In Allium and Lilium roots, saturated aqueous solution of coumarin produced a disruption of the metaphase typical of many benzene derivatives, viz. suppression of the spindle, splitting and shortening of the chromosomes, with retarded division of the centromere. The resultant polyploid nuclei and binucleate cells resumed division when the roots were returned to water.
Saturated solution of parasorbic acid slowed Allium mitosis, but caused no abnormalities.
Both coumarin and parasorbic acid eventually prevented the inception of mitosis, and this suppression of prophases persisted for several hours after removal of the agent.
The results are in agreement with prior evidence that in some configurations the benzene ring disrupts mitosis whereas the lactone ring inhibits growth.
A substance which shows comparable ontogenetic effects throughout the plant and animal kingdoms is valuable in interpreting certain fundamental growth phenomena. Colchicine is such a compound. Parasorbic acid is another, and is of even greater interest since it exhibits a range of effects far wider than that of colchicine.
Parasorbic acid inhibits the growth of Amphibia and mammals (Heaton, 1929), destroys the mesenchyme of planarians (Hauschka, Toennies & Swain, 1945), and in tissue cultures inhibits the mesenchyme of embryonic chick (Medawar, Robinson & Robinson, 1943 ; Kuhn et al. 1943) and of mammals (Drew, 1927). It also inhibits the fermentation of yeast, the germination of pollen, and the germination of seeds (Kuhn et al. 1943 ; Köckemann, 1936). Parasorbic acid fed to rats has been shown to inhibit the Jensen sarcoma (Heaton, 1929), but in vitro evidence as yet is inadequate to demonstrate whether it selectively inhibits malignant fibroblasts (Cornman, 1945). Kuhn and his co-workers have shown that it does not attack Erlich carcinoma in tissue culture.
The action of lactones such as parasorbic acid, in common with that of some antibiotics, may possibly be explained by their ability to react with sulphydryl groups (Cavalito & Haskell, 1945).
In the above experiments the parasorbic acid was obtained from yeast, fruits, and animal tissues. Growth inhibitors are often reported in the literature (Amer. J. Bot. 31; suppl.), which may also prove to be related to parasorbic acid, extending even further its natural range of occurrence.
The related compound, coumarin (see diagram), occurs in a number of plants, including sweet clover. It also has growth and germination-inhibiting powers (Kuhn et al. 1943 ; Nutile, 1945), which Veldstra & Havinga (1943) relate to its antagonism toward naphthaleneacetic acid. It may not be purely an inhibitor, however, inasmuch as Grace (1938) finds it can function as an auxin.
The purpose of the work presented here was to test these two compounds and sorbic acid, the straight-chain isomere of parasorbic, for their influence on mitosis. In addition to whatever of interest is revealed in the mitotic effects themselves, it should help to determine the role of mitotic inhibition in the mechanism by which normal and malignant growth is slowed by these compounds.
Roots growing from bulbs of Allium cepa and LUium longiflorum provided the dividing cells. The roots were sprouted in tap water, then immersed in saturated solutions of coumarin, sorbic acid, or crude parasorbic acid.
Two to eight roots were fixed after hr. exposure and after 9,12, 31 and 45 hr. in tap water following hr. exposure. Zenker’s and Bouin’s fixatives were used, followed by routine paraffin sectioning at 8 and 10 p, and staining in Harris’s or Heidenhain’s iron haematoxylin.
Exact measurements of growth were not made, but it was noted that elongation of onion roots continued only a short time after immersion in the experimental solutions. The vapour from a saturated solution of parasorbic did not prevent the unsubmerged roots from growing into the solution.
Hypertrophic swellings were visible on the coumarin-treated roots after hr., that is, after 21 hr. recovery. Smaller swellings were induced by parasorbic acid after 31 hr.
Cytological effects of coumarin
After hr. the mitotic block was complete. Some telophases with phragmoplast and cell plate and some anaphases remained (see Table 1). All metaphases appeared to be blocked. Abnormalities included telophase cells with two nuclei or double (dumbbell) nuclei, and the typical dispersion and shortening of metaphase chromosomes. These all resulted from the destruction of the metaphase and anaphase spindles.
