|Molecular Vision 2005;
Received 27 October 2004 | Accepted 8 June 2005 | Published 7 July 2005
Differential regulation of melatonin synthesis genes and phototransduction genes in embryonic chicken retina and cultured retinal precursor cells
Virginie Cailleau, Marianne Bernard, Fabrice Morin, Jerome
Guerlotte, Pierre Voisin
UMR CNRS 6187 Institut de Physiologie et Biologie Cellulaires, Poitiers, France
Correspondence to: Pierre Voisin, UMR CNRS 6187 Institut de Physiologie et Biologie Cellulaires, 40 Avenue Recteur Pineau, 86022 Poitiers, France; Phone: (33) 549 45 39 78; FAX: (33) 549 45 40 51; email: email@example.com
Purpose: Photoreceptor differentiation involves the activation of two specific sets of genes; those encoding the proteins of the phototransduction cascade and those encoding the enzymes of the melatonin synthesis pathway, arylalkylamine N-acetyltransferase (AANAT) and hydroxyindole O-methyltransferase (HIOMT). The purpose of the present study was to examine the conditions of AANAT and HIOMT gene activation, relative to that of selected phototransduction markers (α-transducin and opsins), in both in vivo and in vitro differentiating photoreceptors of the chicken retina.
Methods: Neural retina RNA was obtained between embryonic day 7 (E7) and posthatch day 8 (P8) and analyzed on northern blots with cDNA probes to AANAT, HIOMT, visinin, α-transducin, rhodopsin, and the four cone opsins. Cell cultures were prepared from E7 chicken neural retina and incubated for two to four days in vitro, either in basal medium or in serum-supplemented medium or in medium containing an insulin-based supplement. RNA from the cultured cells was analyzed on northern blots as above. Real time RT-PCR was used to confirm in vitro changes in HIOMT and red opsin mRNA levels. The cultured cells were transfected with promoter-reporter plasmids for direct analysis of HIOMT promoter regulation by the dual luciferase method.
Results: The different mRNAs composing the photoreceptor phenotype appeared at E7 (visinin), E10 (α-transducin), E14 (HIOMT), E15 (rhodopsin, red opsin, and green opsin), E16 (AANAT), E17 (blue opsin), and E18 (violet opsin). In the early differentiating cones of the central retina, HIOMT mRNA appeared two days earlier than red opsin and green opsin mRNAs (E12 rather than E14). In cultured embryonic neural retina cells, basal medium was sufficient to activate α-transducin gene transcription, an insulin-based supplement was sufficient to activate HIOMT gene transcription, whereas serum was required for red opsin gene transcription after two days in vitro. All serum batches were able to activate red opsin gene transcription, whereas some of them failed to activate HIOMT gene transcription. Activation of the HIOMT gene promoter by an insulin-based supplement and by serum was confirmed after transfection of chicken embryonic neural retina cells with promoter-reporter plasmids.
Conclusions: Activation of the melatonin synthesis genes in vivo takes place in a time window very close to that of early opsins. However, a 24-48 h lead of HIOMT gene expression over early opsins was clearly observed. Our in vitro experiments indicate that different exogenous signals are required to activate the different genes encoding photoreceptor specific functions. Significantly, marker genes for light sensitivity (red opsin) and for melatonin synthesis (HIOMT) appear to be activated in response to different signals.
In addition to the role they play in vision by affecting the phototransduction cascade, retinal photoreceptors specifically convert the photoperiodic component of ambient light into a day/night rhythm of melatonin synthesis. Melatonin is produced with a large amplitude peak at night and regulates several aspects of retinal physiology, including disk shedding from rod outer segments and dopamine release from amacrine cells [1,2]. Melatonin is synthesized from serotonin in two enzymatic steps catalyzed by arylalkylamine N-acetyltransferase (AANAT, EC 126.96.36.199) and by hydroxyindole O-methyltransferase (HIOMT, EC 188.8.131.52). Photoreceptor specific expression of AANAT and HIOMT has been documented by immunocytochemical and in situ hybridization studies in the chicken pineal and retina [3-10]. Based on these observations, functional differentiation of photoreceptors in embryonic life would appear to rely on the activation of at least two specific sets of genes: those of the phototransduction cascade and those of the melatonin synthesis pathway.
