- Email: [email protected]
Developmental evolution and developmental plasticity of the olfactory epithelium and olfactory skills in Mexican cavefish
Developmental Biology xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology
Original research article
Developmental evolution and developmental plasticity of the olfactory epithelium and olfactory skills in Mexican cavefish Maryline Blin, Eugène Tine, Lydvina Meister, Yannick Elipot, Jonathan Bibliowicz, ⁎ Luis Espinasa1, Sylvie Rétaux Paris-Saclay Institute of Neuroscience, Université Paris-Sud, CNRS UMR9197, Université Paris-Saclay, Avenue de la terrasse, 91198 Gif-sur-Yvette, France
A R T I C L E I N F O
A BS T RAC T
Keywords: Olfaction Organ size Proliferation Neurogenesis Olfactory sensory neuron
The fish Astyanax mexicanus comes in two forms: the normal surface-dwelling (SF) and the blind depigmented cave-adapted (CF) morphs. Among many phenotypic differences, cavefish show enhanced olfactory sensitivity to detect amino-acid odors and they possess large olfactory sensory organs. Here, we questioned the relationship between the size of the olfactory organ and olfactory capacities. Comparing olfactory detection abilities of CF, SF and F1 hybrids with various olfactory epithelium (OE) sizes in behavioral tests, we concluded that OE size is not the only factor involved. Other possibilities were envisaged. First, olfactory behavior was tested in SF raised in the dark or after embryonic lens ablation, which leads to eye degeneration and mimics the CF condition. Both absence of visual function and absence of visual organs improved the SF olfactory detection capacities, without affecting the size of their OE. This suggested that developmental plasticity occurs between the visual and the olfactory modalities, and can be recruited in SF after visual deprivation. Second, the development of the olfactory epithelium was compared in SF and CF in their first month of life. Proliferation, cell death, neuronal lifespan, and olfactory progenitor cell cycling properties were identical in the two morphs. By contrast, the proportions of the three main olfactory sensory neurons subtypes (ciliated, microvillous and crypt) in their OE differed. OMP-positive ciliated neurons were more represented in SF, TRPC2-positive microvillous neurons were proportionately more abundant in CF, and S100-positive crypt cells were found in equal densities in the two morphs. Thus, general proliferative properties of olfactory progenitors are identical but neurogenic properties differ and lead to variations in the neuronal composition of the OE in SF and CF. Together, these experiments suggest that there are at least two components in the evolution of cavefish olfactory skills: (1) one part of eye-dependent developmental phenotypic plasticity, which does not depend on the size of the olfactory organ, and (2) one part of developmental evolution of the OE, which may stem from embryonic specification of olfactory neurons progenitor pools.
1. Introduction The olfactory epithelium (OE) of vertebrates is the external sensory organ devoted to the sense of smell. During development, the OE derives from a thickening called the placode that is formed at the end of gastrulation (Grocott et al., 2012; Whitlock, 2004). The non-neural ectoderm (at the border of the neural plate) or neural ectoderm (inside the neural plate) origin of the placode is currently debated. During neurulation and after, the placode cells undergo
⁎
1
migrations and morphogenesis, transforming a thickened sheet of ectoderm into a multilayered pit, in which proliferation and neurogenesis occur (Breau and Schneider-Maunoury, 2014; Maier et al., 2014; Torres-Paz and Whitlock, 2014). Differentiated olfactory sensory neurons of the OE then project their axons onto the glomeruli of the olfactory bulbs, in the telencephalon, progressively establishing the path of the olfactory nerve (Koide et al., 2009; Shao et al., 2017; Whitlock and Westerfield, 1998). Each olfactory sensory neuron expresses a single GPCR olfactory receptor among the
Corresponding author. E-mail address: [email protected] (S. Rétaux). School of Science, Marist College, New York 12601, USA.
https://doi.org/10.1016/j.ydbio.2018.04.019 Received 31 January 2018; Received in revised form 19 April 2018; Accepted 24 April 2018 0012-1606/ © 2018 Elsevier Inc. All rights reserved.
Please cite this article as: Blin, M., Developmental Biology (2018), https://doi.org/10.1016/j.ydbio.2018.04.019
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
facility. Fish are maintained at 23–26 °C on 12:12 h light: dark cycle and they are fed twice a day with dry and live food. The breeding colonies spawn two or three times per week and generate a highly variable quantity of embryos (a few dozens to a few thousands). Spawning induction and larval care are described elsewhere (Elipot et al., 2014b). Embryos and larvae were raised at 24 °C in embryo medium (EM). SF raised in the dark were kept in thermostaticallycontrolled light-proof tanks. Feeding and daily EM changes were done in the dark. The F1 hybrids used in this paper were the progeny of a female SF x male Pachón cross. Animals were treated according to the French and European regulations for use of animals in research. SR's authorization for use of animals in research including Astyanax mexicanus is 91–116. Paris Center-Sud Ethic Committee authorization numbers are 2012-0055, 2016-36 and 2017-04.
species’ repertoire of odorant receptors and is specialized to detect a single odorant (Korsching, 2009; Shao et al., 2017; reviewed in Miyasaka et al., 2013). The olfactory system is used to detect, discriminate and identify odorant molecules in relation with a variety of adaptive behaviors such as foraging, communication, reproduction or predator avoidance. Although different species show markedly different odor detection capacities, or olfactory specialization, not much is known about the factors determining olfactory sensitivity. Among mammals for example, differences of 5–6 orders of magnitude in the capacity to detect certain odorants have been reported, but no direct correlation between olfactory sensitivity and numbers or densities of olfactory sensory neurons, or the size of olfactory structures, has been reported (reviewed in Wackermannova et al., 2016). On the other hand, some birds such as the nocturnal New Zealand kiwi for which olfaction is of strong behavioral relevance possess enlarged olfactory structures (Corfield et al., 2014), and some fish such as sharks with exceptional olfactory skills present an extra-large OE surface and large olfactory bulbs (Collin, 2012; Tricas et al., 2009). The developmental origin(s) of these sensory specializations, in terms of both olfactory system anatomy and function, are mostly unknown. Here, we addressed the question of the developmental origin of olfactory sensitivity using the comparison between the two morphs of the fish Astyanax mexicanus. The blind and depigmented cavedwelling morphs (cavefish, CF) and their river-dwelling conspecifics (surface fish, SF) have markedly different olfactory capacities and are unique models to study this question. In fish, olfactory responses to relevant odorant cues (amino-acids, nucleotides, pheromones, alarm substance) can be recorded through electroolfactograms, functional imaging or behavioral analyses (Behrens et al., 2014; Caprio et al., 1989; Friedrich and Korsching, 1997; Hara, 1994, 2006; Keller-Costa et al., 2015; Miyasaka et al., 2013; Tricas et al., 2009; Vitebsky et al., 2005; Wakisaka et al., 2017; Whitlock, 2006; Yoshihara, 2008). In an olfactory assay performed in the lab, CF larvae originating from the Pachón cave are able to detect and show attractive response to concentrations as low as 10−10 M of amino-acid, whereas SF larvae can only detect 10−5 M ranges (Hinaux et al., 2016). In the wild, blind CF inhabiting the Subterráneo cave respond to food odors whereas eyed hybrid fish phenotypically resembling SF do not (Bibliowicz et al., 2013). Interestingly, CF from these two caves have larger OEs and nostrils than SF. In Pachón embryos and larvae, the larger sensory organ results from early developmental evolution during gastrulation and neurulation, due to CF-specific modulations of midline signaling from organizer centers (Hinaux et al., 2016; Pottin et al., 2011; Yamamoto et al., 2004). Here, in search for the developmental origins of the enhanced olfactory skills of cavefish, we investigated the relationship between the size of the olfactory organ and olfactory capacities in Astyanax mexicanus. We also analyzed other developmental processes, including developmental plasticity due to loss of vision, and changes in neurogenesis control influencing the neuronal composition of the OE. The data are presented in a “results and discussion” format.
2.2. Behavioral testing Behavioral tests were performed as previously described (Hinaux et al., 2016) in a specially constructed sound- and lightproof room that includes a main compartment for testing and a second computer work station compartment from which recording of the tests was performed with minimal disturbances. All fish were fed 24 h prior to the test with two day-old Artemia and then food was withheld until testing in order to standardize their feeding state. Four one-month-old juveniles were placed in behavioral testing boxes (see Fig. 1C; 9 cm wide x 13 cm long) containing 150 mL embryo medium (EM) and let acclimatize for two hours prior to the test at a temperature of 24 °C and in the dark. SF and CF (or experimental and control animals) were always tested in parallel. Boxes were placed on top of an infrared light box (ViewPoint S.A.). Each test was initiated by simultaneously opening the Luer stoppers of medical solution administration tubing (Baxter, U.K.) to perfuse solutions at 5 mL/min from two reservoirs containing 60 mL of either amino-acid containing EM or EM alone (control). On the EM-perfused side, the flow generated was identical to the flow on the amino acid perfused side. Tests were recorded for 7 min on a Dell work station using ViewPoint imaging software and a DragonFly2 camera equipped with an infrared filter (PointGray). Utilizing a colorimetric test for the quantification of amino acid concentrations (Hinaux et al., 2016), we defined four quadrants of the boxes in which the amino acid concentration was very high, high, low or zero over the duration of the test. Each of these quadrant was attributed a coefficient to calculate a Preference Index Score (PIS) reflecting the attraction of fish (or absence of attraction) to the amino acid source. The PIS for each time point (at 30 s intervals) was the cumulative score of the four fish in the box, where the position of each fish was scored with the values of − 3 (quadrant furthest away from amino acid source), − 1, 1, or 3 (quadrant closest to amino acid source). Thus the maximum and minimum PIS scores are + 12 and − 12, respectively, for a given time point. In order to correct for the initial position of the fish when the amino acid first enters the box at 1.5 min after the start of the experiment, the PIS was reset at zero for this time point and the subsequent PIS values were corrected by subtracting of the initial raw score at 1.5 min. Experiments in which the initial PIS score was > 6 or < −6 were discarded because the correction for initial position of the fish lead to artefactual “false attraction” or “false repulsion” at subsequent time points (Hinaux et al., 2016). Statistical significance of replicate tests was calculated using the non-parametric Mann-Whitney test and was performed using the StatView software. In all figures, n = 1 corresponds to one test, i.e., the cumulative score of 4 fish.
2. Materials and methods 2.1. Fish Laboratory stocks of A. mexicanus SF and CF (Pachón population) were obtained in 2004 from the Jeffery laboratory at the University of Maryland, College Park, MD, and were since then bred in our local
2
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
primers (Omp-Fw: TGCTACAGGGTCTAGGTCGTC; Omp-Rv: GGCCT CGTTCTTCTTGAACATC; Trpc2-Fw: TGCTGGGGGAGTCTATCACA; Trpc2-Rv: CTCGATATAGCGCTGCACCA; OR-52Z1-like -Fw: GGTTAT GATTTGCCTCCAGAAACCGT; OR-52Z1-like -Rv: ACACATGTT GAAAAGGGGATGGTTC). PCR products were sub-cloned in TOPOPCR II vector (Invitrogen) and pGEMt-easy (Promega), respectively, and sequenced for verification. The calcium-binding protein Capsl cDNA clone (Accession: ARA0ABA71YE02EM1) was present in our cDNA library and corresponded to PREDICTED Astyanax mexicanus calcyphosine-like protein (LOC103033765), mRNA. cDNAs were amplified by PCR from plasmids using SP6 and T7 primers, and digoxygenin-labelled riboprobes were synthesized from PCR templates. A manual protocol for whole-mount in situ hybridization was performed. Briefly, embryos were progressively re-hydrated before being incubated over night at 65° in hybridization buffer containing the probe. After stringent washes, the hybridized probes were detected by immunohistochemistry using an alkaline phosphatase-conjugated antibody against digoxygenin (Anti-Digoxigenin-AP, Fab fragments-Sigma, 11093274910) and a NBT/BCIP chromogenic substrate (Sigma, 11681451001). After staining, the dissected embryos were whole-mounted in glycerol and photographed with a Nikon Eclipse E800 microscope. For fluorescent in situ hybridization, probes were detected with a horse radish peroxidase conjugated anti-digoxygenin (Anti-DigoxigeninPOD, Fab fragments-Sigma, 11207733910) and FITC-tyramide as substrate. After DAPI staining, the dissected embryos were wholemounted in Vectashield medium (Vector) and imaged with a confocal microscope (Leica TCS SP8).
Significance was set at p < 0.05 (*). **, ***, and **** correspond to p < 0.01, p < 0.001, and p < 0.0001, respectively. In addition, individual fish were tracked manually during the olfaction behavior assay and the total distance swam during the 7 min test was calculated. 2.3. Lens ablations Ablations of the embryonic SF lens were performed at 32–36 h post fertilization (hpf) using a published protocol with some modifications (Espinasa et al., 2014; Yamamoto and Jeffery, 2002). Larvae were embedded in 2% low melt agarose (Sigma) and then submerged in EM. A glass needle (Sutter Instruments) attached to a micromanipulator was used to cut through the cornea and remove the lens. After surgery, the larvae were released from the agarose block and raised in the fish facility in the dark until analysis at one month of age. For olfactory organ measurement, unilateral ablations were performed with the nonablated eye side serving as an internal control. For behavioral studies, bilateral lens ablations were performed. 2.4. EdU pulse and chase experiments For the proliferation study, a pulse of Edu 10 mg/mL in EM (DMSO 0.01%) was carried out by bath during 30 min (for stages 40hpf and 84hpf) or during 1 h (for stages 7dpf and 21dpf). The larvae were washed in EM and fixed immediately or after a chase in PFA 4%. For the cell cycle experiments, two different chase times (6 h or 30 h) were carried out because we had no prior estimation of the cell cycle length in A. mexicanus at this stage. For the neuronal lifespan experiment, 64hpf embryos were treated by bath with Edu 10 mg/mL in EM (DMSO 0.01%) for 2 h. They were then washed several times in EM and returned to the incubator for growth. A batch of larvae /juveniles was fixed every week up to 5 weeks. Edu-positive cells were detected with the Click-iT™ EdU Alexa Fluor™ 647 (Thermo Fisher Scientific, C10340). Samples were whole-mounted in Vectashield medium (Vector, H1000) and imaged with a confocal microscope (Leica TCS SP8).
