Illuminating the terminal nerve: Uncovering the link between GnRH-1 neuron and olfactory development

Abstract

During embryonic development, the olfactory placode (OP) generates migratory neurons, including olfactory pioneer neurons, cells of the terminal nerve (TN), Gonadotropin-releasing hormone-1 (GnRH-1) neurons, and other uncharacterized neurons. Pioneer neurons from the olfactory placode induce olfactory bulb morphogenesis.

In mice, GnRH-1 neurons appear in the olfactory system around mid-gestation and migrate via the terminal nerve axons to different brain regions. The GnRH-1 neurons are crucial in controlling the hypothalamic-pituitary-gonadal (HPG) axis. Kallmann syndrome is characterized by impaired olfactory system development, defective olfactory bulbs, secretion of GnRH-1, and infertility. The precise mechanistic link between the olfactory system and GnRH-1 development remains unclear. Studies in humans and mice highlight the importance of the Prokineticin-2/Prokineticin-Receptor-2 (Prokr2) signaling pathway in olfactory bulb morphogenesis and GnRH-1 neuronal migration. Prokr2 loss-of-function mutations can cause Kallmann syndrome, and hence the Prokr2 signaling pathway represents a unique model to decipher the olfactory/GnRH-1 connection. We discovered that Prokr2 is expressed in the TN neurons during the critical period of GnRH-1 neuron formation, migration, and induction of olfactory bulb morphogenesis. Single-cell RNA sequencing identified that the TN is formed by neurons distinct from the olfactory neurons. The TN neurons express multiple genes associated with Kallmann Syndrome (KS). Our study suggests that the aberrant development of pioneer/TN neurons might cause the KS spectrum.

Graphical Abstract

The Prokineticin-Receptor-2 (Prokr2) gene is associated with impaired GnRH-1 neuronal migration and olfactory development in Kallmann Syndrome. Through genetic tracing, we found that neurons express Prokr2 in the putative terminal nerve (TN) of rodents. Sorting and sequencing TN cells revealed their distinction from olfactory, vomeronasal and GnRH-1 neurons. Furthermore, the TN expresses multiple genes linked to Kallmann syndrome.

Introduction

Puberty is an essential postnatal developmental process that triggers multiple hormonal changes leading to sexual and reproductive maturity (Dunkel and Quinton, 2014). Signals from the hypothalamus within the forebrain cause downstream initiation and regulation of sexual maturation and fertility (Bourguignon et al., 1992; Uenoyama et al., 2019). The hypothalamic gonadotropin-releasing hormone-1 (GnRH-1) neurons (Gore et al., 1996; Marshall et al., 1993; Skynner et al., 1999; Wray, 2002) act as the main source of the GnRH-1 decapeptide that is critical in initiating the pubertal developmental cascade, maintaining sex hormone production and fertility (Dutlow et al., 1992; Schally et al., 1971).

From birds to mammals (Lovejoy et al., 1992; Murakami et al., 2022; White et al., 1995; Wray et al., 1989), the GnRH-1 neurons are first detectable during embryonic development in various areas of the developing olfactory pit, including respiratory, olfactory, and vomeronasal epithelium (Oka and Matsushima, 1993). In rodents, the GnRH-1 neurons appear to form in a region of the olfactory pit that gives rise to the vomeronasal organ (VNO) as development progresses (Schwanzel-Fukuda and Pfaff, 1989; Wray et al., 1989). However, in higher species, including humans, the region of the olfactory pit where the GnRH-1 neurons form (Casoni et al., 2016) does not generate a functional VNO (Bruintjes and Bleys, 2023a; Garcia-Velasco and Mondragon, 1991; Trotier, 2011). Nonetheless, in humans, like in mice, the GnRH-1 neurons form and migrate from the nasal area to the brain. Once in the brain, most of the GnRH-1 neurons migrate to the hypothalamus. However, several GnRH-1 cells migrate around the olfactory bulb and cortex (Casoni et al., 2016; Cho et al., 2019; Lettieri et al., 2016).

Within the brain, most of the GnRH-1 neurons reside scattered in the preoptic-hypothalamic area (POA) (Casoni et al., 2016). From here, they send their axonal projections to the median eminence and secrete GnRH-1 in the hypophyseal-portal circulation, thus governing the hypothalamic-pituitary-gonadal axis (HPG axis) (Duittoz et al., 2022; Krsmanovic et al., 1993; Pitteloud et al., 2010; Tobet et al., 2001).

The pulsatile release of GnRH-1 prompts pituitary gonadotroph cells to release luteinizing hormone (LH) and follicle-stimulating hormone (FSH) into the bloodstream, activating the gonadal hormone production. The onset of puberty relies on the activation of the HPG axis. However, in humans, crucial GnRH-1-dependent milestones occur much earlier than puberty, with transient activation of the HPG-axis during fetal life and early infancy in a process called “mini puberty” (Abreu and Kaiser, 2016; Forest et al., 1973; Forest et al., 1974; Kuiri-Hanninen et al., 2014; Winter et al., 1976). Notably, the hormonal changes secondary to the activation of the HPG axis are crucial for fertility and brain development (Auyeung et al., 2010; Chachlaki et al., 2022; Hines et al., 2016; Manfredi-Lozano et al., 2022; Renault et al., 2020; Schaadt et al., 2015).

Disruptions in the migration of the GnRH-1 neurons, a defective release of GnRH-1, or absent/aberrant GnRH-1 signaling can lead to various forms of congenital hypogonadotropic hypogonadism (HH) (Christian et al., 1971; Radovick et al., 1991). HH negatively affects sexual maturation, sexual function, social behavior, and fertility (Ravikumar and Crowley, 2010). HH has two main clinical categorizations, Kallmann Syndrome (KS) and normosmic idiopathic HH (niHH) (Rugarli, 1999). In KS, there are different degrees of olfactory deficiencies spanning from a partial reduction in the sense of smell (hyposmia) to a total loss in the ability to perceive odors (anosmia) (Meschede et al., 1994).

Multiple lines of evidence indicate that GnRH-1 neurons migrate from the nose to the brain along specific sets of neurons forming in the nasal area (Norgren et al., 1995; Schwanzel-Fukuda and Pfaff, 1989; Schwanzel-Fukuda and Silverman, 1980; Toba et al., 2008; Wray et al., 1989; Wray et al., 1994; Yoshida et al., 1995). However, the identity of these cells serving as a migratory scaffold has remained unclear (Duittoz et al., 2022; Forni and Wray, 2015; Taroc et al., 2017). Defects in the formation of olfactory bulbs (OBs), such as OB aplasia (absence of OB) or hypoplasia (reduced OB), are common phenotypes in KS patients (Martinez-Mayer and Perez-Millan, 2023). However, within families affected by KS, olfactory defects and HH do not necessarily co-segregate, suggesting that a direct link between anosmia and HH may not always exist (Balasubramanian et al., 2014). In keeping with this notion, we previously showed that in Arx-1 null mutant mice, which cannot form olfactory bulbs (Taroc et al., 2017), the GnRH-1 neurons can still migrate into the brain, while the olfactory sensory neurons (OSNs) and vomeronasal sensory neurons (VSNs) fail to connect to the brain. We revealed that the TN axons, differently from the olfactory neurons, can invade the brain with or without the olfactory bulb. This evidence suggests that the development of the OBs and the connectivity of the olfactory and vomeronasal neurons to the brain are not required for GnRH-1 neuronal migration. On this basis, we and others (Cho et al., 2019) suggested that, in mammals, the GnRH-1 neurons migrate along the axons of pioneer/terminal nerve (TN) neurons (Demski and Schwanzel-Fukuda, 1987).

