure 3—figure supplement 1 and Figure 3—figure supplement 4). Based on this observation, it may seem unlikely that sister M/T cells would have apical dendrites projecting into the same glomerulus. However, this could still be possible because the soma of M/T cells innervating the same glomerulus may be separated from each other up to 450 µm (for M cells) and 350 µm (for T cells) (Liu et al., 2016). To investigate whether sister M/T cells receive synaptic input from the same glomerulus, we tracked their apical dendrites (Figure 4 and Figure 4—figure supplement 1). Among all the labeled M/T cells that we detected (310 cells from 14 putative M/T clones (E10.5) and 74 cells from nine putative M/T clones (E12.5)), we never observed two neurons innervating the same glomerulus, even when their cell bodies were near each other (Figure 4B–E, Figure 4—figure supplement 1E). Nevertheless, it is still possible that, although we did not observe them, there may exist clones of M/T cells in the OB genetically pre-determined to project to the same glomerulus. This scenario could be expected for putative glomeruli responsive to relevant odors for survival, such as those responsive to predators or poisons, which require an innate and hardwired response of avoidance (Kobayakawa et al., 2007; Sosulski et al., 2011). Future experiments analyzing a much larger number of clones than those detected here may reveal the existence of these putative ‘hardwired’ M/T clones.

Figure 4 with 2 supplements see all
Connectivity of clonally related M/T cells when TMX was administered at E10.5.

(A) Confocal images of four sister M/T cells belonging to a putative individual clone in the OB. (B–E) Confocal images of sister M/T cells from four clones, in four different OBs, with their somata close to each other and their apical dendrites innervating different glomeruli. (B’–E’) Schematic representation of the confocal images in B-E. Scale bar in E is 50 µm and applies to A-E.

https://doi.org/10.7554/eLife.46675.015

It is generally thought that the AOB has a preference for innate odorants, and thus, one may anticipate that lineage may regulate the connectivity of AOB projection neurons. However, there is a critical caveat that make it difficult to investigate the relationship between lineage and connectivity in the AOB. Although glomeruli are clearly distinct in the MOB, glomeruli in the AOB are less well defined and more difficult to identify. As indicated above, we observed only a small number of AOBs (four out of 10) that contained more than one (2 or 3) labeled M/T cells. Although the small number of labeled AOB M/T cells does not allow us to draw any firm conclusions, we did not find any M/T cells whose apical dendrites innervated the same glomerulus (Figure 4—figure supplement 2), similar to what we observed in the MOB.

In summary, our results indicate that lineage does not determinate the input connectivity of the apical dendrites of projection neurons in the mammalian OB. This is in contrast to what has been described for projection neurons in the Drosophila antennal lobe (Jefferis et al., 2001) and suggested for pyramidal neurons in the rodent visual cortex (Li et al., 2012). Our results indicate that the sensory input received by M/T cells is regulated by other factors independent of lineage, including random targeting of dendrites towards glomeruli and activity-dependent mechanisms, consistent with previous observations from multiple lines of evidence. First, during early postnatal stages M/T cells have several dendrites (between 3 to 5), and each of these dendrites project into different glomeruli that are close to each other, and immediately above their cell bodies (Hinds, 1968a; Blanchart et al., 2006). Starting approximately 1 week after birth, a process of refinement occurs such that around 90% of M/T cells retain just one apical dendrite and retract all others, and that remaining single apical dendrite branches into a single glomerulus (Malun and Brunjes, 1996; Lin et al., 2000; Matsutani and Yamamoto, 2000; Blanchart et al., 2006). It is important to note that even in full adult animals approximately 10% of mature M/T cells have two apical dendrites that project into two different glomeruli (Lin et al., 2000). Interestingly, the refinement by which M/T cells retain a single dendrite is a process partially dependent on neuronal activity. Olfactory deprivation by naris occlusion retards the refinement of M/T cell dendrites by approximately one week, although eventually the refinement process is accomplished to the same degree as in non-manipulated animals (Matsutani and Yamamoto, 2000). Interestingly, a recent work demonstrated that genetic blocking of action potentials in M/T cells prevented the dendrite refinement process such that even in adult animals the majority of M/T cells have several dendrites projecting into multiple glomeruli (Fujimoto et al., 2019). Finally, recent experiments indicate that activity-dependent mechanisms can direct the projection of M/T cell dendrites into specific glomeruli. For example, sensory odor experience in utero recruits the apical dendrites of M/T cells to the activated glomeruli (Liu et al., 2016). Similarly, genetic ablation of a large set of OSNs results in the absence of a large number of glomeruli in the dorsal OB, and in these animals, some M/T cells located in those regions lacking glomeruli extend their dendrites tangentially for a long distance until they reach a region with glomeruli, where they branch (Nishizumi et al., 2019).

In summary, multiple observations indicate that M/T cells are not committed to project into specific glomeruli. Instead, the available evidence, including the data presented here, suggests a model where progenitor cells give rise to a clone of sister M/T cells that migrate throughout the olfactory bulb such that sister cells disperse independently from each, and their cell bodies do not end up close to each other in specific regions of the bulb. After neuronal migration is completed, immature M/T cells initially grow multiple dendrites that receive synaptic input from multiple glomeruli without any apparent specificity. After a period of refinement regulated, in part, by neuronal activity, most (but not all) M/T cells retain a single dendrite that branches into a single glomerulus. However, the available evidence indicates that any of the multiple apical dendrites displayed by immature M/T cells can be retained, suggesting that M/T cells are not committed to receive synaptic input from any specific glomeruli. Finally, it is curious that the targeting of OSN axons and M/T dendrites toward the glomeruli appears to be regulated by very different mechanisms. Each OSN expresses a single olfactory receptor molecule that instructs its axons to project into a single glomerulus with high specificity (Ressler et al., 1994; Vassar et al., 1994; Mombaerts et al., 1996; Wang et al., 1998). In contrast, the existing evidence suggests that the apical dendrite of M/T cells can project to any glomeruli within a certain distance from the position of their cell bodies, without any apparent specificity. What is the relationship between the OSN axons and the M/T dendrites for synapse formation in the glomeruli? Animals with mutations in the Tbr1 gene that result in the complete loss of M/T cells demonstrate that OSN axons can still reach the OB and converge into glomeruli-like structures in the same location as in wild-type animals (Bulfone et al., 1998). These experiments suggest that the targeting of OSN axons into the OB to form glomeruli does not require the presence of M/T cells. In contrast, the apical dendrites of M/T cells cannot form glomeruli in regions in mice in which a large set of olfactory receptors are genetically ablated, indicating that M/T cells require the presence of OSN axons to target their apical dendrites (Kobayakawa et al., 2007; Nishizumi et al., 2019).