2.7. Statistical analyses Size measurement and cell counts were performed on ImageJ software. Statistical significances were tested using the Mann-Whitney nonparametric test, performed on the StatView software. Values are mean ± SEM. Significance was set at p < 0.05 (*). * *, * ** , and * ** * correspond to p < 0.01, p < 0.001, and p < 0.0001, respectively. 3. Results and discussion
2.5. Immunohistochemistry 3.1. Does olfactory epithelium size impact on olfactory capacities? Immunohistochemistry was performed on whole-mounts following the protocol published by Inoue et al. (Inoue and Wittbrodt, 2011). Primary and secondary antibodies were used at a dilution 1/500. Rabbit anti-Gαolf (Santa Cruz Biotechnology, sc-383), Rabbit antiactivated Caspase 3 (Abcam, Ab13847-25), Mouse anti-HUc/D (LifeTech, A21271), Rabbit anti-S100 (Dako, Z0311) and Rabbit anti-TRPC2 (ThermoFisher Scientific, OST00078G) were used. Secondary antibodies were Goat anti-Mouse IgG (H+L), Alexa Fluor 594 (Thermo Fisher Scientific, R37121) and Goat anti-Rabbit IgG (H+L), Alexa Fluor 488 (Thermo Fisher Scientific, A-11034). Samples were counterstained using DAPI (Sigma, 10236276001), wholemounted in Vectashield medium (Vector, H1000) and imaged with a confocal microscope (Leica TCS SP8).
In sharks, the excellent sense of smell (Tricas et al., 2009) has been associated to the very large size of the olfactory lamellae, following the simple assumption that an increase in surface area will allow for a more effective odor sampling (Theiss et al., 2009). To start testing whether a large OE size, as observed in CF, confers better olfactory sensory abilities, we used a previously designed olfactory test, with amino-acids as odorant molecules. Of note and importantly, this assay allows detecting olfactory-driven responses, and the taste system, which is also expanded in CF (Varatharasan et al., 2009; Yamamoto et al., 2009), is not involved in the behavioral responses (Hinaux et al., 2016). With this test, one month-old SF respond to 10−5 M alanine while one month-old CF respond to 10−10 M alanine (Hinaux et al., 2016). As OE size increases with age due to fish growth (Fig. 1A), here we evaluated the olfactory capacities of older and larger fish. Two months-old SF have larger OEs in absolute size than younger and smaller individuals, and their relative OE size is also larger to 1 month old SF, due to positive allometric growth of the OE (Fig. 1A). However, they showed the same detection threshold for alanine as one month old SF: they did not respond to 10−6 M alanine, nor to 10−7 M (n = 11, not shown).
2.6. In situ hybridization Total RNA from Astyanax embryos was reverse-transcribed using iScript cDNA synthesis kit (BioRad). Partial cDNA sequences for OMP (GenBank ID KP826791.1), TRPC2 (Genbank ID XM_022671326.1; predicted LOC111192950), and OR-52Z1-like (XM_022687171.1; predicted LOC103043790) were amplified by PCR using specific
3
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 1. OE size and olfactory capacities. A, OE growth in SF. Increase in absolute OE size (left) and relative OE size (right) according to growth in SF between 1 month (blue squares) and 2 months (blue triangles). SL, standard length. B, Comparison of relative OE size in the indicated morphs at the indicated ages/sizes. Numbers in brackets on bars indicate the numbers of fish used for quantifications, and data are mean ± sem. Mann-Whitney tests. C, Behavioral set-up. The test was performed in the dark with infrared recordings. Amino acids or EM were perfused on either side of boxes with a semi-partition. The amino acids reached the box 1.5 min after flow was initiated from the perfusion system. The position of the 4 fish was scored before and during odor exposure. Four quadrants were defined in the boxes, in which the amino acid concentration was very high, high, low, or zero over the duration of the test, see also (Hinaux et al., 2016). Each quadrant was attributed a coefficient (+3; +1; −1, −3, respectively) to calculate a preference index, which reflects the fish response (or absence of response) to the perfused amino acid. As an example, the preference index corresponding to the distribution of the fish shown on the schema would be zero. D, Olfactory response of F1 hybrids (female SF x male Pachón cross) to indicated concentrations of Alanine, represented as the preference index as a function of time. The vertical arrow indicates the time when the odorant reaches the box. A positive preference index (bars towards positive values) with significance as compared to zero (asterisks) indicates attraction to the odorant. The conditions and numbers of tests are indicated (n = 1 corresponds to one test, i.e., the cumulative score of 4 fish). Asterisks indicate significant response as compared with zero (Mann-Whitney tests).
deprived modality can be recruited by the remaining, intact sensory modalities. This phenomenon is well documented in the visual neuroscience literature, in which blindness results in the recruitment of visual cortex for somatosensory and auditory processing (Lazzouni and Lepore, 2014 for a review). Here, we sought to test whether high olfactory performance in CF is due to loss of eyes/vision and/or its pleiotropic enlargement of the naris. Indeed, anatomical changes in cranio-facial structures in CF can be divided into eye-dependent and eye-independent processes (Dufton et al., 2012; Yamamoto et al., 2003). An eye-dependent character is the size of the naris in adults: SF which have undergone unilateral lens ablation while embryos and therefore are eyeless on one side of the head as adults, have a wider olfactory pit (+ 12.9%) on the side with a degenerate eye than on the side of the normal eye. Moreover, phenotypic plasticity due to environmental conditions (darkness) may also be involved in the development of CF olfactory performances. To test this possibility, we performed bilateral lens ablations on SF embryos, which results in progressive eye degeneration and mimics the CF blind phenotype (Yamamoto and Jeffery, 2000; Yamamoto and Jeffery, 2002) (Fig. 2A). To distinguish between the potential effects of the loss of eyes, from those due to functional blindness by absence of vision, we raised the lens-ablated SF and their non-ablated SF siblings in complete darkness from the time of embryos collection (usually at about 5–10 h post-fertilization) to the age of one month. Olfactory tests were performed using 10−6 M alanine or Serine stocks, i.e., below the threshold concentration for normal, light-raised SF. Both lens-ablated/
We also tested one-month old F1 hybrids resulting from crosses between a female SF and a Pachón male. F1 hybrids possessed large olfactory epithelia comparable to CF (Fig. 1B) but showed olfactory capabilities comparable to SF. They responded to Alanine 10−3 M (n = 13, not shown) and 10−4 M, but not to 10−5 M (Fig. 1CD). We however noted that F1 hybrids swam very actively all along the test (for Alanine 10−5 M, distance swam: 5.07-fold/SF, p = 0.001; 4.87-fold/CF, p = 0.002, n = 9 each; Mann Whitney tests). This may explain their absence of sustained attraction towards the odorant compartment at SF threshold concentration (10−5 M), due to mixing of the odor in the test box by intense swimming activity. Thus, the olfactory capacities of F1 hybrids were not intermediate between those of their two parents and were far less good than those of CF even though they have comparably large OEs. Taken together, these data suggested that 1) olfactory skills correspond to a recessive or complex genetic trait, 2) OE size is not strictly correlated to olfactory capacities. The larger OE size of CF, providing them with a larger number of olfactory sensory neurons, is not the sole reason for the improved sensory trait. Below we tested two other, non-mutually exclusive developmental mechanisms that may underlie olfactory capacities in CF. 3.2. Cavefish olfactory capacities and developmental plasticity Cross-modal plasticity can occur as a result of decreased or abnormal sensory input, and is well described for example in the mammalian cortex: the cortical regions normally processing the 4
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 2. Abrogation of visual function in SF improves olfactory performance but does not affect OE size. A, Example of unilateral lens ablation performed at embryonic stage and resulting in eye degeneration in SF. Top photos show the eyes of a one month old SF after ablation on the left side. Bottom photos show Gαolf immunofluorescence used to quantify OE size after lens ablation. B, C, Responses of dark-raised and dark-raised/lens-ablated SF to the below-threshold concentration of Serine 10−6M (C) and Alanine 10−6M (D). Both types of visually-deprived SF show significant attractive response, contrary to normal SF. Ages of the tested fish and number of behavioral tests performed are indicated. D, Quantification of eye and OE circumferences, normalized to the standard length, in control (n = 16) and dark-raised (n = 18) one month old SF. Numbers in brackets on bars indicate the numbers of fish used for quantifications. Values are mean ± sem. For eye size p = 0.308 and for OE size p = 0.704 (Mann-Whitney tests, NS). E, Quantification of eye and OE circumferences (arbitrary units) on the lens-ablated side versus the contralateral control side of one month old SF (n = 5). Numbers in brackets on bars indicate the numbers of fish used for quantifications. Values are mean ± sem. * * indicates p = 0.0079. For OE size p = 0.841 (Mann-Whitney tests). F, Photographs of 64hpf larval heads after in situ hybridization for OMP (purple) labeling the OE, in frontal views. SF is on the left, CF is on the right. The scale bar is 100 µm.
difference in eye size between the two morphs is moderate: already at this stage, the OE labelled by expression of OMP (Olfactory Marker Protein) was significantly larger in CF than in SF (Fig. 2F; OE circumference standardized to length (a.u.): 0.414 ± 0.017 in SF versus 0.499 ± 0.043 in CF, n = 5 each, p = 0.008, Mann-Whitney). In sum, this series of experiments suggested two components in the enhanced olfactory capacities already present in one month old cave Astyanax: (1) one part of phenotypic plasticity, which can be recruited in SF in conditions of visual deprivation, and which does not depend on the size of the olfactory organ and (2) one part of evolution of the olfactory system in CF, probably related to OE development, and which is eye-independent.
dark-raised SF and dark-raised SF showed attraction to the odor source (Fig. 2BC). The same results were obtained for the two amino-acids: dark-raised SF responded well to both Alanine 10−6 M (Fig. 2B) and Serine 10−6 M (Fig. 2C). This improved olfactory performance was also maintained in older and larger dark-raised 2 months old individuals (Fig. 2C). These experiments showed that eye- and vision-dependent developmental processes are factors promoting enhanced olfactory skills in A. mexicanus. As elimination of visual function (by raising the fish in the dark) and elimination of the visual organ (by lens ablation) both resulted in improved olfactory performance in SF, we attributed this change to developmental functional phenotypic plasticity between olfactory and visual sensory modalities. Adult SF which have undergone unilateral lens ablation as embryos have a wider olfactory pit on the side with a degenerate eye than on the side of the normal eye (Yamamoto et al., 2003). At the stage assayed in the present study, such a difference in olfactory pit size was not yet established. Fig. 2D–E shows that neither lens ablation nor life in darkness affected the size of the OE, as measured on one month old SF. Thus, improved olfactory capabilities in visually-deprived SF at this age cannot be attributed to an increase in the size of their olfactory organ. This is in line with the above findings showing that OE size by itself is not strictly predictive for olfactory performance. Furthermore, these data supported that in CF, the OE does not grow larger as a consequence of the eye degeneration process, but rather that its size is autonomously developmentally-controlled and proportionately enlarged when compared to SF (Hinaux et al., 2016). This notion was further supported by measurements of OE sizes in SF and CF embryos/ larvae at 2.5 days of development (one day after hatching (Hinaux et al., 2011)), before the onset of eye degeneration in CF, and when the
3.3. Cavefish olfactory capacities and developmental modulation of OE neurogenesis To search for developmental and cellular mechanisms underlying the continued size differences between SF and CF OEs as well as the differences in olfactory skills between the two morphs, we next compared the proliferation, cell death, lifespan and neurogenesis patterns in the developing OEs of the two morphs along development. In all vertebrates, olfactory neurons of the OE are continuously renewed along the life of the animals. Therefore, differences in the rate of cell death and/or the lifespan of olfactory neurons might account for the OE differences between SF and CF. To test the first possibility (cell death), the numbers of activated caspase3-positive apoptotic cells were compared in the OE of the two morphs. At the stages examined, apoptotic cells were detected in low numbers. They corresponded to newly differentiated neurons as they were Hu-positive, and they were 5
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 3. Comparison of apoptosis and neuronal lifetime in SF and CF olfactory epithelium. A, Confocal images of double immunofluorescence against activated caspase-3 (green) and Hu (blue) in SF (left) and CF (right) at 84hpf. DAPI nuclear counterstain is magenta. Z-projections on a 10 µm depth are shown. The dotted lines show the contours of the OE, adjacent to the brain. Green arrows indicate activated Caspase3-positive apoptotic neurons, note also their condensed nuclei. White asterisks point to non-specific labeling on the cilia and mucus of olfactory sensory neurons. Scale bars: 25 µm. B, Quantification of apoptotic cells at different developmental stages in SF (blue bars) and CF (red bars). Mann-Whitney tests (n = 10–28, all NS). C, Confocal images of double labeling for Edu-positive cells (red) and Hu immunofluorescence (blue) in SF (left) and CF (right) after an EdU pulse performed at 64hpf, followed by a 5 weeks chase. Z-projections of frontal views on a 10 µm depth are shown. Red arrows point to EdU-positive nuclei. Blue asterisks indicate the central raphe that has formed at this stage after complex morphogenetic movements of the olfactory neuroepithelium. Insets show high magnification. Note the different scale bars (50 µm) for the 2 morphs, as well as the round versus oval shape of the OE in SF and CF, respectively, as previously reported in (Hinaux et al., 2016). D, Quantification of the number of Edu-positive neurons remaining in the olfactory epithelium after a 5 weeks chase in SF (n = 7) and CF (n = 8). Mann-Whitney tests; ** p = 0.0038.
