Regeneration-dependent conditional gene knockdown (Readyknock) in planarian: Demonstration of requirement for Djsnap-25 expression in the brain for negative phototactic behavior

Authors

  • Tomomi Takano,

    1. Center for Developmental Biology, RIKEN Kobe, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan,
    2. Department of Biology, Graduate School of Science and Technology, Kobe University, Japan and
    Search for more papers by this author
  • Jeremy N. Pulvers,

    1. Center for Developmental Biology, RIKEN Kobe, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan,
    Search for more papers by this author
  • Takeshi Inoue,

    1. Center for Developmental Biology, RIKEN Kobe, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan,
    Search for more papers by this author
  • Hiroshi Tarui,

    1. Genome Resource and Analysis Subunit, Center for Developmental Biology, RIKEN Kobe, Kobe, Japan
    Search for more papers by this author
  • Hiroshi Sakamoto,

    1. Department of Biology, Graduate School of Science and Technology, Kobe University, Japan and
    Search for more papers by this author
  • Kiyokazu Agata,

    1. Center for Developmental Biology, RIKEN Kobe, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan,
    Search for more papers by this author
  • Yoshihiko Umesono

    Corresponding author
    1. Center for Developmental Biology, RIKEN Kobe, 2-2-3 Minatojima-minamimachi, Chuo-ku, Kobe 650-0047, Japan,
      *Author to whom all correspondence should be addressed (present address: Department of Biophysics, Graduate School of Science, Kyoto University, Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan).
      Email: umesono@mdb.biophys.kyoto-u.ac.jp
    Search for more papers by this author

*Author to whom all correspondence should be addressed (present address: Department of Biophysics, Graduate School of Science, Kyoto University, Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan).
Email: umesono@mdb.biophys.kyoto-u.ac.jp

Abstract

Freshwater planarians have a simple and evolutionarily primitive brain structure. Here, we identified the Djsnap-25 gene encoding a homolog of the evolutionarily conserved synaptic protein SNAP-25 from the planarian Dugesia japonica and assessed its role in brain function. Djsnap-25 was expressed widely in the nervous system. To investigate the specific role of Djsnap-25 in the brain, we developed a unique technique of RNA interference (RNAi), regeneration-dependent conditional gene knockdown (Readyknock), exploiting the high regenerative capacity of planarians, and succeeded in selectively eliminating the DjSNAP-25 activity in the head region while leaving the DjSNAP-25 activity in the trunk region intact. These knockdown animals showed no effect on brain morphology or on undirected movement of the trunk itself. Light-avoidance behavior or negative phototaxis was used to quantitatively analyze brain function in the knockdown animals. The results suggested that the DjSNAP-25 activity within the head region is required for two independent sensory-processing pathways that regulate locomotive activity and directional movement downstream of distinct primary sensory outputs coming from the head margin and the eyes, respectively, during negative phototaxis. Our approach demonstrates that planarians are a powerful model organism to study the molecular basis of the brain as an information-processing center.

Introduction

All animals adapt to their surrounding environment by behaving appropriately. The brain has evolved as an information-processing center to deal with a variety of signals coming from the outside. An increase in brain size and complexity during evolution has enabled humans in particular to output a variety of sophisticated behavioral traits. One of the current greatest interests in the field of neuroscience is our understanding of what kinds of molecules were integrated into the ancestral brain lineage to establish the abilities of the brain during evolution.

Planarians are generally believed to be one of the simplest organisms with a central nervous system (CNS). The planarian brain has a bilobed structure with a cortex of nerve cells and a core of nerve fibers connected at the dorsal side of the ventral nerve cords (Agata et al. 1998). There are several experimental advantages in analyzing the role of brain function-related genes in the planarian. Planarians have a high regenerative ability, which enables analysis of the whole process of regeneration of the brain circuitry within 1 week (Inoue et al. 2004). Use of RNA interference (RNAi) enables a large-scale survey of planarian gene function, including the function of CNS-related genes involved in behavior (Sánchez Alvarado & Newmark et al. 1999; Inoue et al. 2004; Reddien et al. 2005). Of crucial importance is the establishment of an assay system to analyze the functional properties of the brain and its regions. When exposed to light, planarians display a distinct light avoidance behavior, known as negative phototaxis, via the light-sensing organ or eyes (MacRae 1964; Carpenter et al. 1974; Asano et al. 1998). This behavioral trait can be quantitatively measured in detail for parameters such as movement speed, distance and direction and therefore is suitable for behavioral analysis (Inoue et al. 2004). Sequential negative phototaxis assays during head regeneration could easily isolate functional properties of the brain from complex behavioral traits of the whole animals (Inoue et al. 2004). Furthermore, our previous studies strongly suggested that the genetic machineries and genetic programs used in planarian brain development are similar to those used in vertebrates (Umesono et al. 1997, 1999; Tazaki et al. 1999; Cebriàet al. 2002a, b, c; Ogawa et al. 2002b; Mineta et al. 2003; Nakazawa et al. 2003). These findings strongly validate the use of the planarian nervous system as a model to study the molecular basis of the brain function, and to elucidate ancestral features of the brain as an information-processing center.