Is there any biological advantage to the dispersion of projection neurons in the OB such that sister M/T cells receive synaptic input from different OSNs? Interestingly, it has been proposed that the M/T cells receiving input from the same glomerulus exhibit a wide diversity in their biophysical properties, and this diversity may be important for neural coding (Padmanabhan and Urban, 2010). In addition, neurons in the piriform cortex receive synaptic input from M/T cells innervating different glomeruli (Miyamichi et al., 2011), whereas M/T cells connected to the same glomerulus project their axons into many different areas of the olfactory cortex (Sosulski et al., 2011; Ghosh et al., 2011). However, the connectivity between M/T cells and the amygdala appears to be more stereotypical than between the M/T cells and other targets in the olfactory cortex (anterior olfactory nucleus, piriform cortex, tenia tecta, olfactory tubercle, cortical amygdala and entorhinal cortex) (Haberly, 2001; Sosulski et al., 2011). Based on these observations, one can speculate that the connectivity between the OB and its targets in the olfactory cortex may occur by two different mechanisms. Genetic factors, including lineage, may contribute to the connectivity between M/T cells and the amygdala, as this brain area is involved in innate behavior responses that may require hardwired connections (Sosulski et al., 2011). In contrast, the connectivity between M/T cells and areas of the olfactory cortex involved in the perception of odors that do not elicit innate behaviors are more plastic and may be regulated by mechanisms independent of lineage, such as random neurite targeting and activity-dependent wiring, among others (Caron et al., 2013; Schaffer et al., 2018). Our results indicating that lineage does not determine the sensory synaptic input of M/T cells raise further questions about the assembly of the olfactory circuits, including the mechanisms regulating the formation of synapses between OSNs and M/T cells, the role that experience may play sculpting the odor representations in the piriform cortex, and whether lineage regulates the connections with the amygdala to trigger innate behaviors.

Materials and methods

Animals

Nestin-CreERT2, Nestin-Cre, Confetti, and Ai9 mice were obtained from Jackson Laboratory. The Nestin-Cre and Nestin-CreERT2 mice can be used to induce the activity of Cre recombinase in neuronal progenitors directly or by the administration of tamoxifen (TMX) into animals, respectively (Tronche et al., 1999; Kuo et al., 2006). The Ai9 mouse is a Cre-dependent reporter that expresses tdTomato fluorescent protein upon cre-mediated recombination (Madisen et al., 2010), while the Confetti mouse is a Cre-dependent reporter that produces four different fluorescent proteins (Snippert et al., 2010). We crossed the Nestin-Cre mouse with the Ai9 mouse and the Nestin-CreERT2 mouse with the Confetti mouse. The resulting transgenic Nestin-Cre::Ai9 and Nestin-CreERT2::Confetti mice were used for the experiments. For the timed pregnancy, the plug date was designated as E0.5 and the day of birth as P0. In all experiments, mice were handled according to the protocols approved by the Caltech Institutional Animal Care and Use Committee (IACUC). Mice colonies were maintained at the animal facility of the California Institute of Technology (Caltech).

Tamoxifen induction

Request a detailed protocol

Tamoxifen (TMX, Sigma T-5648) was dissolved in 37°C pre-warmed corn oil (Sigma C8267) at a concentration of 10 mg/ml. NestinCreERT2::Confetti embryos were induced at E10.5 (embryonic day 10.5) by a single intraperitoneal injection of 1 mg TMX into pregnant females (~40 grams). Animals were euthanized at embryonic day 12 (E12.5) or postnatal day 21 (P21).

Tissue processing, immunohistochemistry, and imaging

Request a detailed protocol

Mouse embryos (E10.5 and E12.5) were fixed by immersion in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS, pH 7.4) at 4°C overnight. Postnatal mice (P7 and P21) were fixed by intracardiac perfusion with 4% PFA in PBS. Brains were then extracted and incubated in 4% PFA at 4°C overnight. Next day, all samples were washed three times, 10 min each, with 0.1 M PBS, pH 7.4. Postnatal mice (P21) brains were embedded into 3% agarose and cut in a vibratome into 60 μm thick sections. Sections were collected sequentially. Embryonic and P7 brains were cut with a cryostat into 20 μm thick sections as previously described (Sánchez-Guardado et al., 2009).

We amplified the signal from fluorescent proteins by performing immunohistochemistry with antibodies against RFP and GFP. Although anti-GFP antibody recognizes nGFP, cYFP and mCFP proteins, we were able to distinguish between them based on the different subcellular location of the proteins (nuclear, cytoplasmic and membrane). In the figures, cells are shown with their original colors from the Confetti cassette, even though the signal from cYFP and mCFP proteins was amplified using the antibody against GFP (Figure 1—figure supplement 1, Figure 2, Figure 2—figure supplement 1). We did not include nGFP+ cells in our analyses because we cannot identify their morphology.

For immunocytochemistry, we incubated the sections for 30 min in blocking solution containing 1% bovine serum albumin in 0.1 M PBS-0.1% Triton X-100 (PBS-T). Sections were incubated overnight with the following primary antibodies diluted into blocking solution: 1:1000 chicken anti-GFP, Aves Labs Cat# GFP-1020 (RRID:AB_10000240), 1:1000 rabbit anti-RFP, Lifespan Cat# LS-C60076-100 (RRID:AB_1514409), 1:1000 rat anti-RFP, ChromoTek Cat# 5f8-100 (RRID:AB_2336064); 1:500 rat anti-Tbr2, Thermo Fisher Scientific Cat# 14-4875-82 (RRID:AB_11042577), 1:10,000 rabbit anti-Tbx21(kind gift from Y. Yoshihara), 1:250 rabbit anti-PAX6, Covance Cat# PRB-278P, (RRID:AB_291612), 1:20 mouse anti-RC2, DSHB Cat# RC2, (RRID:AB_531887). The next day, sections were washed three times, 10 min each, in PBS-T. Later, sections were incubated for 90 min at room temperature with secondary antibodies: Goat anti-chicken IgY Alexa-488 (RRID:AB_2534096), Donkey Anti-Rat IgG Alexa-488 (RRID:AB_2535794), Goat anti-Rabbit IgG Alexa-488 (RRID:AB_143165), Goat anti-Mouse IgG Alexa-488 (RRID:AB_2534069), Goat anti-Rat IgG Alexa-555 (RRID:AB_141733), Goat anti-Rabbit IgG Alexa-555 (RRID:AB_2535850), Goat anti-Rabbit IgG Alexa-647 (RRID:AB_2535812) diluted 1:1500 in blocking solution. Finally, the sections were counterstained with DAPI (D9542, Sigma), mounted sequentially on glass slides and mounted with Fluoromount (F4680, Fluoromount Aqueous Mounting Medium).

Z-stacks images were acquired using 10x, 20x or 40x objectives on a confocal microscope (Zeiss LSM 800). Z-stacks were merged and analyzed using ImageJ and edited with Photoshop (Adobe) software.

3D reconstruction and data analysis

Request a detailed protocol

Each section was analyzed and traced in sequential order from rostral to caudal using Neurolucida and StereoInvestigator software (MBF Bioscience Inc, Williston, VT). The boundaries of the OB and neocortex were traced and used to line up each section with the previous one to form 3D reconstructions. Each labeled cell in the OB or neocortex was tagged with a dot. Blue dots represent mCFP cells, red dots cRFP cells and green dots cYFP.