The relative timing of expression of these two sets of genes during photoreceptor differentiation has not been examined and it is not known whether they are activated in response to similar or different developmental signals. A number of histological studies have shown that photoreceptor differentiation in the chicken retina is a protracted phenomenon, starting when precursors exit mitosis around embryonic day 6 (E6) and extending over two weeks, with stepwise appearance of the different cellular components that characterize the photoreceptor phenotype . A similar picture was derived from gene expression studies, showing that photoreceptor precursors express visinin as soon as they exit mitosis at E6, whereas activation of the opsin genes takes place only eight to ten days later, as observed by in situ hybridization . Studies in mammalian species appear to confirm the stepwise activation of phototransduction genes, as examplified by the sequential expression of transducin, rhodopsin, phosphodiesterase 6, and arrestin over a period of several days or weeks [13-15]. A developmental study of a larger number of photoreceptor markers in the chicken retina would be required to examine whether there are any intermediate steps in photoreceptor specific gene activation between visinin expression at E6 and the activation of the opsin genes around E15. In this context, special interest in the melatonin synthesis genes is warranted because information on the timing of expression of AANAT and HIOMT in the embryonic chicken retina is relatively scarce [16-18], which makes it difficult to compare with the developmental expression of the phototransduction genes. To obtain a better description of these two aspects of photoreceptor differentiation, we simultaneously analyzed the mRNAs encoding HIOMT, AANAT, visinin, α-transducin, rhodopsin, and the four cone opsins in the developing chicken retina. The results indicate that α-transducin and HIOMT expressions start two to four days before opsin expression, thus suggesting intermediate steps in photoreceptor differentiation.
The next question we wished to examine was whether the phototransduction genes and the melatonin synthesis genes were activated by similar signals or by different signals during photoreceptor differentiation. To address this question, we took advantage of a dispersed cell culture of chicken retinal precursors that was previously shown to accelerate the differentiation of red cones [19,20]. In this system, we observed that different culture conditions were required to support HIOMT, α-transducin, and red opsin gene transcription.
Fertilized chicken eggs (Gallus domesticus) were purchased from Rumolo Cie (Quincay, France) and were incubated at 37 °C in an atmosphere of 70% humidity. Embryos were staged according to the method of Hamburger and Hamilton . Newly-hatched chicks were transferred to a hot room maintained at 27 °C, with food and water ad libitum.
Primary retinal cell cultures were prepared as described by Adler . Embryonic day 7 (E7) neural retinas, free of pigment epithelial cells, were dissociated and cultured at 37 °C in 6 well or 24 well polyornithine-coated plates (3x106 cells/well and 6x105 cells/well, respectively), in an atmosphere of 5% CO2 in air. Basal culture medium was M199 (Sigma, Saint Quentin Fallavier, France) supplemented with 100 U/mL penicillin/streptomycin, 2 mM glutamine, and 0.5% bovine serum albumin (BSA). Serum-supplemented medium contained 0.1% BSA and either 10% newborn calf serum (NCS; Invitrogen, Cergy Pontoise, France) or 10% fetal calf serum (FCS; Invitrogen). ITS-supplemented medium contained 0.5% BSA and 1X ITS (1.5 μM insulin, 69 nM transferrin and 38 nM selenium, Invitrogen).
The cDNA probes were produced by PCR amplification of either plasmid clones or total retinal cDNA, with the primer pairs listed in Table 1, purified by agarose gel electrophoresis and silica gel adsorbtion (QIAquick, QIAGEN, Coutaboeuf, France or GENECLEAN, Qbiogene, Illkirch, France) and verified by DNA sequencing (Big Dye, Perkin Elmer, Courtaboeuf, France). The chicken HIOMT probe was produced by PCR amplification of a 518-bp fragment from a previously described Blue-pGEM clone . The chicken AANAT probe was produced by PCR amplification of a 1388-bp fragment from a previously described pBK-CMV clone . The chicken visinin probe was produced by PCR amplification of a 444-bp fragment from total retinal cDNA. The chicken α-transducin probe was a 479-bp fragment of 89% identical sequence for the rod and cone forms (positions 574 to 1052 of the coding region in the cone form), produced by PCR amplification from total retinal cDNA. The presence of a mix of rod and cone forms in the PCR product was verified by DNA sequencing. Hybridization of this probe on northern blots of total retinal RNA sometimes showed two bands around 2 to 2.5 kb, in agreement with the reported sizes of rod and cone forms of α-transducin mRNA . More generally, however, the two forms of α-transducin mRNA were not resolved on our northern blots due to the large quantity of material loaded on the gels. Because the α-transducin probe also showed 82% sequence identity with Gi2 α-subunit, we verified the absence of cross-hybridization on northern blots of total brain RNA. The different chicken opsins probes were produced by PCR amplification of pGEM-T-cloned fragments of the following sizes: rhodopsin 475 bp, red opsin 300 bp, green opsin 475 bp, blue opsin 429 bp, violet opsin 392 bp. Each opsin probe had less than 70% sequence identity with any other opsin and hybridized only with the mRNA of expected size on northern blots: rhodopsin at 1.6 kb (major band) and 3 kb (minor band); red opsin at 1.6 kb; green opsin at 2.8 kb; blue opsin at 3.1 kb, and violet opsin at 3.7 kb .
Northern blot analysis
Neural retinas were obtained at the beginning of day time, from at least 10 animals for each experimental point, dissected free of the pigment epithelium and rapidly frozen in liquid nitrogen as pools. For center to periphery gradient studies, a 4 to 9 mm piece of retina was dissected at the tip of the pecten (area centralis), and another 36 mm piece of retina was dissected from the nasal periphery. Total RNA was extracted as previously described . Tissues were sonicated in a suitable volume (20-50 μl/mg of tissue) of ice-cold LiCl/urea solution (3 M LiCl, 6 M urea). The homogenates were incubated overnight at 4 °C. Total RNA was pelleted by centrifugation at 15000x g for 30 min. Pellets were washed once with 1 volume of LiCl/urea solution and were then resuspended in TE/SDS buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA, pH 8, 0.5% sodium dodecyl sulfate). Total RNA was then purified by extraction with phenol/chloroform and precipitated with ice-cold ethanol. RNA concentrations were measured by the optical density at 260 nm. Total RNA was denatured with 48% formamide and 6.4% formaldehyde at 95 °C for 5 min, and loaded (15 μg per lane) on 1% agarose/0.7 M formaldehyde gels containing 0.7 μg/ml ethidium bromide. After electrophoresis, the RNA was transferred overnight onto nitrocellulose membrane in 10X NaCl/Citrate buffer (1.5 M NaCl, 0.15 M sodium citrate, pH 7). Nonspecific sites were blocked by incubating the membrane for 2 h at 42 °C in the high-stringency hybridization solution (40% formamide, 10% dextran sulfate, 4X NaCl/Cit, 0.02% polyvinylpyrrolidone, 0.02% BSA fraction V, 0.02% Ficoll 400, 20 mM Tris-HCl, pH 7.4, and 0.3 mg/ml denatured salmon sperm DNA).
The cDNA probes described above were labeled to similar specific radioactivities by random priming  with [α-32P]dCTP (3000 Ci/mmol). The chicken 18S rRNA probe was a 21-mer antisense oligonucleotide corresponding to bp 956-936 (GenBank accession number M59389) and was labeled with [α-32P]dCTP (3000 Ci/mmol) using Terminal Deoxynucleotidyl Transferase (Amersham Biosciences, Orsay, France). The prehybridized membranes were incubated overnight at 42 °C with the radiolabelled probes diluted at 1 μCi/ml in high-stringency hybridization solution. Membranes were then washed for 1 h at room temperature in 2X NaCl/Citrate, 0.1% SDS and for 2 h at 52 °C in 0.1X NaCl/Citrate, 0.1% SDS and exposed on the screen of a PhosphorImager (Storm 820, Amersham Biosciences).
Real time RT-PCR
Total RNA (1 μg) from cultured retinal cells was reverse transcribed with M-MLV (Promega, Charbonnieres, France) and random hexamers (Amersham Biosciences) at 37 °C for 2 h and the reaction was terminated by a 1/5 dilution with double-distilled water. Real time PCR was performed on a LightCycler (Roche Applied Science, Meylan, France) according to the manufacturer's instructions. A typical reaction was performed in 10 μl, consisting of 5 μl of a 1/5 diluted RT reaction and 5 μl of SYBR Green Master Mix (Roche Applied Science), containing the specific primer pairs (final 0.5 μM each). The PCR temperature cycle was 95 °C 10 min, followed by 45 cycles of 95 °C 10 s, 65 °C 10 s, 72 °C 20 s, with SYBR Green fluorescence recording at the end of each elongation segment. The PCR reaction was followed by a melting curve analysis: 95 °C 10 s, 60 °C 10 s, linear increase to 95 °C at 0.1 °C/s, with continuous SYBR Green fluorescence recording. The primer pairs used in this study for HIOMT (S518, AS1035), red opsin (S47, AS488) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, S7, AS260; Table 1) were validated as follows: They gave a single PCR product as verified by melting curve analysis, agarose gel electrophoresis, and DNA sequencing, and the distribution of the PCR sigmoids was linear (r was 0.99 to 1) over 5 logs of template concentration with an efficiency of 1.85 to 1.98. The critical cycle of each sigmoid PCR curve was calculated by the LightCycler Software 3.5® (Roche Applied Science) as the PCR cycle corresponding to the maximum of the second derivative. Total cDNA from each cell culture sample was analyzed by real time PCR for HIOMT, red opsin, and GAPDH. Critical cycle values of GAPDH PCR reactions were used for correction of the RNA load in the RT reactions. After this correction, critical cycle differences for HIOMT or for red opsin, between samples cultured in basal medium and samples cultured in either ITS- or NCS-supplemented medium were converted into fold increase over control.
Embryonic retinal cells were seeded at a density of 3x106 cells per well in 6 well plates, one day before transfection. Cells were transfected with promoter-reporter constructs (2 μg/well) containing either 3 kbp or 225 bp of the chicken HIOMT promoter, upstream of the firefly luciferase reporter gene, as previously described , or with the promoterless plasmid pGL3-Basic (Promega). To correct for variations in transfection efficiencies, each experimental group was cotransfected with a plasmid containing the Renilla luciferase reporter gene, under control of the SV40 promoter and enhancer (pRL-SV40, Promega; 1.5 μg/well). Transfections were performed following the CaPO4 method. Briefly, plasmid DNA was diluted in 75 μl of 250 mM CaCl2 and was added dropwise to 75 μl of 2X HeBS (280 mM NaCl, 1.4 mM Na2HPO4, 50 mM HEPES, pH 7.1). Cells were incubated overnight with the precipitate (150 μl/well in 2 ml of culture medium) and were cultured for 24 h in fresh medium before being assayed for luciferase activities. Transfections and assays were performed in triplicate for each experimental group. Basal medium alone (M199 with 0.1% BSA) could not be used in these experiments because it generated a major CaPO4 precipitate that was lethal to the cells. Instead, culture and transfections were performed in basal medium supplemented either with 10% NCS, or with 10% FCS from a batch that did not induce northern blot-detectable levels of HIOMT mRNA (this batch is thereafter termed inactive FCS). Firefly and Renilla luciferase activities, were assayed using the Dual-Glo luciferase assay system (Promega). Light emission was measured on a Beckman LS 6500 counter, on the luminometer program (Beckman Instruments, Gagny, France). The differences in transfection efficiency between the experimental groups were normalized by calculating the ratio between the firefly and Renilla luciferase activities.
Developmental expression of melatoninergic and phototransduction genes in the embryonic chicken retina
Developmental activation of the genes encoding melatonin-synthesizing enzymes and phototransduction proteins was examined by northern blot, at 3-7 days intervals between embryonic day 7 (E7) and posthatch day 8 (P8) and at 1 day intervals between E14 and E18 (Figure 1). Analysis on the wide time scale indicated that visinin mRNA was present from E7 onwards, α-transducin mRNA appeared at E10; HIOMT mRNA at E14, while AANAT and all opsins mRNAs were detected from E18 onwards (Figure 1A). Minor mRNA bands (about 10% of the signal) were observed for HIOMT (at 4 kb) and for rhodopsin (at 3 kb), in agreement with previous reports [20,25,29]. The intensity of these minor bands increased in parallel with the major signal (not illustrated). Analysis on a more resolutive time scale allowed us to accurately define the order of appearance of the mRNAs encoding the melatonin-synthesizing enzymes, relative to the mRNAs encoding the different opsins. HIOMT gene expression at E14 was confirmed, while AANAT mRNA became clearly detectable at E16, with only a faint trace at E15 (Figure 1B). In agreement with a previous in situ hybridization study , rhodopsin, red opsin, and green opsin mRNAs became detectable at E15, while blue opsin mRNA appeared at E17 and violet opsin mRNA only at E18 (Figure 1B). The antecedence of HIOMT gene transcription over early cone opsins (red and green) was confirmed in three independent experiments.
To further analyze this time difference in gene expression, we examined the spatial distribution of HIOMT mRNA in the developing retina, based on the notion that cone photoreceptors differentiate in a center to periphery gradient . When the area centralis and the nasal periphery of the retina were dissected separately, northern blot analysis allowed us to detect HIOMT mRNA in the central retina as early as E12, whereas it became detectable in the peripheral retina at E14 (Figure 2). The previously described center to periphery gradient of red opsin and green opsin expression  could be observed at E14 and E15 (Figure 2). Thus, the data clearly indicated a 48 h lead of HIOMT over red opsin and green opsin in the early-differentiating photoreceptors of the area centralis. The relative precocity of HIOMT gene transcription (Figure 3) would appear to make it a valuable marker to examine the early events controling melatoninergic differentiation of photoreceptors.
Differential regulation of HIOMT, α-transducin, and red opsin gene expression in retinal precursors in vitro
In agreement with previous reports [19,20], retinal precursors dissociated at E7 and cultured for 2-3 days in serum-supplemented medium showed a precocious activation of red opsin gene transcription and an increase in visinin mRNA levels, when compared to in vivo E10 retinas (Figure 4A). With all due caution when comparing in vitro and in vivo situations that do not necessarily reflect the same embryonic stages, our data appeared to confirm that red cones undergo an accelerated differentiation in vitro. Our experiments also confirmed that retinal precursor cells failed to differentiate into other cones or into rods, as indicated by the absence of rhodopsin, green, blue or violet opsins mRNAs (data not shown). In addition, we observed that the cultured cells expressed other components of the photoreceptor phenotype (i.e., the mRNA transcripts of α-transducin and HIOMT genes; Figure 4A). AANAT mRNA was hardly detectable on northern blots after two or three days of culture (data not shown) and therefore was not further investigated.
The different mRNA transcripts composing the photoreceptor phenotype were differentially affected when cells were cultured in serum-free (basal) medium. Red opsin and HIOMT transcripts remained undetectable after two days of culture in basal medium (Figure 4B). In contrast, visinin mRNA was still present at a concentration about half that observed in serum-supplemented cultures (Figure 4B). Moreover, α-transducin gene transcription was clearly activated in cells cultured in basal medium, albeit less efficiently than in the presence of serum (Figure 4B). Cell counts indicated 60% plating efficiency after two days of culture in both basal and serum-supplemented media. Although cell viability dropped rapidly in basal medium (30% after three days and about 10% after four days), cells cultured for two days in basal medium remained capable to activate red opsin and HIOMT gene transcription when transferred to serum-supplemented medium for two additional days (Figure 4B). Together, the data seemed to indicate that photoreceptor precursors cultured for two days in basal medium were viable and actively differentiating (activation of the α-transducin gene), but were lacking a serum factor to activate red opsin and HIOMT gene transcription. Further experiments indicated that different batches of fetal calf serum (FCS) varied strongly in their ability to activate HIOMT gene transcription, although they all proved capable to activate red opsin gene transcription (Figure 5). This was the first hint at a differential regulation of these two genes during photoreceptor differentiation in vitro. In contrast to the variability of FCS batches, newborn calf serum (NCS) consistently activated both HIOMT and red opsin gene transcription (Figure 5).
To further compare HIOMT and red opsin gene transcription in vitro, retinal precursors were cultured in a chemically-defined medium, where serum was replaced by an insulin-based culture supplement (insulin, transferrin, selenium: ITS). After two days of culture in the presence of ITS, HIOMT gene transcription was strongly activated, whereas red opsin mRNA remained undetectable (Figure 6A). Photoreceptor precursors survival in ITS-supplemented medium could be verified by the presence of visinin and α-transducin mRNAs (Figure 6A). In addition, cells cultured for two days in ITS-supplemented medium retained the ability to activate red opsin gene transcription, if further-cultured for two additional days in the presence of NCS (Figure 6A). When the components of the ITS culture supplement were tested separately, transferrin had no effect, whereas insulin (1.5 μM) and selenium (38 nM) both activated HIOMT gene transcription, without apparent additivity (Figure 6B). The effect of insulin was not observed at 15 nM concentration (data not shown).
Real time RT-PCR was used to confirm the effects of ITS and NCS on HIOMT and red opsin mRNA levels. Due to its higher sensitivity, real time RT-PCR detected small amounts of both HIOMT and red opsin mRNAs in cells cultured in basal medium, which allowed for quantitation of the relative increase in mRNA levels evoked by either ITS or NCS. In a first series of five consecutive experiments, this quantitation revealed a 17 fold increase in HIOMT mRNA levels with ITS and a 40 fold increase with NCS, compared to no supplement (Figure 7A). In the same experiments, red opsin mRNA levels showed a 300 fold increase with NCS and no significant increase with ITS (Figure 7A). Another series of eight experiments confirmed this observation, albeit with less dramatic changes in mRNA levels: HIOMT mRNA concentration was increased 9 fold by ITS and 15 fold by NCS, while red opsin mRNA concentration was increased 25 fold by NCS and not increased by ITS (Figure 7B).
Accumulation of HIOMT mRNA in response to NCS or ITS appeared to reflect transcriptional activation rather than mRNA stabilization, because both culture conditions also activated promoter-reporter constructs consisting of the luciferase reporter under control of either a 3 kbp or a 225 bp version of the chicken HIOMT promoter (Figure 8). Due to constraints on transfection efficiency (see Methods), basal conditions in these experiments did not correspond to unsupplemented medium, but to medium supplemented with a batch of inactive FCS (i.e., a batch of FCS that did not induce northern blot-detectable levels of HIOMT mRNA). By comparison with a promoterless plasmid, some HIOMT promoter activity was clearly observed in the presence of inactive FCS. However, transcription of the promoterless plasmid did not change in the presence of NCS or ITS, whereas both supplements increased HIOMT promoter activity about 3 fold (Figure 8).
The present study describes the timing of gene activation for core proteins of the phototransduction cascade (opsins and α-transducin) and for the enzymes of the melatonin synthesis pathway (AANAT and HIOMT), in the developing chicken retina. In addition, our in vitro experiments allowed us to define culture conditions that differentially stimulated α-transducin, HIOMT, and red opsin gene transcription in chicken retina precursor cells. Although AANAT gene transcription could not be analyzed in our short-term cultures, other studies have documented the regulation of this gene by light and circadian oscillators after longer culture times [30,31]. It should also be mentioned that a minor proportion of the AANAT mRNA detected in vivo might be contributed by retinal ganglion cells, because this cell type was recently shown to express low levels of AANAT .
Our in vivo experiments indicated that transcriptional activation of the melatonin synthesis genes takes place in a time window very close to that of early opsins gene activation. However, the antecedence of HIOMT gene expression was clear, especially when the center to periphery gradient of differentiation was taken into account. It is tempting to speculate on a possible relationship between this early onset of HIOMT gene transcription and the detoxifying role played by this enzyme in the bis-retinyl arylalkylamine scenario recently proposed by Klein  to explain the evolution of photoreceptors and pinealocytes.
The timings of expression of HIOMT and α-transducin genes in vivo, between visinin expression at E6 and early opsins at E15 , are of special interest because they define intermediate steps in photoreceptor differentiation. Previous studies in mammals have also shown that transcriptional activation of the genes encoding beta-PDE6, opsins, α-, β-, and γ-transducin, arrestin and recoverin occur at several days intervals during retinogenegis [13-15]. In the chicken retina, a more complete study, including a larger number of photoreceptor markers (phosphodiesterase 6, arrestin, phosducin, peripherin, etc.) would still be needed to precisely delineate all the successive steps of specific gene activation. Nevertheless, our data strengthen the view that photoreceptor differentiation, in terms of specific gene activation, is a stepwise process that extends over several days, as previously noted for morphological traits such as the production of outer segments or the establishment of synaptic contacts . As previously stated by Adler , this protracted differentiation might represent the execution of an endogenously-timed program initiated at the exit of mitosis, or the successive actions of exogenous signals required to activate each functional gene composing the photoreceptor phenotype. Our in vitro experiments suggest that both hypotheses apply, depending on which functional gene is being considered.
The photoreceptor precursor cells placed in culture at E7 appeared to be fated in as much as they continued to express visinin mRNA and were able to activate α-transducin transcription, in the absence of any culture supplement. However, they were unable to activate HIOMT and red opsin gene transcription after two days in basal medium, at least not to a level that would be sufficient for detection by northern blot. The sustained and even increased expression of visinin mRNA in culture is in agreement with previous reports . The activation of α-transducin gene transcription in basal medium is new information, which adds support to previous studies indicating that photoreceptor precursors can achieve a certain degree of phenotypic differentiation in a cell-autonomous manner . The absence of northern blot-detectable red opsin mRNA after culture in basal medium contrasts with previous reports indicating that red opsin-positive cells can be detected by in situ hybridization after four days of culture in serum-free medium . The sensitivity cut-off of northern blot analysis and the shorter culture duration used in our experiments (two days rather than four days) may be responsible for these contrasted results. Indeed, our real time PCR analysis confirmed the presence of small amounts of red opsin mRNA in cells cultured in basal medium, but it also showed that serum causes a dramatic increase in the concentration of this transcript (up to 300 fold). This large increase in red opsin mRNA levels, which appeared to be required for detection on northern blot, would represent a more advanced degree of phenotypic differentiation achieved in response to serum factors. Like red opsin, HIOMT mRNA levels were increased by serum to northern blot-detectable levels. This observation generalizes the regulation of HIOMT gene transcription by serum that we previouly observed in embryonic chicken pineal gland and in human retinoblastoma Y79 cells [26,36].
One unexpected result of the present study is that all batches of serum were able to increase red opsin mRNA levels, whereas only some of them activated HIOMT gene transcription. This is reminiscent of another study on chicken retina precursor cells, indicating that multiple lots of serum had to be screened to obtain a high proportion of photoreceptors and high levels of AANAT activity . These observations stress the sensitivity of retinal precursor cells to serum composition and should to be taken into account in further studies utilizing this culture system. In the present study, the variability among serum batches was meaningful, because it was the first hint at a differential regulation of HIOMT and red opsin gene transcription. This was further supported by the use of an insulin-based culture supplement which activated HIOMT gene transcription without increasing red opsin mRNA levels. The high concentration of insulin required for HIOMT gene activation suggests a nonspecific effect. However, this response was also observed with selenium, which has been characterized as an insulinomimetic in several model systems [37,38] and whose effects have been related to activation of both the PI3K and the MAP kinase pathways, downstream of the insulin receptor . Further studies using insulin-like growth factors and other peptides known to activate the same intracellular signaling pathways should help elucidate the mechanism of HIOMT gene activation. This would improve our knowledge of signal-dependent gene activation in developing chicken photoreceptors, which was first evidenced with the CNTF-dependent activation of green opsin gene expression . Also of interest is the fact that ITS and NCS activated transcription of promoter-reporter constructs containing both long and short versions of the HIOMT promoter. Although no single consensus insulin-response element has been defined, examination of the HIOMT promoter sequence revealed several regions sharing up to 73% identity with DNA elements previously shown to mediate insulin effects in other gene promoters . Further studies on truncated or mutated versions of the HIOMT promoter should lead to the identification of the cis-regulatory elements involved in this response. Comparative studies with other promoters known to be activated in retinal precursor cells, like the IRBP promoter , would also be of interest to elucidate the signaling mechanisms governing functional differentiation of photoreceptors. Ultimately, a comparison with red opsin, transducin, and visinin promoters should be most informative.
Together, the different culture conditions used in the present study have allowed us to selectively activate the transcription of either α-transducin (basal medium), HIOMT (ITS), or red opsin (some lots of FCS), in chicken retinal precursor cells. Although our short-term cultures do not rule out the possibility that photoreceptor precursors might differentiate in a cell-autonomous manner if they were cultured long enough in basal medium, they do indicate that different exogenous signals are involved in activating the different genes encoding photoreceptor specific functions. Significantly, marker genes for light sensitivity (red opsin) and for melatonin synthesis (HIOMT) appear to be activated in response to different signals.
We wish to thank Dr. Pierre Greve and Dr. Brigitte Vannier for their help with some experiments and Mrs. Nathalie Girard and Mr. Gael Epistolin for their technical assistance. This research was supported by a grant from Retina-France.
1. Besharse JC, Dunis DA. Methoxyindoles and photoreceptor metabolism: activation of rod shedding. Science 1983; 219:1341-3.
2. Dubocovich ML. Melatonin is a potent modulator of dopamine release in the retina. Nature 1983-1984; 306:782-4.
3. Voisin P, Guerlotte J, Collin JP. An antiserum against chicken hydroxyindole-O-methyltransferase reacts with the enzyme from pineal gland and retina and labels pineal modified photoreceptors. Brain Res 1988; 464:53-61.
4. Guerlotte J, Voisin P, Brisson P, Faure JP, Collin JP. Synthesis of melatonin by the pineal modified photoreceptors of birds immunocytochemical localization of hydroxyindole-O-methyltransferase. Biol Cell 1988; 64:93-6.
5. Bernard M, Voisin P, Guerlotte J, Collin JP. Molecular and cellular aspects of hydroxyindole-O-methyltransferase expression in the developing chick pineal gland. Brain Res Dev Brain Res 1991; 59:75-81.
6. Greve P, Bernard M, Voisin P, Cogne M, Collin JP, Guerlotte J. Cellular localization of hydroxyindole-O-methyltransferase mRNA in the chicken pineal gland. Neuroreport 1993; 4:803-6.
7. Wiechmann AF, Craft CM. Localization of mRNA encoding the indolamine synthesizing enzyme, hydroxyindole-O-methyltransferase, in chicken pineal gland and retina by in situ hybridization. Neurosci Lett 1993; 150:207-11.
8. Wiechmann AF. Hydroxyindole-O-methyltransferase mRNA expression in a subpopulation of photoreceptors in the chicken retina. J Pineal Res 1996; 20:217-25.
9. Guerlotte J, Greve P, Bernard M, Grechez-Cassiau A, Morin F, Collin JP, Voisin P. Hydroxyindole-O-methyltransferase in the chicken retina: immunocytochemical localization and daily rhythm of mRNA. Eur J Neurosci 1996; 8:710-5.
10. Bernard M, Iuvone PM, Cassone VM, Roseboom PH, Coon SL, Klein DC. Avian melatonin synthesis: photic and circadian regulation of serotonin N-acetyltransferase mRNA in the chicken pineal gland and retina. J Neurochem 1997; 68:213-24.
11. Grun G. The development of the vertebrate retina: a comparative survey. Adv Anat Embryol Cell Biol 1982; 78:1-85.
12. Bruhn SL, Cepko CL. Development of the pattern of photoreceptors in the chick retina. J Neurosci 1996; 16:1430-9.
13. Bowes C, van Veen T, Farber DB. Opsin, G-protein and 48-kDa protein in normal and rd mouse retinas: developmental expression of mRNAs and proteins and light/dark cycling of mRNAs. Exp Eye Res 1988; 47:369-90.
14. He L, Campbell ML, Srivastava D, Blocker YS, Harris JR, Swaroop A, Fox DA. Spatial and temporal expression of AP-1 responsive rod photoreceptor genes and bZIP transcription factors during development of the rat retina. Mol Vis 1998; 4:32 <http://www.molvis.org/molvis/v4/a32/>.
15. Johnson PT, Williams RR, Reese BE. Developmental patterns of protein expression in photoreceptors implicate distinct environmental versus cell-intrinsic mechanisms. Vis Neurosci 2001; 18:157-68.
16. Wainwright SD. Development of hydroxyindole-O-methyltransferase activity in the retina of the chick embryo and young chick. J Neurochem 1979; 32:1099-101.
17. Iuvone PM. Development of melatonin synthesis in chicken retina: regulation of serotonin N-acetyltransferase activity by light, circadian oscillators, and cyclic AMP. J Neurochem 1990; 54:1562-8.
18. Espinar A, Osuna C, Feliu C, Guerrero JM. High activity of retinal N-acetyltransferase in the early development of the chick embryo: independence of lighting conditions. Neurosci Lett 1994; 179:103-6.