The regulated release of neurotransmitters is a fundamental process throughout the animal kingdom. It is known that evolutionarily conserved synaptosome-associated protein 25 kDa (SNAP-25) is involved in regulated exocytosis required for stimulus-dependent neurotransmission (Molnar et al. 2002; Washbourne et al. 2002). However, the roles of the SNAP-25 activity in the primitive brain circuitry are still unknown. Here, we identified a planarian snap-25 homolog gene (Djsnap-25) and carried out functional silencing in the brain circuitry by a unique conditional gene knockdown approach. Subsequently, the knockdown planarians were assayed for alterations in behavior during negative phototaxis. Our findings indicated that Djsnap-25 was indispensable for the functioning of the brain as an information-processing center.

Materials and methods

Organisms

A clonal strain of the planarian Dugesia japonica originating from the Iruma River (Gifu prefecture, Japan), that was kept in autoclaved tap water at 22–25°C was used. In all experiments, planarians of 8–10 mm in length that had been starved for at least 1 week were used.

Whole-mount in situ hybridization

Animals were treated with 2% HCl in 5/8 Holtfreter's solution for 5 min at 4°C and fixed in 5/8 Holtfreter's solution containing 4% paraformaldehyde and 5% methanol for 3 h at 4°C. Hybridization and color detection of digoxygenin (DIG)-labeled RNA probes were carried out as previously described by Umesono et al. (1997). The pBluescript SK(–) containing a 1182-bp cDNA fragment of expressed sequence tag (EST) clone Dj_aH_007_G21 encoding Djsnap-25 was digested with NotI to synthesize an antisense (T7) DIG-RNA probe.

Synthesis of dsRNA

EST clone Dj_aH_007_G21 was digested with HindIII to generate two non-overlapping fragments of 602 bp (5′ region) and 580 bp (3′ region). Double-strand RNA (dsRNA) was synthesized as previously described by Sánchez Alvarado & Newmark (1999). pBluescript SK(–) containing the full sequence or a 5′ fragment was digested with KpnI or NotI to synthesize sense (T3) or antisense (T7) RNAs. The synthesized RNAs were digested with RQI RNase-free DNase (Promega, Madison, WI, USA) for 60 min at 37°C. After extraction with phenol/chloroform, the RNAs were denatured for 20 min at 65°C and annealed for 40 min at 37°C. dsRNA was purified by ethanol precipitation, and suspended in diethylpyrocarbonate (DEPC)-treated H2O. The electrophoretic mobilities of dsRNA and single-strand RNA were assessed in 1.0% agarose gels.

Microinjection and amputation

Intact planarians were injected with dsRNA daily for 3 days by using a Drummond Scientific NanojectII injector (Drummond Scientific, Broomall, PA, USA). Control animals were injected with dsRNA of the enhanced green fluorescent protein (EGFP) gene, which has no homology in the planarian genome. Approximately 3 h after the last injection, the animals were cut transversely into two pieces using a surgical blade for the phototaxis assay.

Quantitative RT-PCR

Head and trunk regions of 10 non-regenerated or 10 head-regenerated planarians injected with Djsnap-25 dsRNA or with EGFP dsRNA as a control (see detailed time schedule in Fig. 4A) were dissolved in sample buffer (125 mm Tris, Bromophenol Blue (BPB) 2 mg/mL, sodium dodecyl sulfate (SDS) 40 mg/mL, 20% glycerol, 4% 2-Me, pH 6.8) and were boiled for 2 min. Isolation of both RNA and protein from these experimental samples was carried out using a PARIS Kit (Ambion #1921, Applied Biosystems, Foster City, CA, USA). First-strand cDNA was synthesized by using a Quantitect Reverse Transcription kit (QIAGEN, Hilden, Germany). Quantitative analysis of the amount of each gene product was carried out as previously described by Ogawa et al. (2002a) using a 7900HT (Applied Biosystems) real-time polymerase chain reaction (PCR) machine. As an internal control, the constitutively transcribed gene GAPDH was amplified. A neuron-specific gene, Djsyt, was analyzed by real-time PCR to investigate the effects of the injected Djsnap-25 dsRNA.

Figure 4.

