Production of nerveless Hydra Attenuata by hydroxyurea treatments

Hydra has often been used as an experimental animal and as a basis for devising theoretical models in developmental biology (Crick, 1970; Wolpert, 1969; Gierer & Meindhardt, 1972). The extensive usage of this animal lies in its simple structure, being composed of 7 basic cell types distributed between 2 tissue layers. Despite its structural simplicity, hydra possesses enormous developmental potentials and for this reason has been used extensively in studying growth, differentiation, morphogenesis and regeneration (see reviews by Burnett, 1961 ; Webster, 1971 ; Shostak, 1974; Bode, 1973).

Hydra has also been subjected to many pharmacological agents and ionic conditions which are aimed specifically at damaging, eliminating, or increasing specific cell types (Ham & Eakin, 1958; Lentz, 1966; Diehl & Burnett, 1964; Campbell, 1976; Bode, Flick & Smith, 1976; Browne & Davis, 1977). The rationalization in these studies was that changes in developmental phenomena may be related to individual cells, and therefore the functions of specific cells could be ascertained. In this connexion, nerve cells have received much attention by the above authors. For the most part, the interest in nerve cells is due to the presence of neurosecretion which has been implicated in such diverse phenomena as growth, sexuality, regeneration and budding (Burnett, Diehl & Diehl, 1964; Bursztajn & Davis, 1974; Bode et al. 1973). Recently, Campbell (1976) has shown that hydra lacking nerve cells can be produced by treatment with colchicine, and he reported that these hydra resemble normal animals in their developmental capabilities (Marcum & Campbell, 1978). This observation is an important one, in that it raises several questions regarding the reported neurosecretory control of various developmental processes.

In an attempt to further elucidate the role of nerve cells, hydroxyurea (Bode et al. 1976) was used to produce animals which have a variety of nerve cell densities, including a new type of nerveless hydra. In this paper we report on some of the morphological and developmental characteristics of hydroxyurea nerveless hydra. Regeneration studies were also performed on normal animals containing their usual complement of nerves and animals which have varied nerve cell numbers and densities including 2 types of nerveless hydra (produced by hydroxyurea or colchicine (Campbell, 1976). By using animals with varied nerve cell densities, information regarding the role of nerve cells in the regenerative process may be obtained.

Hydra attenuata (obtained from Dr Charles David) were cultured in a modified M solution (Muscatine, 1961) consisting of 10⁻³ M CaCl₂, 10⁻³ M NaHCO₃, 10⁻⁴ M KC1 and 10⁻⁴ M MgCL at pH 7 4. Animals were fed Artemia napluii daily and washed several hours later. Nerveless animals lack a feeding response and were hand-fed daily using a mouth micropipette made of plastic tubing attached to a drawn out capillary tube. They were cultured in M solution plus rifampicin (50 mg/1.) to retard bacterial growth.

Hydra were treated with 10⁻² M hydroxyurea (HU) as described by Bode et al. (1976). This treatment consists of 3 cycles of 24 h each in hydroxyurea separated by 12 h in normal culture. Animals were fed normally while in hydroxyurea but were not washed until the end of the 24-h treatment. Treated animals were subsequently cultured in the same manner as normal stocks. Classification of HU experimental animals is designated with day 1 being after 24 h in the drug. Nerveless animals produced from hydroxyurea treatment are designated HU nerveless. Colchicine (C) nerveless animals were obtained by 2 treatments with 0.4% colchicine,2 weeks apart (Campbell, 1976).

Adult animals without visible buds were transected proximal to the whorl of tentacles for regeneration experiments (to animals/60 ×15 mm Petri dish). Thirty to fifty animals were regenerated for each experiment. Regenerations of nerveless animals were performed with groups of 5–10 animals. Experiments were repeated several times (e.g. a total of 310 animals were in the control group), so as to be able to treat groups as a population. For gradients of regeneration, hydra were divided into H1234B56F regions (Wolpert, Hombruch & Clarke, 1974) and 60 animals, each containing a late stage bud, were transected at the desired level (15 animals/60 × 15 mm Petri dish). Regenerates were transferred to clean culture at day 1. All data given are cumulative of the repetitions of the experiments.