Differences in proliferation and neurogenesis might also account for the SF/CF differences. To measure proliferation, EdU incorporation pulses were performed at various stages (between 40hpf and 21dpf) and larvae were fixed immediately at the end of the pulse. EdU-positive cells, corresponding to cells undergoing the S phase of mitosis during the pulse, were counted, and the volumes of the OEs were measured (at these stages, the OE still has a round/ovoid shape that allows accurate volume calculations). The results were expressed as densities, to normalize for OE volume size differences between the two morphotypes and to obtain a proxy for OE proliferation rate. With the exception of 40hpf, the density of EdU-positive proliferative cells was identical in the two morphs at all stages examined from 84hpf to 21dpf (Fig. 4AB). This shows that proliferation is proportionately similar in SF and CF OEs and further suggests that a proportionately similar pool of embryonic progenitors and later stem cells is defined in the olfactory placodes of SF and CF. Of note, the densities of progenitors appear stable between 84hpf and 21dpf on the graph shown in Fig. 4B: this is only due to the fact that the EdU pulse was performed during 30 min for the stages 40hpf and 84hpf, while it lasted 1 h for the stages 7dpf and 21dpf (see Methods). This was designed to label a reasonable number of cells to be counted at these older stages (and therefore to reduce errors), and led to the masking of the actual decrease in progenitors/stem cell densities after embryonic stages, which is indeed expected. Next, the possibility of variations in cell cycle and/or progenitor properties was investigated. EdU pulses were performed at 4dpf, followed by chases of 6 h or 30 h (no prior estimation of the cell cycle length in A. mexicanus at this stage were available). Larvae were fixed and double-labelled for EdU (identifying cells that underwent the S
found in equal amounts in SF and CF (Fig. 3AB and Suppl. Fig. 1). To test the second possibility (lifespan), an EdU pulse was performed at 64hpf, followed by chases of various lengths, from 1 to 5 weeks (Fig. 3CD). After short chases, very numerous EdU-positive cells were observed (not counted), corresponding to proliferative progenitors and to post-mitotic neurons which underwent their final mitosis during the pulse. For longer chases, the number of EdU-positive cells decreased similarly in SF and CF: this corresponded to the dilution of EdU in proliferative progenitors and to the disappearance of post-mitotic neurons. Finally, for 5 weeks chases, less than 30 EdU-positive/Hupositive cells were counted in the OE of both SF and CF (Fig. 3CD). This result suggested that the lifespan of olfactory neurons is similar in the two morphs and is about 4–5 weeks maximum. This lifetime is equivalent to that reported for zebrafish (29 days; Bayramli et al., 2017), but shorter than in the mouse (30–90 days; Mackay-Sim and Kittel, 1991; Santoro and Dulac, 2012). We suggest that the more numerous neurons found in CF after a 5 weeks chase is simply due to the fact that their OE is larger, and therefore that more progenitors were labelled during the pulse (see also below). Contrary to the data presented in the next section, it was not possible at this stage to measure exact OE volumes and calculate cell densities to normalize the data, due to complex morphogenesis of the OE neuroepithelium including the formation of a central raphe (blue asterisks on Fig. 3C). However, in line with our interpretation, after normalization to the total external volume of the OE (i.e., not taking into account the neuroepithelial folds and empty internal spaces), the densities of EdUpositive cells after a 5 weeks chase appeared identical in the two morphs (SF: 0.14 ± 0.04, n = 7; CF: 0.12 ± 0.02, n = 8; MannWhitney test).
6
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 4. Comparison of neurogenesis and progenitors properties in SF and CF olfactory epithelium. A, Confocal images of EdU fluorescent staining (blue) with DAPI nuclear counterstain (magenta) in SF (left) and CF (right) at 84hpf. Z-projections on a 10 µm depth are shown. Scale bars: 25 µm. B, Quantification of the density of Edu-positive cells after a pulse and immediate fixation at various developmental stages in SF (blue) and CF (red) (n = 11–21). Densities are given in number of cells per 105 µm3 volume. Mann-Whitney tests. C, D, Confocal images of EdU fluorescent staining (blue) and PCNA immunofluorescence (green) with DAPI nuclear counterstain (magenta) in SF (left) and CF (right) after an EdU pulse followed by a 6 h (C) or a 30 h (D) chase. Z-projections on a 10 µm depth are shown. Scale bars: 25 µm. E, F, Quantification of the density of PCNA-positive cells at 90hpf and 114hpf and the density of double-labelled EdU+/PCNA+ cells after a 6 h or a 30 h chase in SF (blue) and CF (red) (n = 8 for each). Densities are given in number of cells per 105 µm3 volume.
of EdU+ cells, represented 55 ± 9% in SF (n = 8) and 50 ± 7% in CF (n = 8) after a 30 h chase. This series of experiments suggested that the proliferative properties and behaviors of olfactory neurons progenitors were identical in the OEs of the SF and CF. Finally, we compared the time of appearance and proportions of differentiated olfactory sensory neuron (OSN) cell types between the two morphs. The 3 main OSN types described in fish were studied, including the ciliated neurons (OMP-positive), the microvillous neurons (TRPC2positive) and the crypt cells (S100-positive) (Germana et al., 2004; Hansen et al., 2004; Sato et al., 2005)(Fig. 5A). Kappe neurons, a fourth and sparse subtype of OSN (Ahuja et al., 2014), were not studied. OMP-expressing ciliated neurons were detected at 24hpf (hatching stage) in both CF and SF (about 15–20 cells) (Fig. 5B). The first TRPC2-expressing microvillous neurons were also detectable at 24hpf (Fig. 5C), while the S100-positive crypt cells differentiated later, around 84hpf (Fig. 5D). Thus, no difference in the time-course of OSN differentiation was detected between the two morphs. We next counted neuron numbers and calculated densities for each neuronal subtype. At 84hpf, the density of OMP-positive ciliated neurons was higher in SF (Fig. 5F and K). Conversely, the density of TRPC2-positive microvillous neurons was higher in CF (Fig. 5E and K). In
phase of mitosis during the pulse) and PCNA (Proliferating Cell Nuclear Antigen, labeling all phases of the cell cycle, thus identifying the entire population of cycling progenitors at the time of fixation) (Fig. 4C–G). For both chase times, i.e., at 4dpf and 5dpf, the density of PCNA-positive cells was identical in SF and CF (Fig. 4E), in line with the above results showing identical densities of proliferative progenitors captured in the S phase in the OEs of the two morphs after an Edu pulse (Fig. 4AB). For both chase times also, the densities of double EdU+/PCNA+ cells were identical in SF and CF (Fig. 4F). For the 6 h chase, this population of double-labelled cells likely corresponded to cells which were either finishing their cycle, or had already entered a new cycle. For the 30 h chase, this population likely corresponded only to cells which had re-entered a new cycle. These results suggested that cell cycle kinetics, including cell cycle length, are identical in the two Astyanax morphs. Accordingly, the percentage of cell cycle re-entry and the percentage of cell cycle exit were identical in SF and CF: cells re-entering the cycle, calculated as number of EdU+/PCNA+ doublelabelled cells/total number of EdU+ cells, represented 45 ± 9% in SF (n = 8) and 50 ± 7% in CF (n = 8) after a 30 h chase; and cells having left the cycle, calculated as number of EdU+/PCNA- cells/total number
7
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
8
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 5. Comparison of olfactory sensory neurons differentiation and composition in SF and CF olfactory epithelium. A, Schema depicting the 3 main types of OSNs described in fish. Adapted from (Oka et al., 2012). B, C, Bright field photographs showing in situ hybridization (purple) for OMP (B) and TRPC2 (C) on 24hpf SF and CF in toto on dorsal views. The OE is delineated by a dotted line. Scale bars: 25 µm. D, Confocal images of S100 immunofluorescence staining (blue) with DAPI nuclear counterstain (magenta) in 84hpf SF and CF, on frontal views. Z-projections on a 10 µm depth are shown. Arrows point to S100-positive neurons. Scale bars: 25 µm. E, F, Confocal images of fluorescent in situ hybridization for TRPC2 and OMP (green) with DAPI nuclear counterstain (magenta) in 84hpf SF and CF, on frontal views. Z-projections on a 5 µm depth are shown. Arrows point to expressing neurons. Scale bars: 25 µm. G, H, Bright field photographs showing in situ hybridization (purple) for calciphosine-like (G) and odorant receptor OR-52Z1-like (H) on 84hpf SF and CF in toto on frontal views. Scale bars: 25 µm. I, J, Confocal images of immunofluorescence for S100 (I) and TRPC2 (J, green) with DAPI nuclear counterstain (magenta) in 21dpf SF and CF, on frontal views. Scale bars: 20 µm. K, Quantification of OSN subtypes densities across developmental time in SF (blue) and CF (red). Densities are given in number of cells per 105 µm3 volume. Mann-Whitney tests (n = 8–27 for each point). nd, not determined.
organ, the olfactory epithelium. However, differences in processing in the olfactory bulbs or higher brain regions may also be relevant to explain the differences in olfactory skills between the two morphs, particularly concerning developmental phenotypic plasticity. In early blind human individuals for example, superior performance in odorprocessing tasks were shown to be associated with an activation of the occipital cortex that was not found in sighted subjects (Renier et al., 2013) and to a larger size of the olfactory bulbs (Araneda et al., 2016; Rombaux et al., 2010). In cavefish also, the size of the olfactory bulbs seems larger than in surface fish (Pottin et al., 2011; JB and SR, unpublished data). However the possibility that in cavefish, brain regions which normally receive visual inputs start receiving olfactory inputs has not been investigated yet. In fish, the olfactory and visual systems are functionally linked by the terminal nerve system, or olfacto-retinalis system. It is known that olfactory system stimulations can modify visual responses (Stephenson et al., 2012), but whether the absence of visual inputs can modify the olfactory response has never been directly demonstrated. However, the terminal nerve receives afferences from visual structures (Yamamoto and Ito, 2000), the activity of terminal nerve GnRH neurons can be modulated by visual cues (Ramakrishnan and Wayne, 2009), and GnRH can modulate OSN excitability and responses to odors in amphibians (Eisthen et al., 2000; Park and Eisthen, 2003; Zhang and Delay, 2007). Thus, one possible mechanism for increased olfaction in visually-deprived surface fish and /or in cavefish may be through the terminal nerve system. It will also be essential to understand the underlying neurophysiological mechanisms of improved cavefish olfaction, which may be manifold given the huge difference in olfactory detection skills between the two Astyanax mexicanus morphs. Such mechanisms probably include vision-dependent plasticity processes, that we have unmasked in the present study in SF after visual deprivation, and that may have been fixed rapidly by genetic assimilation in the cavefish lineage after environmental change (Waddington, 1953). Vision-independent developmental evolutionary processes also seem to be involved, including those controlling the embryonic size of the olfactory organ (Hinaux et al., 2016) and its neuronal composition (present paper). The evolution of the G protein-coupled olfactory receptor families, representing about 300 genes in fish (Hashiguchi et al., 2008; Korsching, 2009) and which are among the fastest-evolving genes in the genomes, may contribute as well. Finally in vertebrates, OSN sensitivity and odor processing in the olfactory bulbs can be modulated by diverse sources and neurotransmitters including serotonin (Frings, 1993; Petzold et al., 2009). It might be the case in cavefish, as a mutation in the serotonin degrading enzyme (monoamine oxidase) confers them with high brain serotonin levels (Elipot et al., 2014a). The proportionately large size of the OE in cavefish may partly account for its enhanced amino-acid detection skills, but is not the only determinant. Indeed, old SF or F1 hybrids with large OEs have modest detection thresholds, while visually-deprived SF with unchanged OE size show improved olfactory capacities. Rather than the size of the mature OE itself, the important parameter may be the size and early patterning of the embryonic olfactory placode in cavefish. If the olfactory sensory territory that is gained over adjacent placodes gives rise to progenitor pools with specific neurogenic properties which bias the neuronal composition of the later OE, then this placodal expansion may translate into important functional outcomes.