Regeneration-dependent conditional gene knockdown of Djsnap-25. (A–J) Silencing effect of the Djsnap-25 RNAi at the protein level. Immunohistochemical detection of Dugesia japonica synaptosome-associated protein 25 kDa (DjSNAP-25) in non-regenerated (A, B, E, F) and in head-regenerated (C, D, G, H) animals. Expression in the brain in A, B, C, D, and in the ventral nerve cords (VNCs) of the same animal in E, F, G, H, respectively. All sections were stained with Hoechst no. 33258 (for nuclei, shown in blue). Bar; 100 µm. Quantitative protein analysis by Western blotting in head region (I) and in trunk region (J). Severe reduction of DjSNAP-25 protein was observed only within the newly regenerated head region. (K) Silencing effect of the Djsnap-25 RNAi at the mRNA level with the same samples shown in I. Quantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis of Djsnap-25 mRNA indicated that the expression of Djsnap-25 mRNA was significantly reduced in the head region of both non-regenerated and head-regenerated animals. There was no obvious reduction of the Djsyt expression in either type of animal, indicating that the silencing effect was specific. Each gene expression level was normalized relative to the level of planarian GAPDH expression. Standard deviation from the mean is indicated by error bars. All experiments used EGFP dsRNA-injected animals as a control. (L) Experimental outline. The last day of injection was taken as day 0.

The PCR primers used were as follows:

  • Djsnap-25 forward, 5′-AAGAAGAAGCGGGCAAAGACA-3′;

  • Djsnap-25 reverse, 5′-TTGAACGCCCATTCCGTTT-3′;

  • Djsyt forward, 5′-GGCATTGGTCGGACATGTTG-3′;

  • Djsyt reverse, 5′-TTTCCGGCATTTCTTGGAGG-3′;

  • GAPDH forward, 5′-ACCACCAACTGTTTAGCTCCCTTA-3′;

  • GAPDH reverse, 5′-GATGGTCCATCAACAGTCTTTTGC-3′.

Immunohistochemistry

Animals were treated with 2% HCl in 5/8 Holtfreter's solution for 5 min at 4°C and fixed in 5/8 Holtfreter's solution containing 4% paraformaldehyde and 5% methanol for 3 h at 4°C. Fixed animals were embedded in paraffin and serially sectioned at 8-µm thickness. Samples were incubated with a mouse antirat SNAP-25 monoclonal antibody (Cat.No.111001 clone 71.1 Synaptic Systems, Göttingen, Germany) diluted 1:500 as the primary antibody, and with a secondary sheep antimouse antibody conjugated with alkaline phosphatase diluted 1:1000. Cell nuclei were labeled with Hoechst no. 33258 (Sigma, Saint Louis, MO, USA). For whole-mount immunostaining, animals were fixed in Carnoy's solution for 3 h at 4°C. The samples were incubated with the following primary antibodies: a polyclonal antibody against planarian synaptotagmin (DjSYT) diluted 1:2000, and a monoclonal antibody against visual neurons (VC-1) diluted at 1:10 000, respectively, and with a secondary goat antimouse antibody conjugated with Alexa 488 (A11001, Invitrogen, Carlsbad, CA, USA) or 594 (A11005, Invitrogen) diluted 1:400. The samples were observed with a confocal microscope LSM510 Ver.3.2 (Carl Zeiss, Thornwood, NY, USA). Images were processed using Photoshop software (Adobe, San Jose CA, USA).

Western blotting

Twelve head and trunk pieces of planarians were dissolved in sample buffer (125 mm Tris, BPB 2 mg/mL, SDS 40 mg/mL, 20% glycerol, 4% 2-Me, pH 6.8) by pipetting and were boiled for 2 min. Five microliters of each sample solution was subjected to gel electrophoresis. Gel electrophoresis was carried out in duplicate, and one gel was stained with Coomassie Simply Blue SafeStain (Invitrogen) to estimate the total amount of the protein applied, and the other was used for Western blotting. Western blotting was carried out using the mouse antirat SNAP-25 monoclonal antibody described above diluted 1:1000 as the primary antibody, and a secondary sheep antimouse antibody conjugated with alkaline phosphatase diluted 1:500. A mixture of BCIP/NBT (5-bromo-4-chloro-3′-indolylphosphatase p-toluidine salt and nitro-blue tetrazolium chloride) was used for color development of the alkaline phosphatase.

For protein quantification, the absorbance of the extract at 280 nm was measured by ultraviolet absorption spectrometry. About 6.5 ng of total protein in the head region and 30 ng in the trunk region were subjected to gel electrophoresis. Western blotting was carried out using the mouse antirat SNAP-25 monoclonal antibody described above diluted 1:250 as the primary antibody, and mouse anti-actin monoclonal antibody (Abcam, ab40864, Cambridge, UK) diluted 1:1000 as an internal control, and a secondary goat antimouse antibody conjugated with the horseradish peroxidase diluted 1:2000 (included in T20912 TSA kit, Invitrogen). Amersham ECL Plus (GE Healthcare, Buckinghamshire, UK) was used as a substrate for the detection of the signals using a LAS-3000mini luminescent image analyzer (FUJIFILM, Tokyo, Japan). The DjSNAP-25 expression level was normalized relative to the level of planarian actin expression using MultiGauge standard image analysis software (FUJIFILM).

