The remarkable capability of planarian regeneration is mediated by a group of adult stem cells referred to as neoblasts. Although these cells possess many unique cytological characteristics (e.g. they are X-ray sensitive and contain chromatoid bodies), it has been difficult to isolate them after cell dissociation. This is one of the major reasons why planarian regenerative mechanisms have remained elusive for a long time. Here, we describe a new method to isolate the planarian adult stem cells as X-ray-sensitive cell populations by fluorescence-activated cell sorting (FACS). Dissociated cells from whole planarians were labeled with fluorescent dyes prior to fractionation by FACS. We compared the FACS profiles from X-ray-irradiated and non-irradiated planarians, and thereby found two cell fractions which contained X-ray-sensitive cells. These fractions, designated X1 and X2, were subjected to electron microscopic morphological analysis. We concluded that X-ray-sensitive cells in both fractions possessed typical stem cell morphology: an ovoid shape with a large nucleus and scant cytoplasm, and chromatoid bodies in the cytoplasm. This method of isolating X-ray-sensitive cells using FACS may provide a key tool for advancing our understanding of the stem cell system in planarians.
The high regenerative capability of planarians is known to be supported by adult pluripotent stem cells called ‘neoblasts’ (Wolff 1962; Agata & Watanabe 1999). The neoblasts have been defined by their morphological characteristics (Pedersen 1959). When the planarian body is observed by electron microscopy (EM), cells showing typical undifferentiated morphology can easily be recognized throughout the mesenchymal space of the body. They have minimal cytoplasm with few mitochondria, many free ribosomes, no endoplasmic reticulum, and are characterized by a unique cytoplasmic structure known as the chromatoid body. The chromatoid body, which consists of huge RNA-protein complexes observed as electron-dense particles in EM samples, is a hallmark for the identification of neoblasts (Morita et al. 1969). Interestingly, these cells are specifically eliminated by X-ray irradiation (Wolff & Dubois 1948). X-ray-irradiated planarians lose the capability of regeneration, suggesting that these morphologically undifferentiated cells may be pluripotent stem cells in the planarian (Brønsted 1969; Baguñà 1981). Although the neoblasts are a morphologically homogeneous population and are believed to be the only proliferative cells in the planarian, the pluripotency of these cells has not yet been demonstrated by single-cell transplantation or clonal cell culture analyses.
Recently, several molecular markers for the neoblasts have been identified. The first molecular marker, DjvlgA, was isolated by our group as a candidate gene for a chromatoid body component (Shibata et al. 1999; Kurimoto et al. 2005). We also found that two homologues of FGFR were expressed in X-ray-sensitive cells (Ogawa et al. 1998, 2002). S-phase-specific genes such as PCNA and MCM2 homologues were used to detect proliferating stem cells (Salvetti et al. 2000; Orii et al. 2005). RNA-binding proteins, such as a Pumilio homologue, were also identified as candidates for chromotoid body components (Salvetti et al. 2005). Recently, Sànchez's group showed that in Schmidtea mediterranea, piwi homologues were expressed in neoblasts, and gene-knockdown planarians of one of these homologues lost the regenerative capability (Reddien et al. 2005). However, in Dugesia japonica, a piwi homologue (DjPiwi-1) is expressed in small cells aligned in the dorsal midline of the body (Rossi et al. 2006). No proven pluripotent cell-specific gene has yet been isolated. Thus, electron microscopic observation is still the most reliable method to distinguish neoblasts.
Previous studies focusing on planarian neoblasts have utilized cell populations prepared by low-resolution methods based on centrifugation and filtration (Betchaku 1967; Baguñà 1998). Such methods have not yielded cells amenable to the cellular and molecular analyses of the stem cell system. It is thus necessary to develop a more specific method to isolate stem cell populations in order to elucidate their regulatory mechanisms.