addition at 84hpf, a markedly different distribution of the labelled OSNs was observed inside the OEs of the two morphs. In SF, both ciliated (OMP) and microvillous (TRPC2) OSNs were located at the margin of the OE, forming a ring in frontal views. In CF however, OSNs were distributed at the center of the olfactory cup (Fig. 5EF). This result was confirmed with other, independent differentiated OSN markers, such as calciphosine-like (Capsl, a calcium-binding protein) or OR-52Z1-like, a member of the OR family of odorant receptors (Fig. 5GH). This difference is difficult to interpret in terms of neurogenesis because there was not obvious dissimilarity in the localization of the progenitors between SF and CF: EdUretaining cells and PCNA-positive cells were located at the basis of the epithelium in both cases (Fig. 4). Moreover, this difference in the pattern of OSN distribution was transient. Indeed at 21dpf, neurons were evenly distributed at the surface of the OE, both in SF and CF (Fig. 5IJ). At this later stage also, the density of TRPC2+ microvillous neurons was higher in CF, while the density of S100+ crypt cells was identical in the two morphs (Fig. 5IJK; note that ciliated neurons, representing the major type of OSN (90% in zebrafish) were not counted at 21dpf). In sum, these comparative developmental neuroanatomy data show that the densities and distributions of OSN subtypes vary in the OE of SF and CF. Specifically, the microvillous neurons types are proportionately more represented in CF, possibly at the expense of ciliated neurons. In zebrafish, trout or goldfish, microvillous neurons respond to alimentary odors such as amino-acids, while ciliated neurons are more generalists and also respond to social cues such as bile acids or alarm substance, and crypt cells respond to sexual pheromones (Koide et al., 2009; Sato and Suzuki, 2001; Yoshihara, 2008). The 3 OSN subtypes also express different types of odorant receptors and project to distinct glomeruli in the olfactory bulbs (Ahuja et al., 2013; Hamdani el and Doving, 2007; Hansen et al., 2004; Koide et al., 2009; Oka et al., 2012; Sato et al., 2005; Yoshihara, 2008). In Astyanax mexicanus, OSN specificity towards odor types and their involvement in specific behaviors has not been studied yet, but it is tempting to speculate about the contribution of more numerous, amino-acid-responsive, TRPC2-positive microvillous neurons in the enhanced olfactory skills demonstrated by cavefish for amino-acid detection during behavioral assays. Future studies will aim at functionally testing this hypothesis. Altogether, our comparative analyses of OE development in SF and CF demonstrated that general proliferative properties of progenitors are identical but neurogenic properties differ in the two morphs. In cavefish, progenitors seem specified and fated to generate proportionately more microvillous neurons, pushing back the origins of the SF/CF differences at the level of the early placodal ectoderm. Indeed, cell-type heterogeneity in the zebrafish olfactory epithelium is generated from progenitors within the placodal ectoderm (Aguillon et al., 2018), but the mechanisms underlying the segregation of the various olfactory subtypes from overlapping progenitor pools are unknown. In cavefish, the olfactory placode territory is expanded, probably at the expense of adjacent placodal territories including the presumptive lens region (Hinaux et al., 2016). Therefore, cavefish appear like an excellent evo-devo model to perform lineage studies and fate maps to address the question of the establishment and patterning of olfactory neurons progenitor pools in the early placode. 4. Conclusions To start investigating the developmental origins of the excellent olfactory capacities in CF, we have focused on their external sensory 9
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Koide, T., et al., 2009. Olfactory neural circuitry for attraction to amino acids revealed by transposon-mediated gene trap approach in zebrafish. Proc. Natl. Acad. Sci. USA 106, 9884–9889. Korsching, S., 2009. The molecular evolution of teleost olfactory receptor gene families. Results Probl. Cell Differ. 47, 37–55. Lazzouni, L., Lepore, F., 2014. Compensatory plasticity: time matters. Front. Hum. Neurosci. 8, 340. Mackay-Sim, A., Kittel, P.W., 1991. On the life span of olfactory receptor neurons. Eur. J. Neurosci. 3, 209–215. Maier, E.C., et al., 2014. Sensational placodes: neurogenesis in the otic and olfactory systems. Dev. Biol. 389, 50–67. Miyasaka, N., et al., 2013. Functional development of the olfactory system in zebrafish. Mech. Dev. 130, 336–346. Oka, Y., et al., 2012. Crypt neurons express a single V1R-related ora gene. Chem. Senses 37, 219–227. Park, D., Eisthen, H.L., 2003. Gonadotropin releasing hormone (GnRH) modulates odorant responses in the peripheral olfactory system of axolotls. J. Neurophysiol. 90, 731–738. Petzold, G.C., et al., 2009. Serotonergic modulation of odor input to the mammalian olfactory bulb. Nat. Neurosci. 12, 784–791. Pottin, K., et al., 2011. Restoring eye size in Astyanax mexicanus blind cavefish embryos through modulation of the Shh and Fgf8 forebrain organising centres. Development 138, 2467–2476. Ramakrishnan, S., Wayne, N.L., 2009. Social cues from conspecifics alter electrical activity of gonadotropin-releasing hormone neurons in the terminal nerve via visual signals. Am. J. Physiol. Regul. Integr. Comp. Physiol. 297, R135–R141. Renier, L., et al., 2013. Right occipital cortex activation correlates with superior odor processing performance in the early blind. PLoS One 8, e71907. Rombaux, P., et al., 2010. Increased olfactory bulb volume and olfactory function in early blind subjects. NeuroReport 21, 1069–1073. Santoro, S.W., Dulac, C., 2012. The activity-dependent histone variant H2BE modulates the life span of olfactory neurons. Elife 1, e00070. Sato, K., Suzuki, N., 2001. Whole-cell response characteristics of ciliated and microvillous olfactory receptor neurons to amino acids, pheromone candidates and urine in rainbow trout. Chem. Senses 26, 1145–1156. Sato, Y., et al., 2005. Mutually exclusive glomerular innervation by two distinct types of olfactory sensory neurons revealed in transgenic zebrafish. J. Neurosci. 25, 4889–4897. Shao, X., et al., 2017. Olfactory sensory axons target specific protoglomeruli in the olfactory bulb of zebrafish. Neural Dev. 12, 18. Stephenson, J.F., et al., 2012. Food and conspecific chemical cues modify visual behavior of zebrafish, Danio rerio. Zebrafish 9, 68–73. Theiss, S.M., et al., 2009. Morphological indicators of olfactory capability in Wobbegong sharks (Orectolobidae, Elasmobranchii). Brain Behav. Evol. 73, 91–101. Torres-Paz, J., Whitlock, K.E., 2014. Olfactory sensory system develops from coordinated movements within the neural plate. Dev. Dyn. 243, 1619–1631. Tricas, T.C., et al., 2009. Response of the hammerhead shark olfactory epithelium to amino acid stimuli. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 195, 947–954. Varatharasan, N., et al., 2009. Taste bud development and patterning in sighted and blind morphs of Astyanax mexicanus. Dev. Dyn. 238, 3056–3064. Vitebsky, A., et al., 2005. Isolation and characterization of the laure olfactory behavioral mutant in the zebrafish, Danio rerio. Dev. Dyn. 234, 229–242. Wackermannova, M., et al., 2016. Olfactory sensitivity in mammalian species. Physiol. Res. 65, 369–390. Waddington, C.H., 1953. Genetic assimilation of an acquired character. Evolution 7, 118–126. Wakisaka, N., et al., 2017. An Adenosine Receptor for Olfaction in Fish. Curr. Biol. 27, 1437–1447, (e4). Whitlock, K.E., 2004. A new model for olfactory placode development. Brain Behav. Evol. 64, 126–140. Whitlock, K.E., 2006. The sense of scents: olfactory behaviors in the zebrafish. Zebrafish 3, 203–213. Whitlock, K.E., Westerfield, M., 1998. A transient population of neurons pioneers the olfactory pathway in the zebrafish. J. Neurosci. 18, 8919–8927. Yamamoto, N., Ito, H., 2000. Afferent sources to the ganglion of the terminal nerve in teleosts. J. Comp. Neurol. 428, 355–375. Yamamoto, Y., et al., 2009. Pleiotropic functions of embryonic sonic hedgehog expression link jaw and taste bud amplification with eye loss during cavefish evolution. Dev. Biol. 330, 200–211. Yamamoto, Y., et al., 2003. Development and evolution of craniofacial patterning is mediated by eye-dependent and -independent processes in the cavefish Astyanax. Evol. Dev. 5, 435–446. Yamamoto, Y., Jeffery, W.R., 2000. Central role for the lens in cave fish eye degeneration. Science 289, 631–633. Yamamoto, Y., Jeffery, W.R., 2002. Probing teleost eye development by lens transplantation. Methods 28, 420–426. Yamamoto, Y., et al., 2004. Hedgehog signalling controls eye degeneration in blind cavefish. Nature 431, 844–847. Yoshihara, Y., 2008. Molecular genetic dissection of the zebrafish olfactory system. Results Probl. Cell Differ., 97–119. Zhang, W., Delay, R.J., 2007. Gonadotropin-releasing hormone modulates voltageactivated sodium current and odor responses in Necturus maculosus olfactory sensory neurons. J. Neurosci. Res. 85, 1656–1667.
Acknowledgements We thank Stéphane Père and Victor Simon for care of our Astyanax breeding colony, and Muriel Perron for interesting discussions about progenitor proliferation assays. This Work was supported by an ANR (Agence Nationale pour la Recherche) grant [BLINDTEST] and an « Equipe FRM » grant [DEQ. 20150331745] from the Fondation pour la Recherche Médicale. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ydbio.2018.04.019. References Aguillon, R., et al., 2018. Cell-type heterogeneity in the early zebrafish olfactory epithelium is generated from progenitors within preplacodal ectoderm. Elife 7. Ahuja, G., et al., 2014. Kappe neurons, a novel population of olfactory sensory neurons. Sci. Rep. 4, 4037. Ahuja, G., et al., 2013. Zebrafish crypt neurons project to a single, identified mediodorsal glomerulus. Sci. Rep. 3, 2063. Araneda, R., et al., 2016. Cortical plasticity and olfactory function in early blindness. Front. Syst. Neurosci. 10, 75. Bayramli, X., et al., 2017. Patterned arrangements of olfactory receptor gene expression in zebrafish are established by radial movement of specified olfactory sensory neurons. Sci. Rep. 7, 5572. Behrens, M., et al., 2014. ORA1, a zebrafish olfactory receptor ancestral to all mammalian V1R genes, recognizes 4-hydroxyphenylacetic acid, a putative reproductive pheromone. J. Biol. Chem. 289, 19778–19788. Bibliowicz, J., et al., 2013. Differences in chemosensory response between eyed and eyeless Astyanax mexicanus of the Rio Subterraneo cave. Evodevo 4, 25. Breau, M.A., Schneider-Maunoury, S., 2014. Mechanisms of cranial placode assembly. Int. J. Dev. Biol. 58, 9–19. Caprio, J., et al., 1989. Electro-olfactogram and multiunit olfactory receptor responses to binary and trinary mixtures of amino acids in the channel catfish, Ictalurus punctatus. J. Gen. Physiol. 93, 245–262. Collin, S.P., 2012. The neuroecology of cartilaginous fishes: sensory strategies for survival. Brain Behav. Evol. 80, 80–96. Corfield, J.R., et al., 2014. Anatomical specializations for enhanced olfactory sensitivity in kiwi, Apteryx mantelli. Brain Behav. Evol. 84, 214–226. Dufton, M., et al., 2012. Early lens ablation causes dramatic long-term effects on the shape of bones in the craniofacial skeleton of Astyanax mexicanus. PLoS One 7, e50308. Eisthen, H.L., et al., 2000. Neuromodulatory effects of gonadotropin releasing hormone on olfactory receptor neurons. J. Neurosci. 20, 3947–3955. Elipot, Y., et al., 2014a. A mutation in the enzyme monoamine oxidase explains Part of the Astyanax cavefish behavioral syndrome. Nat. Commun. Apr 10 (5), 3647. Elipot, Y., et al., 2014b. Astyanax transgenesis and husbandry: how cavefish enters the lab. Zebrafish 11 (4), 291–299. Espinasa, L., et al., 2014. Enhanced prey capture skills in Astyanax cavefish larvae are independent from eye loss. Evodevo 5, 35. Friedrich, R.W., Korsching, S.I., 1997. Combinatorial and chemotopic odorant coding in the zebrafish olfactory bulb visualized by optical imaging. Neuron 18, 737–752. Frings, S., 1993. Protein kinase C sensitizes olfactory adenylate cyclase. J. Gen. Physiol. 101, 183–205. Germana, A., et al., 2004. S100 protein-like immunoreactivity in the crypt olfactory neurons of the adult zebrafish. Neurosci. Lett. 371, 196–198. Grocott, T., et al., 2012. The peripheral sensory nervous system in the vertebrate head: a gene regulatory perspective. Dev. Biol. 370, 3–23. Hamdani el, H., Doving, K.B., 2007. The functional organization of the fish olfactory system. Prog. Neurobiol. 82, 80–86. Hansen, A., et al., 2004. Differential distribution of olfactory receptor neurons in goldfish: structural and molecular correlates. J. Comp. Neurol. 477, 347–359. Hara, T.J., 1994. Olfaction and gustation in fish: an overview. Acta Physiol. Scand. 152, 207–217. Hara, T.J., 2006. Feeding behaviour in some teleosts is triggered by single amino acids primarily through olfaction. J. Fish Biol. 68, 810–825. Hashiguchi, Y., et al., 2008. Evolutionary patterns and selective pressures of odorant/ pheromone receptor gene families in teleost fishes. PLoS One 3, e4083. Hinaux, H., et al., 2016. Sensory evolution in blind cavefish is driven by early events during gastrulation and neurulation. Development 143, 4521–4532. Hinaux, H., et al., 2011. A developmental staging table for Astyanax mexicanus surface fish and Pachon cavefish. Zebrafish 8, 155–165. Inoue, D., Wittbrodt, J., 2011. One for all–a highly efficient and versatile method for fluorescent immunostaining in fish embryos. PLoS One 6, e19713. Keller-Costa, T., et al., 2015. Chemical communication in cichlids: a mini-review. Gen. Comp. Endocrinol. 221, 64–74.