Phototaxis assay system

The phototaxis assay was carried out as previously described by Inoue et al. (2004). Seven to ten individual animals were used in a series of phototaxis assays. Every experiment was carried out twice independently to confirm that the results were reproducible.

Statistical analysis

The statistical significance of differences was determined by a two-sided Student's t-test; P-values greater than 0.05 were taken as not significant (NS). The time spent in the target quadrant and the moving distances moved were expressed as means ± SEM from independent animals.

Analysis of the direction of movement

The trajectory of the movement of the planarian during negative phototaxis was recorded at 200-millisecond intervals for a total of 90 s using SMART v2.0 behavior analysis software (Panlab, Barcelona, Spain). The angle of movement was also recorded at each 200-millisecond time point. This angle data representing the direction of movement was used to phenotype the phototactic behavior in detail. Analysis was carried out using the statistical package R (http://www.r-project.org). The data from zones 1 and 2 were used as they represent the initial response to light exposure. The data were binned into 36 intervals of 10 degrees and represented in a circular graph. A grayscale gradient was generated with black representing the highest number of angle counts that fell within that 10-degree interval and white indicating no movement in the direction of that 10-degree range. Therefore, the darker shades represent the directions toward which the planarians showed the greatest movement. The data are the combined results of the individuals examined. A gradient was generated separately for each circular graph. The top of the circle represents closest to the light source and the bottom represents the furthest from it.

Results

Isolation of Snap-25 gene homolog from planarian head EST database

We searched for neural function-related genes in our EST database (50 655 sequences) constructed from a planarian head cDNA library and identified a unique gene closely related to Snap-25. The longest clone, Dj_aH_007_G21, 1.2 kb in length, encodes a 207-amino acid open reading frame (ORF), including all of the highly conserved elements shared with SNAP-25 proteins in Caenorhabditis elegans and in Drosophila, with 57% amino acid identity and slightly lower identity with rat SNAP-25, and was designated as Djsnap-25 (Fig. 1). There is a potential initiator methionine codon upstream of all of the highly conserved elements. The full sequence of the Djsnap-25 cDNA has been submitted to GenBank under Accession number AB260023.

Figure 1.

Alignment of the deduced amino acid sequence of planarian (Dugesia japonica) synaptosome-associated protein 25 kDa (SNAP-25) with those of Caenorhabditis elegans, Drosophila melanogaster and Rattus norvegicus. Identical amino acid residues are shaded gray. Percentages indicate amino-acid identity with planarian SNAP-25. The underline indicates the epitope sequence of the antibody against rat SNAP-25. A possible initial methionine is marked by an asterisk.

Expression of the Djsnap-25 gene in intact planarians

The expression pattern of Djsnap-25 in intact planarians was analyzed by whole-mount in situ hybridization. Djsnap-25 was expressed exclusively in the brain, in putative sensory cells along the body periphery and in the pair of ventral nerve cords (VNCs) along the length of the body (Fig. 2B). Interestingly, Djsnap-25 expression was apparently more limited than that of Djsyt, a planarian synaptotagmin homolog gene (Tazaki et al. 1999), which may be involved in neurotransmitter release in all neurons. Hybridization to cross sections revealed that there was no expression of Djsnap-25 in the photoreceptor cells marked by the expression of Djsyt (Fig. 2C,D; Tazaki et al. 1999). Within the brain, the expression of Djsnap-25 was observed in the main lobes, which are a mass of interneurons (Okamoto et al. 2005), but not in the brain branches, which are believed to form chemosensory organs also marked by Djsyt expression (Fig. 2A,B,E,F). These findings indicate that the DjSNAP-25 activity is required for subsets of neurons.

Figure 2.

Expression pattern of Djsnap-25 mRNA in normal intact planarians. (A, B) Ventral view of whole-mount in situ hybridization in normal adults. Anterior to the left. (A) Neural-specific expression of Djsyt. (B) Neural-specific expression of Djsnap-25. (C–F) Transverse views of the part of the head that includes the brain. Dorsal side is up. (C) Expression of Djsyt in the photoreceptor cells. (D) Expression of Djsnap-25 in the photoreceptor cells. Djsyt but not Djsnap-25 is expressed in the photoreceptor cells (arrowheads). (E) Expression of Djsyt in a section containing the posterior region of the brain. (F) Expression of Djsnap-25 in a section containing the posterior region of the brain. Djsyt but not Djsnap-25 is expressed in the lateral branches of the brain (brackets). Black and white stars indicate the brain and the ventral nerve cord, respectively.

