SEARCH

SEARCH BY CITATION

Keywords:

  • cnidarian;
  • retinoic acid;
  • nitric oxide;
  • neuronal differentiation

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Retinoic acid (RA) and nitric oxide (NO) are known to promote neuronal development in both vertebrates and invertebrates. Retinoic acid receptors appear to be present in cnidarians and NO plays various physiological roles in several cnidarians, but there is as yet no evidence that these agents have a role in neural development in this basal metazoan phylum. We used primary cultures of cells from the sea pansy Renilla koellikeri to investigate the involvement of these signaling molecules in cnidarian cell differentiation. We found that 9-cis RA induce cell proliferation in dose- and time-dependent manners in dishes coated with polylysine from the onset of culture. Cells in cultures exposed to RA in dishes devoid of polylysine were observed to differentiate into epithelium-associated cells, including sensory cells, without net gain in cell density. NO donors also induce cell proliferation in polylysine-coated dishes, but induce neuronal differentiation and neurite outgrowth in uncoated dishes. No other cell type undergoes differentiation in the presence of NO. These observations suggest that in the sea pansy (1) cell adhesion promotes proliferation without morphogenesis and this proliferation is modulated positively by 9-cis RA and NO, (2) 9-cis RA and NO differentially induce neuronal differentiation in nonadherent cells while repressing proliferation, and (3) the involvement of RA and NO in neuronal differentiation appeared early during the evolutionary emergence of nervous systems. © 2010 Wiley Periodicals, Inc. Develop Neurobiol 70: 842–852, 2010


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Cnidarians constitute the first animal phylum in which representatives possess an organized nervous system and neurally supported behaviors. While experimental models such as hydra and the starlet sea anemone (Nematostella vectensis) led to significant advances for cnidarian developmental biology (Bode,1996; Galliot et al.,2006; Fritzenwanker et al.,2007), our understanding of mechanisms underlying neurogenesis and neuronal specification in cnidarians is limited. In hydra nerve cells are known to derive from a group of stem cells called the interstitial cell system (Bode,1996). Cells committed to the neuronal pathway are arrested at the G2 phase and need a signal to divide and differentiate (Schaller et al.,1989). A native neuropeptide, Hym355, was reported to act as a signaling molecule positively regulating neuronal differentiation in hydra (Takahashi et al.,2000). In contrast, native peptides released by epithelial cells have inhibitory effects on neuronal differentiation (Takahashi et al.,1997; Bosch and Fujisawa,2001). These findings beg the question: is cnidarian neuronal differentiation regulated by uniquely evolved signaling molecules, or are there other categories of signaling systems found in more complex animals that are already involved in differentiation of cnidarian nerve cells?

Two signaling molecules that were reported to be important for neuronal development in vertebrates are retinoic acid (RA) and nitric oxide (NO). RA, a physiologically active derivative of vitamin A, is known to play a variety of roles in mammalian neurogenesis, neuronal differentiation and patterning (Sidell et al.,1983; Andrews,1984; Maden et al.,1996; Liu et al.,2001; Wilson et al.,2004; Maden,2007 for review). Two forms of RA, all-trans and 9-cis RA, exert their actions through retinoic acid receptors (RAR) or retinoid X receptors (RXR), members of class II receptors in the nuclear receptor superfamily (Allenby et al.,1993). Other members of Class II nuclear receptors, orphan COUP-TF receptors, are known to repress RA-induced neuronal differentiation in a mammalian cell line (Neuman et al.,1995). Because hydra homologs of COUP-TF receptors are expressed in a hydra neuron population during neurogenesis and are also able to repress RAR/RXR-induced transactivation in mammalian cells (Gauchat et al.,2004), a role for RA in cnidarian neuronal development can be envisaged. Recent studies have produced the first direct evidence of induction of neurite outgrowth by RA in an invertebrate, the mollusc Lymnaea stagnalis (Dmetrichuk et al.,2006). Therefore, widening investigations may lead to findings of a generalized role for RA in invertebrate neuronal differentiation. A hypothesized physiological role for RA in cnidarians was bolstered by (1) evidence that retinoids affect body pattern formation in a hydrozoan (Müller,1984) and (2) the identification of a RXR-like receptor in the jellyfish Tripedalia cystophora showing significant similarity with the human RXR alpha receptor (Kostrouch et al.,1998). More recently, report of the presence of immunoreactive RXR in interstitial cells and neurons of two anthozoans, the sea pansy Renilla köllikeri and the staghorn coral Acropora millepora, adds support for a role of RA in cnidarian neuronal differentiation (Bouzaiene et al.,2007).