Fluorescence-activated cell sorting (FACS) strategies using specific cell surface markers have been applied to isolate near-pure populations of hematopoietic stem cells (HSC) from vertebrate systems, including bone marrow and fetal liver (Spangrude et al. 1988). In addition, a subset of mouse HSC can be isolated from bone marrow as a side population (SP) based on their functional characteristics, that is, their ability to rapidly expel the fluorescent DNA-binding dye Hoechst 33342, rather than by relying on molecular markers (Goodell et al. 1996). The bone marrow SP constitutes 0.05–0.1% of the nucleated cells in the bone marrow, has long-term reconstituting ability and is part of the HSC population (Goodell et al. 1997). In addition, SP cells have been identified in several non-hematopoietic adult tissues, such as skeletal muscle, brain, and mammary gland; indeed, it has been suggested that the SP may correspond to stem cells in the tissue from which they were isolated (Gussoni et al. 1999; Asakura et al. 2002; Murayama et al. 2002; Welm et al. 2002). However, it remains unclear whether SP cells exhibit the entire repertoire of stem cell activity (Kubota et al. 2003). Although there have been several studies on mammalian SP cells from the bone marrow and other tissues, SP cells have not been well studied in invertebrates.
Recently, we reported the successful application of FACS to planarian studies for the isolation of brain neurons (Asami et al. 2002). By the simultaneous use of several fluorescent dyes and FACS-based comparative analysis of fluorescence-labeled cells derived from the head and other body parts, we were able to enrich planarian brain neurons to 80–90% purity, demonstrating that FACS could be used to purify particular planarian cell populations using a simple combination of fluorescent dyes.
In the absence of surface markers for planarian stem cells, we have extended this FACS-based approach and here describe the isolation of X-ray-sensitive cell populations from adult planarians. Classical data demonstrated that neoblasts are irreversibly eliminated after X-ray irradiation, providing a rationale for their identification via comparisons of FACS plots of cells from X-ray-irradiated and non-irradiated planarians. Moreover, combinatorial use of the fluorescent dyes Hoechst 33342, calcein AM and PI allowed us to identify X-ray-sensitive cell populations.
Materials and methods
A clonal strain of the planarian D. japonica (ssp.) was used in this study. At least 10 adults approximately 8 mm in body length were employed for each cell sorting fractionation. To reduce contamination from gut content debris, the animals were starved for more than 1 week before each experiment. To prepare X-ray-irradiated planarians, worms placed on wet filter paper on ice were irradiated with 12 R X-rays using an X-ray generator (SOFTEX B-4; SOFTEX, Tokyo, Japan).
Preparation of dissociated cells
Planarians were cut into three to four fragments on ice with a scalpel. The resultant animal fragments were soaked in Holtfreter's solution diluted 5/8 in distilled water (5/8 Holtfreter). The fragments were cut into smaller pieces and treated with 0.25% (w/v) trypsin (DIFCO) for several minutes at 20°C. The samples were completely dissociated into single cells by gentle pipetting. The cell mixture was then filtered through a 35 µm pore size cell strainer (Becton-Dickinson, Franklin Lakes, NJ, USA) and a 20 µm nylon net filter (Millipore, Billerica, MA, USA) to remove tissue fragments. The procedure for the preparation of dissociated cells is illustrated in Figure 1.
Preparation of single-cell suspensions for fluorescence-activated cell sorting analysis
Single-cell suspensions were characterized by selective staining with fluorescent dyes as follows: Hoechst 33342 (Sigma, St Louis, MO, USA; excitation 351–364 nm; emission, 465 nm) was used at a final concentration of 18 µg/mL to stain nuclear DNA; and calcein AM (Sigma; excitation 485 nm; emission 530 nm), a substrate of intracellular esterase, was used at a final concentration of 0.5 µg/mL to determine the cellular volume of viable cells; propidium iodide (PI; Dojindo, Kumamoto, Japan; excitation 536 nm; emission 617 nm) was used at a final concentration of 1 µg/mL to stain and eliminate dead cells. Single-cell suspensions were incubated at 20°C for 2 h in 5/8 Holtfreter's solution, pelleted by centrifugation at 1500 g for 2 min, and resuspended at an appropriate cell density in 5/8 Holtfreter's solution. Flow cytometric analysis was performed using a FACS Vantage SE triple-laser flow cytometer (Becton-Dickinson).
Electron microscopy of dissociated cells
The cells isolated by FACS were centrifuged at 1800 g for 10 min using a swing-bucket rotor (TOMY; TMS-21). The samples were fixed in 1.2% glutaraldehyde in 0.1 m sodium cacodylate buffer (pH 7.4) for 1 h at 4°C. The samples were then washed for 15 min in 0.1 m sodium cacodylate buffer and postfixed in 2% osmium tetroxide in the same buffer for 1 h at 4°C, and after the wash suspended in 0.1% agarose gel in the same buffer to prevent the cells from becoming dispersed. The samples were dehydrated at 4°C by passage through a series of increasing concentrations of ethanol (70, 80, 90, 95, 100%) and finally by treatment with acetone for 15 min. After infiltration with QY-1 for 30 min, the cells were embedded in Epon 812. Ultra-thin sections were cut with a diamond knife on an ultramicrotome. The sections were stained with 4% uranyl acetate and lead citrate, and observed by transmission electron microscopy.
Total RNA from the cells collected by FACS was prepared using an RNeasy Mini Kit (Qiagen). First-strand cDNA synthesis was performed using a First-strand cDNA synthesis kit (Amersham Biosciences, Piscataway, NJ, USA), and then the cDNA was used for semiquantitative real-time polymerase chain reaction (PCR) analysis. Thirty microliters of real-time PCR mixture including 1× Quantitect SYBR green PCR master mix (Qiagen), each gene-specific primer at 0.3 µm and 1 µL of diluted cDNA template was reacted in an iCycler (Bio-Rad, Hercules, CA, USA). The PCR conditions were as follows: 50°C for 2 min, 95°C for 15 min, 50 cycles of 95°C for 15 s, 60°C for 30 s, 72°C for 1 min; 95°C for 15 s, 60°C for 15 s and 95°C for 15 s for drawing a dissociation curve. PCR analyses of cDNA were performed using the following primers corresponding to planarian sequences:
Planarians were dissociated into single cells and stained with Hoechst 33342, calcein AM and PI. Dissociation and vital staining procedures are summarized in Figure 1. The photographs in Figure 2 show dissociated cells after staining with the three fluorescent dyes. The dissociated cells contained very heterogeneous cell populations (Fig. 2A). Muscle fibers and large intestinal cells could be recognized by their cytological characteristics, but the other cells had an ovoid shape and were not distinguishable. Many cells were killed mechanically in the course of the dissociation procedures and stained in red with PI (Fig. 2D). DNA was stained with Hoechst 33342, and cell debris could be distinguished by the absence of Hoechst 33342 signals (Fig. 2B). Calcein AM stained the vital cytoplasm in green, and calcein AM-positive cells (Fig. 2C) showed a complementary staining pattern to PI-positive cells (Fig. 2D). Fragmented cells could be categorized by visual inspection as Hoechst+/calcein– or Hoechst–/calcein+ cells. Hoechst+/calcein– cells may have lost most of their cytoplasm due to fragmentation. Hoechst–/calcein+ cells may be cytoplasmic fragments that have lost their nucleus (Fig. 2B,C). We collected Hoechst+/calcein+ cells as vital cells using FACS as described below.
Separation of dissociated cells by fluorescence-activated cell sorting
The dissociated cells were analyzed by FACS to determine staining intensities and fractionated into populations based on the intensities of different fluorescent dyes. PI-positive cells comprised approximately 30% of the total and were previously excluded as dead cells. About 60% of the dissociated cells were Hoechst/calcein double positive, and were thought to be vital cells. Hoechst+/calcein– or Hoechst–/calcein+ cells derived from fragmentation comprised about 10% of the total. We also conducted forward-angle light scatter (FSC) and side-angle light scatter (SSC) analyses, which reveal cell size and intracellular complexity, respectively, to exclude non-cellular debris and cytotoxic intestinal granular cells (data not shown).
The intensities of Hoechst and calcein AM staining may reflect the DNA content of the nucleus and the cytoplasmic volume, respectively. The fluorescence of cells was plotted on a graph in which the X-axis represents the relative fluorescence intensity of calcein AM, and the Y-axis the relative fluorescence intensity of Hoechst 33342 (Fig. 3A). We obtained a reproducible profile showing an inverted T-shape. The majority of cells (70%) formed a large cluster, but about 20% of cells were located in a wide area along the Y-axis. These cells were stained with Hoechst 33342 more strongly than the majority of cells, suggesting that they were at the proliferative stage.
Identification of X-ray-sensitive cell fractions by fluorescence-activated cell sorting
We obtained reproducible FACS profiles for both non-irradiated and irradiated animals (Fig. 3A,B). Comparison of these profiles revealed that the vertical region described above, which consisted of cells undergoing proliferation, was specifically eliminated by X-ray irradiation. The highly X-ray-sensitive cells of this region were only weakly stained with calcein AM, indicating that they possessed scant cytoplasm. This population was designated X1 (Fig. 3A). A second distinct X-ray-sensitive region in the profile of non-irradiated planarians showed weak Hoechst 33342 and calcein AM staining, and was designated X2. Whereas the X1 fraction could be specifically gated, we could not create a gate to collect only X-ray-sensitive cells in this second region. About half of the cells in this fraction were X-ray-insensitive cells. Finally, much of the remaining area of the FACS plot corresponding to X-ray-insensitive cells was common to the profiles of both irradiated and non-irradiated planarians. We designated this population XIS (X-ray-insensitive). The relative proportions of the three cell populations X1 : X2 : XIS were about 1:2:6.
Morphological homogeneity of X1 and X2 fraction cells
The morphological traits were examined microscopically and by FACS. Both the X1 and X2 fractions consisted of homogeneous cells in terms of the size and shape of the cells (Fig. 4B,C), as compared to the total cell population before sorting (Fig. 4A). Typically, both X1 and X2 contained cells that were spherical with large nuclei and scant cytoplasm (Fig. 4B′,B″,C′,C″). However, they showed different DNA content and cell size with approximately 10 and 6 µm diameters, respectively (Fig. 4B,C). When analyzed by FSC and SSC, the X1, X2 and total cell populations exhibited distinct features (Fig. 4D–F). Observation of the nuclei of Hoechst-stained cells revealed that the X1 population included around 23% of cells undergoing mitosis; therefore, it seemed most likely that the difference of cell size between X1 cells and X2 cells was due to the phases of their cell cycle. In contrast, the XIS population exhibited a greater degree of both FSC and SSC than either X1 or X2 and exhibited variable morphology, size and shape and some of them exhibited variable intracellular complexity, reflecting their relative heterogeneity (data not shown).
X-ray-sensitive cells in X1 and X2 fractions possessed neoblast morphology
To investigate whether X-ray-sensitive cells in the X1 and X2 fractions really possess neoblast morphology or not, we collected the cells in each fraction and observed their morphology by electron microscopy. In the X1 fraction, about 70% of X1 cells showed typical neoblast morphology: ovoid shape with a large nucleus and scant cytoplasm with chromatoid bodies in the cytoplasm (Fig. 5A). Very few differentiated cells could be observed in this fraction, suggesting that stem cells were highly enriched in the X1 fraction. In the X2 fraction, although many X-ray-insensitive cells were present, a substantial population of cells possessed neoblast morphology (Fig. 5B). EM observation clearly indicated that in addition to the neoblasts, the X2 fraction contained a variety of differentiated cells, which would not be expected to be X-ray sensitive. Therefore, we concluded that X-ray-sensitive cells in the X2 fraction may consist of cells possessing typical neoblast morphology. In contrast to the X1 and X2 fractions, there were no neoblasts in the XIS fraction.
Molecular characterization of X1 and X2 fractions
To characterize each fraction, the expression of marker genes was analyzed by quantitative reverse transcription–polymerase chain reaction. Djpcna and DjMCM2 are known to be proliferating cell markers (Salvetti et al. 2000; Orii et al. 2005). The relative expression rates of these genes strongly supported the notion that proliferating stem cells were enriched in the X1 fraction (Fig. 6, red and orange). In contrast, the expression of neural- and muscle-specific genes (DjPC2, Djsyt, DjMHC-A and DjMHC-B) was strongly detected in the XIS fraction, but not in X1 or X2 (Fig. 6). These data were consistent with the morphological observations described above.
Side population fraction in planarian
We also employed alternative FACS analyses to characterize planarian stem cells. We were interested in the SP cells, since these represent a subset of adult stem cells in vertebrates. The distinctly weak Hoechst staining of the SP cells is due to a multidrug resistance protein (mdr) or mdr-like mediated efflux of the dye, which was reported to be specific to adult stem cells (Goodell et al. 1996). The above FACS analyses were therefore supplemented with an assessment of Hoechst 33342 efflux in a concurrent evaluation of planarian cells. In this way, the planarian's Hoechst 33342-effluxing SP population was identified (Fig. 7A). However, the SP cells were not eliminated by X-ray irradiation at all (Fig. 7B), and thus they were distinct from both X1 and X2 cells. Moreover, SP cells did not overlap with X1 or X2 cells in this FACS analysis (data not shown). It was not possible to collect enough SP cells for electron microscopic observation. However, we were able to characterize the collected SP cells by examining their gene expression. We have not obtained any evidence indicating that the SP fraction contained neoblasts or a stem cell population so far. Indeed, we found that the majority of the SP cells were digestive cells expressing a multidrug ABC transporter (data not shown).