To determine the distribution of DjSNAP-25 protein, normal adults were stained with an anti-SNAP-25 antibody (Mab 71.1, Synaptic System), with an epitope targeting a highly conserved N-terminal region of rat SNAP-25 protein (indicated by an underline in Figure 1). Western blotting analysis showed that this antibody specifically recognized a single band of the expected molecular weight about 25 kDa in extracts prepared from both the head and trunk regions (Fig. 3A). EST database analysis and Western blotting analysis strongly suggest that the planarian genome has a single SNAP-25-related gene. The DjSNAP-25 protein is abundant in axons, and less abundant in neuronal cell bodies (Fig. 3B,C). Within the brain, the staining was observed in the main lobes and also in the brain branches, suggesting that the DjSNAP-25-positive cells located in the main lobes may elongate their axons toward the head margin (Fig. 3B). In the trunk region, DjSNAP-25-positive commissure axons seemed to be in contact with the body wall muscle fibers (Fig. 3C), suggesting that the DjSNAP-25-positive neurons in the trunk region may include motor neurons.

Figure 3.

Expression pattern of Dugesia japonica synaptosome-associated protein 25 kDa (DjSNAP-25) protein in normal intact planarians. (A) Western blotting analysis (right panel) and Coomassie Brilliant Blue (CBB) staining (left panel). A single band (black star) was detected with antirat SNAP-25 antibody. H; extracts from head region, T; extracts from trunk region. (B, C) DjSNAP-25 protein in transverse sections of normal adults. Anterior to the top. (B) Expression of DjSNAP-25 in the brain. The signal is observed in a mass of axons of the brain main lobes and lateral branches. (C) Expression of DjSNAP-25 in the ventral nerve cords (white stars). (B′, C′) Hoechst no. 33258 staining (for nuclei in blue) of the same section as in B, C, respectively. Yellow is a pseudo-color of the Hoechst staining. DjSNAP-25-positive commissure axons innervate the body wall muscle fibers (m). Bar, 100 µm.

Regeneration-dependent conditional gene knockdown (Readyknock)

Silencing of the DjSNAP-25 activity by RNA interference (RNAi) shed much light upon its specific roles in the brain circuitry. However, Djsnap-25 is expressed widely, in the brain and in the VNCs, as well as in putative sensory cells along the body periphery (Fig. 2B). Therefore, careful consideration was given to distinguishing the specific role in the brain circuitry from those in other regions. A previous study concerning body-wall myosin RNAi indicated that the injected dsRNA affects the differentiating myogenic lineage in the regeneration blastema more severely than the pre-existing, terminally differentiated muscle cells at the protein level (Sánchez Alvarado & Newmark et al. 1999). This may be the reason why RNAi in stem cells is more effective, because they do not express any target protein, but RNAi in the terminally differentiated cells is not as effective as they have already abundantly expressed the target protein. Based on these considerations, we believe that DjSNAP-25 activity can be specifically eliminated in the head region when the head region of the RNAi-treated animals has once regenerated. Such a manipulation would exclude defects of the trunk region itself from the knockdown phenotypes. According to this notion, we tested the silencing effects of the Djsnap-25 RNAi. Animals were injected with Djsnap-25 dsRNA and subsequently were either amputated or kept intact. A detailed experimental time-schedule is shown in Figure 4L. We obtained results similar to those of the previous study by Sánchez Alvarado & Newmark 1999. Immunohistochemical analysis indicated that severe reduction of the DjSNAP-25 protein was observed only within the newly formed head region (Fig. 4D) when the RNAi-treated animals regenerated their heads. In contrast, strong signals of the DjSNAP-25 protein were still detected in the pre-existing trunk region of the same animal (Fig. 4H), and also in the knockdown non-regenerated animals (Fig. 4B,F). Next, we carried out quantitative protein analysis by Western blotting. RNA and protein were isolated simultaneously from the same experimental samples using a PARIS Kit. We confirmed that the reduction level of DjSNAP-25 protein was the most prominent in the newly regenerated head region of the Djsnap-25 knockdown animals (about four-fifths reduction relative to the level in the control regenerated head region in Figure 4I). In contrast, Djsnap-25 RNAi was not so effective in the pre-existing tissues at the protein level (Fig. 4I,J). However, at the mRNA level, we observed that the introduction of Djsnap-25 dsRNA into non-regenerated animals could reduce the expression of Djsnap-25 mRNA similarly to that of the head-regenerated animals (Fig. 4K). These observations strongly indicate that the stability of the DjSNAP-25 protein is a major reason for the difference in protein levels between the newly differentiating cells and the terminally differentiated cells. Interestingly, the same phenomenon was also observed in the case of Djsyt (data not shown). Here, we refer to this phenomenon in planarians as regeneration-dependent conditional gene knockdown termed Readyknock (Readyknock is a coined word that means ready-to-knockdown). Regeneration is required to significantly reduce the protein product of the gene.