The morphogenic process leading to the formation of OBs is believed to be triggered by the interaction of early migratory olfactory pioneer neurons with the developing brain (Gong and Shipley, 1995; Wu et al., 2018). However, what the olfactory/pioneer neurons are, and their molecular characterization still needs to be addressed.

The TN was first described in sharks (Demski, 1987) and then reported to exist in heterogeneous forms in many vertebrate species (Demski and Northcutt, 1983; Sonne et al., 2020). Interestingly, the TN has been described in cetaceans, animals that lack olfactory structures such as an olfactory bulb and olfactory epithelium (OE) (Ridgway et al., 1987), and in humans, which lack a functional VNO (Stoyanov et al., 2018). Recent data in chick suggest that the terminal nerve is a transient embryonic neuronal structure (Palaniappan et al., 2019).

The G-protein coupled receptor Prokineticin receptor 2 (Prokr2) is predominantly expressed in the mammalian central nervous system, where it has been described to regulate neurogenesis and neuronal survival (Martinez-Mayer and Perez-Millan, 2023; Ngan and Tam, 2008). Prokr2 and its ligand Prokineticin 2 (Prok2) have been implicated in neuronal migration in the developing brain and the olfactory bulb morphogenesis (Ng et al., 2005). Loss-of-function mutations in Prokr2 and its ligand, Prok2, have been linked to KS and nIHH in humans (Avbelj Stefanija et al., 2012; Balasubramanian et al., 2011; Martinez-Mayer and Perez-Millan, 2023). In line with this, mice homozygous for Prokr2 or Prok2 null mutations have severe atrophy of the reproductive system and hypoplastic olfactory bulbs. These phenotypes recapitulate the human phenotype of KS (Matsumoto et al., 2006). Thus, the Prokineticin system represents a compelling candidate that may help bridge the knowledge gaps pertaining to the relevance of the terminal nerve, OB formation, and GnRH migration.

The effects of defective Prok2-Prokr2 signaling on GnRH-1 neuronal migration are likely non-cell-autonomous, as most of the GnRH-1 neurons do not express Prokr2 (Mohsen et al., 2017). Following Prokr2 expression and lineage tracing in developing rodent embryos, we identified Prokr2 expression in the early olfactory pioneer neurons, of which some appear to form the TN of mice. Through genetic lineage tracing, cell sorting, and single-cell transcriptomics, we provide evidence pointing to the fact that the terminal nerve/olfactory-pioneer neurons are distinct neuronal populations from olfactory/vomeronasal sensory neurons and GnRH-1 neurons. The critical role of pioneer neurons in triggering OB formation and the enriched expression of KS-associated genes in these cells suggests that the pathological pioneer/TN neurons’ development might be a crucial mechanistic link between anosmia and defective GnRH migration in KS.

Results

During embryonic development, Prokr2 is expressed by populations of cells in and emerging from the developing olfactory area.

We performed in-situ hybridization to follow Prokr2 expression during development (E10.5-E15.5) (Fig. 1) when the pioneer and GnRH-1 neurons form and migrate out of the developing olfactory pit. At all ages Prokr2 was found to be expressed by pioneer migratory neurons spanning between the OP to the developing forebrain (FB). Between E12.5 and E15.5, clusters of Prokr2+ cell bodies could be detected outside of the developing vomeronasal area with a pattern of expression similar to that we previously described for the putative terminal nerve marker Robo3 (Duittoz et al., 2022; Taroc et al., 2019). Notably, at these developmental stages, Prokr2 mRNA expression did not appear to be detectable in the forebrain.

Figure 1. Prokr2 mRNA detected in cells of the migratory mass.

A-A’) In situ hybridization at E10.5, coronal section, highlights Prokr2 mRNA expression within cells (Arrowheads) near the olfactory placode (OP) and proximal to the forebrain (FB). B-B’) Prokr2 expressing cells at E11.5 (parasagittal) were found in the olfactory epithelium (OE) and proximal to the FB. C-C’) At E12.5 (parasagittal) Prokr2 is expressed in cells in and migrating out from the developing vomeronasal (VNO). D-D’) At E15.5, Prokr2 expression was detected in cells within the ventral portion of the developing vomeronasal organ and in clusters of cells outside of the VNO. Scale bars in A-A’, B-B’, C, D’, 100 μm; C’, 50 μm; D, 250 μm.

Prokr2^iCre lineage tracing highlights the pioneer migratory neurons.

Mice with null mutations for Prok2 or Prokr2 exhibit defective GnRH-1 migration; however, Prokr2 has been reported not to be expressed by the GnRH-1 neurons (Matsumoto et al., 2006). Based on this, we performed Prokr2Cre lineage tracing experiments to better understand which neuronal populations express Prokr2 between E11.5 and E14.5, stages at which the GnRH-1 neurons migrate to the brain (Gong and Shipley, 1995). In line with what we observed via in-situ hybridization, Prokr2^iCre/R26^tdTomato lineage tracing at E11.5 showed recombination mostly limited to pioneer/TN neurons proximal to the developing ventral telencephalon (Fig. 2A). As suggested by the in-situ hybridization (Fig.1), at these stages no recombination was observed in the forebrain.

Figure 2. Prokr2 tracing highlights TN/pioneer neurons.

A-A”’) Immunofluorescent staining of HuCD (green) reveals a migratory mass (MM) (arrows) of Prokr2 traced neurons (red) at E11.5. These neurons could be the putative terminal nerve or pioneer neurons of the olfactory bulb. A’-A”’ show a magnification of the boxed area. B-B”) At E12.5, GnRH-1 neurons (green) negative for Prokr2 tracing (arrows) can be seen migrating along terminal nerve (TN) axons traced for Prokr2 into the brain (Br). B’ is a magnification of the area within the white box, and B” is a magnification of the boxed area within the grey box. B”) Prokr2 traced neurons (white arrows) and GnRH-1 neurons (grey arrows) can be seen within and leaving the vomeronasal organ (VNO). C) Arrows indicate Prokr2 traced cells in Grueneberg ganglion (GG), respiratory epithelium (RE), migratory mass (MM), vomeronasal organ (VNO), olfactory epithelium (OE), forebrain junction (FBJ) at E13.5. Traced TN projections access the brain ventral to the olfactory bulb (OB), reaching the basal forebrain (BFB). D-E) At E1.5 the GnRH-1 neurons (green, grey arrows) can be seen migrating along the axons of the traced neurons (magenta, white arrows) projecting from the nose to the brain. Scale bars in A’-A”’ 50 μm; A, B-B’, D, E 100 μm; C, 500 μm.

Between E12.5 and E13.5, the Prokr2 tracing highlighted cell bodies and axons of the putative terminal nerve/pioneer neurons (Gong and Shipley, 1995; Taroc et al., 2017) extended from the developing VNO and were found to be invading the ventral telencephalon (Fig. 2B).

Immunostaining against GnRH-1 neurons revealed that the GnRH-1 cells were associated with the ProKr2+ pioneer/terminal nerve (Fig. 2B,D,E). Quantifications at E13.5 revealed that the GnRH-1 neurons were, for the most part (82%) negative for the recombination but migrated along the Prokr2 positive projections. At this stage, traced axons were detected projecting to the basal forebrain (Fig. 2C). Notably, Prokr2 lineage tracing also highlighted chimeric recombination in cells within the RE (putative solitary chemosensory cells), sparse cells in the OE, and within the putative Grueneberg ganglion (GG).