The other signaling molecule of interest, NO, is a gaseous molecule involved in many different physiological processes, including cell proliferation and differentiation, smooth muscle contraction, apoptosis, macrophage activity and neurotransmission (Krumenacker and Murad,2006; Tota and Trimmer, 2007). It is known to play a direct role in inhibiting mammalian neurogenesis during neuronal differentiation but with few exceptions it is not involved directly in neuronal differentiation (Peunova and Enikolopov,1995; Packer et al.,2003; Gibbs,2003; Estrada and Murillo-Carretero,2005). Gaseous transmisson using NO appears to be widespread in the animal kingdom (Moroz,2001). In insects and molluscs NO is known to regulate neurite extension and neuronal migration (Seidel and Bicker,2000; Haase and Bicker,2003; Trimm and Rehder,2004). Although NO was shown to modulate muscle contraction and nitric oxide synthase (NOS) was found to be present in neurons of cnidarians (Moroz et al.,2004; Anctil et al.,2005), there is no evidence so far for its involvement in cnidarian neuronal development.

The evidence of the presence of retinoic acid and nitric oxide signaling systems in cnidarians prompted us to ask whether they participate in neurogenesis in this phylum. We used cell culture techniques to investigate the role of RA and NO in cell proliferation and differentiation, and especially in the process of neuronal differentiation, in the sea pansy Renilla köllikeri. We found that these agents affect cell proliferation and neuronal differentiation differently in adhesion-permissive and nonadhesive culture conditions.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Animals

Colonies of Renilla köllikeri were purchased from, and shipped to Montreal by Marinus Scientific (Garden Grove, CA, USA). They were kept at 16°C in a 120-L tank filled with filtered artificial sea water (ASW, Instant Ocean). Sea water osmolarity was adjusted to 1100 mOsm and pH to 8. Colonies were maintained under a 12 h:12 h light-dark cycle and they were kept unfed. They were processed within three to four days of their arrival.

Cell Culture

Sea pansy colonies were progressively anesthetized in a solution of 0.37 mol/L magnesium chloride. Autozooid polyps were excised from the colony, sliced extensively into minute fragments and large tissue pieces were discarded. The remaining small tissue pieces and cell agglomerates were dissociated at room temperature in 10 mL of culture medium to which 3.1 mmol/L papain and 3.2 mmol/L dithiothreitol were added for 90 min with agitation. The culture medium contained 23 mmol/L NaCl, 0.63 mmol/L KCl, 10 mmol/L MgCl2, 1.1 mmol/L CaCl2, 3.69 mmol/L Na2SO4, 2.38 mmol/L HEPES, and 0.09 mmol/L gentamycin. It was millipore filtered and pH was adjusted to 8. This was followed by three washes with fresh culture medium by centrifugation with a VWR microcentrifuge at 1500 rpm.

To grow dissociated cells 35-mm Nunclon culture dishes with a bottom grid were used. For some of the experiments dishes were first coated with 1 mg/mL poly-L-lysine and allowed to dry overnight. Dissociated cells were plated at a density of 5 × 103 cells per dish (1 mL of cell suspension). Cell cultures were maintained at 11°C in an Echo Therm 30 Chilling incubator. Cells were allowed to grow to 90% confluency over a period of 8–10 days postplating.

Assays of Retinoic Acid and Nitric Oxide

The two isomers of retinoic acid (RA), 9-cis and all-trans, and the NO donors, S-nitroso-N-acetylpenicillamine (SNAP) and amino-3-morpholinyl-1,2,3-oxadiazolium chloride (SIN-1), were purchased from Sigma-Aldrich (Canada). Stock solutions of retinoic acid were prepared using 100% ethanol as solvent to obtain a concentration of 1 mmol/L. NO donors were dissolved in dimethyl sulfoxide (DMSO) to a stock concentration of 0.9 mmol/L.

