Characterization of an organ-specific differentiator substance in the planarian Dugesia etrusca

In a wide variety of animals, many investigators have shown that the differentiated cell population controls the mitotic and tissue determinative processes of the undifferentiated cell population. It is believed that the differentiated cells of the planarian control the course of stem cell-mediated tissue renewal by the secretion of promotor and/or inhibitor substances. One such substance, which has been shown to be present in crude extracts of planarians (Lender, 1955,1956, 1960), controls the differentiation of stem cells into brain tissue. Using Dugesia lugubris and Polycelis nigra, Lender demonstrated that homogenates of heads had a much greater capacity to inhibit brain tissue formation (in animals whose brains had been removed) than homogenates of tails. In addition, he showed that:

1. If the crude extract was centrifuged at 10000 rpm for 20 min, the inhibitory activity was found in the supernatant.

2. If this supernatant was then centrifuged at 20000 rpm for 20 min, the inhibitory activity was found in the pellet.

3. The head homogenate was not inactivated by treatment with 70% alcohol, lyophilization, or heating to 60 °C for 2 min.

4. However, higher temperatures reduced the activity, and boiling for 30 min destroyed it.

5. The substance does not exhibit zoological specificity, since ground brains of Dugesia lugubris, Dugesia gonocephala, or even a minced 9-day embryonic chick brain (Lender & Deutsch, 1961) inhibited the differentiation of the brain in Polycelis nigra.

The brain has been shown to be the first organ to regenerate in an anterior blastema (Wolff & Lender, 1950). If, however, an intact brain were implanted into the anterior region of a decapitated animal, either the implanted brain became the brain of the regenerated head, or the animal regenerated two heads, one containing the implanted brain and one head containing no brain tissue (Wolff, Lender & Ziller-Sengel, 1964). Regeneration of the brain not only determines the anterior-posterior polarity, but has been hypothesized to be the primary organizer which, once formed, begins a series of inductions and differentiations to replace all other missing tissues (Wolff, 1962).

This study attempts to characterize further the diffusible brain inhibitory substance in Dugesia etrusca.

Filtered water is tap water which has been filtered through both a Calgon no. 9 and a Culligan ‘Flavr-Gard’ water filter. Planarian saline is similar to McConnell’s (1967) with the exception that it contains 2·5 ml from each of the two stock salt solutions per liter in distilled water.

Stock cultures of the planarian, Dugesia etrusca, were maintained in filtered tap water at 19 ± 1 °C and fed canine liver biweekly.

All experimental animals were starved for between 1 and 2 weeks prior to use, to insure minimal effects of metabolic variations and of nutritional wastes. The standard size animal used for the assay procedure was 5·0 ± 0·5 mm in length. These animals were of such size that they would not and did not fission during the time course of the experiment. Fully grown fissioning animals were 8-10 mm in length and were chosen for homogenization.

The ‘assay’ planarians were decapitated immediately behind the auricles (see Fig. 1). The remaining tails were placed in shallow Petri dishes or vials to which were added the various fractions to be tested (2·5 ml/animal). The animals were then allowed to regenerate for 9 days (except where noted) before sacrificing with 2% HNO₃. They were then fixed, embedded, serially sectioned (the first 40 transverse sections from the anterior end are used), and stained for nervous tissue using the method of Betchaku (1960). Regenerated brain volumes were measured using a projecting microscope (Bioscope, Inc.) and a compensating polar planimeter (Keuffel & Esser Co.).

A crude homogenate was prepared by homogenizing intact planarians in a Potter-Elvehjem type tissue grinder in sterile distilled water at 0-4 °C. This homogenate was then centrifuged at 10000 g for 20 min. The pellet was resus-pended in sterile filtered water. A portion of the supernatant was then centrifuged at 32000 g for 20 min. Aliquots from each fraction in addition to similar aliquots of sterile filtered water for controls, were placed separately in vials. One decapitated animal (5 mm original length) was then added to each vial. The animals were then assayed as described in section 3.

Varying proportions of the 32000 g pellet (described in section 4) were diluted with sterile filtered water. The protein concentration of each fraction was then determined using the method of Waddell (1956). Bacterial growth in this and succeeding experiments was inhibited by the addition of 10 μg/ml Kanamycin (Bristol Laboratories) to each vial (Lange, 1968). Four animals were assayed per fraction.

A 10000 g supernatant extract was first filtered through a 8-0 μm pore Millipore filter to remove remaining cells, then ultrafiltered using a Dow hollow fiber beaker b/HFU-1 (retains molecules ⩾ 30000 daltons). This ⩾ 30000 M.W. retenate extract was then electrophorezed using a Karler-Misco electrophoresis apparatus. The background electrolyte buffer was 0·1 M phosphate buffer (pH 6·8). Perpendicular to the direction of the buffer flow was a 2 mA-280 V direct current field. Fraction tubes and electrode reservoirs were changed daily during the 72 h period of operation and stored at 2 °C. The fractions and reservoirs were combined into five groups: + reservoir, fractions 1-9 (positive side), fractions 10-13 (ca. neutral), fractions 14-22 (negative side), -reservoir. The fractions were concentrated to 30 ml (at the same time dialyzed against sterilized planarian saline to remove the phosphate buffer) again using the Dow hollow fiber b/HFU-1. These five groups plus two control groups were then assayed for inhibitory activity.

Since previous gel filtration studies (Steele & Lange, unpublished) have shown that the active substance was eluted at the void volume using Sephadex G50 and G75, the molecular weight must be over 80000 daltons. Therefore a 1·6 × 100 column, filled with BioGel A-l-5 m gel filtration media, was equilibrated at room temperature with sterilized planarian saline. The void volume, V₀, was measured using Blue Dextran 2000. The elution volumes, V_e, of four molecular weight markers (Calbiochem) were measured (Fig. 4). The partition coefficients, K._{a v}, between the liquid phase and the gel phase were determined using the formula:

All enzymes used were tested for activity using planarian extracts, and the amounts used were in great excess for each experiment.

Pronase, B grade (Calbiochem no. 537088) was added to a resuspended 32000 g pellet containing a known amount of protein. After 12 h at room temperature the digestion mixture was extracted with cold ethanol (Green & Hughes, 1955) and dialyzed. The original extract, the digested extract, and a sterilized planarian saline control were then assayed for inhibitory activity.

Ribonuclease A (Sigma Chem. Co. no. R4875) was added to a resuspended 32000 g pellet and incubated for 12 h at room temperature. The original extract, the RNase-treated extract, RNase alone, and sterilized planarian saline control were then assayed for inhibitory activity.

DNase I (Worthington Biochem. Corp, no. 2007) was added to a ⩾30000 M.W. extract (section 6) in 10⁻³ M MgCl₂ and incubated 24 h at 37 °C, along with extract alone, DNase alone, and sterilized planarian saline alone controls. After the incubation, the digested extract and the controls were ultrafiltered and dialyzed to remove the MgCl₂. These extracts were then tested for inhibitory activity.

A similar procedure was performed for Lipase (Sigma Chem. Co. no. L2253).

A ⩾ 30000 M.W. extract (section 6) was dialyzed against a solution of 1% glycine for 24 h, then added to an ampholine solution of pH 3·5-10 (LKB, Inc.). This mixture was then added to a linear sucrose gradient in a 440 ml electro-focusing column (LKB 8102). The final ampholine concentration was 1%. After electrofocusing had been achieved, the column was fractionated and the volume, pH, and absorbance (280 nm) of each fraction were measured. The 40 fractions were then combined into 12 groups according to pH, concentrated, and placed separately on a BioGel P-60 column (0·7 × 60 cm) and eluted with pH 7·2 phosphate buffer (1 M-NaCl). The extract proteins in each group were eluted at the exclusion volume while the ampholytes were eluted separately at various later times. After concentration and dialysis against sterilized planarian saline, the extract proteins were tested for inhibitory activity.

All animals placed in the resuspended 10000 g pellet (10 kgP) died after 3-4 days. However, an averge reduction in regenerated brain volume of about 42% was observed for animals placed in the 10000 g supernatant (10 kgS) (see Table 1). A value somewhat higher, although not significantly different (P > 0·05), was observed for the resuspended 32000 g pellet (32 kgP). The slight difference in activities is possibly due to diffusion of the inhibitory substance into the supernatant (32 kgS), since this fraction is slightly lower than the control value.

As the extract concentration increased from 0 to approximately 12·5 μg/ml protein, the volume of the regenerated brain tissue decreased significantly by 40% (Fig. 2). At the highest protein concentration used, an 82·6% decrease in brain volume was observed due to the addition of the extract. A regression line fitted to these data yields the equation

Electrophoretic separation of the ⩾ 30000 M.W. extract into five general classes of electrophoretic mobilities resulted in no significant loss of brain inhibitory activity. Addition of these five groups of substances to regenerating assay animals resulted in the mean regenerated brain volumes shown in Fig. 3. The substances which migrated to the anodal reservoir ( + Res.) inhibited brain formation nearly as efficiently as the unseparated extract (b/HFU-1 Extract). The remaining four groups were not significantly different from control.

Substances which demonstrated significant inhibitory activity were found in elution volumes similar to those of globular proteins whose molecular weight ranged from 2 × 10^{4 5} to 4 × 10⁵ (as determined by the calibration curve, Fig. 4). The regenerated brain volumes of five fractions which should contain higher or lower molecular weight substances did not differ significantly from the control volume. The histogram (Fig. 4, top) also shows that two fractions exhibited significant inhibitory activity. Therefore the apparent molecular weight is estimated at slightly greater than 300000 daltons. It should be noted that this apparent molecular weight may be an aggregate of two or more active molecules, however, such aggregation, had it occurred, had no observable inactivating effect.

As seen in Table 2, the addition of pronase to the extract significantly destroyed the inhibitory activity (P < 0·01). The regenerated brain volumes resulting from decapitated animals exposed to the pronase-treated extract did not differ significantly from decapitated animals exposed to sterilized planarian saline. Although the ‘extract’ brain volume is similar to that of the RNase study the ‘control’ volume is noticeably smaller. This is due to the 7-day regeneration period used instead of a 9-day period. Comparison of the control volumes demonstrates a 46·2% increase in mean brain size resulting from the additional 2-day regeneration period. However, the ‘extract only’ controls did not significantly increase in size during this period. This evidence argues strongly against any form of cytotoxic delay resulting from exposure to the extract.

Addition of RNase to the regeneration media did not significantly alter the resultant regenerated brain volume when compared to control values. Incubation of the extract with RNase for 12 h did not produce significantly different results when compared to the untreated extract. However, the regenerated brain volumes resulting from the extract and RNase-treated extract are significantly different from both the RNase only or control values (P < 0·01) for each of the four cases.

The extract incubated with DNase did not show a significantly different amount of inhibitory activity from that of the untreated extract (Table 2). Addition of only the DNase to the decapitated animals did not produce an effect significantly different from control.

Also, the inhibitory activity of the lipase-treated extract was not significantly different from that of the untreated extract. Lipase alone had no inhibitory effect nor did it differ from the mean control R.B.V.

A nearly linear pH gradient (r² = 0·984) was established during the 48-h electrofocusing period for the range of the ampholine mixture used (i.e. pH 3·5-10). Absorbance measurement (280 nm) of each eluted fraction revealed five regions which could be identified as containing substances at their isoelectric points. These absorbance peaks (pi’s) were at pH 4·3, 4·9, 5·6, 6·3, and 7·8. Before fractionation, visual observation of the electrofocusing column’s contents revealed eight bands or zones. Of the 11 groups tested, the group containing substances whose isoelectric points ranged from pH 4·75 to 5·38 was the only group exhibiting brain inhibitory activity resulting in significantly lower R.B.V.’s (Fig. 5). Most of the substances in this group would correspond to those which exhibited a pl of 4-90 as determined by the absorbance peak. The remaining ten groups were not significantly different from control.

Lender (1955, 1956, 1960) has previously shown that a diffusible brain inhibitory substance exists in several planarian species. This was demonstrated by the excision of the head ganglia only, the addition of head or tail extracts every other day, and the measurement of regenerated brain length after 9 days. The experiments we present in this paper show that one can excise the entire head and still observe a significant amount of brain inhibitory activity in homogenates or extracts. Our use of decapitation demonstrates that existing head tissues are not necessary for the inhibitory action of the substance. This method also has the advantage of being certain that: (a) no former brain tissue or head ganglia exist prior to the extract treatment, (b) repeatable tissue volumes are excised, and (c) no effects from anesthesia are present. In Lender’s experiments the head homogenates were replenished on the 3rd, 5th, and 7th day which, in some cases, completely inhibited brain regeneration. Although the differences in brain size, we present in our data, are not as large in some cases as Lender’s, we have shown that a single exposure to our extracts containing the brain inhibitor immediately after decapitation did produce significant differences. It is also highly probable that once the wound surface has healed (see Morita & Best, 1974), the passage of any such inhibiting substances into the regenerating blastema is severely restricted. Since the entire planarian body contains the brain inhibitory substance, the whole animal was homogenized. While Lender measured brain length, we felt that the measurement of brain volume would reflect more accurately any differences in inhibitory activities present in our various extracts. Preliminary tests using our assay system in Dugesia etrusca showed that when differences in mean brain length were not significant, differences in mean brain volume were highly significant (P < 0·01). All animals in our tests regenerated brains, although a number of animals, especially in the extract-treated groups, died before the 9th day of regeneration and were discarded. Eyespots, or pigmented cupules characteristic of this species, were regenerated in all instances. We did not attempt to further characterize or possibly separate the eye inducer substance(s) from the brain inhibitor substance in this study. Excision of the brain without including the eyes would be technically very difficult in D. etrusca.

Since the inhibitory substance remains in the 10000 g supernatant but will pellet at 32000 g, it is possible to assume that in homogenates, as prepared, it is not bound to cell membrane fragments or large organelles, but could be associated with something approximately the size of a ribosome. Since some of the activity still remains in the 32000 g supernatant, this association may be one in equilibrium with other bound or unbound states. As extract concentration increased, the amount of protein present increased linearly, and the resultant regenerated brain volume decreased. Perhaps each potential brain tissue neoblast binds or removes from the intercellular fluid pool one or more inhibitory substance molecule(s). The data from these experiments cannot resolve whether the action of the inhibitory substance is a persistent or a temporary effect. Perhaps the fact that Lender (1956) found some animals with no brain tissue when he replenished the extract every other day, and we, with our single treatment, did not, could support the idea of a temporary mode of inhibition.

The initial gel filtration studies indicated that (1) the inhibitory substance could be partially purified by gel filtration, and (2) the elution at the void volume of Sephadex G75 meant that the inhibitory substance, free or in its bound (but still active) state, was a comparatively large molecule ( ⩾ 80000 daltons) (Steele & Lange, unpublished results).

In the electrophoretically separated extract the inhibitory substance migrated to the anodal reservoir and quantitatively effected a diminished regenerated brain volume comparable to that of the unseparated extract (b/HFU-1). We can therefore state with a high degree of certainty that the inhibitory substance is a negatively charged molecule. It should be noted that substances which migrated to the cathodal reservoir were promotive with respect to brain regeneration at the 90% confidence level. It follows that this promoting substance(s) would also be ⩾30000 daltons in size.

An apparently opposite situation exists in Hydra littoralis. Lenique & Lundblad (1966a, b), using agar electrophoresis of homogenized stems, reported that electronegative proteins were growth promoters and electropositive proteins were growth inhibitors. Rose (1966) reported similar results in Tubularia. Evidence in higher animals, such as the salamander, the adult frog, and the rodent (Becker, 1961; Becker & Spadaro, 1972), suggest that some form of electrical control over growth and differentiation does exist.

The apparent molecular weight of the inhibitory substance(s), as determined by the calibrated agarose gel column, confirmed the earlier results using Sephadex gels. The apparently high molecular weight (2-4 × 10⁵) observed is that of the substance in its active undenatured state. It is possible that while in sterilized planarian saline, ionic conditions may be such as to facilitate aggregation. If aggregates were present, the aggregation did not decrease the inhibitory powers of the substance. It should be noted that slightly larger R.B.V.’s (higher at the 90% confidence level) were observed for substances in the molecular weight range of 4-8 × 10⁵ daltons. These promoting substances are possibly identical to the electropositive promoting substances observed in the extract electrophoretic separation.

It is suggested by the enzyme digestion studies that the brain inhibitory substance is, at least in part, protein or polypeptide in nature. At least the site(s) necessary for its action is labile to digestion by pronase. Since this substance’s activity is not destroyed by DNase or RNase, it is probably not a nucleoprotein (or any associated nucleic acids are not necessary for its activity as assayed). The inability of pancreatic lipase to alter its activity reduces the probability that it is some form of lipoprotein complex. The possibility still remains that the a and a’ ester linkages of a triglyceride portion are obscured by a protein complex and are not susceptible to pancreatic lipase attack. Also remaining are the possibilities that the inhibitor substance may be a glycoprotein, metalloprotein, or a chromoprotein. Since many of the active extracts used in these experiments were colorless, the possibility of having a pigment prosthetic group (chromoprotein) has a low probability. Since many growth regulating hormones, such as thyroglobulin or FSH in higher animals, are glycoprotein in nature, this class of conjugated proteins remains a strong possibility. Experiments are in progress to explore this likelihood.

The isoelectric point of the inhibitory substance reaffirms the results of the electrophoretic separation. The low pl value indicated suggests that a large number of negative charges on the molecule must be balanced by a high hydrogen ion density before the net charge is zero. Experiments in progress, with a narrower pH range of ampholytes, should give a more accurate determination of the isoelectric point.

The characterization, as presented, does not attempt to make any resolution between the eye inducer substance and brain inhibitor substance, found to exhibit many common properties by Lender (1955,1956, 1960). A highly concentrated extract, purified by centrifugation, ultrafiltration, and electrophoresis, should completely inhibit brain formation. If eye formation is then delayed (compared to control) or absent, a good probability exists that both activities are not possessed by the same substance.

V. E. Steele acknowledges the support of the Rochester A.E.C. Laboratory Graduate Participant Program and C. S. Lange the support of an N1H Research Career Development Award. This paper is based on work performed under contract with the U.S. Energy Research and Development Administration in the Department of Experimental Radiology (Contract No. AT(30-l)-4284) and University of Rochester Biomedical and Environmental Research Project and has been assigned Report No. UR-3490-766.