Based on nerve cell densities, the following 7 classes of animals were used. Classes I and II : 3 day starved (3 day St.) and 3–5 day hydroxyurea (HU) have elevated nerve cell density; Classes III and IV: controls and 10–12 day HU which contain normal cell density; Class V: 23–26 day HU animals which have less than the normal nerve cell density; and Classes VI and VII: nerveless animals produced by colchicine (C) and hydroxyurea (HU). Density in this study refers to number of nerve cells per total number of cells. The head activator of nerve cells has been shown to be a mitogen for all cell types and to affect interstitial cell differentiation (Schaller, 1976a, b). For nerve cells, density as a relation to total cells may be more meaningful than reference to only epithelial cells as has been used for density by others (Bode et al. 1976; Campbell, 1976).

Normal and HU nerveless animals were transected proximal to the ring of tentacles. Hypostomes were relaxed in 2% urethane and photographed with a Zeiss photomicroscope. The area at the level of the tentacles was obtained with a planimeter. Assuming the hypostome to be circular, the circumference is equal to 3.54 × (area)¹.

The distribution of cells was monitored using the maceration technique (David, 1973). At least 3000 cells were counted per sample. The total number of cells was obtained using a phase Neubauer haemocytometer. From this information, the number of each cell type per animal was obtained. Tentacle battery cells and their nematocytes were not included in the quantitations.

Animals were fixed for 1 h in 5% glutaraldehyde in phosphate buffer (Millonig, 1961) without sucrose. The tissue was postfixed for 45 min in 1% osmium tetroxide, dehydrated and embedded in Spurt’s medium. Sections were cut on glass knives on a Huxley ultramicrotome and stained with uranyl acetate for 45 min followed by 30 min in lead citrate. Grids were examined in a RCA 4 electron microscope.

Bode et al. (1976) used hydroxyurea to preferentially eliminate interstitial cells. A period of about 41 days was required for the various cell populations to recover to control values and during this time the percentage of nerve cells remained relatively constant. Their cell counts, however, were performed on head-amputated animals. A previous study had shown the head region to contain about 40% of the nerve cell population (Bode et al. 1973). When we macerated treated whole animals, the number of nerve cells fluctuated in time (Table 1). The immediate increase in the per cent nerve cells (3–5 days) is not due to an increased number but rather to the immediate drop of I-cells and cnidocytes. The actual number of nerve cells drops steadily following treatment. Gland cells behave in a manner similar to nerves in that their numbers also decrease before recovery.

At 23–26 days after hydroxyurea treatment, the cultures contained 2 population of animals. The first is a pale, rarely budding, non-feeding population designated as 23–26 day HU-P. These animals lack a glutathione-feeding response (Loomis, 1955) and are first recognizable by around 20 days after treatment. Behaviourally, they resemble the nerveless hydra produced by colchicine (Campbell, 1976), but maceration studies show that they contain 2% nerve cells (Table 1). The second group is normal-coloured, feeding animals, which eventually recover from treatment. They are designated as 23–26 day HU-R. These animals have approximately half the normal percentage and half the normal number of nerve cells (Table 1).

Starvation was shown to affect nerve cell density (Bode et al. 1973). Animals starved for 3 days contained an increased nerve cell density but were otherwise similar to control (Table 1). If fed, their nerve cell density reverted to normal levels within 24 h.

The 23–26 day HU-P animals resemble the nerveless hydra produced by colchicine treatments (Campbell, 1976). These animals are of normal size (see epithelial cell number, Table 1) and can easily be hand-fed as described earlier. Some of the animals fed in this manner will re-establish a feeding response but many will not. Those animals which remain non-feeding will start to bud within 1 week and the buds from these non-feeding animals are nerveless. Clones derived from individual non-feeding parents were periodically tested for their ability to capture prey and samples were macerated (David, 1973) for cell distributions (Table 1). Macerations showed that if a clone was non-feeding, it was always nerveless. We found that there was heterogeneity in response to hydroxyurea treatments so that some treatments produced more HU-P animals than other treatments. With this method thousands of potential clone-forming animals can be easily produced. The number of nerveless animals obtained is only limited by manpower since they must be hand-fed.