To test the effect of retinoic acid on cultures of dissociated cells, various volumes of stock solutions of RA were added to dishes containing one day old cultures to obtain the desired final concentration. For NO donors, 100 μL of stock solutions was diluted in 5 mL of culture medium to obtain a concentration of 2 × 10−2 mmol/L and different volumes of this dilution were added to one-day old cultures to obtain the desired final concentration. Cultures remained exposed to the test substances for their remaining life (7–10 days). For controls, untreated cultures prepared from the same animal as the treated cultures contained the same amount of vehicle, ethanol or DMSO, present in the treated cultures. In addition, degraded NO donor solutions were used as additional controls to ensure that no by-product was responsible for observed effects when there was no more NO producing ability.

After treatments, the cultures were maintained in the incubator at 11°C in the dark. The cultures were examined daily with a Nikon Eclipse TE300 inverted microscope equipped with phase-contrast and with Hoffman modulation contrast optics. Images were obtained with a Nikon Coolpix 4500 digital camera and processed with Corel PHOTO-Paint 12.

Analysis

To assess the effects of RA or NO donors cell densities and morphologies were compared between control and treated cultures in polylysine-coated and uncoated dishes. In uncoated dishes, the cells were settled on the dish bottom. Densities were evaluated from viewing fields under a 20× objective by counting dedifferentiated and/or redifferentiated cells over areas of 2 mm2 using the bottom grid squares of culture dishes as guide. Values for analysis were cell numbers per mm2. Cell types were identified by matching their shape, size and, if applicable, intracellular characteristics with illustrated descriptions of sea pansy (Germain and Anctil,1988; Pani et al.,1994; Pernet et al.,2004) and other anthozoan cells (Fautin and Mariscal,1991). Elements failing to match these criteria were discarded from the counts. The position of the selected area in each culture dish at the beginning of treatment was scored and cell density was recorded daily in that same area over the duration of treatment (6–8 days) for each dish. Average density values from five separate control and experimental cultures were then recorded and the means ± standard error of means computed. The rates of cell proliferation and cell differentiation were analyzed and displayed using Graph Pad Prism. Bonferroni corrected unpaired t tests were performed to assess significance of differences between cell densities at various doses of active agents or at different days after treatment.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Cell Growth in Culture

When cultured on polylysine-coated or uncoated dishes, untreated cells rapidly dedifferentiated into round or ovoid shapes and for the most part remained thus undifferentiated while their population grew over the useful life of the culture (up to 10 days). In polylysine-coated dishes, cell density of untreated cultures rose steadily to reach a fivefold increase by Days 8–10 [Fig. 1(A)], compared with a twofold increase only in uncoated dishes [Fig. 3(A)].

thumbnail image

Figure 1. Effect of retinoic acid on sea pansy cell proliferation in cultures maintained in poly-L-lysine-coated dishes. A: graph showing the increase of cell density over time in untreated cultures and in cultures treated with 100 μmol/L 9-cis retinoic acid added at culture Day 2. Differences in values between treated and untreated cultures are significant for all sampling days in unpaired t tests (p < 0.001). B: Dose-response curve of the proliferative effect of 9-cis retinoic acid on cell cultures. All between-dose comparisons showed significant differences in cell density in unpaired t tests (p < 0.01). The panel below the graph shows the difference of cell density between representative images of untreated and retinoic acid-treated cultures at Day 6.

Download figure to PowerPoint

In contrast to cells grown in polylysine-coated dishes, some of the cells grown in uncoated dishes, while remaining round or slighly oval, exhibited intracellular morphologies associated with known anthozoan cell types. The most outstanding example in this study is the epitheliomuscular and bioluminescent cells of the sea pansy, characterized when dissociated by an excentric cytoplasm and a large vacuole-like pool of disorganized myofibrils [Fig. 2(A); see Germain and Anctil,1988]. Symbiotic algal cells also grew in these cultures, but they were easily identified by their brown color and discarded from the cell counts.

thumbnail image

Figure 2. Micrographs of sea pansy cells in cultures maintained in uncoated dishes. A: Untreated culture at Day 7 in which only round epitheliomuscular cells (EMC) are identifiable. B: Culture six days after treatment with 100 μmol/L 9-cis retinoic (culture Day 7) in which ciliomotor (CMC), epitheliomuscular (EMC), gland (GC), interstitial (IC), nematocyte (NC), and sensory cells (SC) are identifiable.