Severe reduction of the DjSNAP-25 protein in the head region impaired negative phototaxis

A phototaxis assay (Inoue et al. 2004) was carried out to measure and quantify the brain function of Djsnap-25 knockdown head-regenerated and non-regenerated animals. Control animals were injected with EGFP dsRNA. The knockdown non-regenerated animals exhibited normal negative phototaxis, as shown by all of the quantitative assays carried out (Figs 5A and 6). This is very reasonable since the knockdown non-regenerated animals still retained an amount of the DjSNAP-25 protein similar to that in the normal head-regenerated animals in the head region, as well as in the trunk region (Fig. 4I). A clear behavioral defect in negative phototaxis was observed on depletion of DjSNAP-25 protein in the head region. In the control head-regenerated animals, the first detectable phototactic response was observed on the 4th to 5th day after amputation (Fig. 5A; Inoue et al. 2004). On the 7th day after amputation, the head-regenerated animals moved almost directly away from the light source with great sensitivity (Fig. 5A). However, most of the Djsnap-25 knockdown head-regenerated animals failed to reach the target quadrant (Fig. 5A). The light avoidance behavior of the knockdown planarians was clearly slower and uncoordinated (data not shown). These knockdown animals retained a head structure morphologically indistinguishable from that of the control animals (Fig. 5C,D) and normal expression levels of the Djsyt (Fig. 4K), and also 1020HH and eye53 genes (data not shown), whose activities are required to elicit negative phototaxis during head regeneration (Fig. 4J; Inoue et al. 2004). They also were capable of ordinary movements, including a head-shaking movement. These observations suggested that the behavioral defect in the Djsnap-25 knockdown head-regenerated animals did not simply result from a failure of head regeneration.

Figure 5.

Phototactic response of Djsnap-25 knockdown head-regenerated and non-regenerated animals. (A) Behavioral traits of the knockdown animals during head regeneration or animals left intact. Each colored line indicates the trajectory of an individual planarian. (B) A schematic drawing of the container compartments (zones 1–4) in the phototaxis assay system. The circle indicates the start area and zone 4 is the target quadrant shaded in gray. Morphology of the head structure in control (C) and in the Djsnap-25 knockdown head-regenerated animal with an abnormal trajectory (D). The brain structure (upper panel) of the animal showing the trajectory in the left panel was visualized by staining with an anti-DjSYT antibody. Projection pattern of visual axons (lower panel) in a different animal showing as abnormal trajectory was visualized with VC-1 antibody. Termini of the visual axons are indicated by arrowheads. There are no obvious morphological defects in either the brain or in the photoreceptor cells of the knockdown animals. Bar, 100 µm.

Figure 6.

Quantitative analyses of the Djsnap-25 knockdown head-regenerated and non-regenerated animals during negative phototaxis. (A) Time spent in the target quadrant. (B) Total distance moved from starting area. Error bars indicate SEM. (**P < 0.01; t-test). (C) Locomotive activity of the trunk region. Intact: Whole head-regenerated animals. Trunk: Pre-existing trunk region after head regeneration. Error bars indicate SEM. (*P < 0.05; NS, not significantly different; t-test). (D) Overall angle of direction of movement. The circular graphs indicate the direction of movement, determined as the average for up to 10 individuals. The angle of direction of each individual were calculated from the movement data of zones 1 and 2 and the data were binned into 10-degree intervals. The grayscale gradient indicates the proportion of time spent in moving in that direction, black representing the most. Percentages indicate the ratio of the angle of the direction between the bright-half (upper panel) and the dark-half (lower panel). No defect in negative phototaxis in the knockdown non-regenerated animals was observed.

We first quantified the time spent by the knockdown head-regenerated animals in the target quadrant furthest from the light source at various stages during head regeneration and compared it with that of the control animals. The time spent by the knockdown head-regenerated animals was significantly less than that of the control on the 7th day after amputation, and was the same as that of the control headless animals at day 0 (Fig. 6A).