Most of the nerveless hydra produced by hydroxyurea treatment have basically normal form, being composed of the hypostome with its whorl of tentacles, body column, budding zone, peduncle and basal disk. However, these animals reveal certain structural variations. Nerveless animals tended to have more tentacles than normal. To quantify this phenomenon, cultures of normal and nerveless animals (HU and C) were scored as to the number of tentacles they possessed. The distributions obtained are shown in Fig. 1. Nerveless hydra have a mean tentacle number that is significantly higher than control (P < 0.001). The variances of the nerveless distributions are also significantly different from control (F test, max./min. P < 0.001). Ninety per cent of control animals are found within i tentacle of the mean; 78% of colchicine and 76% hydroxyurea nerveless animals are within 1 tentacle of the mean.

Normal animals demonstrate a rigid control over tentacle numbers as shown by the tight tentacle distribution pattern of Fig. 1 (σ² = 0-38). Nerveless animals have an increased variance as seen by the broadness of their distribution patterns (σ² = 1·54 and 3.03). It appears that nerveless animals have lost or demonstrate a diminished level of control over their tentacle number. They produce a wide range of tentacle numbers including a significant number of animals with 9 or more tentacles. Normal animals rarely have more than 8 tentacles, suggesting the presence of a control mechanism which prevents the production of superfluous tentacles.

Nerveless clones are also characterized by having abnormal animals (Fig. 2). These animals may possess secondary hypostomes at or near the primary hypostome, secondary basal disks, and secondary or more hypostomes resulting from buds which failed to detach. Occasionally buds are found which are attached to their parent distal to their own basal disk instead of at their proximal end. These buds may ultimately separate from the parent. Although the majority of nerveless animals have nearly normal form, the nerveless condition is characterized by increased tentacle number, altered bud production at times, and the presence of abnormal animals. These deviations from normality may be significant in pointing towards a deficiency in control mechanisms in the nerveless state.

Growth studies as measured by increase in hypostome numbers and bud protuberances are shown in Fig. 3. Newer clones (Fig. 3 A) generally show a faster growth rate than older clones (Fig. 3B). Erratic patterns are evident and animals in individual dishes might stabilize for a time and then resume growth. Growth did not seem to be affected by the number of animals per dish and decreasing the number of animals, changing culture medium more frequently, or heavier feeding schedules did not affect stationary clones.

Since the growth rate diminished after several months and since it was easy to establish hydroxyurea clones, individual clones were not maintained for long periods. The cell distributions in Table 1 are cumulative totals and do not reflect the age of a clone. Over the time periods monitored (approximately 3 to 4 months) the distributions remained constant and big I-cells never repopulated the animals. During summer months, however, there were several instances of male sexuality among HU nerveless clones. Maceration showed all stages of sperm differentiation but no nerves or

Regeneration experiments were done to monitor tentacle numbers which appeared to be normally under a strong control system (shown by compact tentacle distribution, Fig. 1). Tentacles are a morphological feature of the dominant head and changes in regeneration pattern might reflect changes in dominant control. The results are shown in Fig. 4 which shows the 5-day distribution pattern. Control animals exhibit a compact distribution with 95% of the regenerates falling within 1 tentacle of the mean (4’8). As animals progress towards nervelessness (HU treated), there is a progressive increase in the variance suggesting a loss of control in the nerveless state with respect to tentacle number. The mean tentacle number is also increased in the nerveless state (5.7 for colchicine and 5.6 for hydroxyurea). Forty per cent of the nerveless regenerates (C and HU) have 4–6 tentacles (compared to 95% for controls), 40% have an increased number of tentacles (7–14) and 20% a decreased number (0–5).

Individual cell percentages fluctuate among the classes of animals (Table 1). A Pearson Correlation Coefficient was determined for each cell type and the mean number of tentacles regenerated (Fig. 4). Nerve and gland cells were the only types to show a significant correlation, r = − 0.91 (P < 0.01). This correlation was obtained with experimentally altered animals and appears to be contrary to the supposed function of nerves containing the isolated head activator (Schaller, 1973). This relationship, to be biologically significant to the normal animal, would predict that transection in a proximal direction (along the decreasing nerve cell gradient) would produce an increase in tentacles and variance. Gradients of regeneration were performed with animals transected so as to produce i-F, 3-F, 4-F, and 5-F regions (Fig. 5). Variances and mean tentacle number regenerated were increased as predicted, except for the 5-F pieces which had a decreased mean. This pattern of regeneration was consistent in 4 repetitions of the experiment.