Download figure to PowerPoint

Cell Growth and Morphogenetic Effects of Retinoic Acid

Preliminary results indicated that the effects of the 9-cis and all-trans isomers of RA are similar. Therefore, only the 9-cis isomer was used in all the experiments presented here. When cultures in polylysine-coated dishes were treated with 9-cis RA, cell proliferation occurred at an increasingly greater rate than in untreated cultures [Fig. 1(A)]. By the tenth day post-treatment cell density in treated cultures had increased by three- to fourfold over that of untreated cultures. The increase in cell density was readily apparent by visual inspection of six-day-old cultures [Fig. 1(A)]. No cell differentiation was apparent under these conditions. The RA-induced accelerated rate of cell proliferation was dose-dependent and the sigmoid curve of the relationship [Fig. 1(B)] was consistent with a receptor-mediated event. A significant rise in cell density was apparent at a 9-cis RA concentration of 1 μmol/L and cell density rose exponentially at higher concentrations until it tended to level off at 100 μmol/L [Fig. 1(B)].

When cells were cultured in uncoated dishes, they began differentiating into various types of epithelium-associated cells after exposure to RA. In untreated cultures cells show few distinctive morphologies [Fig. 2(A)]. In cultures treated with 10–100 μmol/L 9-cis RA, cell morphologies typical of anthozoan epithelium-associated cells (Fautin and Mariscal,1991) appear with increasing frequency, especially sensory cells, nematocytes, gland cells, and ciliary motor cells [Fig. 2(B)]. As the density of undifferentiated cells steadily declined over the life of the RA-treated cultures, the density of differentiated cells rose gradually [Fig. 3(A)]. In contrast, no measurable increase in overall cell density was observed, thus indicating that no cell proliferation occurred in RA-treated, uncoated culture dishes. Sensory cells constituted the greater share of observed epithelial cells, with densities of other epithelial-like cell types being significantly lower to varying degrees than those of sensory cells [Fig. 3(B)].

thumbnail image

Figure 3. Histograms quantifying the effect of 100 μmol/L 9-cis retinoic acid on sea pansy cell differentiation in cultures maintained in uncoated dishes. A: Changes in cell density of untreated (controls) versus treated (undifferentiated and differentiated cells) cultures over time. Note the rise in cell density in untreated cultures and the reversal of relative cell densities between undifferentiated and differentiated cells in treated cultures. Single and double asterisks represent significant differences in cell density compared with untreated cultures at p < 0.05 and p < 0.01, respectively (unpaired t test); ns, not significant (p > 0.05). B: Relative cell density among various types of identifiable differentiated epithelial cells after six days of treatment with retinoic acid. Densities of gland cells, nematocytes and ciliomotor cells are all significantly lower than those of sensory cells (unpaired t tests, p < 0.01).

Download figure to PowerPoint

Cell Growth and Morphogenetic Effects of Nitric Oxide Donors

Both NO donors SIN-1 and SNAP enhanced cell growth in polylysine-coated dishes in a manner similar to retinoic acid. However, cell density increased at alesser rate than in retinoic acid-treated cultures in theeffective concentration range of the NO donors (10–100 μmol/L), remaining around double the cell density of untreated cultures throughout the life of the cultures [Fig. 4(A)]. In contrast, NO donors induced the differentiation of an increasingly larger number of cells into neurons in uncoated dishes. Two days after treatment with 90 μmol/L SIN-1 or SNAP only 25% of cultured cells were identifiable as neurons, whereas 85% of the cultured cells were recognizable neurons six days after treatment [Fig. 4(B)]. No cell type other than neuronal was detectable in treated cultures. Cells in untreated cultures grew in number similarly to uncoated control dishes in retinoic acid experiments [compare Fig. 4(B) with Fig. 3(A)]. Treatments with degraded NO donor solutions yielded the same results as in untreated cultures (not shown).

thumbnail image

Figure 4. Effects of nitric oxide (NO) donors on cell proliferation and neuronal differentiation in sea pansy cell cultures. A: Graph showing the effect of 90 μmol/L SNAP added at Day 2 on the rise of cell density over time in cultures maintained in poly-L-lysine-coated dishes. Differences in values between treated and untreated cultures are significant for all sampling days in unpaired t tests (p < 0.001). B: Histogram showing the effect of 90 μmol/L SIN-1 on sea pansy neuronal differentiation in cultures maintained in uncoated dishes. Note the rise in cell density in untreated cultures and the sharp reversal of relative cell densities between undifferentiated and neuronal cells in treated cultures. Single and double asterisks represent significant differences in cell density compared with untreated cultures at p < 0.05 and p < 0.01, respectively (unpaired t test); ns, not significant (p > 0.05).