10
Contents lists available at ScienceDirect
Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology
Original research article
Developmental evolution and developmental plasticity of the olfactory epithelium and olfactory skills in Mexican cavefish Maryline Blin, Eugène Tine, Lydvina Meister, Yannick Elipot, Jonathan Bibliowicz, ⁎ Luis Espinasa1, Sylvie Rétaux Paris-Saclay Institute of Neuroscience, Université Paris-Sud, CNRS UMR9197, Université Paris-Saclay, Avenue de la terrasse, 91198 Gif-sur-Yvette, France
A R T I C L E I N F O
A BS T RAC T
Keywords: Olfaction Organ size Proliferation Neurogenesis Olfactory sensory neuron
The fish Astyanax mexicanus comes in two forms: the normal surface-dwelling (SF) and the blind depigmented cave-adapted (CF) morphs. Among many phenotypic differences, cavefish show enhanced olfactory sensitivity to detect amino-acid odors and they possess large olfactory sensory organs. Here, we questioned the relationship between the size of the olfactory organ and olfactory capacities. Comparing olfactory detection abilities of CF, SF and F1 hybrids with various olfactory epithelium (OE) sizes in behavioral tests, we concluded that OE size is not the only factor involved. Other possibilities were envisaged. First, olfactory behavior was tested in SF raised in the dark or after embryonic lens ablation, which leads to eye degeneration and mimics the CF condition. Both absence of visual function and absence of visual organs improved the SF olfactory detection capacities, without affecting the size of their OE. This suggested that developmental plasticity occurs between the visual and the olfactory modalities, and can be recruited in SF after visual deprivation. Second, the development of the olfactory epithelium was compared in SF and CF in their first month of life. Proliferation, cell death, neuronal lifespan, and olfactory progenitor cell cycling properties were identical in the two morphs. By contrast, the proportions of the three main olfactory sensory neurons subtypes (ciliated, microvillous and crypt) in their OE differed. OMP-positive ciliated neurons were more represented in SF, TRPC2-positive microvillous neurons were proportionately more abundant in CF, and S100-positive crypt cells were found in equal densities in the two morphs. Thus, general proliferative properties of olfactory progenitors are identical but neurogenic properties differ and lead to variations in the neuronal composition of the OE in SF and CF. Together, these experiments suggest that there are at least two components in the evolution of cavefish olfactory skills: (1) one part of eye-dependent developmental phenotypic plasticity, which does not depend on the size of the olfactory organ, and (2) one part of developmental evolution of the OE, which may stem from embryonic specification of olfactory neurons progenitor pools.
1. Introduction The olfactory epithelium (OE) of vertebrates is the external sensory organ devoted to the sense of smell. During development, the OE derives from a thickening called the placode that is formed at the end of gastrulation (Grocott et al., 2012; Whitlock, 2004). The non-neural ectoderm (at the border of the neural plate) or neural ectoderm (inside the neural plate) origin of the placode is currently debated. During neurulation and after, the placode cells undergo
⁎
1
migrations and morphogenesis, transforming a thickened sheet of ectoderm into a multilayered pit, in which proliferation and neurogenesis occur (Breau and Schneider-Maunoury, 2014; Maier et al., 2014; Torres-Paz and Whitlock, 2014). Differentiated olfactory sensory neurons of the OE then project their axons onto the glomeruli of the olfactory bulbs, in the telencephalon, progressively establishing the path of the olfactory nerve (Koide et al., 2009; Shao et al., 2017; Whitlock and Westerfield, 1998). Each olfactory sensory neuron expresses a single GPCR olfactory receptor among the
Corresponding author. E-mail address: [email protected] (S. Rétaux). School of Science, Marist College, New York 12601, USA.
https://doi.org/10.1016/j.ydbio.2018.04.019 Received 31 January 2018; Received in revised form 19 April 2018; Accepted 24 April 2018 0012-1606/ © 2018 Elsevier Inc. All rights reserved.
Please cite this article as: Blin, M., Developmental Biology (2018), https://doi.org/10.1016/j.ydbio.2018.04.019
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
facility. Fish are maintained at 23–26 °C on 12:12 h light: dark cycle and they are fed twice a day with dry and live food. The breeding colonies spawn two or three times per week and generate a highly variable quantity of embryos (a few dozens to a few thousands). Spawning induction and larval care are described elsewhere (Elipot et al., 2014b). Embryos and larvae were raised at 24 °C in embryo medium (EM). SF raised in the dark were kept in thermostaticallycontrolled light-proof tanks. Feeding and daily EM changes were done in the dark. The F1 hybrids used in this paper were the progeny of a female SF x male Pachón cross. Animals were treated according to the French and European regulations for use of animals in research. SR's authorization for use of animals in research including Astyanax mexicanus is 91–116. Paris Center-Sud Ethic Committee authorization numbers are 2012-0055, 2016-36 and 2017-04.
species’ repertoire of odorant receptors and is specialized to detect a single odorant (Korsching, 2009; Shao et al., 2017; reviewed in Miyasaka et al., 2013). The olfactory system is used to detect, discriminate and identify odorant molecules in relation with a variety of adaptive behaviors such as foraging, communication, reproduction or predator avoidance. Although different species show markedly different odor detection capacities, or olfactory specialization, not much is known about the factors determining olfactory sensitivity. Among mammals for example, differences of 5–6 orders of magnitude in the capacity to detect certain odorants have been reported, but no direct correlation between olfactory sensitivity and numbers or densities of olfactory sensory neurons, or the size of olfactory structures, has been reported (reviewed in Wackermannova et al., 2016). On the other hand, some birds such as the nocturnal New Zealand kiwi for which olfaction is of strong behavioral relevance possess enlarged olfactory structures (Corfield et al., 2014), and some fish such as sharks with exceptional olfactory skills present an extra-large OE surface and large olfactory bulbs (Collin, 2012; Tricas et al., 2009). The developmental origin(s) of these sensory specializations, in terms of both olfactory system anatomy and function, are mostly unknown. Here, we addressed the question of the developmental origin of olfactory sensitivity using the comparison between the two morphs of the fish Astyanax mexicanus. The blind and depigmented cavedwelling morphs (cavefish, CF) and their river-dwelling conspecifics (surface fish, SF) have markedly different olfactory capacities and are unique models to study this question. In fish, olfactory responses to relevant odorant cues (amino-acids, nucleotides, pheromones, alarm substance) can be recorded through electroolfactograms, functional imaging or behavioral analyses (Behrens et al., 2014; Caprio et al., 1989; Friedrich and Korsching, 1997; Hara, 1994, 2006; Keller-Costa et al., 2015; Miyasaka et al., 2013; Tricas et al., 2009; Vitebsky et al., 2005; Wakisaka et al., 2017; Whitlock, 2006; Yoshihara, 2008). In an olfactory assay performed in the lab, CF larvae originating from the Pachón cave are able to detect and show attractive response to concentrations as low as 10−10 M of amino-acid, whereas SF larvae can only detect 10−5 M ranges (Hinaux et al., 2016). In the wild, blind CF inhabiting the Subterráneo cave respond to food odors whereas eyed hybrid fish phenotypically resembling SF do not (Bibliowicz et al., 2013). Interestingly, CF from these two caves have larger OEs and nostrils than SF. In Pachón embryos and larvae, the larger sensory organ results from early developmental evolution during gastrulation and neurulation, due to CF-specific modulations of midline signaling from organizer centers (Hinaux et al., 2016; Pottin et al., 2011; Yamamoto et al., 2004). Here, in search for the developmental origins of the enhanced olfactory skills of cavefish, we investigated the relationship between the size of the olfactory organ and olfactory capacities in Astyanax mexicanus. We also analyzed other developmental processes, including developmental plasticity due to loss of vision, and changes in neurogenesis control influencing the neuronal composition of the OE. The data are presented in a “results and discussion” format.
2.2. Behavioral testing Behavioral tests were performed as previously described (Hinaux et al., 2016) in a specially constructed sound- and lightproof room that includes a main compartment for testing and a second computer work station compartment from which recording of the tests was performed with minimal disturbances. All fish were fed 24 h prior to the test with two day-old Artemia and then food was withheld until testing in order to standardize their feeding state. Four one-month-old juveniles were placed in behavioral testing boxes (see Fig. 1C; 9 cm wide x 13 cm long) containing 150 mL embryo medium (EM) and let acclimatize for two hours prior to the test at a temperature of 24 °C and in the dark. SF and CF (or experimental and control animals) were always tested in parallel. Boxes were placed on top of an infrared light box (ViewPoint S.A.). Each test was initiated by simultaneously opening the Luer stoppers of medical solution administration tubing (Baxter, U.K.) to perfuse solutions at 5 mL/min from two reservoirs containing 60 mL of either amino-acid containing EM or EM alone (control). On the EM-perfused side, the flow generated was identical to the flow on the amino acid perfused side. Tests were recorded for 7 min on a Dell work station using ViewPoint imaging software and a DragonFly2 camera equipped with an infrared filter (PointGray). Utilizing a colorimetric test for the quantification of amino acid concentrations (Hinaux et al., 2016), we defined four quadrants of the boxes in which the amino acid concentration was very high, high, low or zero over the duration of the test. Each of these quadrant was attributed a coefficient to calculate a Preference Index Score (PIS) reflecting the attraction of fish (or absence of attraction) to the amino acid source. The PIS for each time point (at 30 s intervals) was the cumulative score of the four fish in the box, where the position of each fish was scored with the values of − 3 (quadrant furthest away from amino acid source), − 1, 1, or 3 (quadrant closest to amino acid source). Thus the maximum and minimum PIS scores are + 12 and − 12, respectively, for a given time point. In order to correct for the initial position of the fish when the amino acid first enters the box at 1.5 min after the start of the experiment, the PIS was reset at zero for this time point and the subsequent PIS values were corrected by subtracting of the initial raw score at 1.5 min. Experiments in which the initial PIS score was > 6 or < −6 were discarded because the correction for initial position of the fish lead to artefactual “false attraction” or “false repulsion” at subsequent time points (Hinaux et al., 2016). Statistical significance of replicate tests was calculated using the non-parametric Mann-Whitney test and was performed using the StatView software. In all figures, n = 1 corresponds to one test, i.e., the cumulative score of 4 fish.
2. Materials and methods 2.1. Fish Laboratory stocks of A. mexicanus SF and CF (Pachón population) were obtained in 2004 from the Jeffery laboratory at the University of Maryland, College Park, MD, and were since then bred in our local
2
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
primers (Omp-Fw: TGCTACAGGGTCTAGGTCGTC; Omp-Rv: GGCCT CGTTCTTCTTGAACATC; Trpc2-Fw: TGCTGGGGGAGTCTATCACA; Trpc2-Rv: CTCGATATAGCGCTGCACCA; OR-52Z1-like -Fw: GGTTAT GATTTGCCTCCAGAAACCGT; OR-52Z1-like -Rv: ACACATGTT GAAAAGGGGATGGTTC). PCR products were sub-cloned in TOPOPCR II vector (Invitrogen) and pGEMt-easy (Promega), respectively, and sequenced for verification. The calcium-binding protein Capsl cDNA clone (Accession: ARA0ABA71YE02EM1) was present in our cDNA library and corresponded to PREDICTED Astyanax mexicanus calcyphosine-like protein (LOC103033765), mRNA. cDNAs were amplified by PCR from plasmids using SP6 and T7 primers, and digoxygenin-labelled riboprobes were synthesized from PCR templates. A manual protocol for whole-mount in situ hybridization was performed. Briefly, embryos were progressively re-hydrated before being incubated over night at 65° in hybridization buffer containing the probe. After stringent washes, the hybridized probes were detected by immunohistochemistry using an alkaline phosphatase-conjugated antibody against digoxygenin (Anti-Digoxigenin-AP, Fab fragments-Sigma, 11093274910) and a NBT/BCIP chromogenic substrate (Sigma, 11681451001). After staining, the dissected embryos were whole-mounted in glycerol and photographed with a Nikon Eclipse E800 microscope. For fluorescent in situ hybridization, probes were detected with a horse radish peroxidase conjugated anti-digoxygenin (Anti-DigoxigeninPOD, Fab fragments-Sigma, 11207733910) and FITC-tyramide as substrate. After DAPI staining, the dissected embryos were wholemounted in Vectashield medium (Vector) and imaged with a confocal microscope (Leica TCS SP8).
Significance was set at p < 0.05 (*). **, ***, and **** correspond to p < 0.01, p < 0.001, and p < 0.0001, respectively. In addition, individual fish were tracked manually during the olfaction behavior assay and the total distance swam during the 7 min test was calculated. 2.3. Lens ablations Ablations of the embryonic SF lens were performed at 32–36 h post fertilization (hpf) using a published protocol with some modifications (Espinasa et al., 2014; Yamamoto and Jeffery, 2002). Larvae were embedded in 2% low melt agarose (Sigma) and then submerged in EM. A glass needle (Sutter Instruments) attached to a micromanipulator was used to cut through the cornea and remove the lens. After surgery, the larvae were released from the agarose block and raised in the fish facility in the dark until analysis at one month of age. For olfactory organ measurement, unilateral ablations were performed with the nonablated eye side serving as an internal control. For behavioral studies, bilateral lens ablations were performed. 2.4. EdU pulse and chase experiments For the proliferation study, a pulse of Edu 10 mg/mL in EM (DMSO 0.01%) was carried out by bath during 30 min (for stages 40hpf and 84hpf) or during 1 h (for stages 7dpf and 21dpf). The larvae were washed in EM and fixed immediately or after a chase in PFA 4%. For the cell cycle experiments, two different chase times (6 h or 30 h) were carried out because we had no prior estimation of the cell cycle length in A. mexicanus at this stage. For the neuronal lifespan experiment, 64hpf embryos were treated by bath with Edu 10 mg/mL in EM (DMSO 0.01%) for 2 h. They were then washed several times in EM and returned to the incubator for growth. A batch of larvae /juveniles was fixed every week up to 5 weeks. Edu-positive cells were detected with the Click-iT™ EdU Alexa Fluor™ 647 (Thermo Fisher Scientific, C10340). Samples were whole-mounted in Vectashield medium (Vector, H1000) and imaged with a confocal microscope (Leica TCS SP8).