The results of regeneration and normal tentacle distributions suggest that nerve cells function, in part, in regulating tentacle number. Another interpretation is that tentacle numbers are correlated with body column size. Nerveless animals are often bloated and the column diameter is increased in the 3-F and 4-F regions. To test the column size possibility, column circumference and tentacle number were obtained for normal and nerveless animals. HU nerveless animals showed a significant relationship (r = 0.73, P < 0.001) when animals with 5–12 tentacles were scored. Control animals, having 5–9 tentacles, showed no such relationship (r = 0.09). In the nerveless condition, hypostomal size appears to be a factor in the number of tentacles produced, but normally other factors must take precedence.

Tissues from the various classes of animals were examined with the electron microscope to determine any structural changes which may be related to developmental properties. Nerveless tissue appears normal except for the lack of I-cells and their products (Table 1). With nerve cells absent, it seems possible that the structural cells (epitheliomuscular and/or digestive) could produce some type of morphologically recognizable material which in turn would mimic the activities of the morphogenetic substances produced by nerves. No structural differences of these cell types were seen as compared to control. If nerveless animals contain the head activator (Schaller, 1973), then it must be packaged differently and, consequently, not distinguishable. The only cell type that appeared to be affected by hydroxyurea was the nerve cell. Following treatment, there was an increase in lysosomal-like structures (Fig. 6) with some cells also having lipid-like droplets. An increase in degenerating nerves was also seen (Fig. 6 B) and although there were no quantitative studies, the nerve cells seem to contain less neurosecretory product.

We were unable to produce nerveless Hydra viridis or H. pseudoligactis with colchicine (Campbell, 1976) and therefore attempted the hydroxyurea method on these species. With the concentration used for H. attenuata, H. viridis was apparently unaffected and recovered quickly while H. pseudoligactis was extremely sensitive and produced non-feeding animals soon and in greater number than H. attenuata. Treated H. pseudoligactis have very little behavioural responses and do not contract when poked as HU-P and nerveless H. attenuata do. This lack of contraction makes them extremely difficult to feed and maintain in culture. Preliminary electron-microscopical studies showed normal myonemes and myonemal junctions so that the basis for this lack of contraction does not appear to be due to anatomical abnormalities. A more detailed study of the myonemal junctions and myofilaments is in process. On extended culturing the animals tended to resorb their tentacles resulting in abnormal forms. Buds sometimes contained some nerves but it is not known whether nerveless stocks could be produced. The extreme difficulty in feeding and the tendency towards abnormality in form prohibit the use of H. pseudoligactis in this respect.

There are now several methods for producing hydra which lack nerve cells. The methods include treatment with colchicine or colcemid (Campbell, 1976), gamma irradiation (Fradkin, Hakis & Campbell, 1977), genetic manipulations (Sugiyama & Fujisawa, 1978a, b) and treatment with hydroxyurea as described in this paper. All methods are aimed at the elimination of stem cells (interstitial cells), so that with the loss of precursor cells, no nerve cell differentiation can occur. Behaviourally, morphologically and developmentally, hydroxyurea nerveless animals are similar to those produced by colchicine (Campbell, 1976) and by genetic means (Sugiyama & Fujisawa, 1978 a, b). The characteristics of nerveless hydra are due solely to their condition and do not appear to be related to the method of obtaining them.

Nerveless animals possess nearly normal form and all developmental capabilities. However, close examination reveals certain differences between normal and nerveless hydra. The tentacle distribution patterns of nerveless hydra (hydroxyurea and colchicine) are broad, as observed in culture and in regeneration studies (Figs. 1, 4). They have significantly increased means and variances. In contrast, normal animals have compact distribution patterns which are demonstrative of a high level of pattern regulation. The increase in means and variances in nerveless animals may represent a loss of normal control mechanisms.

Ham & Eakin (1958) found an increase in the number of tentacles regenerated following treatment with a variety of neuropharmacological agents. They suggested that the increase was due to interference with a normal tentacle number regulatory mechanism which they attributed to nerve cells. In their chimera studies, Sugiyama & Fujisawa (1978a, b) showed that reintroduction of I-cells reverted the increased tentacle numbers of their nerveless nf-i strain (x≄ ∼ 10) to the parental sf-i value (x≄∼7). In this study, a correlation (r = −0.91) was found between nerve cells and tentacles. In the nerveless condition, body column circumference was also correlated with tentacles but this correlation was not present in normal animals. Though size appears to be a factor, other levels of control must be present. It is suggested that one such control level may be associated with the nerve cells. Since nerveless animals have more tentacles than normal animals, inhibition appears to play a major role in the control of tentacle morphogenesis, as it does in other systems (Spiegelman, 1945).

Nerveless animals produced by hydroxyurea treatment contain approximately 0.5% big interstitial cells which never differentiated into nerves or nematocytes. Marcum & Campbell (1978) suggest that there may be a critical density requirement of interstitial cells for differentiation to occur. Over the summer months there were several instances of male sexuality among the nerveless clones. Maceration showed the presence of gametes but no other I-cell products, and this observation raises questions as to the multipotency of the stem cell. David & Murphy (1977) showed that the interstitial stem cell is multipotent with respect to the production of nerves and nematocytes. These sexual animals suggest that there may be either a second stem cell line for gamete differentiation or that the signals to the usual stem cell for nerve and nematocyte differentiation are missing.

In growth studies, HU nerveless clones increased in numbers erratically (Fig. 3) which was similar to results with colchicine-derived clones (unpublished results). Generally, older clones increased at a slower rate than young clones and erratic patterns occurred eventually. The exact cause of this phenomena is not known but it may be related to the ability to digest food or an ageing response. Sugiyama + Fujisawa (1978a) have also noted a decreased growth rate in the H. magmpapillata nerveless clones and have also suggested an ageing effect. Through extensive chimera production with numerous mutants, Sugiyama & Fujisawa (1978b) reported that the chimera growth rate resembled the parental sf-i strain and not the mutant strains. Their interpretation was that nerves have little to do with growth rate. However, the chimeras resembled the parental sf-1 and not the nerveless nf-1 from which they had been derived, and which had been previously shown to deviate from the parent [sf-1] (Sugiyama & Fujisawa, 1978a). Reintroduction of nerves therefore reverted growth characteristics to normal values. Growth rate deviations in the nerveless condition may be due to the absence of nerve cells. This would be in accord with the action of the isolated head activator of nerves (Schaller, 1973, 1976a).

Nerveless clones, hydroxyurea and colchicine, also possess significant numbers of animals with abnormal body form. These animals have secondary or supernumerary hypostomes and/or basal disks. The hypostome and basal disk are dominant regions in the classical sense (Huxley & DeBeers, 1934; Shostak, 1974). One feature of a dominant region is that it should inhibit production of like structures. Nerveless animals possess secondary dominant regions in areas where such redundant structures should be inhibited. We are currently investigating this loss of dominant control (Sacks & Davis, 1978). Redundant dominant regions are also produced by abnormalities in the budding process, in that buds sometimes fail to detach. These results on budding are similar to the results obtained following colcemid treatment (Shostak & Tammariello, 1969).

In all characteristics examined, nerveless animals showed deviations from normality. These differences point towards altered levels of developmental control in the nerveless condition and suggest where nerve cells may be functioning. That nerveless animals possess the full complement of developmental capabilities points to the basic role of the structural epithelial cells. The abnormalities present in all types of nerveless hydra are important because they demonstrate where normal control mechanisms are altered.

Following hydroxyurea treatment, animals show a transient decrease in regenerative ability (3–5 and 10–12 day HU animals, P < 0.001). This coincides with an increased level of inhibition found in grafting experiments (Sacks & Davis, in preparation). The effect, which disappears by 23–26 days, suggests that there may be a leaking of inhibitor during this time period. If nerve cells are the source of morphogens (Schaller & Gierer, 1973; Berking, 1977), then a leaking effect is consistent with the increased degeneration seen in electron-microscopic preparations. The isolated head inhibitor and activator are antagonistic (Schaller, 1976b; Berking, 1977) and Berking (1977) has shown that the inhibitor can suppress budding even in a 15-fold excess of activator.