Download figure to PowerPoint

Cells in untreated cultures remained largely spherical or ovoid throughout the culture period [Fig. 5(A)] whereas the majority of cells displayed a smaller soma and grew fine extensions (neurites) in NO-treated cultures [Fig. 5(B)]. The nerve cells were mostly unipolar or bipolar and their long neurites showed swellings (varicosities) along their entire length [Fig. 5(C)]. Multipolar neurons were also observed, and the neurites of some of these neurons appear to cross over each other, bifurcate and contact non-neuronal cells [Fig. 5(D)]. These cells resemble the norepinephrine- and RFamide-immunoreactive neurons forming the nerve net in the mesoglea of the sea pansy, particularly near the endoderm (Pani et al., 1995; Pernet et al.,2004).

thumbnail image

Figure 5. Micrographs of neurons differentiating six days after treatment with 90 μmol/L SIN-1 in cultures maintained in uncoated dishes. A: Untreated culture at Day 7 in which cells of various sizes possess round or ovoid shapes. B: Treated culture at Day 7 (six days post-treatment) in which numerous neuron soma with long, intersecting neurites are visible and relatively few round cells remain. Scale bar in B applies also to A. C: Enlarged view of a bipolar neuron with elongating, varicose neurite. D: Enlarged view highlighting two neurons (NC), one of which has a neurite crossing over other neurites (arrow) and bifurcating to make contact with non-neuronal cells (arrowheads).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Our results, using the sea pansy as a cnidarian experimental model, show that 9-cis RA and NO donors potentiate cell proliferation and induce differentiation under different culture conditions. All these agents enhance cell proliferation but fail to induce morphogenesis in polylysine-coated dishes. In contrast, while 9-cis RA and NO donors both induce neuronal differentiation in uncoated dishes, they differ in their morphogenetic products, epithelial (including sensory) cells for RA and exclusively neurons for NO donors. Together with previous evidence of RA receptors (Bouzaiene et al.,2007) and NO signaling systems (Anctil et al.,2005), these results support a role for RA and NO in regulating adhesion-dependent cell proliferation and adhesion-independent neuronal differentiation in the sea pansy. It remains to be determined to which extent and in which manner the observed in vitro effects of these morphogens are played out in vivo.

Validation of Cell Culture Method

Several attempts to establish long-term cultures of cnidarian cells have met with obstacles or have altogether failed (Schmid et al.,1999 for review). Frank et al. (1994) reported successful establishment of continuous cell cultures and of cell lines in colonial anthozoans, including octocorallians other than the sea pansy, but the notorious in vivo plasticity of cnidarian cells and the presence of parasites and symbiont cells in cnidarian tissues cast doubt on the identity of cells proliferating over long time periods (Schmid et al.,1999). This demonstrates the need to rely on short-term primary cultures for studies intended to examine pathways and mechanisms of cell differentiation such as in this study.

The first successful attempt at establishing a primary culture of cnidarian neurons was made in a jellyfish (a hydrozoan) in which nerve-rich tissue can be found (Przysiezniak and Spencer,1989). The authors used a homogenate of the jellyfish's own extracellular matrix (ECM) in the abundant mesogleal jelly as substrate for adhesion of the cultured neurons and in these conditions the cultures survived for up to two weeks in ASW (Przysiezniak and Spencer,1989). They and other authors (Schmid et al.,1999) tested other substrate media than ECM, including polylysine, but only ECM appeared satisfactory. However, while cell survival depends on attachment to substrate, cell proliferation was inhibited under such conditions (Schmid et al.,1999). In addition, no anthozoan possesses nerve-rich tissues or can yield sufficient ECM material for practical use as substrate for cell attachment in culture.