The locomotive ability of the knockdown head-regenerated animals was investigated, especially to assess whether the RNAi-induced phenotype was due to a defect in the pre-existing trunk itself or not. We measured quantitatively the total distance moved from the starting point and compared it with that of control animals. In control animals, the total distance moved increased from the 4th day to the 7th day after amputation, correlating with the recovery of photoresponse (Fig. 6B). In contrast, the total distance moved of knockdown head-regenerated animals (7th day) was significantly shorter than that of the control (Fig. 6B). Next, the locomotive ability of the pre-existing trunk itself at day 7 was analyzed. The regenerated head structure of the knockdown animals was amputated, and the total distance moved was measured. There was no significant difference between the locomotive ability of the knockdown pre-existing trunk itself and that of the pre-amputated knockdown animals (Fig. 6C). We detected only a minor defect in the locomotive ability in the knockdown pre-existing trunk itself (Fig. 6C). These observations strongly suggest that the major defect in negative phototaxis results from dysfunction of the head region, but not from dysfunction in the trunk region in the knockdown head-regenerated animals.

Finally, the overall angle direction of movement of the knockdown animals was analyzed to investigate whether the lack of DjSNAP-25 impairs the correct linear movement away from the light source that is seen in control animals. In the control animals (day 5–7), the majority of movement was directed away from the light source (Fig. 6D). However, Djsnap-25 knockdown head-regenerated animals were clearly less able to orient their movement in the correct linear direction away from the light source (Fig. 6D). We divided the overall angle of the direction of movement into two directions, a light-half direction (180 degrees) and a dark-half direction (opposite 180 degrees) and calculated the ratio between movements toward them. In the control animals, 92.1% of movement was directed to the dark-half on average from day 5–7, indicating a full maturation of the functional head. In contrast, Djsnap-25 knockdown head-regenerated animals showed 70.8% of movement directed to the dark-half on average from day 5–7, similar to the data of animals having a functionless or less functional head from day 0–4 (67.1% for the control, 70% for the knockdown on average). The lack of expression of Djsnap-25 in the planarian eyes strongly suggests that this phenotype is not caused by primary sensory defects.

Therefore, we conclude that loss of DjSNAP-25 activity in the head region results in a loss of brain function as an information-processing center at various levels, such as the time required to correctly avoid light and to move in a direct linear fashion away from the light source.

Functional dissection of DjSNAP-25 activity within the head region

The fact that there were no defects in the knockdown non-regenerated animals during negative phototaxis indicates that the pre-existing active DjSNAP-25 protein was stable in the head region at least for the assay period (Fig. 6). We then examined where Djsnap-25 expression is necessary to elicit correct negative phototaxis within the head region using the Readyknock.

We first analyzed whether the head margin itself is required to elicit correct negative phototaxis. When the head margin was eliminated in normal animals, a behavioral defect was observed (Fig. 7A). These animals showed a significant reduction of total distance moved similar to that of the Djsnap-25 knockdown head-regenerated animals (Fig. 7A), but with the difference that the former retained a normal photo-response (Fig. 7C). Elimination of the tail margin in normal animals resulted in normal behavior, indicating that the behavioral defect was the head-margin-specific (Fig. 7A). This phenotype may result from mechanosensory defects, because the head margin is enriched in mechanosensory neurons that project their axons to the peripheral region of the brain (Pigon et al. 1974; Nakazawa et al. 2003). On the other hand, surgical ablation of the pair of eyes resulted in a loss of negative phototaxis without any locomotive defects (Fig. 7A). These findings indicate that the head margin itself regulates the locomotive activity independently of the primary photosensory information.

Figure 7.

Phototactic response of eye-ablated or head-margin-ablated animals, and of Djsnap-25 knockdown partially head-regenerated animals. (A) Analysis of time spent and of total distance moved in tail-margin-ablated, head-margin-ablated and eye-ablated animals (both on the 1st day after amputation), and Djsnap-25 knockdown head-regenerated animals (7th day). Error bars indicate SEM. (**P < 0.005; *P < 0.05; NS not significantly different; t-test). (B–E) Regeneration-dependent conditional gene knockdown analysis of Djsnap-25 within the head region. Time spent in the target quadrant (B) and overall angle of direction of movement (C) during head-margin regeneration, and during regeneration of the head anterior subregion (including the head margin and the pair of eyes) in (D) and (E), respectively. Error bars indicate SEM.

We next analyzed the role of the Djsnap-25 expression in the head margin during negative phototaxis. When the Readyknock of Djsnap-25 was carried out only in the head margin, these animals recovered normal phototaxis at day 2 after amputation (Fig. 7B). It is likely that the DjSNAP-25 activity within the head margin may not be required for the locomotive activity. Even when the ablation field was extended to the posterior end of the eye in the knockdown animals, the animals seemed to recover normal phototaxis at day 5 after amputation, in contrast to the effect of knockdown of the whole head region (Fig. 7D,E). These observations suggest that loss of the DjSNAP-25 activity in the remaining posterior brain region may result in both locomotive and directional movement defects during negative phototaxis, which is regulated by two distinct types of primary sensory information coming from the head margin and the eyes, respectively (Fig. 8).

Figure 8.

Summary of the functional dissection of the Dugesia japonica synaptosome-associated protein 25 kDa (DjSNAP-25) activity eliciting negative phototaxis within the head region. Djsnap-25-positive cell bodies (dark red circles), axons (bright red), Djsnap-25-negative cell bodies in the eyes (dark gray circles) and in the brain branches (bright gray circles). White lines associated with the eyes indicate the visual axons. The Djsnap-25-positive brain region (boxed) is the best candidate to process distinct sensory information coming from the eyes or the head-margin, both of which are required for negative phototaxis. (+) or (–) indicates whether there is a requirement for the DjSNAP-25 activity for negative phototaxis.

Discussion

Regeneration-dependent conditional gene knockdown (Readyknock) in planarians

In this study, we addressed the molecular basis of the brain circuitry in propagating a distinct behavioral reaction to stimulation in the planarian. To investigate the molecular brain circuitry during negative phototaxis, we first searched a head-derived EST database for neural function-related genes in the planarian that are expressed in the brain cells, but not in the photoreceptor cells, and thereby identified a planarian gene encoding a protein closely related to the synaptosome-associated protein of 25 kDa (SNAP-25), which plays roles in a number of neuronal functions, including axonal growth, dendrite formation, and fusion of synaptic vesicles with the membrane (Hepp & Langley 2001). Djsnap-25 is expressed widely, in the brain and in the VNCs, as well as in putative sensory cells along the body periphery. To distinguish the specific role in the brain circuitry from those in other regions, we selectively eliminated the DjSNAP-25 activity in the head region while leaving the DjSNAP-25 activity in the trunk region unaffected by RNAi, making use of the high regenerative ability of planarians. Planarians have a population of pluripotent stem cells throughout the body that gives rise to all cell types after injury or aging (Agata 2003; Sánchez Alvarado & Kang 2005). When the head region of the RNAi-treated animals once regenerates, the effect of RNAi, especially at the protein level, seems to be stronger in cells newly differentiating from the stem cells (that do not express any differentiated marker genes), than in the terminally differentiated cells (since they already contain abundant expressed differentiated marker genes). The protein stability of the RNAi-target gene explains the difference in expression level between the newly differentiating cells and the terminally differentiated cells, and thus we refer to this phenomenon in planarians as regeneration-dependent conditional gene knockdown termed Readyknock. After successful head region-specific knockdown of the SNAP-25 protein, the light avoidance behavior of the planarians was like that of headless animals, uncoordinated and essentially random, and the trunk itself retained normal undirected movement, although knockdown of Djsnap-25 had no effect on head or brain morphology. These findings clearly indicate that DjSNAP-25 activity in the head region is required for the functioning of the brain as an information-processing center in planarians. In addition, we conclude that Readyknock is simple and easy, and a powerful approach for silencing gene activity in a body position-dependent manner in the planarian.

Functional dissection of the planarian brain circuitry

Our observations suggested that the DjSNAP-25 activity is required for two independent sensory-processing pathways that regulate locomotive activity and directional movement downstream of distinct primary sensory outputs coming from the head margin and the eyes during negative phototaxis, respectively. Recently, DjDSCAM, a planarian DSCAM homolog gene, was identified and shown to be specifically expressed in the brain main lobes, but not in the VNCs (Fusaoka et al. 2006). DjDSCAM knockdown planarians showed disorganization of the neural network in the brain and severe defects in locomotive activity, but normal recognition of light stimulation (Fusaoka et al. 2006). It is likely that the formation of the DjSNAP-25-dependent locomotive circuitry is dependent on the DjDSCAM activity, but that of the DjSNAP-25-dependent directional movement circuitry is not, suggesting that they may be mutually independent pathways in the brain. The results of the Readyknock suggested that the Djsnap-25-positive cells in the posterior brain region may function in forming both sensory-processing pathways during negative phototaxis.

So far, however, it is still difficult to pin down the activity of DjSNAP-25 in the brain itself, as the knockdown planarians that failed to elicit negative phototaxis have defects in both the brain and the VNCs just beneath the newly formed brain (Fig. 2F). Further studies combining the Readyknock technique with microlaser cell ablation techniques would enable the elimination of subsets of Djsnap-25-positive cells within the brain and the investigation of effects on phototactic behavior.

Acknowledgements

This work was supported by a Grant-in-Aid for Creative Scientific Research to K. A. (17GS0318), a Grant-in-Aid for Scientific Research on Priority Areas to K. A., and the Grant for Biodiversity Research of the 21st Century COE (A14). We thank Elizabeth Nakajima for critically reading the manuscript, Michael Royle for naming regeneration-dependent conditional gene knockdown Readyknock and all our laboratory members for helpful discussions.

Ancillary