The distribution of the nearest neighbor distance (NND) was calculated using Matlab based on the cell coordinates of our 3D reconstruction created in Neurolucida software. NND was calculated by identifying the shortest straight path between labeled cells using the Euclidean distance. The NND was represented as cumulative percentage (average ± standard deviation) of the clones analyzed in the OB (n = 14) and neocortex (n = 6) (Figure 3E). In addition, we generated a dataset of random simulations based on the same number of M/T cells detected in our experiments (n = 310). The random dataset was generated based on the external plexiform layer (EPL) volume from one of the OBs analyzed. Using Matlab, we randomized eight times the number of cells of each OB clone in the EPL volume (n = 112 simulations). Using the same procedure, we randomized 17 times the number of pyramidal neurons of each neocortical clone (n = 102 simulations). The volume of one neocortex clone, representative of the average, was used as a volume boundary. Data are presented as average ± standard deviation, and statistical differences in the clone distribution were determined using two-way analysis of variance (ANOVA).

The division of the OB into dorsal and ventral domains was based on the expression of the NQO1 and OCAM markers (Figure 3—figure supplement 1) based on the previous published results (see Figure 7 in Cho et al., 2007 and Figure 1 in Imamura et al., 2011). The results were analyzed using unpaired two-tailed t-test.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
    Cell lineage in the rat cerebral cortex: a study using retroviral-mediated gene transfer
    1. J Price
    2. L Thurlow
    (1988)
    Development 104:473–482.
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61

Decision letter

  1. Stephen Liberles
    Reviewing Editor; Harvard Medical School, United States
  2. Catherine Dulac
    Senior Editor; Harvard University, United States
  3. Stephen Liberles
    Reviewer; Harvard Medical School, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Lineage does not regulate the connectivity of projection neurons in the mouse olfactory bulb" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Reviewing Editor and Catherine Dulac as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Stephen Liberles (Reviewer #1).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

All reviewers noted that the work was interesting and the data generally convincing, but also indicated that additional experiments were needed to support principal conclusions. Reviewer comments are attached in their entirety below, but a few specific points are highlighted to help focus your revision efforts.

1) All reviewers asked for more validation that all mitral cells could be labeled by the approach. This is important given that dedicated subsets of mitral cells might be relevant for innate behaviors, and that such neurons might be born at different times, in different locations (dorsal/ventral), or not be marked in Nestin-Cre mice.

2) Reviewers 2/3 also ask for analysis of central projections, (reviewer 3 mentions the cortical amygdala as it has been implicated in innate behaviors). Are the central projections of different clonal populations indistinguishable from global mitral cell projections across higher order olfactory nuclei?

3) Reviewers also inquire about sister cells within the AOB, and whether sisters can be distributed among both the AOB and MOB. Is the logic underlying glomerulus connectivity of AOB sister cells distinct?

Please see all comments below.

Reviewer #1:

The authors use a genetic approach to visualize small groups of olfactory projection neurons with a putatively clonal origin. They find that neurons from the same putative clone can project to different glomeruli, and as such, they are likely relevant for different odor responses. These findings argue against lineage-based determinism underlying mitral cell identity. Given the large number of glomeruli, and the ability of the olfactory map to expand and contract with receptor number, lineage-based models do not seem likely from the get-go; that said, I am not sure whether lineage models have formally been challenged in the mouse. Certainly, the notion that glomerulus innervation occurs through an at least partially stochastic mechanism is an intuitive and attractive model, and the authors present data that is at least consistent with this idea. These findings are also consistent with recent work indicating that the receptor-guided spatial map in the olfactory bulb is discarded in the piriform cortex. It should be noted that it is possible that neuron specification occurs after time points of clonal expansion measured here, or in smaller subsets of neurons not detectable here (as the authors discuss).

I only have one technical point. The authors should provide additional evidence that clonal populations are being analyzed. They report ~20 projection neurons labeled per olfactory bulb, and 16/29 olfactory bulbs with labeled cells. Is only one clone (and thus one fluorophore) detected in all ~20 neurons from each of these 16 bulbs? In a table, provide the exact distribution of mitral cells with different fluorophores from each mouse as this information is essential for assessing the extent of clonality and whether the assumption at the end of the second paragraph of subsection “Labeling of progenitors of OB projection neurons” is valid. Conclusions depend on an absence of multiple independent labeling events.

Reviewer #2:

Summary:

Using a genetic strategy to sparsely label neural progenitors, Sánchez-Guardado and Lois found that sister olfactory bulb mitral/tufted cells innervate multiple glomeruli distributed across the olfactory bulb. This paper has relevance to the general problem of understanding how the bulb wires up, and implications for the organization of projections connecting the bulb to the cortex.

Experimental Concerns:

1) Previous work, such as Imamura et al., (2011, 2015), has shown that early-generated mitral cells (E10) cells are distributed more in dorsal domain of the olfactory bulb, whereas late-generated mitral (E12~13) cells are preferentially distributed in ventral domain. Given this, it is important to examine multiple timepoints for tamoxifen injection and ask if there are any timing-dependent changes in clonal distribution of mitral/tufted cells (e.g. test additional embryonic timepoints in addition to E10.5).

2) It is not clear if all the mitral/tufted cells are generated from Nestin-positive RGCs. This is important because otherwise the results shown here may miss subpopulations of more restricted or distributed sister mitral/tufted cells. Using Nestin-Cre or giving more tamoxifen can answer this.

3) The small number of clones (typically 1) labeled at a single time-point makes it hard to interpret their results as actually representative throughout the olfactory bulb. They authors should quantify or demonstrate to what extent their clones uniformly cover the olfactory bulb. In addition, more images of the clones themselves would be useful.

4) In addition to asking about the bulb, this work affords the opportunity to look at connectivity and organization in cortex. Given that prior work (e.g. Miyamichi et al., 2011; Sosulski et al., 2011) suggests that M/T cells innervating a single glomerulus project diffusely across olfactory regions, the authors could provide an explanation for these findings by examining the projection patterns of sister M/T cells, which they should be able to do by looking at downstream olfactory bulb targets of the clones labeled in this study.

5) The authors should examine if sister mitral/tufted cells were distributed across dorsal and ventral domains (as labeled with Nqo1 and OCAM, respectively) or they were specific to either domain. This could provide some information about how the distribution of sister cells is regulated.

6) The methods for their randomized NND analysis are unclear. The authors stated that "the distances were generated randomly with a normal distribution 230 between the longest and shortest distances observed between M/T cells." However, a normal distribution is specified by a mean and variance, neither of which are given. Furthermore, their randomization methods differ from those described previously (e.g. Gao et al., 2014), which aim to randomly sample within a given volume.

7) Is there any spatial bias in the position of labeled progenitor cells by this experiment? It would be better to show the quantification for this to make sure they labeled more or less throughout the entire pool of the progenitors in embryonic brain.