2.7. Statistical analyses Size measurement and cell counts were performed on ImageJ software. Statistical significances were tested using the Mann-Whitney nonparametric test, performed on the StatView software. Values are mean ± SEM. Significance was set at p < 0.05 (*). * *, * ** , and * ** * correspond to p < 0.01, p < 0.001, and p < 0.0001, respectively. 3. Results and discussion
2.5. Immunohistochemistry 3.1. Does olfactory epithelium size impact on olfactory capacities? Immunohistochemistry was performed on whole-mounts following the protocol published by Inoue et al. (Inoue and Wittbrodt, 2011). Primary and secondary antibodies were used at a dilution 1/500. Rabbit anti-Gαolf (Santa Cruz Biotechnology, sc-383), Rabbit antiactivated Caspase 3 (Abcam, Ab13847-25), Mouse anti-HUc/D (LifeTech, A21271), Rabbit anti-S100 (Dako, Z0311) and Rabbit anti-TRPC2 (ThermoFisher Scientific, OST00078G) were used. Secondary antibodies were Goat anti-Mouse IgG (H+L), Alexa Fluor 594 (Thermo Fisher Scientific, R37121) and Goat anti-Rabbit IgG (H+L), Alexa Fluor 488 (Thermo Fisher Scientific, A-11034). Samples were counterstained using DAPI (Sigma, 10236276001), wholemounted in Vectashield medium (Vector, H1000) and imaged with a confocal microscope (Leica TCS SP8).
In sharks, the excellent sense of smell (Tricas et al., 2009) has been associated to the very large size of the olfactory lamellae, following the simple assumption that an increase in surface area will allow for a more effective odor sampling (Theiss et al., 2009). To start testing whether a large OE size, as observed in CF, confers better olfactory sensory abilities, we used a previously designed olfactory test, with amino-acids as odorant molecules. Of note and importantly, this assay allows detecting olfactory-driven responses, and the taste system, which is also expanded in CF (Varatharasan et al., 2009; Yamamoto et al., 2009), is not involved in the behavioral responses (Hinaux et al., 2016). With this test, one month-old SF respond to 10−5 M alanine while one month-old CF respond to 10−10 M alanine (Hinaux et al., 2016). As OE size increases with age due to fish growth (Fig. 1A), here we evaluated the olfactory capacities of older and larger fish. Two months-old SF have larger OEs in absolute size than younger and smaller individuals, and their relative OE size is also larger to 1 month old SF, due to positive allometric growth of the OE (Fig. 1A). However, they showed the same detection threshold for alanine as one month old SF: they did not respond to 10−6 M alanine, nor to 10−7 M (n = 11, not shown).
2.6. In situ hybridization Total RNA from Astyanax embryos was reverse-transcribed using iScript cDNA synthesis kit (BioRad). Partial cDNA sequences for OMP (GenBank ID KP826791.1), TRPC2 (Genbank ID XM_022671326.1; predicted LOC111192950), and OR-52Z1-like (XM_022687171.1; predicted LOC103043790) were amplified by PCR using specific
3
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 1. OE size and olfactory capacities. A, OE growth in SF. Increase in absolute OE size (left) and relative OE size (right) according to growth in SF between 1 month (blue squares) and 2 months (blue triangles). SL, standard length. B, Comparison of relative OE size in the indicated morphs at the indicated ages/sizes. Numbers in brackets on bars indicate the numbers of fish used for quantifications, and data are mean ± sem. Mann-Whitney tests. C, Behavioral set-up. The test was performed in the dark with infrared recordings. Amino acids or EM were perfused on either side of boxes with a semi-partition. The amino acids reached the box 1.5 min after flow was initiated from the perfusion system. The position of the 4 fish was scored before and during odor exposure. Four quadrants were defined in the boxes, in which the amino acid concentration was very high, high, low, or zero over the duration of the test, see also (Hinaux et al., 2016). Each quadrant was attributed a coefficient (+3; +1; −1, −3, respectively) to calculate a preference index, which reflects the fish response (or absence of response) to the perfused amino acid. As an example, the preference index corresponding to the distribution of the fish shown on the schema would be zero. D, Olfactory response of F1 hybrids (female SF x male Pachón cross) to indicated concentrations of Alanine, represented as the preference index as a function of time. The vertical arrow indicates the time when the odorant reaches the box. A positive preference index (bars towards positive values) with significance as compared to zero (asterisks) indicates attraction to the odorant. The conditions and numbers of tests are indicated (n = 1 corresponds to one test, i.e., the cumulative score of 4 fish). Asterisks indicate significant response as compared with zero (Mann-Whitney tests).
deprived modality can be recruited by the remaining, intact sensory modalities. This phenomenon is well documented in the visual neuroscience literature, in which blindness results in the recruitment of visual cortex for somatosensory and auditory processing (Lazzouni and Lepore, 2014 for a review). Here, we sought to test whether high olfactory performance in CF is due to loss of eyes/vision and/or its pleiotropic enlargement of the naris. Indeed, anatomical changes in cranio-facial structures in CF can be divided into eye-dependent and eye-independent processes (Dufton et al., 2012; Yamamoto et al., 2003). An eye-dependent character is the size of the naris in adults: SF which have undergone unilateral lens ablation while embryos and therefore are eyeless on one side of the head as adults, have a wider olfactory pit (+ 12.9%) on the side with a degenerate eye than on the side of the normal eye. Moreover, phenotypic plasticity due to environmental conditions (darkness) may also be involved in the development of CF olfactory performances. To test this possibility, we performed bilateral lens ablations on SF embryos, which results in progressive eye degeneration and mimics the CF blind phenotype (Yamamoto and Jeffery, 2000; Yamamoto and Jeffery, 2002) (Fig. 2A). To distinguish between the potential effects of the loss of eyes, from those due to functional blindness by absence of vision, we raised the lens-ablated SF and their non-ablated SF siblings in complete darkness from the time of embryos collection (usually at about 5–10 h post-fertilization) to the age of one month. Olfactory tests were performed using 10−6 M alanine or Serine stocks, i.e., below the threshold concentration for normal, light-raised SF. Both lens-ablated/
We also tested one-month old F1 hybrids resulting from crosses between a female SF and a Pachón male. F1 hybrids possessed large olfactory epithelia comparable to CF (Fig. 1B) but showed olfactory capabilities comparable to SF. They responded to Alanine 10−3 M (n = 13, not shown) and 10−4 M, but not to 10−5 M (Fig. 1CD). We however noted that F1 hybrids swam very actively all along the test (for Alanine 10−5 M, distance swam: 5.07-fold/SF, p = 0.001; 4.87-fold/CF, p = 0.002, n = 9 each; Mann Whitney tests). This may explain their absence of sustained attraction towards the odorant compartment at SF threshold concentration (10−5 M), due to mixing of the odor in the test box by intense swimming activity. Thus, the olfactory capacities of F1 hybrids were not intermediate between those of their two parents and were far less good than those of CF even though they have comparably large OEs. Taken together, these data suggested that 1) olfactory skills correspond to a recessive or complex genetic trait, 2) OE size is not strictly correlated to olfactory capacities. The larger OE size of CF, providing them with a larger number of olfactory sensory neurons, is not the sole reason for the improved sensory trait. Below we tested two other, non-mutually exclusive developmental mechanisms that may underlie olfactory capacities in CF. 3.2. Cavefish olfactory capacities and developmental plasticity Cross-modal plasticity can occur as a result of decreased or abnormal sensory input, and is well described for example in the mammalian cortex: the cortical regions normally processing the 4
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 2. Abrogation of visual function in SF improves olfactory performance but does not affect OE size. A, Example of unilateral lens ablation performed at embryonic stage and resulting in eye degeneration in SF. Top photos show the eyes of a one month old SF after ablation on the left side. Bottom photos show Gαolf immunofluorescence used to quantify OE size after lens ablation. B, C, Responses of dark-raised and dark-raised/lens-ablated SF to the below-threshold concentration of Serine 10−6M (C) and Alanine 10−6M (D). Both types of visually-deprived SF show significant attractive response, contrary to normal SF. Ages of the tested fish and number of behavioral tests performed are indicated. D, Quantification of eye and OE circumferences, normalized to the standard length, in control (n = 16) and dark-raised (n = 18) one month old SF. Numbers in brackets on bars indicate the numbers of fish used for quantifications. Values are mean ± sem. For eye size p = 0.308 and for OE size p = 0.704 (Mann-Whitney tests, NS). E, Quantification of eye and OE circumferences (arbitrary units) on the lens-ablated side versus the contralateral control side of one month old SF (n = 5). Numbers in brackets on bars indicate the numbers of fish used for quantifications. Values are mean ± sem. * * indicates p = 0.0079. For OE size p = 0.841 (Mann-Whitney tests). F, Photographs of 64hpf larval heads after in situ hybridization for OMP (purple) labeling the OE, in frontal views. SF is on the left, CF is on the right. The scale bar is 100 µm.
difference in eye size between the two morphs is moderate: already at this stage, the OE labelled by expression of OMP (Olfactory Marker Protein) was significantly larger in CF than in SF (Fig. 2F; OE circumference standardized to length (a.u.): 0.414 ± 0.017 in SF versus 0.499 ± 0.043 in CF, n = 5 each, p = 0.008, Mann-Whitney). In sum, this series of experiments suggested two components in the enhanced olfactory capacities already present in one month old cave Astyanax: (1) one part of phenotypic plasticity, which can be recruited in SF in conditions of visual deprivation, and which does not depend on the size of the olfactory organ and (2) one part of evolution of the olfactory system in CF, probably related to OE development, and which is eye-independent.
dark-raised SF and dark-raised SF showed attraction to the odor source (Fig. 2BC). The same results were obtained for the two amino-acids: dark-raised SF responded well to both Alanine 10−6 M (Fig. 2B) and Serine 10−6 M (Fig. 2C). This improved olfactory performance was also maintained in older and larger dark-raised 2 months old individuals (Fig. 2C). These experiments showed that eye- and vision-dependent developmental processes are factors promoting enhanced olfactory skills in A. mexicanus. As elimination of visual function (by raising the fish in the dark) and elimination of the visual organ (by lens ablation) both resulted in improved olfactory performance in SF, we attributed this change to developmental functional phenotypic plasticity between olfactory and visual sensory modalities. Adult SF which have undergone unilateral lens ablation as embryos have a wider olfactory pit on the side with a degenerate eye than on the side of the normal eye (Yamamoto et al., 2003). At the stage assayed in the present study, such a difference in olfactory pit size was not yet established. Fig. 2D–E shows that neither lens ablation nor life in darkness affected the size of the OE, as measured on one month old SF. Thus, improved olfactory capabilities in visually-deprived SF at this age cannot be attributed to an increase in the size of their olfactory organ. This is in line with the above findings showing that OE size by itself is not strictly predictive for olfactory performance. Furthermore, these data supported that in CF, the OE does not grow larger as a consequence of the eye degeneration process, but rather that its size is autonomously developmentally-controlled and proportionately enlarged when compared to SF (Hinaux et al., 2016). This notion was further supported by measurements of OE sizes in SF and CF embryos/ larvae at 2.5 days of development (one day after hatching (Hinaux et al., 2011)), before the onset of eye degeneration in CF, and when the
3.3. Cavefish olfactory capacities and developmental modulation of OE neurogenesis To search for developmental and cellular mechanisms underlying the continued size differences between SF and CF OEs as well as the differences in olfactory skills between the two morphs, we next compared the proliferation, cell death, lifespan and neurogenesis patterns in the developing OEs of the two morphs along development. In all vertebrates, olfactory neurons of the OE are continuously renewed along the life of the animals. Therefore, differences in the rate of cell death and/or the lifespan of olfactory neurons might account for the OE differences between SF and CF. To test the first possibility (cell death), the numbers of activated caspase3-positive apoptotic cells were compared in the OE of the two morphs. At the stages examined, apoptotic cells were detected in low numbers. They corresponded to newly differentiated neurons as they were Hu-positive, and they were 5
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 3. Comparison of apoptosis and neuronal lifetime in SF and CF olfactory epithelium. A, Confocal images of double immunofluorescence against activated caspase-3 (green) and Hu (blue) in SF (left) and CF (right) at 84hpf. DAPI nuclear counterstain is magenta. Z-projections on a 10 µm depth are shown. The dotted lines show the contours of the OE, adjacent to the brain. Green arrows indicate activated Caspase3-positive apoptotic neurons, note also their condensed nuclei. White asterisks point to non-specific labeling on the cilia and mucus of olfactory sensory neurons. Scale bars: 25 µm. B, Quantification of apoptotic cells at different developmental stages in SF (blue bars) and CF (red bars). Mann-Whitney tests (n = 10–28, all NS). C, Confocal images of double labeling for Edu-positive cells (red) and Hu immunofluorescence (blue) in SF (left) and CF (right) after an EdU pulse performed at 64hpf, followed by a 5 weeks chase. Z-projections of frontal views on a 10 µm depth are shown. Red arrows point to EdU-positive nuclei. Blue asterisks indicate the central raphe that has formed at this stage after complex morphogenetic movements of the olfactory neuroepithelium. Insets show high magnification. Note the different scale bars (50 µm) for the 2 morphs, as well as the round versus oval shape of the OE in SF and CF, respectively, as previously reported in (Hinaux et al., 2016). D, Quantification of the number of Edu-positive neurons remaining in the olfactory epithelium after a 5 weeks chase in SF (n = 7) and CF (n = 8). Mann-Whitney tests; ** p = 0.0038.