Immunostaining against the microtubule-associated protein (Map2) at E13.5 revealed strong reactivity in the cells of the MM. As described for other cell types, we found expression in the cell soma and the axonal initial segments (Gumy et al., 2017; Teng et al., 2001). Notably, almost no Map2 immunoreactivity was found in the newly formed neurons of the OE and VNO.

Double immunostaining against Map2 and tdTomato on traced Prokr2cre animals highlighted that 85% of the Map2 positive cells were also positive for the tracing (Fig. 3A,B).

Figure 3. Prokr2 traced neurons express Map2 and are positive for cCasp3.

A-B”) Double immunostaining of Map2 (green) with Prokr2 tracing (red) at E13.5 shows that both Prokr2 traced neurons, positive for Map2 expression (white arrows), within the VNO (traced with dotted lines) and migrating out of the VNO. B-B” show magnifcations of the boxed area in A. Map2 immunolabelling was found in neurons of the migratory mass (MM) and in neurons of the olfactory bulb (OB) and forebrain (FB) Arrowheads indicate co-expression of Map2 and Prokr2 traced neurons. C-C”’) Immunostaining for Cleaved Caspase 3 (green) on Prokr2 traced (magenta) animals at E13.5 reveals that many apoptotic cells are in the forebrain junction (FBJ), including Prokr2 traced- cells (grey arrows) and Prokr2 traced+ cells (white arrows). C’-C”’ are magnifications of the boxed area in C. Scale bars in A-A’, B-B”, C, 100 μm; C’-C”’, 50 μm.

Interestingly, Map2 RNA expression and immunoreactivity of the TN/pioneer neurons appears to be more comparable to that of the forebrain neurons rather than to the one of the OSN/VSNs.

Prokr2 Traced Cells Undergo Apoptosis Proximal to the Forebrain at E13.5

A recent report described massive apoptosis for the cells of the nasal migratory mass in chicks (Palaniappan et al., 2019). By analyzing traced animals between E13.5-E15.5, we observed detectable cleaved Caspase 3 immunoreactivity at E13.5, mostly in cells proximal to the forebrain at the putative forebrain junction (FBJ). Double immunostaining revealed that around 12% of the traced pioneer/TN cells at E13.5 were apoptotic (Fig. 3C).

These data suggest that in rodents, TN/pioneer neurons undergo cell loss proximal to the brain as described in Chick.

Single Cell RNA Sequencing from Whole Noses

To further explore the identity of the neurons of the migratory mass/terminal nerve in the developing nasal area, we performed single-cell RNA sequencing (Sc-RNAseq) from whole embryonic mouse noses of C57 WT animals at E14.5. From these, we acquired ~35,000 cells. Data from single cells were then grouped into clusters based on Uniform Manifold Approximation and Projections (UMAPs) (Hao et al., 2021) (Fig. 4A). Unsupervised clustering allowed us to identify nasal mesenchyme (Alx1 positive), respiratory epithelia (Krt19, Isl1 positive), and Tubb3/Ascl1/Elavl3/Foxg1/Pax6 positive clusters belonging to neurogenic regions of the olfactory pit (Fig. 4B–I). Once we identified the different cellular groups, we re-clustered the data of cells belonging to the OP (~ 1,700 cells) without including the nasal mesenchyme (Fig. 5A). Based on gene expression enrichment and embryonic ISH expression data from the digital atlas Genepaint (Visel et al., 2004), we identified respiratory epithelium (Krt14, Krt19, Tfap2a positive), stem cells (Ki-67, Hes1, Pax6, Sox2 positive), neurogenic cells (Ki-67, Ascl1 Neurog1, Neuro1), and differentiating neurons (Tubb3, Elavl3 positive) (Fig. 5C–W).

Figure 4. Single cell RNA-sequencing of embryonic noses.

A) UMAP showing cell clusters isolated from E14.5 embryonic noses. A large cluster of cells were identified as mesenchyme, while a smaller, separate cluster was identified to belong to the olfactory pit, both neurogenic and non-neurogenic portions. B-I) Feature plots highlighting the expression of various marker genes used to group clusters into different populations of cells within the nose.

Figure 5. TN/pioneer neurons have a distinct transcriptome from other neurons in the nose.

A) UMAP showing clusters present within the olfactory pit of embryonic E14.5 noses. B) Volcano plot showing the enriched genes of the putative migratory pioneer/TN neuron cluster (right) compared to the enriched genes of the cluster identified as differentiated neurons (left).C-N) Feature plots for marker genes used to identify the different regions of the olfactory area. O-W) In situ hybridization data at E14.5 from GenePaint showing the expression of select genes at E14.5. X-AC) Feature plots showing the enriched genes within cells of the putative pioneer/TN neurons’ cluster. AD-AE) ISH for Ndn and Pcsk1n shows their mRNA expression in migratory cells (arrows) outside the VNO.

Following the expression enrichment for previously proposed TN markers such as Robo3, Peripherin, Cntn2/Tag-1 (Fig. 5X–Z) (Casoni et al., 2012; Schwarting et al., 2004; Taroc et al., 2017; Wray et al., 1994), we also identified a small cluster of putative TN/pioneer neurons. This cluster of cells was found to be highly enriched for two genes, Proprotein Convertase Subtilisin/Kexin Type 1 Inhibitor (Pcsk1n) expression and Necdin (Ndn) (Fig. 5AB, AC). In situ hybridization from the digital atlas Genepaint (Visel et al., 2004), showed a pattern of expression in migratory cells of the putative TN (Fig. 5AD, AE). By performing differential gene expression profiling between the putative cells of the TN and the rest of the neurons of the olfactory pit, we identified several enriched genes in our cluster of interest (Fig. 5B). This included Peripherin (Prph), Cntn2/Tag1, Pcsk1n and Ndn.

Single Cell RNA Sequencing pioneer/TN neurons from Prokr2iCre traced embryos.

To get a better understanding of the precise genetic/molecular identity of the Prokr2^iCre/R26^tdTomato traced cells, we dissociated the noses of E14.5 Prokr2^iCre/ R26^tdTomato embryos and performed fluorescence-activated cell sorting (FACS) to isolate the traced neurons. We selected this stage because at E14.5, the GnRH-1 neurons actively migrate to the brain, there are many Prokr2 traced neurons in the nasal areas, and differentiated olfactory and vomeronasal sensory neurons can be found in their respective epithelia. Moreover, at this stage, it is easy to separate the nasal area from the bran without cellular contaminations.

We analyzed 1,050 sorted cells. By integrating the Sc-RNASeq data of sorted and non-sorted nasal cells we observed that part of the sorted cells was grouped in a cluster that formed a continuum with the previously identified unsorted putative PN/TN cluster (Fig. 6A, B).

Figure 6. FACS sorted Prokr2^iCre/ R26^tdTomato embryos have a heterogenous population of TN/pioneer neurons.