8) No discussion is made of AOB M/T cell clones. Do single clones contain sister cells in both the MOB and AOB?

9) Show all data points for Figure 3A and show some quantification of the distributions for all clones.

10) The citation in this sentence “This scenario could be expected for putative glomeruli responsive to relevant odors for survival, such as those responsive to predators or poisons, which require an innate and hardwired response of avoidance (Sosulski et al., 2011).” does not make sense.

Conceptual concerns:

1) Lack of discussion with respect to the timing OSN development: If it's known that OSNs innervate glomeruli after mitral/tufted cell differentiate and migrate to their positions in the olfactory bulb (e.g. Blanchart et al., 2006), would one necessarily expect that sister mitral cells innervate the same glomerulus?

2) If bulb neurons migrate tangentially, wouldn't one expect the clones to be distributed more than those in cortex? Likewise, as the authors state, since it's already known that the olfactory bulb has higher levels of direct neurogenesis rather than indirect neurogenesis via intermediate progenitors (Cárdenas et al., 2018), the smaller clone size of the olfactory bulb already suggests that these regions have different modes of development. Likewise, it remains unclear whether using a single time point for tamoxifen injection labels progenitors at equivalent stages of potency/development across these separate regions.

3) The authors did not show if sister mitral/tufted cells are connected to each other or connect to the same cells other than olfactory sensory neurons. The authors state that they "investigated whether lineage plays a role in the connectivity of mitral and tufted cells," however by connectivity here they only refer to connections with OSNs innervating a single glomerulus. Especially given the extent of intrabulbar connectivity, as well as recent reports of biases in connectivity among clonally-derived neocortical neurons, if the authors would like to claim that "lineage does not regulate the connectivity of projection neurons in the mouse olfactory bulb," which their Title reads, it is not sufficient to only show that sister mitral/tufted cells do not extend dendrites to the same glomerulus.

Reviewer #3:

In this report, the authors attempted to address an interesting issue in the mouse olfactory system, i.e., whether the lineage of mitral/tufted (M/T) cells determines the connectivity to their partner glomeruli within the olfactory bulb (OB). By using an elegant transgenic system, the authors sparsely labeled M/T cell progenitors to study whether sister M/T cells connect to the same glomerulus. They found that clonally-related M/T cells do not necessarily synapse with olfactory sensor neurons (OSNs) expressing the same odorant receptor. Based on this observation, the authors conclude that cell lineage does not determine the synaptic connectivity of M/T-cells with OSN axons.

In the paper, the experiments are well-designed and the results are mostly clear. As this paper presents basically negative results, adding some positive data as to a role of cell lineage in forming M/T-cell circuits would strengthen the paper. It will be quite interesting to examine whether the sister M/T cells send their axons to the same area in the olfactory cortex (OC). It will also strengthen the paper if the authors could discuss more about the connectivity of M/T cells regarding possible mechanisms that mediate partner matching with glomeruli. Specific comments are as follows:

1) In Figure 4, some examples are shown for sister M/T cells connecting to different glomeruli. Are there any differences in sister-cell distribution between the innate and non-innate OB regions?

2) Is it possible to determine when and where the sister cells are derived during embryonic development?

3) Even if sister cells do not connect to the same glomerulus, are there any shared characteristics and common features in their gene expression (particularly for axon guidance molecules), OC projection (particularly to the amygdala), and firing patterns?

4) It is worth examining M/T cells in the accessory OB where all M/T-cell circuits are hard-wired to mediate innate pheromone responses?

5) In the last part of Discussion section, the authors list interesting future questions. This paper would be significantly strengthened if any results could be added regarding these questions, particularly for connectivity to the amygdala.

https://doi.org/10.7554/eLife.46675.021

Author response

All reviewers noted that the work was interesting and the data generally convincing, but also indicated that additional experiments were needed to support principal conclusions. Reviewer comments are attached in their entirety below, but a few specific points are highlighted to help focus your revision efforts.

We thank the editors and reviewers for their comments on the manuscript. We have performed the additional experiments they suggested and made other changes to the manuscript and figures to address their concerns. Finally, we have also made smaller changes to text to improve clarity and readability.

1) All reviewers asked for more validation that all mitral cells could be labeled by the approach. This is important given that dedicated subsets of mitral cells might be relevant for innate behaviors, and that such neurons might be born at different times, in different locations (dorsal/ventral), or not be marked in Nestin-Cre mice.

a) Are there different subsets of M/T cells born at different times with different characteristics?

We initially focused at embryonic day 10.5 because according to the existing literature, this is the time in which the majority of M/T cells are born (Hinds, 1968a, 1968b; Blanchart et al., 2006; Kim et al., 2011; Imamura et al., 2011). In our original submission, all data were derived from labeling E10.5 progenitors. The reviewers raised the question of whether clonally related cells born from later progenitors may produce sister M/T cells that innervate the same glomerulus. We analyzed a later time point of tamoxifen administration (E12.5) and have observed that sister M/T cells born from this E12.5 progenitors do not receive synaptic input from the same glomerulus. These new data are included in a new figure (Figure 4—figure supplement 1).

b) Are there progenitors that produce subsets of M/T cells preferentially destined to occupy the dorsal or ventral regions of the bulb?

We have analyzed the distribution of clones of M/T cells born both at E10.5 and E12.5. We have observed that at both of these time points there are clones preferentially located in the dorsal domain, in the ventral domain and clones with no preference for either dorsal or ventral domains. Moreover, when all clones are analyzed together, we did not observe any preference for distribution in the dorsal or ventral domains. These new data are now included in a new figure (Figure 3—figure supplement 1 (E10.5), and Figure 3—figure supplement 4 (E12.5))

c) Are all M/T cells labeled in Nestin-cre mice?

We agree that these are important questions and we have performed additional experiments to address them. Below we will reply to these questions in order:

To address this question, we have performed two complementary experiments.

First, according to the literature, progenitors for M/T cells are located in the rostralmost region of the forebrain, also called the presumptive olfactory bulb. As suggested by one of the reviewers we crossed the Nestin-cre mouse with the Ai9 mouse strain, a highly sensitive loxP-cre reporter that expresses tdTomato upon cre mediated recombination. When we analyzed these animals at E10.5 we observed that all neuronal progenitors (identified by the expression of Pax6 and RC2 markers) are also positive for tdTomato. This observation indicates that the Nestin-cre can activate recombination of a loxP-reporter in all neuronal progenitors located in the presumptive OB at the time when we labeled the clones of M/T cells that we analyzed in the OB. These new data are presented in a new figure (Figure 1—figure supplement 2)

Second, we analyzed the expression of the tdTomato marker in the OB in a postnatal day 7 mice carrying the Nestin-cre and Ai9 reporter alleles. We observed that all M/T cells analyzed (identified by the expression of the M/T cells marker Tbx21) are also positive for the tdTomato marker. These new data are presented in a new figure (Figure 1—figure supplement 3)

2) Reviewers 2/3 also ask for analysis of central projections, (reviewer 3 mentions the cortical amygdala as it has been implicated in innate behaviors). Are the central projections of different clonal populations indistinguishable from global mitral cell projections across higher order olfactory nuclei?