Differences in proliferation and neurogenesis might also account for the SF/CF differences. To measure proliferation, EdU incorporation pulses were performed at various stages (between 40hpf and 21dpf) and larvae were fixed immediately at the end of the pulse. EdU-positive cells, corresponding to cells undergoing the S phase of mitosis during the pulse, were counted, and the volumes of the OEs were measured (at these stages, the OE still has a round/ovoid shape that allows accurate volume calculations). The results were expressed as densities, to normalize for OE volume size differences between the two morphotypes and to obtain a proxy for OE proliferation rate. With the exception of 40hpf, the density of EdU-positive proliferative cells was identical in the two morphs at all stages examined from 84hpf to 21dpf (Fig. 4AB). This shows that proliferation is proportionately similar in SF and CF OEs and further suggests that a proportionately similar pool of embryonic progenitors and later stem cells is defined in the olfactory placodes of SF and CF. Of note, the densities of progenitors appear stable between 84hpf and 21dpf on the graph shown in Fig. 4B: this is only due to the fact that the EdU pulse was performed during 30 min for the stages 40hpf and 84hpf, while it lasted 1 h for the stages 7dpf and 21dpf (see Methods). This was designed to label a reasonable number of cells to be counted at these older stages (and therefore to reduce errors), and led to the masking of the actual decrease in progenitors/stem cell densities after embryonic stages, which is indeed expected. Next, the possibility of variations in cell cycle and/or progenitor properties was investigated. EdU pulses were performed at 4dpf, followed by chases of 6 h or 30 h (no prior estimation of the cell cycle length in A. mexicanus at this stage were available). Larvae were fixed and double-labelled for EdU (identifying cells that underwent the S
found in equal amounts in SF and CF (Fig. 3AB and Suppl. Fig. 1). To test the second possibility (lifespan), an EdU pulse was performed at 64hpf, followed by chases of various lengths, from 1 to 5 weeks (Fig. 3CD). After short chases, very numerous EdU-positive cells were observed (not counted), corresponding to proliferative progenitors and to post-mitotic neurons which underwent their final mitosis during the pulse. For longer chases, the number of EdU-positive cells decreased similarly in SF and CF: this corresponded to the dilution of EdU in proliferative progenitors and to the disappearance of post-mitotic neurons. Finally, for 5 weeks chases, less than 30 EdU-positive/Hupositive cells were counted in the OE of both SF and CF (Fig. 3CD). This result suggested that the lifespan of olfactory neurons is similar in the two morphs and is about 4–5 weeks maximum. This lifetime is equivalent to that reported for zebrafish (29 days; Bayramli et al., 2017), but shorter than in the mouse (30–90 days; Mackay-Sim and Kittel, 1991; Santoro and Dulac, 2012). We suggest that the more numerous neurons found in CF after a 5 weeks chase is simply due to the fact that their OE is larger, and therefore that more progenitors were labelled during the pulse (see also below). Contrary to the data presented in the next section, it was not possible at this stage to measure exact OE volumes and calculate cell densities to normalize the data, due to complex morphogenesis of the OE neuroepithelium including the formation of a central raphe (blue asterisks on Fig. 3C). However, in line with our interpretation, after normalization to the total external volume of the OE (i.e., not taking into account the neuroepithelial folds and empty internal spaces), the densities of EdUpositive cells after a 5 weeks chase appeared identical in the two morphs (SF: 0.14 ± 0.04, n = 7; CF: 0.12 ± 0.02, n = 8; MannWhitney test).
6
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 4. Comparison of neurogenesis and progenitors properties in SF and CF olfactory epithelium. A, Confocal images of EdU fluorescent staining (blue) with DAPI nuclear counterstain (magenta) in SF (left) and CF (right) at 84hpf. Z-projections on a 10 µm depth are shown. Scale bars: 25 µm. B, Quantification of the density of Edu-positive cells after a pulse and immediate fixation at various developmental stages in SF (blue) and CF (red) (n = 11–21). Densities are given in number of cells per 105 µm3 volume. Mann-Whitney tests. C, D, Confocal images of EdU fluorescent staining (blue) and PCNA immunofluorescence (green) with DAPI nuclear counterstain (magenta) in SF (left) and CF (right) after an EdU pulse followed by a 6 h (C) or a 30 h (D) chase. Z-projections on a 10 µm depth are shown. Scale bars: 25 µm. E, F, Quantification of the density of PCNA-positive cells at 90hpf and 114hpf and the density of double-labelled EdU+/PCNA+ cells after a 6 h or a 30 h chase in SF (blue) and CF (red) (n = 8 for each). Densities are given in number of cells per 105 µm3 volume.
of EdU+ cells, represented 55 ± 9% in SF (n = 8) and 50 ± 7% in CF (n = 8) after a 30 h chase. This series of experiments suggested that the proliferative properties and behaviors of olfactory neurons progenitors were identical in the OEs of the SF and CF. Finally, we compared the time of appearance and proportions of differentiated olfactory sensory neuron (OSN) cell types between the two morphs. The 3 main OSN types described in fish were studied, including the ciliated neurons (OMP-positive), the microvillous neurons (TRPC2positive) and the crypt cells (S100-positive) (Germana et al., 2004; Hansen et al., 2004; Sato et al., 2005)(Fig. 5A). Kappe neurons, a fourth and sparse subtype of OSN (Ahuja et al., 2014), were not studied. OMP-expressing ciliated neurons were detected at 24hpf (hatching stage) in both CF and SF (about 15–20 cells) (Fig. 5B). The first TRPC2-expressing microvillous neurons were also detectable at 24hpf (Fig. 5C), while the S100-positive crypt cells differentiated later, around 84hpf (Fig. 5D). Thus, no difference in the time-course of OSN differentiation was detected between the two morphs. We next counted neuron numbers and calculated densities for each neuronal subtype. At 84hpf, the density of OMP-positive ciliated neurons was higher in SF (Fig. 5F and K). Conversely, the density of TRPC2-positive microvillous neurons was higher in CF (Fig. 5E and K). In
phase of mitosis during the pulse) and PCNA (Proliferating Cell Nuclear Antigen, labeling all phases of the cell cycle, thus identifying the entire population of cycling progenitors at the time of fixation) (Fig. 4C–G). For both chase times, i.e., at 4dpf and 5dpf, the density of PCNA-positive cells was identical in SF and CF (Fig. 4E), in line with the above results showing identical densities of proliferative progenitors captured in the S phase in the OEs of the two morphs after an Edu pulse (Fig. 4AB). For both chase times also, the densities of double EdU+/PCNA+ cells were identical in SF and CF (Fig. 4F). For the 6 h chase, this population of double-labelled cells likely corresponded to cells which were either finishing their cycle, or had already entered a new cycle. For the 30 h chase, this population likely corresponded only to cells which had re-entered a new cycle. These results suggested that cell cycle kinetics, including cell cycle length, are identical in the two Astyanax morphs. Accordingly, the percentage of cell cycle re-entry and the percentage of cell cycle exit were identical in SF and CF: cells re-entering the cycle, calculated as number of EdU+/PCNA+ doublelabelled cells/total number of EdU+ cells, represented 45 ± 9% in SF (n = 8) and 50 ± 7% in CF (n = 8) after a 30 h chase; and cells having left the cycle, calculated as number of EdU+/PCNA- cells/total number
7
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
8
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Fig. 5. Comparison of olfactory sensory neurons differentiation and composition in SF and CF olfactory epithelium. A, Schema depicting the 3 main types of OSNs described in fish. Adapted from (Oka et al., 2012). B, C, Bright field photographs showing in situ hybridization (purple) for OMP (B) and TRPC2 (C) on 24hpf SF and CF in toto on dorsal views. The OE is delineated by a dotted line. Scale bars: 25 µm. D, Confocal images of S100 immunofluorescence staining (blue) with DAPI nuclear counterstain (magenta) in 84hpf SF and CF, on frontal views. Z-projections on a 10 µm depth are shown. Arrows point to S100-positive neurons. Scale bars: 25 µm. E, F, Confocal images of fluorescent in situ hybridization for TRPC2 and OMP (green) with DAPI nuclear counterstain (magenta) in 84hpf SF and CF, on frontal views. Z-projections on a 5 µm depth are shown. Arrows point to expressing neurons. Scale bars: 25 µm. G, H, Bright field photographs showing in situ hybridization (purple) for calciphosine-like (G) and odorant receptor OR-52Z1-like (H) on 84hpf SF and CF in toto on frontal views. Scale bars: 25 µm. I, J, Confocal images of immunofluorescence for S100 (I) and TRPC2 (J, green) with DAPI nuclear counterstain (magenta) in 21dpf SF and CF, on frontal views. Scale bars: 20 µm. K, Quantification of OSN subtypes densities across developmental time in SF (blue) and CF (red). Densities are given in number of cells per 105 µm3 volume. Mann-Whitney tests (n = 8–27 for each point). nd, not determined.
organ, the olfactory epithelium. However, differences in processing in the olfactory bulbs or higher brain regions may also be relevant to explain the differences in olfactory skills between the two morphs, particularly concerning developmental phenotypic plasticity. In early blind human individuals for example, superior performance in odorprocessing tasks were shown to be associated with an activation of the occipital cortex that was not found in sighted subjects (Renier et al., 2013) and to a larger size of the olfactory bulbs (Araneda et al., 2016; Rombaux et al., 2010). In cavefish also, the size of the olfactory bulbs seems larger than in surface fish (Pottin et al., 2011; JB and SR, unpublished data). However the possibility that in cavefish, brain regions which normally receive visual inputs start receiving olfactory inputs has not been investigated yet. In fish, the olfactory and visual systems are functionally linked by the terminal nerve system, or olfacto-retinalis system. It is known that olfactory system stimulations can modify visual responses (Stephenson et al., 2012), but whether the absence of visual inputs can modify the olfactory response has never been directly demonstrated. However, the terminal nerve receives afferences from visual structures (Yamamoto and Ito, 2000), the activity of terminal nerve GnRH neurons can be modulated by visual cues (Ramakrishnan and Wayne, 2009), and GnRH can modulate OSN excitability and responses to odors in amphibians (Eisthen et al., 2000; Park and Eisthen, 2003; Zhang and Delay, 2007). Thus, one possible mechanism for increased olfaction in visually-deprived surface fish and /or in cavefish may be through the terminal nerve system. It will also be essential to understand the underlying neurophysiological mechanisms of improved cavefish olfaction, which may be manifold given the huge difference in olfactory detection skills between the two Astyanax mexicanus morphs. Such mechanisms probably include vision-dependent plasticity processes, that we have unmasked in the present study in SF after visual deprivation, and that may have been fixed rapidly by genetic assimilation in the cavefish lineage after environmental change (Waddington, 1953). Vision-independent developmental evolutionary processes also seem to be involved, including those controlling the embryonic size of the olfactory organ (Hinaux et al., 2016) and its neuronal composition (present paper). The evolution of the G protein-coupled olfactory receptor families, representing about 300 genes in fish (Hashiguchi et al., 2008; Korsching, 2009) and which are among the fastest-evolving genes in the genomes, may contribute as well. Finally in vertebrates, OSN sensitivity and odor processing in the olfactory bulbs can be modulated by diverse sources and neurotransmitters including serotonin (Frings, 1993; Petzold et al., 2009). It might be the case in cavefish, as a mutation in the serotonin degrading enzyme (monoamine oxidase) confers them with high brain serotonin levels (Elipot et al., 2014a). The proportionately large size of the OE in cavefish may partly account for its enhanced amino-acid detection skills, but is not the only determinant. Indeed, old SF or F1 hybrids with large OEs have modest detection thresholds, while visually-deprived SF with unchanged OE size show improved olfactory capacities. Rather than the size of the mature OE itself, the important parameter may be the size and early patterning of the embryonic olfactory placode in cavefish. If the olfactory sensory territory that is gained over adjacent placodes gives rise to progenitor pools with specific neurogenic properties which bias the neuronal composition of the later OE, then this placodal expansion may translate into important functional outcomes.
addition at 84hpf, a markedly different distribution of the labelled OSNs was observed inside the OEs of the two morphs. In SF, both ciliated (OMP) and microvillous (TRPC2) OSNs were located at the margin of the OE, forming a ring in frontal views. In CF however, OSNs were distributed at the center of the olfactory cup (Fig. 5EF). This result was confirmed with other, independent differentiated OSN markers, such as calciphosine-like (Capsl, a calcium-binding protein) or OR-52Z1-like, a member of the OR family of odorant receptors (Fig. 5GH). This difference is difficult to interpret in terms of neurogenesis because there was not obvious dissimilarity in the localization of the progenitors between SF and CF: EdUretaining cells and PCNA-positive cells were located at the basis of the epithelium in both cases (Fig. 4). Moreover, this difference in the pattern of OSN distribution was transient. Indeed at 21dpf, neurons were evenly distributed at the surface of the OE, both in SF and CF (Fig. 5IJ). At this later stage also, the density of TRPC2+ microvillous neurons was higher in CF, while the density of S100+ crypt cells was identical in the two morphs (Fig. 5IJK; note that ciliated neurons, representing the major type of OSN (90% in zebrafish) were not counted at 21dpf). In sum, these comparative developmental neuroanatomy data show that the densities and distributions of OSN subtypes vary in the OE of SF and CF. Specifically, the microvillous neurons types are proportionately more represented in CF, possibly at the expense of ciliated neurons. In zebrafish, trout or goldfish, microvillous neurons respond to alimentary odors such as amino-acids, while ciliated neurons are more generalists and also respond to social cues such as bile acids or alarm substance, and crypt cells respond to sexual pheromones (Koide et al., 2009; Sato and Suzuki, 2001; Yoshihara, 2008). The 3 OSN subtypes also express different types of odorant receptors and project to distinct glomeruli in the olfactory bulbs (Ahuja et al., 2013; Hamdani el and Doving, 2007; Hansen et al., 2004; Koide et al., 2009; Oka et al., 2012; Sato et al., 2005; Yoshihara, 2008). In Astyanax mexicanus, OSN specificity towards odor types and their involvement in specific behaviors has not been studied yet, but it is tempting to speculate about the contribution of more numerous, amino-acid-responsive, TRPC2-positive microvillous neurons in the enhanced olfactory skills demonstrated by cavefish for amino-acid detection during behavioral assays. Future studies will aim at functionally testing this hypothesis. Altogether, our comparative analyses of OE development in SF and CF demonstrated that general proliferative properties of progenitors are identical but neurogenic properties differ in the two morphs. In cavefish, progenitors seem specified and fated to generate proportionately more microvillous neurons, pushing back the origins of the SF/CF differences at the level of the early placodal ectoderm. Indeed, cell-type heterogeneity in the zebrafish olfactory epithelium is generated from progenitors within the placodal ectoderm (Aguillon et al., 2018), but the mechanisms underlying the segregation of the various olfactory subtypes from overlapping progenitor pools are unknown. In cavefish, the olfactory placode territory is expanded, probably at the expense of adjacent placodal territories including the presumptive lens region (Hinaux et al., 2016). Therefore, cavefish appear like an excellent evo-devo model to perform lineage studies and fate maps to address the question of the establishment and patterning of olfactory neurons progenitor pools in the early placode. 4. Conclusions To start investigating the developmental origins of the excellent olfactory capacities in CF, we have focused on their external sensory 9
Developmental Biology xxx (xxxx) xxx–xxx
M. Blin et al.
Koide, T., et al., 2009. Olfactory neural circuitry for attraction to amino acids revealed by transposon-mediated gene trap approach in zebrafish. Proc. Natl. Acad. Sci. USA 106, 9884–9889. Korsching, S., 2009. The molecular evolution of teleost olfactory receptor gene families. Results Probl. Cell Differ. 47, 37–55. Lazzouni, L., Lepore, F., 2014. Compensatory plasticity: time matters. Front. Hum. Neurosci. 8, 340. Mackay-Sim, A., Kittel, P.W., 1991. On the life span of olfactory receptor neurons. Eur. J. Neurosci. 3, 209–215. Maier, E.C., et al., 2014. Sensational placodes: neurogenesis in the otic and olfactory systems. Dev. Biol. 389, 50–67. Miyasaka, N., et al., 2013. Functional development of the olfactory system in zebrafish. Mech. Dev. 130, 336–346. Oka, Y., et al., 2012. Crypt neurons express a single V1R-related ora gene. Chem. Senses 37, 219–227. Park, D., Eisthen, H.L., 2003. Gonadotropin releasing hormone (GnRH) modulates odorant responses in the peripheral olfactory system of axolotls. J. Neurophysiol. 90, 731–738. Petzold, G.C., et al., 2009. Serotonergic modulation of odor input to the mammalian olfactory bulb. Nat. Neurosci. 12, 784–791. Pottin, K., et al., 2011. Restoring eye size in Astyanax mexicanus blind cavefish embryos through modulation of the Shh and Fgf8 forebrain organising centres. Development 138, 2467–2476. Ramakrishnan, S., Wayne, N.L., 2009. Social cues from conspecifics alter electrical activity of gonadotropin-releasing hormone neurons in the terminal nerve via visual signals. Am. J. Physiol. Regul. Integr. Comp. Physiol. 297, R135–R141. Renier, L., et al., 2013. Right occipital cortex activation correlates with superior odor processing performance in the early blind. PLoS One 8, e71907. Rombaux, P., et al., 2010. Increased olfactory bulb volume and olfactory function in early blind subjects. NeuroReport 21, 1069–1073. Santoro, S.W., Dulac, C., 2012. The activity-dependent histone variant H2BE modulates the life span of olfactory neurons. Elife 1, e00070. Sato, K., Suzuki, N., 2001. Whole-cell response characteristics of ciliated and microvillous olfactory receptor neurons to amino acids, pheromone candidates and urine in rainbow trout. Chem. Senses 26, 1145–1156. Sato, Y., et al., 2005. Mutually exclusive glomerular innervation by two distinct types of olfactory sensory neurons revealed in transgenic zebrafish. J. Neurosci. 25, 4889–4897. Shao, X., et al., 2017. Olfactory sensory axons target specific protoglomeruli in the olfactory bulb of zebrafish. Neural Dev. 12, 18. Stephenson, J.F., et al., 2012. Food and conspecific chemical cues modify visual behavior of zebrafish, Danio rerio. Zebrafish 9, 68–73. Theiss, S.M., et al., 2009. Morphological indicators of olfactory capability in Wobbegong sharks (Orectolobidae, Elasmobranchii). Brain Behav. Evol. 73, 91–101. Torres-Paz, J., Whitlock, K.E., 2014. Olfactory sensory system develops from coordinated movements within the neural plate. Dev. Dyn. 243, 1619–1631. Tricas, T.C., et al., 2009. Response of the hammerhead shark olfactory epithelium to amino acid stimuli. J. Comp. Physiol. A Neuroethol. Sens. Neural Behav. Physiol. 195, 947–954. Varatharasan, N., et al., 2009. Taste bud development and patterning in sighted and blind morphs of Astyanax mexicanus. Dev. Dyn. 238, 3056–3064. Vitebsky, A., et al., 2005. Isolation and characterization of the laure olfactory behavioral mutant in the zebrafish, Danio rerio. Dev. Dyn. 234, 229–242. Wackermannova, M., et al., 2016. Olfactory sensitivity in mammalian species. Physiol. Res. 65, 369–390. Waddington, C.H., 1953. Genetic assimilation of an acquired character. Evolution 7, 118–126. Wakisaka, N., et al., 2017. An Adenosine Receptor for Olfaction in Fish. Curr. Biol. 27, 1437–1447, (e4). Whitlock, K.E., 2004. A new model for olfactory placode development. Brain Behav. Evol. 64, 126–140. Whitlock, K.E., 2006. The sense of scents: olfactory behaviors in the zebrafish. Zebrafish 3, 203–213. Whitlock, K.E., Westerfield, M., 1998. A transient population of neurons pioneers the olfactory pathway in the zebrafish. J. Neurosci. 18, 8919–8927. Yamamoto, N., Ito, H., 2000. Afferent sources to the ganglion of the terminal nerve in teleosts. J. Comp. Neurol. 428, 355–375. Yamamoto, Y., et al., 2009. Pleiotropic functions of embryonic sonic hedgehog expression link jaw and taste bud amplification with eye loss during cavefish evolution. Dev. Biol. 330, 200–211. Yamamoto, Y., et al., 2003. Development and evolution of craniofacial patterning is mediated by eye-dependent and -independent processes in the cavefish Astyanax. Evol. Dev. 5, 435–446. Yamamoto, Y., Jeffery, W.R., 2000. Central role for the lens in cave fish eye degeneration. Science 289, 631–633. Yamamoto, Y., Jeffery, W.R., 2002. Probing teleost eye development by lens transplantation. Methods 28, 420–426. Yamamoto, Y., et al., 2004. Hedgehog signalling controls eye degeneration in blind cavefish. Nature 431, 844–847. Yoshihara, Y., 2008. Molecular genetic dissection of the zebrafish olfactory system. Results Probl. Cell Differ., 97–119. Zhang, W., Delay, R.J., 2007. Gonadotropin-releasing hormone modulates voltageactivated sodium current and odor responses in Necturus maculosus olfactory sensory neurons. J. Neurosci. Res. 85, 1656–1667.
Acknowledgements We thank Stéphane Père and Victor Simon for care of our Astyanax breeding colony, and Muriel Perron for interesting discussions about progenitor proliferation assays. This Work was supported by an ANR (Agence Nationale pour la Recherche) grant [BLINDTEST] and an « Equipe FRM » grant [DEQ. 20150331745] from the Fondation pour la Recherche Médicale. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ydbio.2018.04.019. References Aguillon, R., et al., 2018. Cell-type heterogeneity in the early zebrafish olfactory epithelium is generated from progenitors within preplacodal ectoderm. Elife 7. Ahuja, G., et al., 2014. Kappe neurons, a novel population of olfactory sensory neurons. Sci. Rep. 4, 4037. Ahuja, G., et al., 2013. Zebrafish crypt neurons project to a single, identified mediodorsal glomerulus. Sci. Rep. 3, 2063. Araneda, R., et al., 2016. Cortical plasticity and olfactory function in early blindness. Front. Syst. Neurosci. 10, 75. Bayramli, X., et al., 2017. Patterned arrangements of olfactory receptor gene expression in zebrafish are established by radial movement of specified olfactory sensory neurons. Sci. Rep. 7, 5572. Behrens, M., et al., 2014. ORA1, a zebrafish olfactory receptor ancestral to all mammalian V1R genes, recognizes 4-hydroxyphenylacetic acid, a putative reproductive pheromone. J. Biol. Chem. 289, 19778–19788. Bibliowicz, J., et al., 2013. Differences in chemosensory response between eyed and eyeless Astyanax mexicanus of the Rio Subterraneo cave. Evodevo 4, 25. Breau, M.A., Schneider-Maunoury, S., 2014. Mechanisms of cranial placode assembly. Int. J. Dev. Biol. 58, 9–19. Caprio, J., et al., 1989. Electro-olfactogram and multiunit olfactory receptor responses to binary and trinary mixtures of amino acids in the channel catfish, Ictalurus punctatus. J. Gen. Physiol. 93, 245–262. Collin, S.P., 2012. The neuroecology of cartilaginous fishes: sensory strategies for survival. Brain Behav. Evol. 80, 80–96. Corfield, J.R., et al., 2014. Anatomical specializations for enhanced olfactory sensitivity in kiwi, Apteryx mantelli. Brain Behav. Evol. 84, 214–226. Dufton, M., et al., 2012. Early lens ablation causes dramatic long-term effects on the shape of bones in the craniofacial skeleton of Astyanax mexicanus. PLoS One 7, e50308. Eisthen, H.L., et al., 2000. Neuromodulatory effects of gonadotropin releasing hormone on olfactory receptor neurons. J. Neurosci. 20, 3947–3955. Elipot, Y., et al., 2014a. A mutation in the enzyme monoamine oxidase explains Part of the Astyanax cavefish behavioral syndrome. Nat. Commun. Apr 10 (5), 3647. Elipot, Y., et al., 2014b. Astyanax transgenesis and husbandry: how cavefish enters the lab. Zebrafish 11 (4), 291–299. Espinasa, L., et al., 2014. Enhanced prey capture skills in Astyanax cavefish larvae are independent from eye loss. Evodevo 5, 35. Friedrich, R.W., Korsching, S.I., 1997. Combinatorial and chemotopic odorant coding in the zebrafish olfactory bulb visualized by optical imaging. Neuron 18, 737–752. Frings, S., 1993. Protein kinase C sensitizes olfactory adenylate cyclase. J. Gen. Physiol. 101, 183–205. Germana, A., et al., 2004. S100 protein-like immunoreactivity in the crypt olfactory neurons of the adult zebrafish. Neurosci. Lett. 371, 196–198. Grocott, T., et al., 2012. The peripheral sensory nervous system in the vertebrate head: a gene regulatory perspective. Dev. Biol. 370, 3–23. Hamdani el, H., Doving, K.B., 2007. The functional organization of the fish olfactory system. Prog. Neurobiol. 82, 80–86. Hansen, A., et al., 2004. Differential distribution of olfactory receptor neurons in goldfish: structural and molecular correlates. J. Comp. Neurol. 477, 347–359. Hara, T.J., 1994. Olfaction and gustation in fish: an overview. Acta Physiol. Scand. 152, 207–217. Hara, T.J., 2006. Feeding behaviour in some teleosts is triggered by single amino acids primarily through olfaction. J. Fish Biol. 68, 810–825. Hashiguchi, Y., et al., 2008. Evolutionary patterns and selective pressures of odorant/ pheromone receptor gene families in teleost fishes. PLoS One 3, e4083. Hinaux, H., et al., 2016. Sensory evolution in blind cavefish is driven by early events during gastrulation and neurulation. Development 143, 4521–4532. Hinaux, H., et al., 2011. A developmental staging table for Astyanax mexicanus surface fish and Pachon cavefish. Zebrafish 8, 155–165. Inoue, D., Wittbrodt, J., 2011. One for all–a highly efficient and versatile method for fluorescent immunostaining in fish embryos. PLoS One 6, e19713. Keller-Costa, T., et al., 2015. Chemical communication in cichlids: a mini-review. Gen. Comp. Endocrinol. 221, 64–74.
10