A) UMAP of olfactory pit clusters isolated from wild-type unsorted E14.5 noses. B) UMAP with unsorted (grey) and sorted (magenta) cells merged. The box indicates cells of the terminal nerve (TN). C-C’) Feature plot of Bcl11b in the sorted cells, shown to be enriched in the sorted premigratory pioneer/TN neurons. C’ is a magnification of the box in C. D-D”’) Bcl11b immunostaining (green) on E13.5 Prokr^iCre traced animals reveals that traced migratory neurons (magenta) are negative for Bcl11b expression (white arrowheads) and resident pre-migratory within the VNO (traced with dotted lines) and olfactory epithelium (OE) are positive for Bcl11b expression (black arrowheads). D’-D”’ are magnifications of the are within the grey box and D””-D””” are magnifications of the area within the white box. Red blood cells are marked with an asterisk. E) Volcano plot comparing enriched genes of the sorted migratory cells with all unsorted neurons. The UMAP represents the cell clusters compared. F) Volcano plot comparing enriched genes of the two largest sorted clusters: the putative migratory pioneer/TN neuron cluster (right) and the putative pre-migratory pioneer/TN neuron cluster (left). The UMAP represents the cell clusters compared. Scale bars in D’-D””” 50 μm; D 100 μm.

In addition to this, a second larger group of Prokr2^iCre/ R26^tdTomato sorted cells formed a separate cluster that was found to be enriched for olfactory system markers such as OMP, Calb2 and Mef2b and Bcl11b differently from the sorted migratory pioneer/TN cluster. (Jin et al., 2019; Margolis, 1972; Sammeta et al., 2007) (Fig. 6C, D, E, F). By performing immunolabelling against Bcl11b, we noticed that the near totality of the Prokr2^iCre/ R26^tdTomato traced migratory neurons were negative for Bcl11b expression (85%), while Bcl11b expression appeared to be restricted to neurons within the developing olfactory and vomeronasal epithelia (Fig. 6D–D”””). Based on this, we classified the Prokr2^iCre/ R26^tdTomato traced neurons positive for Bcl11b (Bcl11b+) as resident neurons, and the Bcl11b negative (Bcl11b-) migratory neurons.

Comparing the transcriptome of the migratory Prokr2+/TN cells to all the other neurons of the olfactory pit at E14.5 (Fig. 6E), we found the enrichment of 26 genes previously linked to niHH and KS (Table 1). Notably, this list includes Sema3e and Nrp1, which are integral to the migration of the GnRH-1 neurons during embryonic development (Stamou and Georgopoulos, 2018).

Table 1.

Genes enriched in the terminal nerve associated with Kallmann Syndrome or normosmic idiopathic hypogonadotropic hypogonadism.

Gene Symbol	Description	p_value	avg_log2FC	Reference
Prokr2	Prokineticin Receptor 2	8.05E-72	3.88512539	(Matsumoto et al., 2006) (Wen et al., 2019)
Ndn	Necdin, MAGE Family Member	5.17E-20	3.76935476	(Beneduzzi et al., 2011a)
Gnrh1	Gonadotropin Releasing Hormone 1	5.90E-38	2.319258908	(Cattanach et al., 1977) (Taroc et al., 2020b)
Snrpn	Small Nuclear Ribonucleoprotein Polypeptide N	2.66E-16	2.129445932	(Grumbach, 2005)
Otx2	Orthodenticle Homeobox 2	2.65E-05	1.471612839	(McCabe et al., 2011)
Sema3e	Semaphorin 3E	1.11E-67	1.458050388	(Cariboni et al., 2015)
Rmst	Rhabdomyosarcoma 2 Associated Transcript	5.76E-30	1.369694514	(Stamou et al., 2020)
Brd2	Bromodomain Containing 2	2.51E-05	0.627323002	(Afonso Lopes et al., 1995) (Firouzi et al., 2019)
Rrm2b	Ribonucleotide Reductase Regulatory TP53 Inducible Subunit M2B	3.30E-07	0.576550261	(Pitceathly et al., 2012)
Sos1	SOS Ras/Rac Guanine Nucleotide Exchange Factor 1	2.25E-05	0.572023445	(Hadziselimovic et al., 2010)
Lmna	Lamin A/C	8.36E-05	0.539290642	(de Andrade et al., 2020)
Rras	RAS Related	2.07E-08	0.525486022	(van der Kaay et al., 2016)
Cpe	Carboxypeptidase E	0.00177753	0.517399217	(Alsters et al., 2015)
Tubb3	Tubulin Beta 3 Class III	0.000213151	0.48773623	(Chew et al., 2013) (Balasubramanian et al., 2015)
Nrp1	Neuropilin 1	2.27E-05	0.481371612	(Cariboni et al., 2011)
Glce	Glucuronic Acid Epimerase	3.66E-06	0.45641838	(Miraoui et al., 2013)
Stub1	STIP1 Homology And U-Box Containing Protein 1	0.006983989	0.426330669	(Davis et al., 2020)
Rab3gap2	RAB3 GTPase Activating Non-Catalytic Protein Subunit 2	0.000255373	0.418184055	(Sun and Wang, 2021)
Mras	Muscle RAS Oncogene Homolog	0.006374165	0.39834121	(Mikec et al., 2022)
Aip	Aryl Hydrocarbon Receptor Interacting Protein	0.003416918	0.337276059	(Gaspar et al., 2023)
Las1l	LAS1 Like Ribosome Biogenesis Factor	0.000286302	0.321928095	(Gocz et al., 2022)
Rit1	Ras Like Without CAAX 1	0.00038789	0.308631272	(Deng et al., 2020)
Prkar2a	Protein Kinase CAMP-Dependent Type II Regulatory Subunit Alpha	0.020431747	0.276596856	(Matsumoto et al., 2006) (Negri and Ferrara, 2018)
Ptpn11	Protein Tyrosine Phosphatase Non-Receptor Type 11	0.018164779	0.273760812	(Patti et al., 2023)
Fgf9	Fibroblast Growth Factor 9	0.007879595	0.268214645	(Krejci et al., 2009)
Magel2	MAGE Family Member L2	6.18E-17	0.257700962	(Mercer and Wevrick, 2009)

Known Kallmann Syndrome or normosmic idiopathic hypogonadotropic hypogonadism candidate genes enriched in the terminal nerve including their gene symbol, description, p-value, average log fold change, and reference to relevant literature.

Gene expression enrichment in the Prokr2 traced pioneer/terminal nerve neurons.

We and others previously described that mature and maturing GnRH-1 neurons express the transcription factor Islet1 as they migrate from the nose to the brain (Causeret et al., 2023; Palaniappan et al., 2019; Shan et al., 2020; Taroc et al., 2020a). By comparing the transcriptome of the sorted migratory Prokr2+ cells of the terminal nerve to the unsorted putative TN, we observed substantial differences in gene expression, suggesting the existence of two main populations of migratory neurons, the Isl1+ and Prokr2+ neurons.

The transcription factor Isl1 was significantly enriched in the Prokr2- migratory neuron cluster when compared to the migratory Prokr2+ cells. In contrast, the transcription factor Lhx2 was found to be enriched in the cells of the sorted Prokr2iCre+ migratory neurons when compared to the Isl1+ Prokr2- migratory cells (Fig. 7B).

Figure 7. Prokr2 traced neurons express Lhx2 but not Isl1.

A-A”’) Prokr2cre tracing at E13.5 with Isl1 immunofluorescent staining reveals two populations of migratory cells leaving the vomeronasal organ (VNO) (traced with dotted lines): cells positive for Isl1 (green, white arrowheads) and cells negative for Isl1 that are positive for Prokr2 tracing (magenta, clear arrowheads). A’-A”’ are magnifications of the boxed area in A. B) Volcano plot comparing the enriched genes of the Prokr2+ cells (right) and the Isl1+ cells (left). C-C”’) Lhx2 immunostaining (red) on E13.5 Prokr2cre traced embryos reveals that many traced neurons (magenta) of the olfactory epithelium (OE), VNO (traced with dotted lines), and migratory mass (MM) express Lhx2 (white arrows). C’-C”’ are magnifications of the boxed area in C. Scale bars in A-A”’, C-C”’, 100 μm.