Indeed, this is an interesting question, and we included a brief discussion of this possibility in our original submission. The reason why we only mentioned this as a possibility, but we decided not to present any data in this regard is because with the currently available transgenic mice it is not possible to reliably address this issue. Recently, several works have traced the final destination of axons originating from M/T cells locally labeled in the OB (Sosulski et al., 2011, Ghosh et al., 2011, Igarashi et al., 2012). Although there are transgenic mice that can be used to selectively label neocortical progenitors (such as Emx1-CreERT2), currently there are no transgenic mice capable of selective labeling of M/T progenitors. Because of this limitation, to label progenitors of M/T cells we used the Nestin-CreERT2 mice that labels all neuronal progenitors throughout the brain. Therefore, when we induce cre recombination in these mice we label, not only M/T progenitors but also many other progenitors in other brain regions, including the olfactory cortex and other brain regions with neurons whose axons project into the olfactory cortex. When we attempted to analyze the trajectories of axons in the olfactory cortex we observed some local neurons labeled in the olfactory cortex (see Author response image 1F, illustrating a labeled neuron in the olfactory cortex). These neurons have axons that extended locally within the olfactory cortex, and this makes it extremely challenging to discern whether a given labeled axon in the olfactory cortex belongs to a M/T cell from the OB, or to local neurons in the olfactory cortex. Thus, we did not attempt to trace the destination of axons in the olfactory cortex because we could not unambiguously identify the cells from which they originated. As an example, we attach confocal images from two brains as examples of axons in the olfactory cortex that illustrate this challenge.

Author response image 1
Axons in olfactory cortical regions.

(A-F) Confocal images of the axons in the lateral olfactory tract (LOT), in two different brains, at different anterior-posterior levels of the brain in coronal sections. (A-C) Axons in the LOT at the anterior olfactory nucleus and (D-F) piriform cortex in coronal sections from a brain with M/T cells labeled in the OB. Scale bar in F is 100 µm and applies to A-F. Orientation of brain: D, dorsal; M, medial.

3) Reviewers also inquire about sister cells within the AOB, and whether sisters can be distributed among both the AOB and MOB. Is the logic underlying glomerulus connectivity of AOB sister cells distinct?

This is a very interesting question, especially because the AOB is thought to process smells involved in innate behaviors, which could be genetically programmed. In our initial submission we decided to exclusively focus on the relationship between lineage and connectivity for M/T cells in the MOB, because identifying the connectivity of dendrites in the MOB is unambiguous, but the anatomical characteristics of the AOB make it a less reliable system where to study this question. In the MOB, M/T cells have a single apical dendrite that projects into a single, well-defined glomerulus. The AOB has a critical caveat to study questions related to synaptic connectivity because its glomeruli are not well defined, and thus, are difficult to identify. In addition, we also observed that whereas the number of cells per clone in the MOB was (on average) 20 cells, for the AOB the clone size was much smaller (between 1-3 per AOB). Taking into account these limitations we now include these data in the revised manuscript. However, we think that given these caveats we thought that it would be important to temper any interpretation, as follows: “Although the small number of labeled AOB M/T cells does not allow us to draw any firm conclusions, we did not find any M/T cells whose apical dendrites innervated the same glomerulus (Figure 4—figure supplement 2), similar to what we observed in the MOB”.

Please see all comments below.

Reviewer #1:

[…] I only have one technical point. The authors should provide additional evidence that clonal populations are being analyzed. They report ~20 projection neurons labeled per olfactory bulb, and 16/29 olfactory bulbs with labeled cells. Is only one clone (and thus one fluorophore) detected in all ~20 neurons from each of these 16 bulbs? In a table, provide the exact distribution of mitral cells with different fluorophores from each mouse as this information is essential for assessing the extent of clonality and whether the assumption at the end of the second paragraph of subsection “Labeling of progenitors of OB projection neurons” is valid. Conclusions depend on an absence of multiple independent labeling events.

We agree that this is a valuable addition to the ms., and we have now included all these data in two new supplementary figures (Figure 3—figure supplement 1 and Figure 3—figure supplement 4). In these figures we show all the clones obtained by labeling at E10.5 and E12.5, and we indicate the following: (i) the 3D reconstructions of the OBs illustrating the distribution of each labeled M/T cell through the bulb, including the potential bias towards a dorsal or ventral domains, (ii) the number of cells labeled in each putative clone, and (iii) the fluorophore of each clone, including those bulbs that contain cells labeled with 2 fluorophores.

Reviewer #2:

[…] Experimental Concerns:

1) Previous work, such as Imamura et al., (2011, 2015), has shown that early-generated mitral cells (E10) cells are distributed more in dorsal domain of the olfactory bulb, whereas late-generated mitral (E12~13) cells are preferentially distributed in ventral domain. Given this, it is important to examine multiple timepoints for tamoxifen injection and ask if there are any timing-dependent changes in clonal distribution of mitral/tufted cells (e.g. test additional embryonic timepoints in addition to E10.5).

As requested by this reviewed we have performed new experiments in which TMX was administered at E12.5. These new data are now shown in a new figure (Figure 3—figure supplement 4). Again, as suggested by the reviewer we have analyzed the distribution of the labeled M/T cells to investigate for a possible bias towards the dorsal or ventral regions of the OB from the clones labeled at E10.5 and E12.5. These new data are shown in two figures: Figure 3—figure supplement 1 and Figure 3—figure supplement 4.

We have observed that at both of these time points there are clones preferentially located in the dorsal regions, in the ventral regions and clones with no preference for either dorsal or ventral regions. Moreover, when all E10.5 clones are analyzed together, we do not observe any clear preference for distribution in the dorsal or ventral regions. For the clones labeled at E12.5 there is tendency towards labeled M/T cells to occupy more ventral regions of the OB (consistent with Imamura et al. data), but the statistical analysis reveals that these differences do not reach significance.

Imamura et al., (2011) show that E10 generated M/T cells have a bias towards the rostral regions of the OB, in an approximate ratio of 1 (dorsal) to 0.75 ventral. Similarly, the M/T cells generated at E12 have an approximate bias of 0.6 (dorsal) to 1 (ventral). In their experiments, they labeled M/T progenitors with thymidine analogs (brdu, IDU and CiDU), and presumably their samples included thousands of cells which allowed them to perform a robust statistical analysis. For our experimental design it was crucial to label a small number of clones per bulb (ideally a single clone per bulb), which resulted in a small number of total cells labeled, even combining all the animals studied. As mentioned above, our analysis of progenitors labeled at E12.5 may suggest that those late-generated clones could have a bias ventral as demonstrated by Imamura et al., (2011). We would like to point out that the main objective of clonal analysis is to identify lineage relationships between cells, or their distribution within a single clone. However, the small sample numbers provided by clonal analysis are not well suited to provide information about the distribution of all the neurons in a cohort born in a given day, a type of data for which the birthday dating method used by Imamura et al., (2011) is best suited.