Polylysine-coated culture surfaces have been known for several decades to provide an adequate substrate for cell attachment and especially strong adhesion for neurons without inhibiting neurite elongation (Varon,1979). Day and Lenhoff (1981) reported that hydra cells adhered well to polylysine and we were also able to successfully culture sea pansy cells in dishes coated with poly-L-lysine, but in agreement with Przysiezniak and Spencer (1989) cells failed to undergo neuronal differentiation under such conditions. A larger concentration of polylysine (1 mg/mL) than those used for vertebrate cell cultures (0.1 mg/mL) was necessary to ensure cell adhesion, yet contrary to vertebrate cultured cells (Varon,1979) no apparent toxicity to sea pansy cells was noted at such a concentration. It is not clear why a high level of polylysine is needed and why sea pansy cells are resistant to its toxic effect at such a level, but the unusual lipid environment of cnidarian cell membranes (Schetz and Anderson,1993) and the divergent molecular structure of cnidarian ECM proteins (Knack et al.,2008; Magie and Martindale,2008) may at least partly account for these properties.

In our hands dissociated sea pansy cells plated in either uncoated or polylysine-coated dishes and in the absence of enriched media or trophic factors were able to proliferate while showing little or no evidence of differentiation within the life time of the cultures. This allowed us to design experiments during which morphogenetic agents could be tested and their effects on cell growth or morphological phenotype unambiguously observed.

Retinoic Acid and Nitric Oxide Promote Proliferation of Adherent Cells

Our results demonstrate that 9-cis RA and NO potentiate cell proliferation without inducing differentiation in sea pansy cells allowed to adhere to polylysine-coated dish bottoms. This is consistent with studies on vertebrate cell lines showing that both RA (Henion and Weston,1994; Nabeyrat et al.,1998; Wohl and Weiss,1998) and NO (Ulibarri et al.,1999; Mejia-Garcia and Paes-de-Carvalho,2007) can promote proliferation and spreading of adherent cells under specific conditions. It is clear that substrate adhesion is a determining factor in the repression of differentiation of sea pansy cells exposed to RA or NO, because we have also shown that cells cultured without substrate coating and showing little or no substrate adhesion respond to these agents by differentiating into specific cell types (see below).

The concentrations of 9-cis RA used in the majority of experiments in this study are substantially higher than those used in experiments with vertebrate cell cultures (Henion and Weston,1994), and therefore appear to fall outside the physiological range. The endogenous levels of retinoids in cnidarians are unknown, so it is not possible to determine what is the physiological range for these basal metazoans. RA binds in the nanomolar range to a jellyfish RXR fusion protein expressed in bacteria (Kostrouch et al.,1998), but effective concentrations for binding RXR in jellyfish membranes or for eliciting a cell response have not been reported. In another basal metazoan, the sponge Suberites domuncula, a canal regression was induced by RA in the 1–50 micromolar range (Wiens et al.,2003). Even if the RA concentrations used in our experiments could be higher than the endogenous levels, healthy indicators such as growth and differentiation, not adverse or deleterious effects, were observed, thus suggesting that RA was acting physiologically.

The similar proliferative effects of 9-cis RA and NO donors raise the possibility that RA acts through the mediation of a nitrergic pathway. RA stimulation of NO production was previously reported in mammalian endothelial cells (Achan et al.,2002; Urano et al.,2005) and in a neuroblastoma cell line (Ghigo et al.,1998). Even if a functional link between RA and NO is discovered, the process by which cell growth is promoted and differentiation repressed in our experimental conditions remains to be examined. One possibility is that the proliferative effects of these agents are mediated through trophic factor transduction pathways that may also interfere with differentiation signals. This would require that trophic factors are released from the cultured cells in the medium upon adhesion. There are several candidates identified in cnidarians, such as cnidarian homologues of vascular endothelial growth factor and receptor (Seipel et al.,2004), fibroblast growth factor (Matus et al.,2007) and insulin growth factor (Steele et al.,1996).

Retinoic Acid and Nitric Oxide Differentially Induce Neuronal Differentiation in Nonadherent Cells

We have shown that sea pansy cells cultured on a non-adhesive substrate differentiate into sensory and other epithelium-associated cells when exposed to 9-cis RA and selectively into neuronal cells when exposed to NO donors. The RA results are consistent with numerous studies demonstrating the role of retinoids in vertebrate epithelial differentiation (Shapiro,1986 for review), including sensory epithelia (Lefevre et al.,1993; Raz and Kelley,1999). In contrast, the NO results do not concur with the lack of direct involvement of nitrergic pathways in mammalian neuronal differentiation (Estrada and Murillo-Carretero,2005), but are consonant with invertebrate studies demonstrating a direct role in neuronal differentiation, especially neurite outgrowth and navigation (Bicker,2005).