Because of their apparent ability to regenerate any tissue, planarian adult stem cells represent a powerful model for the study of postnatal regenerative mechanisms and stem cell regulatory systems (Agata et al. 2003). Despite the many advantages of using these cells as a resource for research on the regulatory systems of stem cells, substantive molecular and cellular investigations on planarian stem cells have not yet been achieved. We reasoned that this difficulty might best be overcome by the isolation and purification of the stem cells in large numbers. We accordingly developed a FACS-based approach (Asami et al. 2002) and here report the method for the isolation of planarian stem cells.
Using this method, we have shown that two subpopulations (X1 and X2) of cells corresponding to distinct regions of a FACS plot were markedly reduced following X-ray irradiation. Since such irradiation ablates stem cell activity, this provides evidence that these populations include adult stem cells. We confirmed by EM observation after cell sorting that X-ray-sensitive cells of both fractions possessed typical neoblast morphology (Fig. 5). Unfortunately we could not create a gate to collect only X-ray-sensitive cells in the second region (the X2 fraction), whereas the X1 fraction could be easily gated based on its distinct Hoechst staining pattern. The analysis further suggested that the X1 population contained cells undergoing proliferation and that the X2 population contained non-proliferating cells, as indicated by their intensity of nuclear dye staining. The transcriptional profile of X1 cells was consistent with that of typical proliferating cells (Fig. 6).
In early attempts to establish a protocol for the preparation of dissociated cells, we treated the dissociated cells with a solution containing 30 µg/mL Type II-O chicken egg white trypsin inhibitor (Sigma) for several minutes and thereby successfully obtained dissociated cells. This treatment was important to protect the cells against endogenous intestinal proteases. However, we subsequently established a protocol whereby we can obtain vital cells without the use of the trypsin inhibitors by quick handling of the dissociated cells. Here we described the recent protocol without the use of the trypsin inhibitors in Materials and methods. However, we recommend the use of the trypsin inhibitor rather than trypsin when one tries to prepare dissociated planarian cells for the first time.
As a prelude to characterizing X2, we conducted an alternative FACS-based approach which made use of the phenomenon of Hoechst 33342 efflux (Fig. 7). Such ability is characteristic of the side population (SP) and is exhibited by somatic stem cells in several systems (Goodell et al. 1996, 1997). FACS analysis of cells derived from dissociated planarians following Hoechst 33342 staining revealed a pattern typical of the SP (Fig. 7A). However, the SP did not correspond to major parts of either X1 or X2 populations and was not sensitive to X-ray irradiation. Recently, SP cells have been isolated from a variety of tissues, including non-hematopoietic ones (Gussoni et al. 1999; Asakura et al. 2002; Murayama et al. 2002; Welm et al. 2002). However, it is not clear in general whether SP cells possess stem cell activity, with the possible exception of HSC. Indeed, SP cells of mouse testicular origin are clearly distinct from spermatogonial stem cells (SSC), as judged by the differences of the expression patterns of cell surface markers MHC-I, Thy-1 and c-kit (Kubota et al. 2003). These distinctions have led to the conclusion that SP sorting by FACS might not be a universal method for the isolation of stem cell populations. The planarian cell samples analyzed here were more complex than many other samples, such as HSC. This could explain the relative difficulty of applying standard profiling of SP cells to the samples derived from whole planarians and further analyses will be required to define the characteristics of planarian SP cells.
In conclusion, FACS sorting is a powerful method to collect planarian adult stem cell populations. We expect that this approach may facilitate the characterization of adult pluripotent stem cells in planarians. However, we have not succeeded in obtaining their proliferation under either in vivo or in vitro conditions after sorting. We have started systematic screening of genes specifically expressed in these cells.
We thank all our laboratory members for helpful discussions and for technical guidance, and Elizabeth Nakajima, Shigeru Kuratani and Neal Rao for critical reading of the manuscript. This work was supported by Junior Research Associates (JRA) of RIKEN (MA), Special Coordination Funds for Promoting Science and Technology (KA), and Grants-in-Aid for Creative Research (KA) and Scientific Research on Priority Areas (KA).