2) It is not clear if all the mitral/tufted cells are generated from Nestin-positive RGCs. This is important because otherwise the results shown here may miss subpopulations of more restricted or distributed sister mitral/tufted cells. Using Nestin-Cre or giving more tamoxifen can answer this.

Of the 2 strategies suggested by this reviewer, we chose to use the Nestin-cre x Ai9 because increasing the dose of TMX into pregnant females triggers abortions (see Danielian et al., 1998; Ved et al., 2019). As suggested by this reviewer, we crossed the driver Nestin-Cre mouse with the reporter Ai9 mouse. In postnatal animals (P7), all M/T cells in the Nestin-cre::Ai9 mice were labeled in the MOB (Figure 1—figure supplement 3).

3) The small number of clones (typically 1) labeled at a single time-point makes it hard to interpret their results as actually representative throughout the olfactory bulb.

1a) Labeling of single clones per bulb:

Labeling a single clone per bulb is necessary to increase the likelihood that all the sister cells labeled belong to a single clone. In our experiments we used the Confetti mice because this strain can produce clones in 4 different distinguishable markers. With this strain, even if 2 clones are labeled in an olfactory bulb and each clone is labeled with a different color, one can infer that these are 2 independent clones, originating from individual progenitors. However, if several clones are labeled per olfactory bulb (for example, 5 clones in one bulb) some of the clones will be labeled with the same color thus making it impossible to discern how many individual clones are labeled.

1b) Labeling at different time points:

As suggested by this reviewer we have performed additional experiments to label clones at a later time point, E12.5. We have observed that, as expected, the number of cells per clone is smaller at E12.5 compared with E10.5. However, the characteristics of the clones are similar in these 2 time points. As we mentioned in response to point 1 from the reviewer, both at E10.5 and E12.5 we find clones where the sister cells are mostly located ventrally or dorsally, but we also find clones where the sister cells are distributed throughout all the bulb regions without any clear bias. In aggregate, when all clones are analyzed together, we do not observe any clear preference for distribution in the dorsal or ventral regions. These new data are shown in two new figures (Figure 3—figure supplement 1 and Figure 3—figure supplement 4).

1c) Are the clones observed representative?:

We mention in the discussion that it is theoretically possible that there may exist some clones where the sister cells project to the same glomerulus, that those clones may be rare and that we could we have not been able to label those putative clones in our experiments. However, the analysis of the clones that we have studied reveals several commonalities: (i) the M/T cells are not concentrated in defined regions of the bulb, and instead, they are distributed throughout large volumes, (ii) none of the sister M/T cells in one clone project to the same glomerulus, (iii) the size of the clones ranges between 12 and 33 M/T cells, and between 3 and 22 M/T cells, when labeling at E10.5 and E12.5, respectively.

They authors should quantify or demonstrate to what extent their clones uniformly cover the olfactory bulb. In addition, more images of the clones themselves would be useful.

As indicated above we have performed this analysis, and included these data in 2 new figures where we incorporated the 3D reconstruction of the 14 clones detected in our E10.5 experiment (Figure 3—figure supplement 1) and the 9 clones in the E12.5 experiment (Figure 3—figure supplement 4).

4) In addition to asking about the bulb, this work affords the opportunity to look at connectivity and organization in cortex. Given that prior work (e.g. Miyamichi et al., 2011; Sosulski et al., 2011) suggests that M/T cells innervating a single glomerulus project diffusely across olfactory regions, the authors could provide an explanation for these findings by examining the projection patterns of sister M/T cells, which they should be able to do by looking at downstream olfactory bulb targets of the clones labeled in this study.

As indicated in the general comment to all reviewers, this is an interesting question, and we included a brief discussion of this possibility in our original submission. The reason why we only mentioned this as a possibility but we decided not to present any data in this regard is because with the currently available transgenic mice it is not possible to reliably address this issue. Recently, several works have traced the final destination of axons originating from M/T cells locally labeled in the OB (Sosulski et al., 2011, Ghosh et al., 2011, Igarashi et al., 2012). Although there are transgenic mice that can be used to selectively label neocortical progenitors (such as Emx1-CreERT2), currently there are no transgenic mice capable of selective labeling of M/T progenitors. Because of this limitation, to label progenitors of M/T cells we used the Nestin-CreERT2 mice that labels all neuronal progenitors throughout the brain. Therefore, when we induce cre recombination in these mice we label, not only M/T progenitors but also many other progenitors in other brain regions, including the olfactory cortex and other brain regions with neurons whose axons project into the olfactory cortex. When we attempted to analyze the trajectories of axons in the olfactory cortex we observed some local neurons labeled in the olfactory cortex (see reviewer Figure 1F, illustrating a labeled neuron in the olfactory cortex). These neurons have axons that extended locally within the olfactory cortex, and this makes it extremely challenging to discern whether a given labeled axon in the olfactory cortex belongs to a M/T from the OB, or to local neurons in the olfactory cortex. Thus, we did not attempt to trace the destination of axons in the olfactory cortex because we could not unambiguously identify the cells from which they originated. As an example, we attach confocal images from two brains as examples of axons in the olfactory cortex that illustrate this challenge.

5) The authors should examine if sister mitral/tufted cells were distributed across dorsal and ventral domains (as labeled with Nqo1 and OCAM, respectively) or they were specific to either domain. This could provide some information about how the distribution of sister cells is regulated.

We analyzed the distribution of M/T cells in the OB based on the dorsal and ventral domains as revealed by the NQO1 and OCAM markers (Gussing and Bohm, 2004; Yoshihara et al., 1997). These data are now presented in a new figure (Figure 3—figure supplement 1L and Figure 3—figure supplement 4J). We did not observe any preference for any given OB domain when we analyzed all the clones distribution together (Figure 3—figure supplement 1L and Figure 3—figure supplement 4J).

6) The methods for their randomized NND analysis are unclear. The authors stated that "the distances were generated randomly with a normal distribution 230 between the longest and shortest distances observed between M/T cells." However, a normal distribution is specified by a mean and variance, neither of which are given. Furthermore, their randomization methods differ from those described previously (e.g. Gao et al., 2014), which aim to randomly sample within a given volume.

We apologize for this mistake, because in our original submission we randomize the data based on an assumed normal distribution of the cell location. As the reviewer indicates, this is not appropriate, and we have now corrected this mistake and we recalculated the data based on their location within a volume (in 3D). In the update Figure 3E, the random data were generated based on the volume of the external plexiform layer of one of the OB analyzed, and the volume of a neocortex clone with a volume closer to the average neocortical clone volume. To illustrate how we calculated the randomization of the data, we have included a 3D diagram in Author response image 2. Using these new criteria, we observed that the results are identical with those presented in the original submission, which were calculated with different criteria): (i) there are no differences when comparing the characteristics of the different “experimental” clones found in the olfactory bulb, and (ii) there are significant differences when the “experimental” clones are compared with the “simulated” clones generated by randomizing the “experimental” dataset, both in the OB and neocortex

Author response image 2
Volume for random dataset in the olfactory bulb and neocortex.

(A-C) 3D reconstruction from clones observed in OB (A-B) and neocortex (D) on our E10.5 experiment. (D-F) 3D reconstruction used to generate the random data set in the OB (D-E) and the neocortex (F).

7) Is there any spatial bias in the position of labeled progenitor cells by this experiment? It would be better to show the quantification for this to make sure they labeled more or less throughout the entire pool of the progenitors in embryonic brain.

The information that can be obtained from clonal analysis regards the relationship between the sister cells that are derived from a single progenitor. The main objective of our study was to investigate whether “sister” M/T cells receive synaptic input from a single glomerulus, and our data indicates that they do not. However, as in any clonal analysis experiment, we cannot say anything about the position of the specific progenitor that generated those mature cells, because those progenitors disappear when they differentiate.

Based on the current literature, it is generally assumed that the M/T cell progenitors for the olfactory bulb are located in the rostral part of the forebrain, in the so-called presumptive olfactory bulb. Crossing the driver mouse Nestin-Cre with the reporter mouse Ai9, we observed at E10.5 that progenitors tdTomato+ co-expressed the progenitor markers RC2 and Pax6 in the presumptive OB (Figure 1—figure supplement 2). Based on these results, any putative progenitor cells in the OB primordium has the potential to be labeled by the TMX administration time. However, as mentioned above, when we analyze the characteristics of the mature M/T cells in a clone we cannot infer anything regarding the position of the progenitor from which they originated.

8) No discussion is made of AOB M/T cell clones. Do single clones contain sister cells in both the MOB and AOB?

The main objective of our study is to investigate the relationship between lineage and connectivity, and accordingly we decided to focus our analysis on the connectivity of M/T cells in the main olfactory bulb. However, we agree that lineage relationships between the main and accessory olfactory bulb is an interesting question, and we now include this new paragraph in the main text of the manuscript to discuss this issue:

“Our experiments were designed to investigate the relationship between lineage and connectivity in the main olfactory bulb (MOB). […] Further experiments will be required to clarify these questions.”

9) Show all data points for Figure 3A and show some quantification of the distributions for all clones.

As suggested by this reviewer, we now include all data points in this plot (Figure 3A). We did the same for the new E12.5 data (Figure 3—figure supplement 4B)

10) The citation in this sentence “This scenario could be expected for putative glomeruli responsive to relevant odors for survival, such as those responsive to predators or poisons, which require an innate and hardwired response of avoidance (Sosulski et al., 2011).” does not make sense.

In the Discussion section we wrote: “This scenario could be expected for putative glomeruli responsive to relevant odors for survival, such as those responsive to predators or poisons, which require an innate and hardwired response of avoidance (Sosulski et al., 2011).” The reason why we decided to include this reference is because the take-home message from the Sosulski article is that the connectivity between the olfactory bulb and the olfactory cortex varies from animal to animal, presumably to process smells whose interpretation has to be learned. In contrast, the connectivity between the bulb and the amygdala is stereotyped, and this is thought to mediate smells associated with innate behaviors (such as predator or poison avoidance). For example, in the Sokulski et al., (2011) Abstract, they mention “The identification of a distributive pattern of projections to the piriform and stereotyped projections to the amygdala provides an anatomical context for the generation of learned and innate behaviours.” We thought that it was appropriate to give credit to a work that studied these questions, and emphasized how differences in connectivity could explain the differences between learned and innate behaviors.

Conceptual concerns:

1) Lack of discussion with respect to the timing OSN development: If it's known that OSNs innervate glomeruli after mitral/tufted cell differentiate and migrate to their positions in the olfactory bulb (e.g. Blanchart et al., 2006), would one necessarily expect that sister mitral cells innervate the same glomerulus?

We agree that these are important issues and we have now extended the Discussion section to address these questions (see subsection “Synaptic input of sister M/T cells”).

2) If bulb neurons migrate tangentially, wouldn't one expect the clones to be distributed more than those in cortex?

The fact that immature M/T cells migrate tangentially does not necessarily mean that sister M/T cells would be dispersed. For example, one could imagine a scenario in which a group of sister M/T cells migrate long distances tangentially to reach a specific domain of the bulb, but that all those sister would end up close to each other.

Likewise, as the authors state, since it's already known that the olfactory bulb has higher levels of direct neurogenesis rather than indirect neurogenesis via intermediate progenitors (Cárdenas et al., 2018), the smaller clone size of the olfactory bulb already suggests that these regions have different modes of development. Likewise, it remains unclear whether using a single time point for tamoxifen injection labels progenitors at equivalent stages of potency/development across these separate regions.

As requested by this reviewer we performed a new experiment where we administered TMX to a 12 days pregnant female (E12.5). We have found that on average the number of cells per clone at E12.5 (around 9 cells) is smaller than at E10.5 (around 22 cells). These data suggest that in our experiments we are labeling the same M/T progenitors at E10.5 and E12.5, but that at a later time points those progenitors have already produced much of its progeny. (see Figure 3—figure supplement 4).

3) The authors did not show if sister mitral/tufted cells are connected to each other or connect to the same cells other than olfactory sensory neurons.

We would like to point out that our main interest was to investigate the possible relationship between lineage and sensory synaptic input, which we believe is a central question in developmental neuroscience. The relationship between lineage and synaptic connectivity has been mostly studied in just two systems: (a) the Drosophila antennal lobe where it has been shown that lineage determines the synaptic input of the principal neurons (the equivalent of M/T cells), and (b) the pyramidal neurons in the neocortex, where it has been suggested that sister neurons are preferentially connected to each other. To further investigate the possible relationship between lineage and synaptic input, we focused on the connectivity of M/T cells because the anatomical organization of the olfactory bulb, with its well-defined glomeruli is ideally suited to study this question.

The authors state that they "investigated whether lineage plays a role in the connectivity of mitral and tufted cells," however by connectivity here they only refer to connections with OSNs innervating a single glomerulus. Especially given the extent of intrabulbar connectivity, as well as recent reports of biases in connectivity among clonally-derived neocortical neurons, if the authors would like to claim that "lineage does not regulate the connectivity of projection neurons in the mouse olfactory bulb," which their Title reads, it is not sufficient to only show that sister mitral/tufted cells do not extend dendrites to the same glomerulus.

We agree with this comment. With our experiments we can only conclude that sister M/T cells do not receive sensory input from the same glomeruli. Accordingly, we have changed the Title of the manuscript which now reads as follows: “Lineage does not regulate the sensory synaptic input of projection neurons in the mouse olfactory bulb”.

Reviewer #3:

[…] In the paper, the experiments are well-designed and the results are mostly clear. As this paper presents basically negative results, adding some positive data as to a role of cell lineage in forming M/T-cell circuits would strengthen the paper. It will be quite interesting to examine whether the sister M/T cells send their axons to the same area in the olfactory cortex (OC).