The induced differentiations occur in a time-dependent manner, with a gradual reversal of dominance from undifferentiated to differentiated morphologies. It is striking that while control cultures displayed a twofold increase in density, the size of cell populations remained stable during the RA- and NO-induced differentiation processes, suggesting that the morphogens inhibit growth while inducing differentiation. This is consistent with reports showing that both retinoids (Sidell,1981; Sidell et al.,1983) and NO (Peunova and Enikolopov,1995; Kuzin et al.,1996) suppress cell growth during differentiation of neuronal and other cell lines. Termination of cell division is required also in cnidarians to engage in processes of cell differentiation (Bode,1996), including neurodifferentiation (Hager and David,1997). Retinoids appear to coordinate growth arrest with neuronal differentiation in mammalian target cells through distinct RXR/RAR receptor signaling pathways (Cheung et al.,1996; Giannini et al.,1997). NO, on the other hand, appears to specifically mediate the growth suppressing effect of nerve growth factors (Peunova and Enikolopov,1995) and is linked to the onset of differentiation only as a prerequisite. Without a better knowledge of retinoid receptor and nitrergic signaling it will be difficult to unravel the mechanisms by which these morphogens affect cell proliferation and differentiation in this cnidarian. Although 9-cis RA and NO clearly commit sea pansy cells to different fates, it remains possible that the RA-induced arrest of cell proliferation is mediated by NO.

The occurrence of morphogen-induced differentiation only in nonadhesive conditions suggests that the absence of adhesion activates signals that make the dedifferentiated cells responsive to the morphogenetic actions of RA and NO. The nature of the hypothesized signals is suggested by studies showing that jellyfish striated muscle cells remain differentiated when the integrity of their ECM is preserved, but dedifferentiate and undergo transdifferentiation when the ECM is removed or inactivated (Schmid and Reber-Müller,1995). Dedifferentiation in jellyfish involves disassembly of cytoskeletal components, hence the round cell shapes in culture, and is followed by DNA replication, which the cell proliferations we observed are assumed to reflect. Protein kinase C signaling may also be involved (Kurz and Schmid,1991). In jellyfish transdifferentiation into smooth muscle cells and neurons follows dedifferentiation and cell division without any further experimental treatment. In the sea pansy cells remain largely dedifferentiated and proliferate until exposure to 9-cis RA or NO when redifferentiation is initiated.

The differential effects of RA and NO suggest that epithelium-associated cell commitment and mesogleal neuron commitment are subjected to distinct signaling pathways. The majority of epithelium-associated cells in this study (sensory neurons, nematocytes, gland cells) are known in hydra to be derived from interstitial cells, a form of multipotent stem cells (Bode,1996 for review). In hydra all neurons are associated with the epithelial monolayers (ectoderm and endoderm) and the mesogleal jelly, which is sandwiched by the two epithelial layers, is acellular. In contrast, the mesoglea of anthozoans such as the sea pansy is thicker and contains cells including neurons resembling the ganglion cells of hydra (Fautin and Mariscal,1991; Pani et al., 1995; Pernet et al.,2004). Thus RA appears to specifically target interstitial cells for commitment to epithelium-associated cell lineages in the sea pansy, whereas NO selectively targets precursor cells of mesogleal neurons. Mesogleal neurons are known to form nerve nets in the sea pansy (Pernet et al.,2004) and the neurite extensions of the differentiating mesogleal-like neurons in this study are seen to cross each other [Fig. 5(D)] as expected from cnidarian nerve-net organizations (Minobe et al.,1995).

Evolutionary Significance

To our knowledge this is the first report of effects of a retinoid and of NO on cell proliferation, neuronal differentiation and neurite outgrowth in a cnidarian. This suggests that RA and NO are important morphogens in cnidarians as they are in vertebrates. As neurons are considered to have evolved first in cnidarians, the implication is that these morphogens were functionally integrated into morphogenetic signaling early in the evolutionary emergence of nervous systems.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The authors thank Meriem Bouzaiene for her advice on technical matters.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES