We report the presence of two Pax6-related genes, Pax6A and Pax6B, which are highly conserved in two planarian species Dugesia japonica and Girardia tigrina (Platyhelminthes, Tricladida). Pax6A is more similar to other Pax6 proteins than Pax6B, which is the most divergent Pax6 described so far. The planarian Pax6 homologs do not show any clear orthology to the Drosophila duplicated Pax6 genes, eyeless and twin of eyeless, which suggests an independent Pax6 duplication in a triclad or platyhelminth ancestor. Pax6A is expressed in the central nervous system of intact planarians, labeling a subset of cells of both cephalic ganglia and nerve cords, and is activated during cephalic regeneration. Pax6B follows a similar pattern, but shows a lower level of expression. Pax6A and Pax6B transcripts are detected in visual cells only at the ultrastructural level, probably because a limited amount of transcripts is present in these cells. Inactivation of both Pax6A and Pax6B by RNA-mediated gene interference (RNAi) inhibits neither eye regeneration nor eye maintenance, suggesting that the genetic network that controls this process is not triggered by Pax6 in planarians.

An evolutionary conserved complex network of different signaling pathways and nuclear factors regulates brain and eye specification (Kurata et al., 2000; Callaerts et al., 2001; Kumar and Moses, 2001). One striking illustration of such genetic conservation is the Pax6 gene, which is structurally and functionally characterized in a large number of triploblastic metazoans from planarians to humans (Quiring et al., 1994; Chisholm and Horvitz, 1995; Czerny and Busslinger, 1995; Zhang and Emmons, 1995; Loosli et al., 1996; Glardon et al., 1997; Tomarev et al., 1997; Glardon et al., 1998; Callaerts et al., 1999). Pax6 represents one of the nine members of the Pax class, characterized by the presence of two conserved DNA-binding domains, the paired domain and the homeodomain (Callaerts et al., 1997). The comparative molecular characterization of Pax6 in different animal phyla supports the hypothesis that this protein has a conserved function in a variety of developmental processes, ranging from regionalization to cell type specification, in eye and central nervous system development (Quiring et al., 1994; Kurusu et al., 2000; Pratt et al., 2000). In some vertebrates, Pax6 is also related to the development of head sensory organs or other structures such as pituitary or pancreas (Walther and Gruss, 1991; Turque et al., 1994). Pax6 genes from various animal phyla are capable of inducing formation of ectopic eyes in Drosophila (Gehring and Ikeo, 1999). In frogs, Pax6 misexpression can also induce ectopic eyes (Chow et al., 1999). This finding indicates that Pax6 is a master control gene for eye development, and also suggests a common genetic program for making eyes, despite the great variety of visual structures found in the animal kingdom. Two Pax6 genes involved in eye development have been found in zebrafish (Nornes et al., 1998). The presence of a second Pax6 gene arisen by gene duplication during insect evolution is also described in holometabolous insects, the two Drosophila genes eyeless (ey) and twin of eyeless (toy) being the best-studied examples. ey and toy share a similar pattern of expression in the developing visual system, and have non-redundant functions, indicating that they have been recruited with different roles into the eye genetic pathway (Czerny et al., 1999).

Planarians are flatworms well known for their exceptional regenerative capabilities. They are free-living members of the Platyhelminthes, considered a basal phylum of the Lophotrocozoa clade (Aguinaldo et al., 1997; Carranza et al., 1997; Bayascas et al., 1998). The study of the role of Pax6 genes during eye and central nervous system (CNS) regeneration and maintenance will be of interest for understanding the different ways that this genetic network evolved in the animal kingdom. Planarians are one of the simplest organisms with cephalization, defined by the presence of two dominant cephalic ganglia connected by commissural connections and followed by two ventral nerve cords. The nerve cords run the entire length of the body and are regularly connected by commissures producing a small concentration of neurons at the crossing points (Baguñá and Ballester, 1978; Rieger et al., 1991). Light perception by special photosensitive cells occurs in planarians. Photoreceptor structures, which are capable of detecting light and shadow, are grouped into eyespots defining a simple ancestral visual system that consists of bipolar retinal neurons whose dendrites project into a cup-shaped structure composed of pigment cells (Kishida, 1967; Sakai et al., 2000). Although the planarian eyes cannot focus the light pattern into images, as they have no focusing lens, they serve essentially the same function, receiving and transducing light into neuronal signals, as eyes do in all metazoans.

We report an extensive search for Pax6 genes in Platyhelminthes that has led to the characterization of two related Pax6 genes in the planarian species Dugesia japonica (DjPax6A and DjPax6B) and Girardia tigrina [GtPax6A and GtPax6B, the latter previously named DtPax6 (Callaerts et al., 1999)]. The Pax6A and Pax6B sequences share high similarity and comparable expression patterns in both species, suggesting that they originated by duplication in a triclad or, possibly, in a platyhelminth ancestor. Both genes are detected in the adult CNS – a pattern of expression shared with all Pax6 genes described so far – and are activated during regeneration, Pax6A being more expressed than Pax6B. Although no specific transcripts were detectable in the photoreceptors by whole-mount in situ hybridization, the presence of low levels of Pax6A and Pax6B mRNA could be demonstrated in the eye cells by electron microscope in situ hybridization. Inactivation of both Pax6 genes by RNAi neither prevents eye formation during regeneration nor inhibits the eye expression pattern of the planarian genes Gtsix-1 or Gtops, supporting the hypothesis that the function of Pax6A and Pax6B is not essential in planarian eye regeneration.

Animals

Planarians used in this work belong to the species Dugesia japonica, clonal strain GI (Orii et al., 1993), and to an asexual race of Girardia tigrina, formerly classified as Dugesia(G) tigrina, collected in the Calders River (Barcelona, Spain). Animals were cultured as described previously (Salvetti et al., 1998; Saló and Baguñà, 1984) and starved for 2 weeks before being used in the experiments. Regenerating fragments were obtained by transverse amputation at the pre-pharyngeal or post-pharyngeal level. Anterior regeneration corresponds to a tail regenerating a head, while posterior regeneration corresponds to a head regenerating a tail. Finally, lateral regeneration was obtained by sagittal amputation.

Isolation and characterization of planarian genes related to Pax6

A GtPax6A PCR fragment of 111 bp was amplified with two degenerate primers. The primers correspond to the amino acid sequences LEKEFER (sense strand) and QVWFSNR (antisense strand) of the homeodomain. A λGt10 cDNA library from regenerating G. tigrina was screened according to Garcia-Fernandez et al. (Garcia-Fernandez et al., 1991), with the 111 bp PCR fragment as a probe. One positive clone was found corresponding to the 3′ Pax6A gene end (Pax6A3′). A 5′ RACE strategy was used to elongate the sequence at the 5′ end, obtaining a second clone (GtPax6A790) that contained 790 bp from the paired box to the homeobox of the Pax6A sequence. The GtPax6A sequence has been deposited to the EMBL/GenBank with the Accession Number AY028904. DjPax6A was directly amplified from the sequence deposited in the EMBL/Genbank (Accession Number, AB017632), as described by Rossi et al. (Rossi et al., 2001). A fragment (clone DjPax6B-520 bp) of the second Pax6 gene, DjPax6B, was amplified in D. japonica (Rossi et al., 2001). A sense-strand degenerate primer was designed, taking advantage of the high sequence similarity in the paired domain between DjPax6A and the Pax6 gene, here referred to as GtPax6B, previously characterized in G. tigrina (Callaerts et al., 1999). The antisense primer corresponded to the GtPax6B amino acid region TLFGYN. A 5′/3′RACE strategy was used to further characterize the DjPax6B sequence, deposited at the EMBL/Genbank with the Accession Number AJ311310. The sequence-specific antisense primer, corresponding to the amino acid region SKPRVATN was used to amplify selectively the 5′ region. The DjPax6B 3′ region was obtained with the sequence-specific sense primer, corresponding to INTWPPTS. The PCR products were TA-cloned using the pGEM-T easy vector (Promega). All clones were sequenced by automated fluorescent cycle sequencing (ABI).

Phylogenetic analysis of the Pax family homeodomains

The phylogenetic tree of the homeodomains and their flanking regions sequences was inferred using the CLUSTALX package. Sequences were aligned with CLUSTALX software. Evolutionary distances were calculated using Kimura’s equation (Nei and Koehn, 1983), and used for phylogenetic tree construction by the Neighbor-joining method. Sequences were obtained from the EMBL/ GenBank.

In situ hybridization experiments

All sense and antisense digoxigenin-labeled RNA probes used for in situ hybridization experiments were made using the RNA in vitro labeling kit (Roche). Whole-mount in situ hybridization was carried out on intact and regenerating planarians according to Agata et al. (Agata et al., 1998). Cryosections after whole-mount in situ hybridization were performed as described by Pineda et al. (Pineda et al., 2000). In situ hybridization on sections was carried out as described by Kobayashi et al. (Kobayashi et al., 1998). Hybridization took place at 55°C for 24-60 hours. The final probe concentration was 0.1 ng/μl. In some hybridized sections, 20 μg/ml DAPI was added to detect nuclei. Transmission electron microscope (TEM) in situ hybridization was performed on specimens of D. japonica, fixed with 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 1 hour at 4°C. After dehydration in a graded series of methanol (each step for 30 minutes at –20°C) specimens were embedded in Unicryl resin and polymerized for 72 hours at 4°C under u.v. light. Ultrathin sections, obtained with a diamond knife on Ultracut Reichert-Jung microtome, were placed on Formvar-carbon coated nickel grids and were incubated face down on a drop containing the hybridization buffer (50% formamide, 10% dextran sulphate, 4×saline sodium citrate, 400 μg/ml salmon testis DNA and 8 ng/μl of antisense or sense probes) for 4 hours at 37°C. After hybridization, ultrathin sections were washed in phosphate buffer, pre-incubated in 1% bovine serum albumin and then incubated in anti-digoxigenin antibody (1:40 in phosphate buffer) conjugated to gold particles of 10 nm in diameter, for 30 minutes. Grids were stained with uranyl acetate and lead citrate and observed with a Jeol 100 SX transmission electron microscope. Controls were performed using both sense probes and RNAse treatment (100 μg/ml of RNAse A for 90 minutes at 37°C) (Le Guellec et al., 1991) before the hybridization step.

The following clones were used as probes for in situ hybridization experiments.

D. japonica: DjPax6A, DjPax6B 520 bp, Dj18S (central region of 18S rDNA, about 1.1 kb) Djops 480 bp (Accession number, AJ421264) and Djsyt (Tazaki et al., 1999).

G. tigrina: GtPax6A3′ and GtPax6A790 were used in non-injected animals, GtPax6A790, Gtsix-1 so-5′ and so-3′.2 (Accession number, AJ251661), and Gtops p-250 (accession no. AJ251660) were used in dsRNA-injected animals.

Microinjection of double-strand RNA (dsRNA) and analysis of endogenous transcripts

Double-stranded RNA was synthesized as described by Sanchèz-Alvarado and Newmark (Sanchèz-Alvarado and Newmark, 1999). In G. tigrina Pax6A3′ clone and a ClaI-HindIII fragment of 400 bp (GtPax6B3′clone) of GtPax6B were used for dsRNA synthesis. In D. japonica dsRNA was synthesized from DjPax6A 1600 bp and DjPax6B 520 bp clones. Planarians were injected with 1010-1011 molecules of each dsRNA preparation or with an equimolar mixture of both Pax6A and Pax6B dsRNA. After the first injection, performed just after amputation, further injections were performed every 1 or 3 days, using a Drummond Scientific (Broomall, PA) Nanoject injector. Control injections were performed with water or β-Gal dsRNA. Injected planarians were kept at 17°C. Injected G. tigrina specimens were fixed at different regeneration times and processed for whole-mount in situ hybridization.

Total RNA was isolated from injected planarians after 1 to 3 days of regeneration and semi-quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) analysis was performed according to Bayascas et al. (Bayascas et al., 1998). Control reactions were performed identically in the absence of reverse transcriptase. Specific oligonucleotides from Pax6A and Pax6B were used to ascertain the reduction of Pax6A and Pax6B endogenous transcripts. Each couple of primers was designed from two regions, one internal and the other external to the sequence used for dsRNA synthesis. As an internal control, the ubiquitous transcripts of the homeobox gene Dth2 (Garcia-Fernandez et al., 1993) and the eye-specific Gtsix1 and Gtops transcripts (Pineda et al., 2000; Pineda et al., 2001), were amplified in G. tigrina. Two specific primers for the constitutively transcribed elongation factor gene DjEF2 were used for control amplifications in D. japonica.

Primer sequences used for PCR were as follows:

GtPax6A reverse, 5′-GAAGCTTCTGTTTCTGTTTTAGAG-3′;

GtPax6A forward, 5′-CGTACTTCGTTTTCGACAGATCAA-3′;

GtPax6B reverse, 5′-TCGCTTCTTTTGTTGTACAGTTTG-3′;

GtPax6B forward, 5′-CGGACTTCATTTACAAATGATCAG-3′;

Dth-2 reverse, 5′-TGGGAGACCGTTCTTTATCGTCAAC-3′;

Dth-2 forward, 5′-CCAATGCTAGTAATGATCCGCGTAT-3′;

Gtops reverse, 5′-GGACAGATACTTTGTTATCGCTCA-3′;

Gtops forward, 5′-TAACAAAATTCCCGATGTACATTC-3′;

Gtsix-1 reverse, 5′-AACGGCTCGGGATTTTTCTTTAAA-3′;

Gstix-1 forward, 5′-ATATGGTCTCTTCCACCTTGCCAA-3′;

DjPax6A forward, 5′-CCAAATCTTTCGCAATCTTC-3′;

DjPax6A reverse, 5′-CAATAAGTATCAAATACGTTACA-3′;

DjPax6B forward, 5′-CATCAATACATGGCCGCCTACAA-3′;

DjPax6B reverse, 5′CGTCTCCATTTTGCTCTGCGATT-3′;

DjEF2 forward, 5′-TTAATGATGGGAAGATATGTTG-3′; and

DjEF2 reverse 5′- GTACCATAGGATCTGATTTTGC-3′

For each PCR reaction the concentration of cDNA, primers and the number of cycles used were optimized with the aim of observing a quantifiable signal within the linear range of the amplification, according to both the putative abundance of each mRNA amplified and to the size of the corresponding PCR product.

Immunohistochemistry

Whole-mount immunohistochemistry of regenerating G. tigrina was carried out as described by Cebrià et al. (Cebrià et al., 1997). Rabbit anti-FMRFamide (Diasorin) was used as a primary antibody at a 1:100 dilution, and a fluorescein-conjugated goat anti-rabbit (IgG; Sigma) was used as a secondary antibody at a 1:50 dilution. The samples were examined with a epifluorescence microscope (Axiophot, Zeiss) and with a confocal laser scanning microscope (Leica Lasertechnik, Heidelberg).

Cloning of planarian Pax6A and Pax6B genes

Using a PCR approach and first strand cDNA derived from planarian heads, two Pax6 genes, Pax6A and Pax6B, were isolated in G. tigrina and D. japonica. Three types of sequences, which corresponded to three of the four families of Pax genes, Pax2-5-8, Pax1-9 and Pax4-6, were found in G. tigrina using two different sets of degenerate primers directed against two regions of the paired box (GRPLP and WEIRD) and of the homeobox (LEKEFER and QVWFSNR). The subsequent screening of a cDNA library using Pax4-6 fragment as a probe, allowed us to isolate a second Pax6 gene, GtPax6A, in addition to the previously described Pax6 gene (DtPax6) (Callaerts et al., 1999), here renamed GtPax6B. Two degenerate primers were selected on the basis of a comparative sequence analysis of DtPax6 and DjPax6A, a Pax6 clone previously isolated in D. japonica (Accession Number, AB017632). These primers allowed us to identify a second Pax6 gene in D. japonica, DjPax6B. DjPax6A and GtPax6A, as well as DjPax6B and GtPax6B deduced proteins, show extensive sequence identity even outside the conserved DNA-binding boxes (Fig. 1A).

Characterization and phylogenetic relationships of Pax6A and Pax6B proteins

Both Pax6A and Pax6B show a significant similarity in the paired domain and homeodomain with vertebrate and invertebrate Pax6 proteins. Comparison of the paired domain of Pax6A with those of published Pax6 proteins shows that it is 88% identical, and maintains most (six out of eight) of the residues that typify the Pax6 family. In Pax6B paired domain the degree of sequence conservation is lower, 77-78%, and only four out of eight Pax6-specific amino acids are conserved (Fig. 1B). The homeodomain in Pax6A has a higher sequence identity to the corresponding region of vertebrate Pax6 (85%), maintaining nine out of 10 specific amino acids, while in Pax6B it shows a lower value (74%), where only five out of 10 conserved amino acids are present. Moreover, four conserved residues preceding the homeodomain, as well as five out of seven distinctive amino acids following the homeodomain, are conserved in Pax6A, but not in Pax6B (Fig. 1C). The length of the linker region between the paired domain and homeodomain is highly variable in different invertebrate as well as vertebrate Pax6 proteins. It spans 224 amino acids in Drosophila ey and 108 amino acids in toy (Quiring et al., 1994; Czerny et al., 1999), whereas it is only 72 amino acids in length in sea urchin Pax6 (Czerny and Busslinger, 1995). Planarian Pax6 proteins contain a long linker region. In fact, the linker sequence consists of 138 amino acids in Pax6A and it is 171 amino acids long in Pax6B. Moreover, as commonly found in Pax6 proteins of other organisms (Loosli et al., 1996), Pax6A linker sequence is typified by the presence of some conserved amino acids (MYDKLSLLSGQ). A similar arrangement of amino acids has not been found in Pax6B.

The evolutionary relationships between the planarian Pax6A and Pax6B and Pax proteins of other organisms were inferred from the homeodomain and their flanking regions (Fig. 2). We can observe the closest similarity to Pax6 protostomate representatives, clustering the planarian Pax6A with Drosophila ey and toy, although with low bootstrap values. Pax6B shows the lowest similarity and is at the base of the Pax6 cluster. Such results indicate that Pax6A and Pax6B are bona fide Pax6 members.

Expression pattern of Pax6A and Pax6B in intact planarians

The expression pattern of Pax6A and Pax6B was analyzed by whole-mount in situ hybridization in intact adult specimens of both species. The presence of Pax6A transcripts was observed in the cephalic ganglia and in the nerve cords that cross the entire organism ventrally. (Fig. 3A-G). Pax6B showed a similar pattern, albeit with a lower level of expression (data not shown). In G. tigrina, the presence of additional GtPax6A transcripts was also detected in a non-cephalic parenchyma region close to the eosinophilic secretory cells, which give rise to the ventral marginal adhesive zone (Fig. 3A,F-H). This latter expression pattern suggests that GtPax6A has a derived role in the formation of this zone, an observation that deserves further investigation.

To investigate in detail the distribution of Pax6-expressing cells in the planarian CNS, we performed in situ hybridization directly on paraffin wax-embedded sections of intact D. japonica specimens (Fig. 4). The complete set of nerve cells at corresponding section levels was identified by the expression of the planarian neural marker synaptotagmin, Djsyt (Fig. 4C-H). Expression of DjPax6A was detected in a subset of cells of the CNS, but Pax6 transcripts were not detected in the eye cells using this procedure (Fig. 4I-P). Conversely, electron microscope in situ hybridization showed expression of both DjPax6A and DjPax6B in the perikaryon, the dendrites and the rhabdomeric region of the photoreceptor cells (Fig. 5A-C). We also observed expression in the perinuclear cytoplasm of pigment cells (Fig. 5D). Positive controls with the ribosomal 18S riboprobe Dj18S showed that both the nucleolus and the endoplasmic reticulum of the cells were labeled (Fig. 5E). In addition, the Dj18S hybridization signal was seen in the rhabdomeric projections of photoreceptor cells, thus supporting the possibility of translational activity in this region (Fig. 5G). Transcripts from the planarian synaptotagmin, DjSyt, were specifically detected in the perikaryon of nerve cells (Fig. 5H). No signal was observed with DjPax6A or DjPax6B sense-strand probes (Fig. 5F), or with RNAse treatment followed by hybridization using DjPax6A and DjPax6B antisense-strand probes.

Expression pattern of Pax6A and Pax6B during regeneration

Planarians have a great regenerative plasticity. After experimental decapitation, a complete head including brain and eyespots is formed during anterior or cephalic regeneration. We examined the expression pattern of the planarian Pax6-related genes during regeneration by whole-mount in situ hybridization. After 1-2 days, regenerating animals showed two intense spots of Pax6A expression near the region where new fibers emerge from the amputated old nerve cords (Fig. 6A). After 3-6 days of regeneration, the hybridization signal was localized in an arch-shaped structure of the blastema, where neoblasts fated to become nerve cells intermingle with the bent fibers to regenerate the cephalic ganglia (Fig. 6B,C). The arch-shaped expression pattern was particularly clear in G. tigrina, probably owing to morphological differences in the ganglia of the two species. At later stages of regeneration – from 3 to 15 days – we observed a gradual spreading of the Pax6A-positive tissue consequent to the growth of the cephalic ganglia (Fig. 6D-F). After that, the expression decreased to the basal level observed in the adult CNS (Fig. 3). In posterior regeneration a slight increase of Pax6A transcripts was observed in the region where the new nerve cords would differentiate from the edge of the old-sectioned nerve cords (Fig. 6G). Finally, after sagittal amputation, we observed Pax6A expression during lateral regeneration of the missing cephalic ganglion and nerve cord. Increased Pax6A expression was detected in the regenerating CNS area (Fig. 6H). This differential expression gradually decreased to the basal level as regeneration proceeded in a similar way to that observed during anterior regeneration (Fig. 6I). Owing to the weak Pax6B expression, almost undetectable under the experimental conditions used, there was no perceptible difference with Pax6A transcript distribution during regeneration.

Reduction of the levels of endogenous Pax6A and Pax6B by RNA interference (RNAi)

As both planarian Pax6 genes are expressed in the CNS and eye cells, we aimed to determine the function of Pax6 in the maintenance or the regeneration pattern of CNS and eye. Consequently, we injected Pax6A and Pax6B dsRNA into the postblastema region in regenerating D. japonica and G. tigrina, as well as in the head region of intact animals. An equimolar mixture of dsRNA of Pax6A/Pax6B was also injected in another experiment, in order to prevent a possible gene redundancy effect. Injected adults did not reveal any gross morphological or behavioral abnormalities in any condition, and regeneration of the eyespots was also observed during cephalic blastema formation. Inspection of the photoreceptor cells by TEM showed a normal rhabdomeric organization of these structures. We also compared the number of photoreceptors in the eye of dsRNA-injected animals and water-injected controls of both species, after 15 days of regeneration. We counted about 15-20 cells in transverse adjacent sections at the eye level, hybridized with the photoreceptor-specific molecular marker opsin (Djops and Gtops clones), and also stained with DAPI to facilitate the quantification of photoreceptor cells. Although the number of photoreceptors depends on the species, planarian size and light exposition average, we did not detect any significant difference in the number of these cells between dsRNA-injected planarians and controls (Fig. 7A,B).

To assay for the reduction effects of dsRNA on the cognate mRNA, the level of Pax6A and Pax6B transcripts in injected animals was analyzed by comparative RT-PCR and whole-mount in situ hybridization. Total RNA was extracted from planarians injected with Pax6A and Pax6B dsRNA, as well as from animals injected with dsRNA Pax6A/Pax6B mixture. In both species, RT-PCR analysis showed a strong reduction in the level of expression of the corresponding endogenous mRNA, when compared with β-Gal- or water-injected controls (Fig. 8A). A similar strong reduction of endogenous RNA was observed after RNAi experiments with Gtops (Fig. 8A) and Gtsix-1 genes (data not shown), which produce a loss-of-function phenotype (Pineda et al., 2000; Pineda et al., 2001).

Whole-mount in situ hybridization did not reveal GtPax6A mRNA in animals injected with dsRNA Pax6A or Pax6A/Pax6B mixture (Fig. 8D,E). By contrast, water- or β-Gal-injected head-regenerating fragments showed the typical pattern of GtPax6A mRNA expression (Fig. 6; Fig. 8B,C). These results demonstrate that RNAi interferes with the normal accumulation of endogenous transcripts of both Pax6 genes in planarians. However, this elimination does not produce morphological defects in the planarian eye phenotype. No gross neuroanatomical alterations were observed during cephalic ganglia regeneration by immunohistochemical staining with a general nerve cell marker, the anti-FMRFamide neuropeptide (Fig. 9A-F). However, the lack of specific neuronal cell markers prevented us from detecting neuronal cell fate changes due to the Pax6 loss of function.

To determine the possible effects of exogenous Pax6A and Pax6B dsRNA on the activation of other genes expressed in the planarian eye, we performed whole-mount in situ hybridization on injected animals using Gtsix-1 and Gtops (Pineda et al., 2000; Pineda et al., 2001) as probes. No appreciable delay or reduction in the level of Gtsix-1 mRNA expression was observed in the cephalic blastema during eye determination or differentiation, with respect to water-injected animals (Fig. 10A-F). Similar experiments performed with the planarian opsin gene Gtops again did not reveal any spatio-temporal change of the expression level during eye regeneration (Fig. 10G-I).

Planarians have two Pax6 genes

We report that the genome of two planarian species belonging to different genera (D. japonica and G. tigrina) has two genes, Pax6A and Pax6B, that encode distinct paired domain- and homeodomain-containing proteins. The DNA-binding domain sequence analysis of both Pax6A and Pax6B unambiguously identifies them as members of the Pax6 family. While Pax6A appears more closely related to Pax6 proteins of other organisms, Pax6B can be considered the most diverged Pax6 characterized so far (Callaerts et al., 1997; Callaerts et al., 1999). In fact it shares only 77-78% identity in the paired domain and 74% identity in the homeodomain of the vertebrate/mammalian Pax6 sequence, while in Pax6A the identities are 88% and 85%, respectively. Pax6A also possesses a conserved Pax6-specific motif of eleven amino acids (Loosli et al., 1996) in the linker region that is not present in Pax6B. With the exception of the Pax6 homologs ey and toy, which are found in Drosophila and other holometabolous insects (Czerny et al., 1999), and a Pax6 duplication reported in zebrafish by Nornes et al. (Nornes et al., 1998), no duplicated Pax6 genes have been described in other organisms so far. The phylogenetic analysis of the homeodomain and flanking regions with representative Pax genes supports a weak but closer clustering of planarian Pax6A to the duplicated Drosophila ey and toy, compared with the Lophotrocozoa clade representatives such as the nemertine LsPax6 or the mollusc LoPax6. Such weak clustering could be produced by a long branch attraction effect between Drosophila and planarians. According to its structural divergence, Pax6B is located outside the main Pax6 clustering in the tree. Although the amino acids that determine DNA-binding specificity are well conserved in Pax6B, half (nine out of 18) of the invariant residues present in the paired and homeodomain are not conserved, thus representing a notable exception to the notion that Pax6 is a highly conserved transcription factor. Ectopic expression of this gene in Drosophila fails to produce eyes (Callaerts et al., 1999). From an evolutionary point of view, we can hypothesize that gene duplication produced two Pax6 paralogs in a triclad or platyhelminth ancestor. After that, the two genes evolved with different selective pressures, probably resulting from different functional constraints. Nowadays Pax6A would conserve most of the structural characteristics of the ancestral Pax6, and consequently, higher similarity to Pax6 proteins of other organisms. In evolutionary times, mutations accumulating in the duplicated gene generated the structural differences present in Pax6B.

Pax6A and Pax6B are expressed in intact planarians and activated during regeneration

Both Pax6 planarian genes are expressed in the CNS of adult organisms, Pax6A being more strongly expressed than Pax6B. The presence of Pax6B transcripts along the anteroposterior planarian axis, which was barely detected by in situ hybridization, was confirmed by RT-PCR experiments (Callaerts et al., 1999). Both genes are expressed in a subset of cells located along the entire CNS. The time course of Pax6A expression in cephalic regeneration clearly demonstrated activation of this gene during cephalic ganglia formation. Similarly, increased production of Pax6B transcripts during regeneration was demonstrated in G. tigrina (Callaerts et al., 1999). Regeneration of the new cephalic ganglia requires the presence of the old nerve cords. This process has been followed by immunoreactivity to the molluscan cardioactive peptide (FMRFamide). New neural fibers emerge very early from the sectioned old nerve cords and reach the blastema, then bend transversally and fuse, producing a commissure, where cephalic ganglia will differentiate (Reuter et al., 1996) (Fig. 9A). Early activation of Pax6 at the level of cells located near the old nerve cords that reach the blastema suggests pivotal functions of these genes in the formation of neural structures in these organisms. A role for Drosophila ey in axon pathway selection during embryogenesis has recently been proposed by Noveen et al. (Noveen et al., 2000). In addition, it has been reported that severe defects in adult brain structures that are essential for vision, olfaction and the coordination of locomotion, are detectable in eyeless mutant Drosophila (Callaerts et al., 2001). As planarians are considered to be close relatives of primitive animals that acquired the CNS, further study of the role of Pax6 genes during CNS regeneration will be of interest for understanding the evolution of the genetic program which triggers brain formation in higher organisms.

Owing to the low expression level, Pax6A and Pax6B transcripts were not detected in the eye cells in either intact or regenerating planarians by conventional in situ hybridization on paraffin sections with digoxigenin-labeled riboprobes. However, DtPax6B expression in the eye cells was detected with a more sensitive in situ hybridization method using radioactive riboprobes (Callaerts et al., 1999). TEM in situ hybridization also revealed Pax6A and Pax6B mRNA in eye cells (Fig. 5) (Callaerts et al., 1999). These transcripts were distributed both in pigment eye cells and in different subcellular compartments of photoreceptors, i.e. throughout the perikaryon, and in the rhabdomeres. The presence of Pax6 transcripts has recently been monitored by competitive RT-PCR in adult human lens epithelium, cornea and monkey retina (Zhang et al., 2001). Moreover, Pax6 expression persists in amacrine and ganglion cells of the mature retina (Ashery-Padan and Gruss 2001; Marquardt et al., 2001). On the whole, these results support a role of Pax6 in the maintenance of eye cells.

Reduction of Pax6A and Pax6B endogenous transcripts by RNAi indicates that both genes are functionally dispensable in eye regeneration

In many organisms, the Pax6 transcription factor is critical for eye formation, as well as in patterning the CNS (Quiring et al., 1994; Kurusu et al., 2000; Pratt et al., 2000). As the basic functioning of the eyes in capturing photons and transmitting the information to the brain is similar in all animals, the presence of Pax6 transcripts in light-sensitive cells and pigment cells of planarian eye strongly suggested a conserved role of both Pax6 genes in this structure. Pax6 is considered very ancient and it has been indicated that the ancestral role of this gene was to construct a light-sensitive unit by direct regulation of opsin expression (Sheng et al., 1997; Pichaud et al., 2001). The primitive eye of basal metazoans such as planarians is the most suitable model system for studying Pax6 ancient function(s) in visual structures (Gehring and Ikeo, 1999).

The use of dsRNA to disrupt gene expression is a powerful method of achieving RNA interference in planarians (Sanchèz-Alvarado and Newmark, 1999). Thus, complete loss of eye has been obtained after Gtsix-1 dsRNA injection in planarians regenerating a head (Pineda et al., 2000). Moreover RNAi-mediated depletion of opsin mRNA also leads to the loss of phototactic behavior in these animals (Pineda et al., 2001).

Our experiments using dsRNA synthesized by Pax6A and Pax6B provide strong evidence that RNAi acts by decreasing endogenous cognate mRNA levels. The reduction of these gene products was comparable with that obtained for the corresponding RNAs in Gtsix-1 or Gtops RNAi experiments, which give rise to abnormal eye phenotypes. Despite the drastic RNAi-induced reduction in Pax6A and Pax6B transcripts, we did not observe any gross morphological alterations of the CNS in intact planarians or during regeneration. Moreover, eyes formed without any sign of defects during head regeneration and contained several photoreceptors comparable with those found in the eye of water-injected controls. A simple interpretation of these results is that the genetic network that controls eye formation in planarians is not triggered by Pax6 genes. The low level of Pax6 expression in the eye cells opens an alternative hypothesis, that both planarian Pax6 genes control eye cell fate decisions by a dose-independent mechanism, insensitive to the RNAi-induced transcript reduction. We favor the former possibility, as gene dose appears to be a fundamental requirement for the activity of Pax transcription factors, Pax6 being one of the best documented examples (Schedl et al., 1996; Van Raamsdonk and Tilghman, 2000). An exhaustive search for other Pax6 did not yield any additional Pax6 gene in either planarian species, making an eye-specific Pax6 highly improbable. If Pax6A and Pax6B genes are not crucial for eye induction during regeneration in planarians, we can speculate that other genes play a related role and substitute or compensate Pax6 action in the regulatory eye network. In this respect, the planarian Gtsix-1 gene, which is essential for eye regeneration (Pineda et al., 2000), could represent a putative candidate. The finding that Pax6 is not expressed during ganglionic photoreceptors (Joseph cells and organs of Hesse) development in Amphioxus (Glardon et al., 1998), the demonstration that Rx, but not Pax6, is essential for the formation of retinal progenitor cells in mice (Zhang et al., 2000), and our data on a Pax6-independent eye regeneration process in two species of planarians support the hypothesis that more than one molecular pathway can generate functional visual cells. Moreover, the recent hypothesis that other factors acting in parallel to Pax6 during retinal development can compensate Pax6 function (Ashery-Padan and Gruss, 2001) supports the possibility that several alternative combinations could give rise to the same phenotypic structure. The selection of an alternative combination could be promoted by a peculiar developmental scenario, i.e. blastemal regeneration. In this context, the Pax6-independent eye regeneration can be considered a remarkable example of such flexibility. Further analysis of the role of Pax6 during planarian eye development could contribute to establishing similarities between the processes of eye development and eye regeneration.

Fig. 1.

Amino acid sequence comparison of Pax6A and Pax6B from D. japonica and G. tigrina. (A) Multiple alignment of DjPax6A, GtPax6A, and DjPax6B, GtPax6B sequences. Amino acids that are identical both in Pax6A and Pax6B are shown in blue; amino acids conserved between DjPax6A and GtPax6A are indicated in green; amino acids conserved between DjPax6B and GtPax6B are indicated in red; differing amino acids are shown in black. The paired domain, the conserved motif found in the linker region, and the homeodomain are boxed. The missing sequence in the 5′ end of GtPax6A is indicated by dots; the introduced gaps are indicated by dashes. (B,C) Sequence comparison of the paired domain (B), and the homeodomain (C) of the planarian Pax6 proteins to the Pax6 paired domain and homeodomain of different species. Only Homo sapiens and Mus musculus Pax6 are shown for vertebrates. Drosophila eyless and toy are included as invertebrate Pax6. The shaded bars indicate Pax6-specific amino acids. Percentages of sequence identity (%I), determined by comparison with Homo sapiens and Mus musculus Pax6, are indicated at the end of each line. The structure of paired domain and homeodomain are shown on the top in B and C, respectively. Identical amino acids are indicated by dots. The incomplete sequence of GtPax6A is indicated by a dashed line. Non conserved amino acid residues are indicated in bold. The homeodomain flanking regions are shown in C.

Fig. 1.

Amino acid sequence comparison of Pax6A and Pax6B from D. japonica and G. tigrina. (A) Multiple alignment of DjPax6A, GtPax6A, and DjPax6B, GtPax6B sequences. Amino acids that are identical both in Pax6A and Pax6B are shown in blue; amino acids conserved between DjPax6A and GtPax6A are indicated in green; amino acids conserved between DjPax6B and GtPax6B are indicated in red; differing amino acids are shown in black. The paired domain, the conserved motif found in the linker region, and the homeodomain are boxed. The missing sequence in the 5′ end of GtPax6A is indicated by dots; the introduced gaps are indicated by dashes. (B,C) Sequence comparison of the paired domain (B), and the homeodomain (C) of the planarian Pax6 proteins to the Pax6 paired domain and homeodomain of different species. Only Homo sapiens and Mus musculus Pax6 are shown for vertebrates. Drosophila eyless and toy are included as invertebrate Pax6. The shaded bars indicate Pax6-specific amino acids. Percentages of sequence identity (%I), determined by comparison with Homo sapiens and Mus musculus Pax6, are indicated at the end of each line. The structure of paired domain and homeodomain are shown on the top in B and C, respectively. Identical amino acids are indicated by dots. The incomplete sequence of GtPax6A is indicated by a dashed line. Non conserved amino acid residues are indicated in bold. The homeodomain flanking regions are shown in C.

Fig. 2.

Phylogenetic tree of Pax genes. The tree was derived from the homeodomain and flanking regions of representatives from different Pax classes (indicated on the right), using the Neighbor-joining method. Sequences reported in this paper are in bold. Bootstrap percentage values (1000 replicates) are shown over the corresponding nodes. All branch lengths are proportional to the distances between sequences. The tree was rooted using Hydra PaxB as an outgroup. The analysis includes: human Pax 3, 4, 6 (Aniridia) and 7; mouse Pax3, 4, 6 and 7; zebrafish Pax6; sea urchin suPax6; chordate PmPax6 and Amphi-Pax6; mollusc Lo-Pax6; Drosophila ey and toy; C. elegans Cevab-3; nemertine LsPax6; planarian GtPax6A and GtPax6B, DjPax6A and DjPax6B; and cnidarian PaxC and Hl PaxB.

Fig. 2.

Phylogenetic tree of Pax genes. The tree was derived from the homeodomain and flanking regions of representatives from different Pax classes (indicated on the right), using the Neighbor-joining method. Sequences reported in this paper are in bold. Bootstrap percentage values (1000 replicates) are shown over the corresponding nodes. All branch lengths are proportional to the distances between sequences. The tree was rooted using Hydra PaxB as an outgroup. The analysis includes: human Pax 3, 4, 6 (Aniridia) and 7; mouse Pax3, 4, 6 and 7; zebrafish Pax6; sea urchin suPax6; chordate PmPax6 and Amphi-Pax6; mollusc Lo-Pax6; Drosophila ey and toy; C. elegans Cevab-3; nemertine LsPax6; planarian GtPax6A and GtPax6B, DjPax6A and DjPax6B; and cnidarian PaxC and Hl PaxB.

Fig. 3.

Expression of Pax6A mRNA in an intact planarian, as detected by whole-mount in situ hybridization. (A) Dorsal view of GtPax6A expression in G. tigrina. (B-G) Anteroposterior sequence of some representative transverse cryosections of whole-mount depicted in A. GtPax6A is expressed in the cephalic ganglia and the nerve cords. Presence of GtPax6A transcripts can also be observed in the lateroposterior region, close to the dorsoventral border. (H) Nomarski view of a higher magnification of a transverse cryosection of the whole-mount in A, showing the localization of GtPax6A-labelled cells (arrowheads) in the lateroventral marginal region, close to the eosinophilic secretory cells (arrows). Scale bars: 0.5 mm.

Fig. 3.

Expression of Pax6A mRNA in an intact planarian, as detected by whole-mount in situ hybridization. (A) Dorsal view of GtPax6A expression in G. tigrina. (B-G) Anteroposterior sequence of some representative transverse cryosections of whole-mount depicted in A. GtPax6A is expressed in the cephalic ganglia and the nerve cords. Presence of GtPax6A transcripts can also be observed in the lateroposterior region, close to the dorsoventral border. (H) Nomarski view of a higher magnification of a transverse cryosection of the whole-mount in A, showing the localization of GtPax6A-labelled cells (arrowheads) in the lateroventral marginal region, close to the eosinophilic secretory cells (arrows). Scale bars: 0.5 mm.

Fig. 4.

In situ hybridization and camera lucida drawings of transverse paraffin sections from the cephalic region of D. japonica. (A) The CNS in an intact planarian, which is composed of a mass of nerve cells, the cephalic ganglia (cg) and a pair of ventral nerve cords (nc). I, II and III indicate paraffin section levels. (B) The planarian eye. pc, pigment cell; phc, photoreceptor cell. (C-E) Anteroposterior sequence of transverse paraffin sections, visualized after in situ hybridization with Djsyt. (F-H) Camera lucida drawings of sections in C-E, illustrating the various morphological structures. (I-M) Anteroposterior sequence of transverse paraffin sections, visualized after in situ hybridization with DjPax6A. (N-P) Camera lucida drawings of sections in I-M, illustrating the various morphological structures. The nerve cell marker Djsyt labels all nerve cells, including the photoreceptors, while DjPax6A is expressed only in a subset of nerve cells. No detectable DjPax6A expression is observed in visual cells. Scale bars: 0.05 mm.

Fig. 4.

In situ hybridization and camera lucida drawings of transverse paraffin sections from the cephalic region of D. japonica. (A) The CNS in an intact planarian, which is composed of a mass of nerve cells, the cephalic ganglia (cg) and a pair of ventral nerve cords (nc). I, II and III indicate paraffin section levels. (B) The planarian eye. pc, pigment cell; phc, photoreceptor cell. (C-E) Anteroposterior sequence of transverse paraffin sections, visualized after in situ hybridization with Djsyt. (F-H) Camera lucida drawings of sections in C-E, illustrating the various morphological structures. (I-M) Anteroposterior sequence of transverse paraffin sections, visualized after in situ hybridization with DjPax6A. (N-P) Camera lucida drawings of sections in I-M, illustrating the various morphological structures. The nerve cell marker Djsyt labels all nerve cells, including the photoreceptors, while DjPax6A is expressed only in a subset of nerve cells. No detectable DjPax6A expression is observed in visual cells. Scale bars: 0.05 mm.

Fig. 5.

TEM in situ hybridization on D. japonica. (A-D) DjPax6A antisense-strand RNA; (F) DjPax6A sense-strand RNA; (E,G) Dj18S antisense-strand RNA; (H) DjSyt antisense-strand RNA. (A) Low magnification of the pigment cup ocellus showing some nuclei of photosensitive cells. The box indicates the figure shown in B. (B) Enlargement of A with clusters of gold particles (arrows) on the cytoplasm. (C) Clusters of gold particles (arrows) on the dendrite and the rhabdomeric region of a photosensitive cell. (D) A cluster of gold particles on the perinuclear cytoplasm of a pigment cell. (E) Clusters of gold particles (arrows) on the nucleolus and the endoplasmic reticulum of a cell. (F) No cluster of gold particles is visible on the cytoplasm of a pigment cell or in the rhabdomeric region of a photoreceptor cell. (G) A large cluster of gold particles (arrows) on the rhabdomeric region of a photoreceptor cell. (H) A cluster of gold particles (arrow) on the perikaryon of a photoreceptor cell. d, dendritic region of the photosensitive cell; n, nucleus; nu, nucleolus; p, pigment cup; pc, pigment cell; r, rhabdomeres. Scale bars: 5 μm in A; 0.5 μm in B-H).

Fig. 5.

TEM in situ hybridization on D. japonica. (A-D) DjPax6A antisense-strand RNA; (F) DjPax6A sense-strand RNA; (E,G) Dj18S antisense-strand RNA; (H) DjSyt antisense-strand RNA. (A) Low magnification of the pigment cup ocellus showing some nuclei of photosensitive cells. The box indicates the figure shown in B. (B) Enlargement of A with clusters of gold particles (arrows) on the cytoplasm. (C) Clusters of gold particles (arrows) on the dendrite and the rhabdomeric region of a photosensitive cell. (D) A cluster of gold particles on the perinuclear cytoplasm of a pigment cell. (E) Clusters of gold particles (arrows) on the nucleolus and the endoplasmic reticulum of a cell. (F) No cluster of gold particles is visible on the cytoplasm of a pigment cell or in the rhabdomeric region of a photoreceptor cell. (G) A large cluster of gold particles (arrows) on the rhabdomeric region of a photoreceptor cell. (H) A cluster of gold particles (arrow) on the perikaryon of a photoreceptor cell. d, dendritic region of the photosensitive cell; n, nucleus; nu, nucleolus; p, pigment cup; pc, pigment cell; r, rhabdomeres. Scale bars: 5 μm in A; 0.5 μm in B-H).

Fig. 6.

Expression of GtPax6A mRNA in regenerating G. tigrina, as detected by whole-mount in situ hybridization. (A-F) Dorsal view of fragments regenerating a head. Anterior is towards the top. (A) Activation of GtPax6A after 2 days of regeneration is detected as two hybridization spots located in the region close to the sectioned old nerve cords, where new cephalic ganglia are forming. (B) After 3 days of regeneration, GtPax6A-positive spots merge and follow the fibers emerging from the old nerve cords. (C-F) After 6, 8, 11 and 15 days of regeneration, GtPax6A expression becomes broader and follows the regenerating cephalic ganglia. (G) Dorsal view of a posterior regeneration. Six days after cutting, two labeling spots in the area corresponding to the regenerating nerve cords are observed. (H,I) Dorsal view of a lateral regeneration. An increased level of GtPax6A expression is observed where a cephalic ganglion is regenerating. A basal expression is detected in the corresponding non-regenerating ganglion. Scale bars: 0.5 mm.

Fig. 6.

Expression of GtPax6A mRNA in regenerating G. tigrina, as detected by whole-mount in situ hybridization. (A-F) Dorsal view of fragments regenerating a head. Anterior is towards the top. (A) Activation of GtPax6A after 2 days of regeneration is detected as two hybridization spots located in the region close to the sectioned old nerve cords, where new cephalic ganglia are forming. (B) After 3 days of regeneration, GtPax6A-positive spots merge and follow the fibers emerging from the old nerve cords. (C-F) After 6, 8, 11 and 15 days of regeneration, GtPax6A expression becomes broader and follows the regenerating cephalic ganglia. (G) Dorsal view of a posterior regeneration. Six days after cutting, two labeling spots in the area corresponding to the regenerating nerve cords are observed. (H,I) Dorsal view of a lateral regeneration. An increased level of GtPax6A expression is observed where a cephalic ganglion is regenerating. A basal expression is detected in the corresponding non-regenerating ganglion. Scale bars: 0.5 mm.

Fig. 7.

Expression of Djops in regenerating D. japonica after injection with Pax6A/Pax6B dsRNA mixture, visualized by in situ hybridization on transverse paraffin sections at the eye level. (A,B) Nomarski images of Djops-expressing photoreceptors in a dsRNA-injected animal (A) and in a water-injected control (B), after 15 days of regeneration. Scale bar: 0.015 mm.

Fig. 7.

Expression of Djops in regenerating D. japonica after injection with Pax6A/Pax6B dsRNA mixture, visualized by in situ hybridization on transverse paraffin sections at the eye level. (A,B) Nomarski images of Djops-expressing photoreceptors in a dsRNA-injected animal (A) and in a water-injected control (B), after 15 days of regeneration. Scale bar: 0.015 mm.

Fig. 8.

Effects of GtPax6A and GtPax6B dsRNA injection in regenerating G. tigrina. (A) Visualization of comparative RT-PCR experiments. Relative levels of endogenous transcripts in water-injected controls and in GtPax6A, GtPax6B dsRNA-injected animals are shown. Reduction of the Gtops mRNA level after Gtops dsRNA injection is shown as a comparison. Expression of the homeobox gene Dth2 is unaffected by GtPax6A and GtPax6B RNAi experiments. (B-E) Dorsal view of regenerating heads visualized by GtPax6A whole-mount in situ hybridization. (B-C) Control animals injected with water. The typical arch-shaped GtPax6A expression is observed in cephalic regeneration after 6 (B) or 9 (C) days from cutting. (D,E) After injection with GtPax6A/GtPax6B dsRNA mixture, no GtPax6A hybridization signal is detected in regenerating animals after 6 (D) and 9 (E) days from cutting. Regenerating eyespots can be observed both in the controls and in planarians injected with GtPax6A/GtPax6B dsRNA mixture. Scale bars: 0.5 mm.

Fig. 8.

Effects of GtPax6A and GtPax6B dsRNA injection in regenerating G. tigrina. (A) Visualization of comparative RT-PCR experiments. Relative levels of endogenous transcripts in water-injected controls and in GtPax6A, GtPax6B dsRNA-injected animals are shown. Reduction of the Gtops mRNA level after Gtops dsRNA injection is shown as a comparison. Expression of the homeobox gene Dth2 is unaffected by GtPax6A and GtPax6B RNAi experiments. (B-E) Dorsal view of regenerating heads visualized by GtPax6A whole-mount in situ hybridization. (B-C) Control animals injected with water. The typical arch-shaped GtPax6A expression is observed in cephalic regeneration after 6 (B) or 9 (C) days from cutting. (D,E) After injection with GtPax6A/GtPax6B dsRNA mixture, no GtPax6A hybridization signal is detected in regenerating animals after 6 (D) and 9 (E) days from cutting. Regenerating eyespots can be observed both in the controls and in planarians injected with GtPax6A/GtPax6B dsRNA mixture. Scale bars: 0.5 mm.

Fig. 9.

Dorsal view of cephalic ganglia regeneration in G. tigrina after injection of GtPax6A/GtPax6B dsRNA mixture, visualized by whole-mount neuropeptide FMRFamide immunoreactivity pattern at different regenerative stages. (A,C) After 5 (A) and 9 days (C) of regeneration, new transversal commissures produced from the amputated old nerve cords and differentiating the proximal cephalic ganglia (arrows) can be observed in water-injected controls. The distal part of the new cephalic ganglia is organizing an arch-shaped structure that connects the nerve cords (arrowhead). (B,D) No abnormal cephalic ganglia regeneration is apparent in corresponding head-regenerating fragments subsequent to injection of GtPax6A/GtPax6B dsRNA mixture. Regenerating eyes appear as small black spots of similar size in both C and D. (E,F) After 20 days from cutting, a complete regeneration of the cephalic ganglia can be observed. No differences in the CNS pattern can be seen between a water-injected control (E) and an dsRNA-injected planarian (F). Scale bars: 0.5 mm.

Fig. 9.

Dorsal view of cephalic ganglia regeneration in G. tigrina after injection of GtPax6A/GtPax6B dsRNA mixture, visualized by whole-mount neuropeptide FMRFamide immunoreactivity pattern at different regenerative stages. (A,C) After 5 (A) and 9 days (C) of regeneration, new transversal commissures produced from the amputated old nerve cords and differentiating the proximal cephalic ganglia (arrows) can be observed in water-injected controls. The distal part of the new cephalic ganglia is organizing an arch-shaped structure that connects the nerve cords (arrowhead). (B,D) No abnormal cephalic ganglia regeneration is apparent in corresponding head-regenerating fragments subsequent to injection of GtPax6A/GtPax6B dsRNA mixture. Regenerating eyes appear as small black spots of similar size in both C and D. (E,F) After 20 days from cutting, a complete regeneration of the cephalic ganglia can be observed. No differences in the CNS pattern can be seen between a water-injected control (E) and an dsRNA-injected planarian (F). Scale bars: 0.5 mm.

Fig. 10.

Expression of GtSix-1 and Gtops in regenerating G. tigrina after injection with Pax6A/Pax6B dsRNA mixture, visualized by whole-mount in situ hybridization. Dorsal view of head regenerating fragments after 3, 6 and 9 days from cutting. (A-C) Water-injected controls hybridized with Gtsix-1. (D-F) GtPax6A/GtPax6B dsRNA mixture-injected organisms hybridized with GtSix1. (G-I) GtPax6A/GtPax6B dsRNA mixture-injected organisms hybridized with Gtops. (A,D) A faint GtSix1 hybridization signal is visible after 3 days of regeneration both in the controls and in injected animals. Later on, GtSix1 mRNA is clearly visualized at the eye level. No differences are detected between controls (B,C) and injected animals (E,F). (G-I) A normal expression pattern of Gtops mRNA can also be detected in animals injected with GtPax6A/GtPax6B dsRNA. Scale bars: 0.5 mm.

Fig. 10.

Expression of GtSix-1 and Gtops in regenerating G. tigrina after injection with Pax6A/Pax6B dsRNA mixture, visualized by whole-mount in situ hybridization. Dorsal view of head regenerating fragments after 3, 6 and 9 days from cutting. (A-C) Water-injected controls hybridized with Gtsix-1. (D-F) GtPax6A/GtPax6B dsRNA mixture-injected organisms hybridized with GtSix1. (G-I) GtPax6A/GtPax6B dsRNA mixture-injected organisms hybridized with Gtops. (A,D) A faint GtSix1 hybridization signal is visible after 3 days of regeneration both in the controls and in injected animals. Later on, GtSix1 mRNA is clearly visualized at the eye level. No differences are detected between controls (B,C) and injected animals (E,F). (G-I) A normal expression pattern of Gtops mRNA can also be detected in animals injected with GtPax6A/GtPax6B dsRNA. Scale bars: 0.5 mm.

We are grateful to M. Domínguez and F. Casares for helpful comments on the manuscript. We thank anonymous referees for their comments, which helped to improve the manuscript. We thank Dr Hidefumi Orii for providing us with D. japonica GI clonal strain, and with Dj18S and DjEF2 clones, and also Robin Rycroft for checking the English. This work was supported by a grant from the DGICYT to E. S. (Ministerio de Educación y Ciencia, Spain, PB98-1261-C02-01), from MURST-Italy (Cofinanziamento Programmi di Ricerca di Interesse Nazionale) to R. B. D. P. and M. M. are the recipients of two ‘Formación de Personal Investigador’ fellowships from Universitat de Barcelona and from Ministerio de Educación y Ciencia, respectively.

Agata, K., Soejima, Y., Kato, K., Kobayashi, C., Umesono, Y. and Watanabe, K. (
1998
). Structure of the planarian central nervous system (CNS) revealed by neuronal cell markers.
Zool. Sci
.
15
,
433
-440.
Aguinaldo, A., Turbeville, J., Linford, L., Rivera, M., Garey, J., Raff, R. and Lake, J. (
1997
). Evidence for a clade of nematodes, arthropods and other moulting animals.
Nature
387
,
489
-493.
Ashery-Padan, R. and Gruss, P. (
2001
). Pax6 lights-up the way for eye development.
Curr. Opin. Cell Biol
.
13
,
706
-714.
Baguñá, J. and Ballester, R. (
1978
). The nervous system in planarians: peripheral and gastrodermal plexuses, pharynx innervation, and the relationship between central nervous system structure and the acelomate organization.
J. Morphol
.
155
,
237
-252.
Bayascas, J. R., Castillo, E. and Saló, E. (
1998
). Platyhelminthes have a Hox code differentially activated during regeneration, with genes closely related to those of spiralian protostomes.
Dev. Genes Evol
.
208
,
467
-473.
Callaerts, P., Halder, G. and Gehring, W. J. (
1997
). Pax6 in development and evolution.
Annu. Rev. Neurosci
.
20
,
483
-532.
Callaerts, P., Muñoz-Marmol, A. M., Glardon, S., Castillo, E., Sun, H., Li, W. H., Gehring, W. J. and Saló, E. (
1999
). Isolation and expression of a Pax-6 gene in the regenerating and intact Planarian Dugesia(G)tigrina.
Proc. Natl. Acad. Sci. USA
96
,
558
-563.
Callaerts, P., Leng, S., Clements, J., Benassayag, C., Cribbs, D., Kang, Y. Y., Walldorf, U., Fischbach, K. F. and Strauss, R. (
2001
). Drosophila Pax-6/eyeless is essential for normal adult brain structure and function.
J. Neurobiol
.
46
,
73
-88.
Carranza, S., Baguñà, J. and Riutort, M. (
1997
). Are the Platyhelminthes a Monophyletic Primitive group? An assessment using 18S rDNA sequences.
Mol. Biol. Evol
.
14
,
485
-497.
Cebrià, F., Vispo, M., Newmark, P., Bueno, D. and Romero R. (
1997
). Myocite differentiation and body wall muscle regeneration in the planarian Girardia tigrina.
Dev. Genes Evol
.
207
,
306
-316.
Chisholm, A. D. and Horvitz, H. R. (
1995
). Patterning of the Caenorhabditis elegans head region by the Pax-6 family member vab-3.
Nature
377
,
52
-55.
Chow, R. L., Altmann, C. R., Lang, R. A. and Hemmati-brivanlou, A. (
1999
). Pax6 induces ectopic eyes in a vertebrate.
Development
126
,
4213
-4222.
Czerny, T. and Busslinger, M. (
1995
). DNA-binding and transactivation properties of Pax-6: three amino acids in the paired domain are responsible for the different sequence recognition of Pax-6 and BSAP (Pax-5).
Mol. Cell. Biol
.
15
,
2858
-2871.
Czerny, T., Halder, G., Kloter, U., Souabni, A., Gehring, W. J., Busslinger, M. (
1999
). twin of eyeless, a second Pax-6 gene of Drosophila, acts upstream of eyeless in the control of eye development.
Mol. Cell
3
,
297
-307.
Garcia-Fernàndez, J., Baguñá, J. and Saló, E. (
1991
). Planarian homeobox genes: cloning, sequence analysis and expression.
Proc. Natl. Acad. Sci. USA
88
,
7338
-7342.
Garcia-Fernàndez, J., Baguñà, J. and Saló, E. (
1993
). Genomic organization and expression of the planarian homeobox genes Dth-1 and Dth-2.
Development
118
,
241
-253.
Gehring, W. J. and Ikeo, K. (
1999
). Pax-6 mastering eye morphogenesis and eye evolution.
Trends Genet
.
15
,
371
-377.
Glardon, S., Callaerts, P., Halder, G. and Gehring, W. J. (
1997
). Conservation of Pax-6 in a lower chordate, the ascidian Phallusia mammillata.
Development
124
,
817
-825.
Glardon, S., Holland, L. Z., Gehring, W. J. and Holland, N. D. (
1998
). Isolation and developmental expression of the amphioxus Pax-6 gene (AmphiPax-6): insights into eye and photoreceptor evolution.
Development
125
,
2701
-2710.
Kishida, Y. (
1967
). Electron microscopic studies on the planarian eye. Fine structures of the normal eye.
Sci. Rep. Kanazawa Univ
.
12
,
75
-110.
Kobayashi, C., Watanabe, K. and Agata, K. (
1998
). The process of pharynx regeneration in planarians.
Dev. Biol
.
211
,
27
-38.
Kumar, J. P. and Moses, K. (
2001
). EGF receptor and Notch signaling act upstream of Eyeless/Pax6 to control eye specification.
Cell
104
,
687
-697.
Kurata, S., Go, M. J., Artavanis-Tsakonas, S. and Gehring,W. J. (
2000
). Notch signaling and the determination of appendage identity.
Proc. Natl. Acad. Sci. USA
97
,
2117
-2122.
Kurusu, M., Nagao, T., Walldorf, U., Flister, S., Gehring, W. J. and Furukubo-Tokunaga, K. (
2000
). Genetic control of development of the mushroom bodies, the associative learning centers in the Drosophila brain, by the eyeless, twin of eyeless, and dachshund genes.
Proc. Natl. Acad. Sci. USA
97
,
2140
-2144.
Le Guellec, D., Frappart, L. and Desprez, P. Y. (
1991
). Ultrastructural localization of mRNA encoding for the receptor in human breast cell cancer line BT20 by in situ hybridization.
J. Histochem. Cytochem
.
39
,
1
-6.
Loosli, F., Kmita-Cunisse, M. and Gehring, W. J. (
1996
). Isolation of a Pax-6 homolog from the ribbonworm Lineus sanguineus.
Proc. Natl. Acad. Sci. USA
93
,
2658
-2663.
Marquardt, T., Ashery-Padan, R., Andrejewski, N., Scardigli, R., Guillemot, F. and Gruss, P. (
2001
). Pax6 is required for the multipotent state of retinal progenitor cells.
Cell
105
,
43
-55.
Nei, M. and Koehn, R. (
1983
). Evolution of genes and proteins. In The Neutral Theory of Molecular Evolution (ed. M. Nei and R. Koehn), pp. 208-233. Sunderland, MA: Sinauer Associates.
Nornes, S., Clarkson, M., Mikkola, I., Pedersen, M., Bardsley, A., Martinez, J. T., Krauss, S. and Johansen, T. (
1998
). Zebrafish contains two pax6 genes involved in eye development.
Mech. Dev
.
77
,
185
-196.
Noveen, A., Daniel, A. and Hartenstein, V. (
2000
). Early development of the Drosophila mushroom body: the roles of eyeless and dachshund.
Development
127
,
3475
-3488.
Orii, H., Agata, K. and Watanabe, K. (
1993
). POU-domain genes in planarian Dugesia japonica: the structure and expression.
Biochem. Biophys. Res. Commun
.
192
,
1395
-1402.
Pichaud, F., Treisman, J. and Desplan, C. (
2001
). Reinventing a common strategy for patterning the eye.
Cell
105
,
9
-12.
Pineda, D., Gonzalez, J., Callaerts, P., Ikeo, K., Gehring, W. J. and Saló, E. (
2000
). Searching for the prototypic eye genetic network: sine oculis is essential for eye regeneration in planarians.
Proc. Natl. Acad. Sci. USA
97
,
4525
-4529.
Pineda, D., Gonzalez, J., Marsal, M. and Saló, E. (
2001
). Evolutionary conservation of the initial eye genetic pathway in planarians.
Belg. J. Zool
.
131
,
85
-90.
Pratt, T., Vitalis, T., Warren, N., Edgar, J. M., Mason, J. O. and Price, D. J. (
2000
). A role of Pax6 in the normal development of dorsal thalamus and its cortical connections.
Development
127
,
5167
-5178.
Quiring, R., Walldorf, U., Kloter, U. and Gehring, W. J. (
1994
). Homology of the eyeless gene of Drosophila to Small eye gene in mice and Aniridia in humans.
Science
265
,
785
-789.
Reuter, M., Sheiman, I. M., Gustafsson, M. K. S., Halton, D. W., Maule, A. G. and Shaw, C. (
1996
). Development of the nervous system in Dugesia tigrina during regeneration after fission and decapitation.
Invert. Reprod. Dev
.
29
,
199
-211.
Rieger, R., Tyler, M., S., Smith III, J. P. S. and Rieger, G. R. (
1991
). Platyhelminthes: Turbellaria.
In Microscopic Anatomy of Invertebrates
(ed. F. W. Harrison and B. J. Bogitsch), pp. 7-140. New York: Wiley-Liss.
Rossi, L., Batistoni, R., Salvetti, A., Deri, P., Bernini, F., Andreoli, I., Falleni, A. and Gremigni, V. (
2001
). Molecular aspects of cell proliferation and neurogenesis in planarians.
Belg. J. Zool
.
131
,
5
-9.
Sakai, F., Agata, K., Orii, H. and Watanabe, K. (
2000
). Organization and regeneration ability of spontaneous supernumerary eyes in planarians – eye regeneration field and pathway selection by optic nerves.
Zool. Sci
.
17
,
375
-381.
Saló, E. and Baguñà, J. (
1984
). Regeneration and pattern formation in planarians. I. The pattern of mitosis in anterior and posterior regeneration in Dugesia (G) tigrina, and a new proposal for blastema formation.
J. Embryol. Exp. Morphol
.
83
,
63
-80.
Salvetti, A., Batistoni, R., Deri, P., Rossi, L. and Sommerville, J. (
1998
). Expression of DjY1, a protein containing a cold shock domain and RG repeat motifs, is targeted to sites of regeneration in planarians.
Dev. Biol
.
201
,
217
-229.
Sánchez-Alvarado, A. and Newmark, P. A. (
1999
). Double-stranded RNA specifically disrupts gene expression during planarian regeneration.
Proc. Natl. Acad. Sci. USA
96
,
5049
-5054.
Schedl, A., Ross, A., Lee, M., Engelkamp, D., Rashbass, P., Van Heyningen, V. and Hastle, N. (
1996
). Influence of Pax6 gene dosage on development: overexpression causes severe eye abnormalities.
Cell
86
,
71
-82.
Sheng, G., Thouvenot, E., Schmucker, D., Wilson, D. S. and Desplan, C. (
1997
). Direct regulation of rhodopsin 1 by Pax6/eyeless in Drosophila: evidence for a conserved function in photoreceptors.
Genes Dev
.
11
,
1122
-1131.
Tazaki, A., Gaudieri, S., Ikeo, K., Gojobori, T., Watanabe, K. and Agata, K. (
1999
). Neural network in planarian revealed by an antibody against planarian synaptotagmin homologue.
Biochem. Biophys. Res. Commun
.
260
,
426
-432.
Tomarev, S. I., Callaerts, P., Kos, L., Zinovieva, R., Halder, G., Gehring, W. J. and Piatigorsky, J. (
1997
). Squid Pax-6 and eye development.
Proc. Natl. Acad. Sci. USA
94
,
2421
-2426.
Turque, N., Plaza, S., Radvanyi, F., Carriere, C. and Saule, S. (
1994
). Pax-QNR/Pax6, a paired box and homeobox-containing gene expressed in neurons, is also expressed in pancreatic endocrine cells.
Mol. Endocrinol
.
8
,
929
-938.
Van Raamsdonk, C. D. and Tilghman, S. M. (
2000
). Dosage requirement and allelic expression of Pax6 during lens placode formation.
Development
127
,
5439
-5448.
Walther, P. and Gruss, P. (
1991
). Pax6, a murine paired-box gene is expressed in the developing CNS.
Development
113
,
1435
-1449.
Zhang, Y. and Emmons, S. W. (
1995
). Specification of sense-organ identity by a Caenorhabditis elegans Pax-6 homologue.
Nature
377
,
55
-59.
Zhang, L., Mathers, P. H. and Jamrich, M. (
2000
). Function of Rx, but not Pax6, is essential for the formation of retinal progenitor cells in mice.
Genesis
28
,
135
-142.
Zhang, W., Cveklova, K., Oppermann, B., Kantorow, M. and Cvekl, A. (
2001
). Quantitation of Pax6 and Pax6(5a) transcript levels in adult human lens, cornea, and monkey retina.
Mol. Vis
.
7
,
1
-5.