19. Tamres AN, Belecky-Adams TL, Adler R. Microenvironmental regulation of visual pigment expression in the chick retina. Invest Ophthalmol Vis Sci 1999; 40:S457.
20. Adler R, Tamres A, Bradford RL, Belecky-Adams TL. Microenvironmental regulation of visual pigment expression in the chick retina. Dev Biol 2001; 236:454-64.
21. Hamburger V, Hamilton HL. Series of embryonic chicken growth. J Morphol 1951; 88:49-52.
22. Adler R. Preparation, enrichment and growth of purified cultures of neurons and photoreceptors from chicks embryos and from normal and mutant mice. Vol. 2. In: Conn PM, editor. Methods in neurosciences. San Diego: Academic Press; 1990. p. 134-50.
23. Voisin P, Guerlotte J, Bernard M, Collin JP, Cogne M. Molecular cloning and nucleotide sequence of a cDNA encoding hydroxyindole O-methyltransferase from chicken pineal gland. Biochem J 1992; 282:571-6.
24. Kasahara T, Okano T, Yoshikawa T, Yamazaki K, Fukada Y. Rod-type transducin alpha-subunit mediates a phototransduction pathway in the chicken pineal gland. J Neurochem 2000; 75:217-24.
25. Okano T, Kojima D, Fukada Y, Shichida Y, Yoshizawa T. Primary structures of chicken cone visual pigments: vertebrate rhodopsins have evolved out of cone visual pigments. Proc Natl Acad Sci U S A 1992; 89:5932-6.
26. Grechez-Cassiau A, Greve P, Guerlotte J, Collin JP, Voisin P. Hydroxyindole-O-methyltransferase gene expression in the pineal gland of chicken embryo: development of messenger RNA levels and regulation by serum. Brain Res Dev Brain Res 1995; 88:204-11.
27. Feinberg AP, Vogelstein B. A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1983; 132:6-13.
28. Bernard M, Dinet V, Voisin P. Transcriptional regulation of the chicken hydroxyindole-O-methyltransferase gene by the cone-rod homeobox-containing protein. J Neurochem 2001; 79:248-57.
29. Bernard M, Guerlotte J, Cogne M, Greve P, Collin JP, Voisin P. Transcriptional regulation of hydroxyindole O-methyltransferase in the chicken pineal gland: day/night changes and long-term effects of light and darkness. Biochem J 1993; 290:661-4.
30. Ivanova TN, Iuvone PM. Melatonin synthesis in retina: circadian regulation of arylalkylamine N-acetyltransferase activity in cultured photoreceptor cells of embryonic chicken retina. Brain Res 2003; 973:56-63.
31. Haque R, Alonso-Gomez AL, Chaurasia SS, Iuvone PM. Diurnal regulation of arylalkylamine N-acetyltransferase activity in chicken retinal cells in vitro: analysis of culture conditions. Mol Vis 2003; 9:52-9 <http://www.molvis.org/molvis/v9/a9/>.
32. Garbarino-Pico E, Carpentieri AR, Contin MA, Sarmiento MI, Brocco MA, Panzetta P, Rosenstein RE, Caputto BL, Guido ME. Retinal ganglion cells are autonomous circadian oscillators synthesizing N-acetylserotonin during the day. J Biol Chem 2004; 279:51172-81.
33. Klein DC. The 2004 Aschoff/Pittendrigh lecture: Theory of the origin of the pineal gland--a tale of conflict and resolution. J Biol Rhythms 2004; 19:264-79.
34. Adler R. Plasticity and differentiation of retinal precursor cells. Int Rev Cytol 1993; 146:145-90.
35. Adler R. A model of retinal cell differentiation in the chick embryo. Prog Retin Eye Res 2000; 19:529-57.
36. Bernard M, Voisin P, Klein DC. Hydroxyindole-O-methyltransferase in Y-79 cells: regulation by serum. Brain Res 1996; 727:118-24.
37. Stapleton SR, Garlock GL, Foellmi-Adams L, Kletzien RF. Selenium: potent stimulator of tyrosyl phosphorylation and activator of MAP kinase. Biochim Biophys Acta 1997; 1355:259-69.
38. Heart E, Sung CK. Insulin-like and non-insulin-like selenium actions in 3T3-L1 adipocytes. J Cell Biochem 2003; 88:719-31.
39. Stapleton SR. Selenium: an insulin-mimetic. Cell Mol Life Sci 2000; 57:1874-9.
40. Villafuerte BC, Phillips LS, Rane MJ, Zhao W. Insulin-response element-binding protein 1: a novel Akt substrate involved in transcriptional action of insulin. J Biol Chem 2004; 279:36650-9.
41. Boatright JH, Buono R, Bruno J, Lang RK, Si JS, Shinohara T, Peoples JW, Nickerson JM. The 5' flanking regions of IRBP and arrestin have promoter activity in primary embryonic chicken retina cell cultures. Exp Eye Res 1997; 64:269-77.