Notably, both Isl1 and Lhx2 have been recently reported to be expressed by GnRH and migratory neurons of the developing terminal nerve ganglion of chick (Palaniappan et al., 2019). However, Isl1 immunostaining on Prokr2^iCre traced mice visualized that the Isl1+ immunoreactivity and the Prokr2+ lineage tracing highlight two distinct migratory neuronal populations in rodents (Fig. 7A). In fact, 89% of the Prokr2 traced neurons were negative for Isl1 expression. This is in line with what we observed after GnRH and ProKR2 quantifications. Immunostaining against Lhx2 showed that many (63%) traced neurons were positive for Lhx2 expression (Fig. 7C).

The Prokr2 cells in the vomeronasal area.

In line with previous observations (Pitteloud et al., 2007) Prokr2 lineage tracing labeled neurons within the developing VNO at E13.5 and E15.5. However, at these stages, immunostaining against the VSN markers AP-2ε and Meis2 (Lin et al., 2018; Taroc et al., 2020b) showed that 90% of Prokr2 traced neurons do not express the VSNs’ markers (Fig. 8A–C). These data suggest that Prokr2 tracing highlights cell populations distinct from the VSNs and GnRH-1 neurons (Fig 8A,B). However, analysis at E18.5 found sparse Prokr2-traced cells within the VNO, and many of the Prokr2-traced neurons seen at previous stages were no longer present in the VNO (Fig. 8C).

Figure 8. Prokr2 traced neurons are distinct from VSNs during embryonic development.

A-C) Developmental time-course of the Prokr2 expression within the VNO from E13.5-E18.5. Immunostaining for VSN markers: AP2-ε (blue) and Meis2 (green) against Prokr2 lineage tracing (magenta) reveals that Prokr2 traced neurons are prevalent within the VNO. Arrows indicate Prokr2 traced neurons that are negative AP2-ε and Meis2 expression. The amount of neurons that are Prokr2 traced and negative for AP2-ε and Meis2 expression within the VNO appears to decrease over time. The boxed region indicates magnifications in the proceeding panels. D-D”’) Prokr2iCre tracing at P10. Coronal sections of the VNO immunoassayed for AP2-ε and Meis2 reveal that the Prokr2 traced neurons present in postnatal animals are positive for the two VSN markers. Arrows indicate Prokr2 traced neurons that are positive for AP2-ε and Meis2 expression. D’-D”” are magnifications of the area within the box of D. Scale bars in A-C, 100 μm; C’-C””, 50 μm.

Next, we looked at Prokr2 tracing at postnatal stage P10. Coronal sections of the VNO reveal that Prokr2 traced neurons accounted for ~1% of the VSNs (Fig. 8D) positive for either AP-2ε or Meis2 expression. These data suggest that the majority of Prokr2 positive cells in the vomeronasal anlage are pioneer/migratory cell types that are distinct from the nonmigratory olfactory and vomeronasal sensory neurons.

Additionally, traced cells were found in the main olfactory epithelium (MOE) at P10. When paired with double immunofluorescence for odorant receptor neuron’s dorsal marker I (Nqo1) and lateral/ventral marker (Ncam2) (Bozza et al., 2009; Enomoto et al., 2019), we found that ~95% of the traced cells were positive for either marker. These data suggest the existence of heterogeneous cell populations of Prokr2+ cells within the MOE.

Validation of Terminal Nerve markers

To further validate the identity of the Prokr2 traced cells of the putative TN, we performed immunolabeling of Calbindin-1 (Calb1) (Fig. 9A–C), a gene that we found to be enriched in the sorted Prokr2+ cells. Calb1 is a calcium-binding protein previously reported in migratory neurons following a similar pathway to that of the GnRH-1 neurons (Honma et al., 2004; Murakami et al., 2022). In mice, at E13.5, the GnRH-1 neurons were detected juxtaposed to Calb1 positive neurons. However, only a few (~16%) of GnRH-1 neurons were found to be Calb1+ (Fig. 9A–A”’). In line with this, performing Calb1/Isl1 double immunostaining, we observed that 86% of the Calb1+ cells are negative for Isl1 (Fig. 9B–B”’). At E13.5, Calb1-expressing cells (Fig.9C) were found in the VNO with a pattern similar to that of the Prokr2 expressing/traced cells (Fig.1,3). In line with this Calb1 immunostaining on Prokr2iCre traced embryos (Fig. 9C’–C”’) revealed Calb1+ expression in most (70%) of the traced cells.

Figure 9. Calb1 is expressed by Prokr2 traced neurons but not Isl1 or GnRH-1 neurons.

A-A”’) C57 BL WT E13.5 immunofluorescent staining for Isl1 (green) and Calb1 (magenta). Both Isl1 positive (clear arrowhead) and Calb1 positive (white arrowhead) cells can be seen leaving the VNO (traced with dotted lines), as two separate populations of migratory cells. A’-A”’ are magnifications of the boxed region in A. B-B”’) Immunofluorescent staining for GnRH-1 (green) (clear arrowhead) shows two distinct populations of migratory neurons Isl1 positive (Including the GnRH-1 neurons) and Calb1 positive (white arrowhead). C) Immunohistochemistry of Calb1 reveals many neurons within the VNO positive for Calb1 expression at E13.5(black arrows). C’-C”’) Prokr2iCre tracing reveals co-expression of Calb1 with Prokr2 traced neurons (white arrows) in the VNO (traced with dotted lines), autofluorescent red blood cells (asterisks). Scale bars in A-A”’, B, 100 μm; B’-B”’, C-C”’, 50 μm.

These data further support the existence of various populations of migratory neurons and suggest that Calb1 can be used as marker for the terminal nerve neurons.

Discussion

The olfactory placode undergoes invagination and gives rise to various populations of migratory cells (Causeret et al., 2023; Fornaro et al., 2003; Palaniappan et al., 2019; Taroc et al., 2020a; Valverde et al., 1993). These migratory neurons, known as pioneer neurons or neurons of the migratory mass (Fornaro et al., 2003; Maier and Gunhaga, 2009), delaminate from the placodal tissue and spread between the placode and the brain (Miller et al., 2010). In rodents, including mice, the GnRH-1 neurons, which are among the first neuronal cells to form during embryonic development, are found in the nasal migratory mass (Fornaro et al., 2003; Wray et al., 1989). The olfactory placode develops into the main olfactory system, consisting of the main olfactory epithelium and the main olfactory bulb in the brain. Some animal species also have an accessory olfactory system, comprising the vomeronasal organ and the accessory olfactory bulb. Still, humans lack a functional accessory olfactory system, although vestigial vomeronasal structures in developing human fetuses have been reported (Bruintjes and Bleys, 2023b; Casoni et al., 2016; Garciavelasco and Mondragon, 1991). Despite this, no accessory olfactory bulb has been identified in humans and there is no evidence for neuronal connections between the accessory olfactory system in humans (D’Aniello et al., 2017).

The prevailing model suggests that the migration of GnRH-1 neurons depends on proper olfactory development, considering the developmental defects observed in the olfactory bulbs, tracts, and sulci of KS patients (Knorr et al., 1993; Yu et al., 2022). However, evidence from mice indicates that GnRH-1 neuronal migration can occur despite a defective olfactory bulb development (Taroc et al., 2017). Interestingly, GnRH-1 neuronal migration and fertility have also been reported in human patients with arrhinia, a condition characterized by the complete absence of olfactory structures (Delaney et al., 2020).

To investigate the cellular composition of the GnRH-1 migratory track and its relation to olfactory defects in KS, researchers have explored the role of the terminal nerve or nervus terminalis, a cranial nerve that guides GnRH-1 neurons to the brain (Jennes, 1987; Jin et al., 2019; Pena-Melian et al., 2019; Schwanzel-Fukuda et al., 1989; Schwanzel-Fukuda and Pfaff, 1989; Schwanzelfukuda and Silverman, 1980; Taroc et al., 2017; Wirsigwiechmann, 1990). However, the presence of a link between the terminal nerve, GnRH-1 neuronal migration, and olfactory defects in KS still needs to be fully understood.

It is noteworthy that olfactory bulb absence or hypoplasia is common in KS patients, including those with Prokr2 and Prok2 gene mutations (Matsumoto et al., 2006; Pitteloud et al., 2007). Previous studies have shown that early pioneer neurons from the olfactory placode induce the olfactory bulb morphogenesis (Gong and Shipley, 1995; Tufo et al., 2022).

Our lineage tracing results show that Prokr2 is expressed by pioneer/terminal nerve neurons, and the migratory neurons contacting the brain before olfactory bulb formation. Moreover, in line with previous studies(Mohsen et al., 2017) most GnRH-1 neurons resulted in negative for Prokr2 but associated to axons of Prokr2-positive neurons (Fig. 2B, D, E).

The absence of detectable Prokr2 recombination in the GnRH-1 neurons and in the the cells forming the olfactory bulb between E11.5 and E15 (Fig. 2) strongly suggests that GnRH and olfactory bulb defects in Prokr2 null mice (Matsumoto et al., 2006; Pitteloud et al., 2007) are secondary to impaired development of pioneer/terminal nerve neurons.

Furthermore, the GnRH neurons invade the ventral forebrain before olfactory bulb formation along Prokr2-traced neuronal fibers (Fig. 2B).

Together, these data suggest that the TN migration (Fig.1,2), GnRH-1 invasion of the forebrain, and the induction of olfactory bulb (Gong and Shipley, 1995; Tufo et al., 2022) are developmentally related events. Based on this we propose that the Prokr2+ pioneer/terminal nerve neurons could have the dual function of triggering olfactory bulb formation and guiding GnRH-1 neuronal migration to the basal forebrain.

In situ hybridization experiments and lineage tracing revealed that, in the developing olfactory area, Prokr2 is expressed by migratory neurons and by limited subsets of cells in the developing vomeronasal and olfactory anlage.

Analysis of Prokr2 lineage tracing in the developing vomeronasal organ showed that most of the Prokr2-traced neurons were negative for VSN markers AP-2ε and Meis2 (Fig. 8). However, analysis at postnatal stages showed the existence of heterogeneous VSNs and OSNs positive for Prokr2 tracing.

However, our findings (Fig. 8A–D) suggest that most Prokr2+ cells originating in the olfactory placode are neurons that migrate out of the sensory epithelia as embryonic development progresses.

Single Cell RNA Sequencing Reveals the Transcriptome of Prokr2 Cells

To better understand the cellular identity of Prokr2-expressing cells in the developing nasal area, we performed FACS sorting and single-cell sequencing of Prokr2iCre/R26tdTomato traced cells from embryonic noses. The analysis revealed that Prokr2-traced neurons were mainly distributed in two clusters (Fig. 6D), one mostly positive for Bcl11b expression and another negative for Bcl11b. Immunostaining against Bcl11b confirmed that migratory Prokr2+ pioneer/TN neurons were negative for Bcl11b, while the Prokr2-traced neurons positive for Bcl11b were within olfactory and vomeronasal epithelia. However, the postnatal analysis highlighted a minimal contribution of Prokr2+ cells to the olfactory and vomeronasal sensory lineage (Fig. 8D).

Single-cell transcriptomic analysis and immunolabelling also highlighted that the nasal migratory mass of rodents comprises two main types of migratory neurons, Prokr2+, and Isl1+ neurons. Previous studies have shown that the Isl1 migratory neurons are, for the most part, immature or maturing GnRH-1 neurons (Shan et al., 2020; Taroc et al., 2020a).

The pioneer neurons, terminal nerve cells, or more vastly the so called migratory mass (MM) has been explored for many years (Demski and Schwanzel-Fukuda, 1987). Differently from what we observed in mice, previous studies in chick have indicated that the TN of chick comprises three cell types (Isl1+; GnRH1+ and Lhx2+) (Palaniappan et al., 2019). Since in mice Isl1+ cells are for the most part mature or maturing GnRH, it is possible that the discrepancies between mice and chick might reflect differences in the timing of Isl1 and GnRH expression as these migratory cells mature. In line with what described in chick we observed that the majority of the Prokr2+ cells also express Lhx2 (Fig. 7C). What the developmental and functional relationships between the Isl1 and Prokr2+ cells are needs to be further explored.

Beyond basic developmental biology, the translational relevance

Transcriptomic analysis of the Prokr2+ Pioneer/TN cells revealed an enriched expression of the putative Grueneberg ganglia marker Pde2a (Zhi Huang, 2018), Pcsk1n, Tmem130, and Resp18. In addition to these, in the Migratory Prokr2+ cells, we identified the expression of several previously reported antigens for the putative terminal nerve/pioneer neurons, such as Calbindin1 (Fujiwara et al., 1997) (Fig. 9), Lhx2 (Palaniappan et al., 2019) (Fig. 7), Peripherin (Tobet et al., 2003), Cntn2/Tag1, Robo3. Notably, in the migratory Prokr2+ Pioneer/TN neurons, we also identified the expression of many of the candidate genes associated with KS and normosmic idiopathic hypogonadotropic hypogonadism (niHH), including Sema3e (Cariboni et al., 2007; Messina and Giacobini, 2013; Oleari et al., 2019), Prokr2, Glce (Liu and Zhi, 2022), and Necdin (ndn) (Beneduzzi et al., 2011b) (Table 1). Understanding if mutations in KS-related genes have a cell-autonomous effect on TN development and, therefore on GnRH-1 migration must be evaluated in the future.

Studies in chick (Palaniappan et al., 2019) have highlighted that the TN/migratory mass undergoes massive cell death as development proceeds. By following apoptosis at E13.5 we observed similar cellular behavior in rodents (Fig. 3C). These data suggest that the TN might function as a transient developmental structure.

A technical limitation of our study is that we investigated the TN lineage by using a constitutive Prokr2Cre line that, after E14.5, does not allow us to discriminate between neurons of the TN and other brain neurons as Prokr2Cre recombination becomes widespread in various brain regions. This limitation prevented us from further analyzing the TN connectivity or cell migration into the brain. To address this, new inducible Cre mouse models are needed to restrict Prokr2 recombination to the developmental window during TN formation.

The study provided evidence for the close relationship between the terminal nerve and GnRH-1 neurons and presented the enriched transcriptome of putative TN clusters. Lineage tracing techniques allowed the visualization of the developing TN as distinct from other neuronal populations within the nose. To fully understand TN development and its relation to KS and niHH, the identified TN markers should be further utilized in future studies.

By identifying the Prokr2+ cells as a distinct cell type from the GnRH, olfactory, and vomeronasal neurons, we set the foundation for a new chapter in the investigation of the cellular and molecular etiology of pathologies affecting olfactory and GnRH development.

Future investigations utilizing the newly discovered TN markers, identification of mutations in the TN-enriched genes, and refined mouse models will allow us to better understand the role of this nerve in normal and pathologic olfactory and sexual development.

Materials and Methods

Animals

The Prokr2^iCre mice were previously described (Mohsen et al., 2017) and ordered from the same source. R26tdTomato were maintained on a C57 BL/6J background, the mouse line (stock #007914) can be found at JAX. Prokr2^iCre mice were genotyped using PCR analysis from the following primers: Prokr2creF: TCCCCACGGTAGTTGTGAAG; Prokr2creR: ATTGGTTGGTGTGGTTTGCAG; Prokr2cremt: CAGCAGGTTGGAGACTTTCCT. Animals were euthanized using CO₂, followed by cervical dislocation. Experiments used mice of either sex. All animal experiments were completed in accordance with the guidelines of the Animal Care and Use Committee (IUACUC) at the University at Albany, SUNY.

Tissue Preparation

Embryos were collected from time-mated females where the observation of the vaginal plug was determined to be E0.5. Embryos collected were fixed in 3.7% Formaldehyde/PBS at 4°C for 2 h. Samples were then submerged in 30% sucrose overnight, then embedded, and frozen in O.C.T. (Tissue-TeK) and stored at −80°C. Embedded embryos were cryosectioned using CM3050S Leica cryostat and collected on Superfrost plus slides (VWR) at 14–20 μm thickness for immunofluorescent staining.

ISH

Digoxigenin-labeled cRNA probes were prepared via in vitro transcription (DIG RNA labeling kit; Roche Diagnostics, Basel, Switzerland) from the Prokr2 template available from Genepaint (https://gp3.mpg.de/viewer/setInfo/MH1790/12). ISH was performed as described (Forni et al., 2013) and visualized with an alkaline phosphatase conjugated anti-DIG (1:1000), and NBT/BCIP developer solution (Roche Diagnostics).

Immunohistochemistry

Primary antibodies and dilutions used in this study: mouse (biotin)-α-HuC/D (1:1000. Invitrogen), goat-α-RFP (1:1000, Rockland), SW rabbit-α-GnRH-1 (1:6000, Susan Wray, NIH), rabbit-α-Map2 (1:200, Cell Signaling Technology), rabbit-α-Cleaved Caspase 3 (1:1000, Milipore), mouse-α-RFP (1:100, Origene), rat-α-Bcl11b (1:500, Biolegend), rabbit-α-Isl1 (1:500, Abcam), rabbit-α-Lhx2 (Bioss Antibodies, 1:250), goat-α-AP-2ε (1:500, R&D Systems), rabbit-α-Meis2 (1:1000, Abcam), mouse-α-Calb1 (1:200, Santa Cruz Biotechnology), goat-α-Nqo1 (1:700, Abcam), rabbit-α-NCAM2 (1:500, Proteintech Group). Antigen retrieval was used for experiments utilizing rabbit-α-Map2, rabbit-α-Cleaved Caspase 3, rabbit-α-Isl1, mouse-α-RFP, rat-α-Bcl11b, goat-α-AP-2ε, rabbit-α-Meis2 and mouse-α-Calb1 submerging slides in a citric acid solution heated to 95°C for 15 minutes and then cooling for 15 minutes before continuing with standard immunofluorescence procedure. For immunofluorescent staining, secondary antibodies selected for the correct species were conjugated with Alexa Fluor-488, Alexa Flour-594, or Alexa Flour-680 (Invitrogen and Jackson Laboratories). 4′,6′-diamidino-2-phenylindole (1:3000; Sigma-Aldrich) was used to counterstain sections and coverslips were mounted using Fluorogel l (Electron Microscopy Sciences, Hatfield, PA, USA). Confocal Microscopy image acquisition was taken on a LSM 710 microscope (Zeiss, Oberkochen, Germany). Epifluorescence images were acquired on a DM4000 B LED fluorescence microscope equipped with a Leica DFC310 FX camera. Image analysis and quantifications were done using FIJI/ImageJ software. Each staining presented was replicated on three different animals.

Antibody	Source	Dilution	Reference
Mouse (biotin) anti-HuC/D	Molecular Probes a-21272	1:200	(Katreddi et al., 2022)
Goat anti-RFP	Rockland 200-101-379	1:1000	(Chen et al., 2017)
SW Rabbit anti-GnRH-1	Susan WRAY, NINDS/NIH	1:6000	(Wray et al., 1988)
Rabbit anti-Map2	Cell Signaling Technology #4542	1:200	(Li et al., 2023).
Rabbit anti-Cleaved Caspase 3	Millipore AB3623	1:1000	(Zhang et al., 2021)
Mouse anti-RFP	Origene TA180084	1:100	(Katreddi et al., 2022)
Rat anti-Bcl11b	Biolegend 650604	1:500	(Visel et al., 2004)
Rabbit anti-Isl1	abcam ab20670	1:500	(Qin et al., 2020)
Rabbit anti-Lhx2	Bioss Antibodies bs-11200R	1:250	(Palaniappan et al., 2019)
Goat anti-AP-2ε	R&D Systems AF5060	1:500	(Lin et al., 2022)
Rabbit anti-Meis2	abcam ab73164	1:1000	(Lin et al., 2022)
Mouse anti-Calb1	Santa Cruz SC365360	1:200	(Chaudhari et al., 2021)
Goat anti-Nqo1	Abcam ab2346	1:700	(Ihara et al., 2016)
Rabbit anti-NCAM2	proteinTech Group 13850-1-AP	1:500	(Bozza et al., 2009)

Antibody Characterization

HuC/HuD (HuC/D)

The monoclonal HuC/D antibody was isolated from a patient with paraneoplastic encephalomyelitis with mouse as the host species. A previous study in our lab utilized this antibody for IHC (Katreddi et al., 2022), and our staining is consistent, specifically labelling neurons within the olfactory area.

Goat anti-red fluorescent protein (RFP)

This polyclonal antibody was derived from the mushroom polyp coral Discosoma with mouse as the host species. Western blotting labels a specific band at 27 kDA (manufacturer information). Previous work with this antibody has shown specific staining of Cre drive, tdTomato positive cells in the olfactory epithelium (Chen et al., 2017). Negative control experiments using this antibody in our lab confirm no labeling present in embryonic wild-type mouse tissue.

SW Gonadotropin releasing hormone-1 (GnRH-1)

This antibody was created by coupling GnRH to four different carrier proteins and using an immunization schedule on the host, rabbit, which was boosted using alternating antigen-carrier complexes. Preabsorbtion of this antibody with synthetic GnRH resulted in no immunopositive cells both in vivo and in vitro (Wray et al., 1988). Studies in our lab confirm the correct staining of the GnRH-1 neurons in the developing olfactory system as detailed in previous studies (Wray, 2010).

Microtubule-associated protein 2 (Map2)

This polyclonal antibody was produced by immunizing the host species, rabbit, with a synthetic peptide corresponding to carboxy terminal residues of human Map2. Map2 isoforms were confirmed to have specific bands via western blot at 280,270,70 and 75 kDa (manufacturer information). A previous study used this antibody to label Map2 in neural progenitor cells (Li et al., 2023). We confirmed Map2 expression to correlate with embryonic ISH expression data of Map2 from the online database Genepaint (Visel et al., 2004).

Cleaved Caspase 3

This polyclonal antibody was produced by cleavage of a specific synthetic peptide sequence with rabbit as the host species. A western blot of a non-apoptotic HeLa cell extract and an apoptotic HeLa cell extract found a specific band present at 17 kDa only for the apoptotic cell extract (manufacturer information). A recent study designed a protocol for detecting apoptotic neurons using this antibody with negative control wild type mice compared to aged mice (Zhang et al., 2021).

Mouse anti-red fluorescent protein (RFP)

This monoclonal antibody was produced using an RFP-monomer tagged recombinant protein expressed in E. coli with mouse as the host species. Western blotting cells transfected for the RFP-monomer were found to have a specific band at 35 kDa (manufacturer information). This antibody was used in a previous study done by our (Katreddi et al., 2022) and wild-type tissue stained for this antibody was not found to have any activity.

Bcl11 transcription factor B (Bcl11b)

This clone of the Bcl11b antibody has since been discontinued. This monoclonal antibody was generated from human Bcl11b with rat as the host species. From ISH expression data Bcl11b is seen to be highly enriched in the olfactory epithelium, matching the expression pattern present in our data.

Islet-1 (Isl1)

This polyclonal antibody was produced from a synthetic peptide corresponding to the human Islet1 amino acid sequence conjugated to keyhole limpet haemocyanin with rabbit as the host species. Similar to previous work (Qin et al., 2020) and ISH expression data (Visel et al., 2004), our staining showed the same expression pattern of Isl1 in the brain and migratory neurons.

LIM homeobox 2 (Lhx2)

This polyclonal antibody was generated via conjugation of keyhole limpet haemocyanin to synthetic peptide derived from human Lhx2 with rabbit as the host species. The expression seen in our labeled tissue match with ISH expression patterns (Visel et al., 2004) and previous work in in the developing chick olfactory system (Palaniappan et al., 2019), Lhx2 is found to prevalent in the migratory mass of the nasal area.

Transcription factor AP-2 epsilon (AP-2ε)

This polyclonal antibody was generated from recombinant human AP-2 epsilon derived from E. coli with goat as the host species. Western blot of a human cell line and human placenta tissue targeted for AP-2ε both saw a specific band at 38 kDa (manufacturer information). This antibody was used in a previous paper in our lab (Lin et al., 2022) with a very strong expression pattern in the vomeronasal organ that matches with our data and ISH data.

Meis Homeobox 2 (Meis2)

This polyclonal antibody was generated from a synthetic peptide corresponding to human Meis2 conjugated to keyhole limpet haemocyanin with rabbit as the host species. The expression of this antibody was confirmed form similar results to a previous studies done by our lab (Lin et al., 2022) (Katreddi et al., 2022).

Calbindin 1 (Calb1)

This monoclonal antibody was generated from the the amino acid sequence near the N-terminus of Calb1 of human origin with mouse as the host species. Western blot of mouse brain, human kidney and rat brain tissue extracts against Calb1 showed a specific band at 28 kDa (manufacturer information). This antibody was used in a previous study to label the Purkinje cells in the cerebellar coretex (Chaudhari et al., 2021). Calb1 was found to label migratory neurons within the olfactory area matching with available ISH expression data.

NAD(P)H quinone dehydrogenase 1 (Nqo1)

This polyclonal antibody was generated from a synthetic peptide corresponding to the C terminal amino acid sequence of human Nqo1 with goat as the host species. Western blot of cell lysate against Nqo1 showed a single band at 30 kDa (manufacturer information). This antibody was used in previous work to label olfactory sensory neurons (Ihara et al., 2016), which matches the expression pattern in our data of the main olfactory epithelium.

Neural cell adhesion molecule 2 (Ncam2)

This polyclonal antibody was generated from an Ncam2 fusion protein with rabbit as the host species. Western blotting human brain tissue against Ncam2 showed a single specific band at 125 kDa (manufacturer information). Previous work has detailed Ncam2 expression within the olfactory epithelium (Bozza et al., 2009) which is what is also seen in our data.

Single-Cell RNA Sequencing

Noses were excised from E14.5 wild-type mouse embryos and 1 ProKR2iCre/R26tdTomato embryo at the same stage. Dissections were done under a dissecting microscope keeping embryos in cold, sterile PBS. Dissected noses were then further dissociated into single cells using dissociation solution (Neural Isolation Enzyme with papain resuspened in Neural Basal Media, with 0.5 mg/mL Collagenase A, 1.5 mM L-cysteine and 100 U/mL DNAse I). Single cells were then preserved in cell freezing media (10%FBS, 90%DMSO) at −80 °C. Tissue samples from the embryos were used for sex determination via PCR and (Jarid1 F: CTGAAGCTTTTGGCTTTGAG; Jarid1 REV: CCACTGCCAAATTCTTTGG) and for genotyping.

After identifying embryos of the same sex (only males were used) and of the desired genotype, single cell stocks were quickly defrosted in a 37°C water bath, spun down, washed, and resuspended using warm 10% FBS in the neurobasal medium. Cells were quantified using a T20 Automated Cell Counter (BioRad) and diluted to attain a cell concentration of ~700–1200 cells/mL. Only samples with a viability equal or above 90% were used for the next steps.

The single cell suspension of the ProKR2iCre/R26tdTomato embryonic nose was sent to SingulOmics (NY, USA) to perform FACS, generating sorted and unsorted datasets. While the wild-type animals were processed in house. Single-cell sequencing was performed following the 10x Chromium Next GEM Single Cell 3’ Reagent Kit v3.1 protocol using the 10x Genomics Chromium Controller within the RNA institute. Libraries generated using the 10x protocol were then submitted to be sequenced at the Center of Functional Genomics (UAlbany) using the Illumina NextSeq2000 system. Before sequencing, quality control was performed on libraries by analyzing fragment size, only libraries with a fragment size between 200 and 500 bp were used in sequencing. For all datasets, sequencing depth was above 20,000,000 reads.

Data analysis was done using FASTQ files first run through the 10X CELL RANGER pipeline to align each cell’s genome to a cell-specific barcode into specific output files that can be fed into the R Package ‘Seurat’ (Satija et al., 2015). The output files were filtered through CELL RANGER, which removes barcodes that are interpreted as background noise. ‘Seurat’ was then used to do QC analysis to remove cells expressing more than 9000 genes limiting the number of dying cells analyzed; this filter was set to eliminate cells with mitochondrial read rate > 5% and expressed < 200 genes. Cell cycle regression was used to limit cell clustering based on cell cycle stage. The function SCTransform was used in the normalization process to remove variability within the datasets.

Uniform manifold approximation and projection (UMAP) plots were used to visualize clustering. Clusters were identified based on known markers for the different neuronal types such as the stem cells, neuronal progenitors, differentiated neurons and TN. Each cluster of interest went through the QC filters stated previously before generating lists of enriched genes and further sub setting.

Key Points.

1. Pioneer or terminal nerve neurons play a crucial role in initiating the development of the olfactory bulbs. We found that the Prokineticin Receptor-2 gene, associated with Kallmann syndrome, is expressed by the olfactory pioneer/terminal nerve neurons.

2. We genetically traced, isolated, and conducted Single-cell RNA sequencing on terminal nerve neurons of rodents. This analysis revealed a significant enrichment of gene expression related to Kallmann syndrome.

3. Our study indicates that the investigation of Pioneer/terminal nerve neurons should be a pivotal focal point for comprehending developmental defects affecting olfactory and GnRH-1 systems.

Data Availability

The Single-cell RNA sequencing datasets used are available through the Gene Expression Omnibus (GEO) database, under GEO accession number GSE234871 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234871).