As indicated in the general comment to all reviewers, this is an interesting question, and we included a brief discussion of this possibility in our original submission. The reason why we only mentioned this as a possibility, but we decided not to present any data in this regard is because with the currently available transgenic mice it is not possible to reliably address this issue. Recently, several works have traced the final destination of axons originating from M/T cells locally labeled in the OB (Sosulski et al., 2011, Ghosh et al., 2011, Igarashi et al., 2012). Although there are transgenic mice that can be used to selectively label neocortical progenitors (such as Emx1-CreERT2), currently there are no transgenic mice capable of selective labeling of M/T progenitors. Because of this limitation, to label progenitors of M/T cells we used the Nestin-CreERT2 mice that labels all neuronal progenitors throughout the brain. Therefore, when we induce cre recombination in these mice we label, not only M/T progenitors but also many other progenitors in other brain regions, including the olfactory cortex and other brain regions with neurons whose axons project into the olfactory cortex. When we attempted to analyze the trajectories of axons in the olfactory cortex we observed some local neurons labeled in the olfactory cortex (see reviewer Figure 1F, illustrating a labeled neuron in the olfactory cortex). These neurons have axons that extended locally within the olfactory cortex, and this makes it extremely challenging to discern whether a given labeled axon in the olfactory cortex belongs to a M/T from the OB, or to local neurons in the olfactory cortex. Thus, we did not attempt to trace the destination of axons in the olfactory cortex because we could not unambiguously identify the cells from which they originated. As an example, we attach confocal images from two brains as examples of axons in the olfactory cortex that illustrate this challenge.

It will also strengthen the paper if the authors could discuss more about the connectivity of M/T cells regarding possible mechanisms that mediate partner matching with glomeruli.

We agree that this is an important issue, and have added a long paragraph discussing this issue immediately after presenting our data (see subsection “Synaptic input of sister M/T cells”).

Specific comments are as follows:

1) In Figure 4, some examples are shown for sister M/T cells connecting to different glomeruli. Are there any differences in sister-cell distribution between the innate and non-innate OB regions?

This is an interesting question. Some works have suggested that the dorsal and ventral regions of the olfactory bulb preferentially process innate and learned odorants, respectively. We have analyzed the distribution of M/T cells clones born both at E10.5 and E12.5. We have observed that at both of these time points there are clones preferentially located in the dorsal domain, in the ventral domain, and clones with no preference for either dorsal or ventral domains. Moreover, when all clones are analyzed together, we do not observe any preference for M/T cells distribution in the dorsal or ventral domains. These new data are now included in a new figure (Figure 3—figure supplement 1 (E10.5), and Figure 3—figure supplement 4 (E12.5))

2) Is it possible to determine when and where the sister cells are derived during embryonic development?

The information that can be obtained from clonal analysis regards the relationship between the sister cells that are derived from a single progenitor. The main objective of our study was to investigate whether sister M/T cells receive synaptic input from a single glomerulus, and our data indicates that they do not. However, as in any clonal analysis experiment, we cannot say anything about the position of the specific progenitor that generated those mature cells, because those progenitors disappear as such when they differentiate.

3) Even if sister cells do not connect to the same glomerulus, are there any shared characteristics and common features in their gene expression (particularly for axon guidance molecules), OC projection (particularly to the amygdala), and firing patterns?

We would like to point out that our main interest was to investigate the possible relationship between lineage and synaptic input, which we believe is a central question in developmental neuroscience. This question has been mostly studied in just two systems: (a) the Drosophila antennal lobe where it has been shown that lineage determines the synaptic input of the principal neurons (the equivalent of M/T cells), and (b) the pyramidal neurons in the neocortex, where it has been suggested that sister neurons are preferentially connected to each other. To further investigate the possible relationship between lineage and synaptic input, we focused on the connectivity of M/T cells because the anatomical organization of the olfactory bulb, with its well-defined glomeruli is ideally suited to study this question.

We agree that investigating whether sister M/T cells share other properties (gene expression, axonal projections or firing patterns) are interesting questions, and we mention some of these scenarios in our Discussion section. However, we believe that those questions, although interesting, are not the main question of this work.

4) It is worth examining M/T cells in the accessory OB where all M/T-cell circuits are hard-wired to mediate innate pheromone responses?

This is a very interesting question, especially because the AOB is thought to process smells involved in innate behaviors, which could be genetically programmed. In our initial submission we decided to exclusively focus on the relationship between lineage and connectivity for M/T cells in the MOB, because identifying the connectivity of dendrites in the MOB is unambiguous, but the anatomical characteristics of the AOB make it a less reliable system where to study this question. In the MOB, M/T cells have a single dendrite that projects into a single, well-defined glomerulus. The AOB has a critical caveat to study questions related to synaptic connectivity because its glomeruli are not well defined, and thus, are difficult to identify. In addition, we also observed that whereas the number of cells per clone in the MOB was (on average) 20 cells, for the AOB we clone size was much smaller (between 1-3 per AOB). Taking into account these limitations we now mention these data in the revised manuscript as follows: “Although the small number of labeled AOB M/T cells does not allow us to draw any firm conclusions, we did not find any M/T cells whose apical dendrites innervated the same glomerulus (Figure 4—figure supplement 2), similar to what we observed in the MOB”.

5) In the last part of Discussion section, the authors list interesting future questions. This paper would be significantly strengthened if any results could be added regarding these questions, particularly for connectivity to the amygdala.

We agree, and we attempted to perform this experiment. However, as we explained above the currently available transgenic do not allow us to reliably trace the projection of the axons of sister M/T into the olfactory cortex and amygdala.

https://doi.org/10.7554/eLife.46675.022

Article and author information

Author details

  1. Luis Sánchez-Guardado

    Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5598-8608
  2. Carlos Lois

    Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    clois@caltech.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7305-2317

Funding

NIH Office of the Director (RO1MH116508)

  • Luis sanchez-guardado

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We are grateful to Walter G Gonzalez and Antuca Callejas for comments on the manuscript.

Ethics

Animal experimentation: In all experiments, mice were handled according to the mice protocol (#1709) approved by the Institutional Animal Care and Use Committee (IACUC) of the California Institute of California.

Senior Editor

  1. Catherine Dulac, Harvard University, United States

Reviewing Editor

  1. Stephen Liberles, Harvard Medical School, United States

Reviewer

  1. Stephen Liberles, Harvard Medical School, United States

Publication history

  1. Received: March 8, 2019
  2. Accepted: August 26, 2019
  3. Accepted Manuscript published: August 27, 2019 (version 1)
  4. Version of Record published: September 13, 2019 (version 2)

Copyright

© 2019, Sánchez-Guardado and Lois

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 798
    Page views
  • 192
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Cell Biology
    2. Developmental Biology
    Ying Zhu et al.
    Research Article Updated
    1. Developmental Biology
    2. Neuroscience
    Casey Paquola et al.
    Research Article