After 5 hr. the picture was that of a complete colchicine effect. Blocked metaphases predominated, and prophases were the only other mitotic stages present. The metaphase chromosomes had shortened further (Pl. 10, fig. 2), and in some cases had split to form the A-configuration typical with delayed centromere division, or the ‘skipairs’ of complete splitting (Pl. 10, fig. 3).
During these first 5 hr. of treatment, reversion to interphase progressed more slowly than the evolution of prophases, resulting in an accumulation of metaphases. The blocked metaphase chromosomes became packed into dense masses (Pl. 10, fig. 4, lower right) before forming giant interphase nuclei. This aggregation was the only morphological variance from the colchicine response wherein the chromosomes remain scattered. Nuclei with projections or lobes were frequent, but unequal nuclei or karyomeres were rare, and only found after a period of recovery (Pl. 10, fig. 5), during which a second abortive division may have produced a fragmented nucleus. Reversion of blocked anaphases produced double nuclei connected by bridges of varying thickness (Pl. 10, fig. 4) and binucleated cells. These cells with doubled chromatin content were enlarged to varying degrees.
Coumarin slowly suppressed the initiation of mitoses, with the result that after hr. only an occasional prophase or a blocked metaphase with extremely short, thick chromosomes could be found.
The figures in Table 1 record all mitoses in the middle sections, totalling 100 μ, of a typical root from each group (in the last column a block 700μ thick was counted to compensate for the scarcity of figures). These numbers cannot be regarded as a statistical approach, inasmuch as the counts vary with the size and mitotic activity of the individual root. Nevertheless, the trend in the proportions of mitotic phases is unmistakable. Once treatment has begun, the metaphases exceed the late prophases, while anaphases and telophases decrease and disappear after 5 hr. These later stages are numerically replaced by binucleate cells, a convincing demonstration that they did not complete the mitotic cycle, but reverted to interphase without completing anaphase separation (dumbbell-shaped fused nuclei) or partitioning of the cell (binucleate). It may be significant that the number of binucleate cells progressively decreased during treatment and recovery, suggesting that fusion and rounding up had produced one symmetrical nucleus.
Recovery in onion
Onion roots treated for hr. and subsequently returned to tap water did not renew mitotic activity within 9 hr. After 21 hr., many diploid mitoses were to be found. After 45 hr., polyploid (Pl. 10, fig. 6) and binucleate cells had also resumed division, many of them with poorly developed spindles and lagging or scattered chromosomes. These secondary irregularities may have been the cause of the unequal nuclei in such hypertrophied cells as that shown in Pl. 10, fig. 6. At this time some roots still showed no mitotic activity, and the pycnotic nuclei and highly acidophil cytoplasm of some cells suggested a dead or dying condition.
The course of events in lily roots was similar to that in onions. Blockage of mitosis was complete after 2 hr., and after 3 hr. an accumulation of metaphases was observed along with the other concomitants of interrupted and reverting mitoses: short chromosomes divided into parallel pairs (Pl. 10, figs. 8, 9, lower), tetrapioid chromosome groups probably derived from anaphases (Pl. 10, fig. 7), giant lobed (Pl. 10, fig. 10, upper right) or double nuclei (Pl. 10, fig. 10, lower left), bridged nuclei, and binucleate cells. All showed varying degrees of irregularity in the nuclear outline, reflecting a moderate scattering of chromosomes. Blocked metaphases reverted to interphase via densely massed chromosomes (Pl. 10, fig. 9, upper) as in the onion. After 6 hr. the suppression of mitosis was evidenced by a decrease to a few prophases and blocked metaphases. Recovery of the lily cells has not been studied.
Cytological effects of parasorbic acid in the onion
Parasorbic acid has little in common with coumarin in its effects on the onion-root mitosis except that counts show a slight excess of metaphases which may represent an accumulation of metaphases or a relatively more rapid decrease of prophases (see Table 1). This altered ratio of mitotic phases was found after and 5 hr. The metaphase chromosomes were noticeably shortened (Pl. 10, fig. 1). It must be emphasized, however, that normal anaphases and telophases persisted and these phases occurred in all layers, so unequal penetration cannot explain their persistence. There were no transitional phenomena indicating a reversion of metaphases to interphases. In roots exposed for longer periods, no giant nuclei were found ; and in roots allowed to recover, no tetrapioid mitoses could be detected. It would seem that the progress of mitosis had been slowed but not stopped, and that the spindle mechanism continued to function. One hesitates to ascribe so slight an effect to the parasorbic acid, inasmuch as only a crude preparation was available, and in a saturated solution enough impurity may have been present to alter the speed of mitosis.
Few mitoses remained after 5 hr. exposure, and only a rare prophase or metaphase after . An occasional prophase was to be found after 9 hr. recovery. Mitoses in appreciable numbers did not reappear until 31 hr. in tap water had elapsed. At this time all stages were visible in the more vacuolated cells at the base of the meristem, while prophases and a few metaphases and anaphases were found in the apex. After 14 hr. more, mitoses of all stages were proceeding normally throughout the proliferative tissue.
Effects of sorbic acid in the onion
Saturated sorbic acid immediately stopped growth and caused the roots to become flaccid. Mitoses in these roots were normal in all parts of the meristematic region after 6 hr. exposure, and later after 5 hr. recovery. The excessive vacuolization of the cytoplasm was similar in appearance to that caused by poor fixation. Indeed, since acidity of even half-saturated sorbic acid is pH 4-47 (glass electrode), its only effect upon the cell was probably an immediate coagulation. Cytological data for the half- and quarter-saturated acid solutions are not yet available, but at quarter-saturation the roots lost their turgidity within 5 hr.
Coumarin takes its place among the natural mitotic poisons, foremost among which are the alkaloids and the essential oils. Others of these aromatic substances are anethol (Lefèvre, 1940), apiol (Gavaudan & Gavaudan, 1940), thymol (Villars, 1940), and methyl methylanthranilate (Simonet Sc Igolen, 1940). In connexion with the problem of a plant’s protecting its growth processes from its own poison, it would be interesting to ascertain whether clover, anise, parsley, thyme and the orange, like Colchicum (Comman, 1942), are sensitive to high concentrations of these active agents contained in their own cells, and whether there is any cross-immunity. Such an investigation might suggest an explanation of the fact that benzene derivatives are able to produce identical effects in a cell (benzene and colchicine represent the extremes), while a slight change in a molecule (colchicine to colchiceine by loss of a methyl group) produces a thousand-fold change in potency (Ludford, 1936).
Coumarin and parasorbic acid differ by one benzene nucleus, and as has been repeatedly shown, the benzene molecule is particularly efficacious in blocking mitosis. The germination-inhibiting properties, on the other hand, appear to depend upon the lactone ring, since both compounds prevent seed germination.
Animal cells respond to parasorbic acid and in a particularly interesting manner. In tissue culture, for instance, the fibroblasts are inhibited while epithelial cells continue to proliferate. Now in well-nourished tissue cultures the fibroblasts outgrow epithelial cells, so one might suspect that parasorbic acid acted by disrupting mitosis, thereby damaging the more frequently dividing cells. The above observations show that such is not the case, if we may generalize from plant cells (onion and lily root meristems have repeatedly been shown to be efficient detectors of poisons which also block animal mitosis). It remains possible, however, that there may be something significant in its preventing the onset of karyokinesis, although this is a property of most mildly toxic agents.
Mitotic poisons have also been suggested for the chemotherapeutic treatment of cancer. It may prove valuable to determine whether the therapeutic value of parasorbic acid, as demonstrated with the Jensen rat sarcoma, could be enhanced by endowing the molecule with anti-mitotic activity, guided by the pattern of the coumarin molecule.
I am indebted to Dr John J. Biesele and to Dr C. W. Metz for suggestions regarding this manuscript. The lily bulbs were kindly supplied by Dr S. L. Emsweller of the U.S. Department of Agriculture. All three compounds were generously contributed by Dr Werner Bergmann.
EXPLANATION OF PLATE 10
Fig. 1. Onion parasorbic acid, 5 hr. exposure. Short disoriented chromosomes of the slowed metaphase.
Figs. 2-4. Onion, coumarin, 5 hr. exposure. Fig. 2. Shortened, moderately scattered metaphase chromosomes with incipient split. No spindle remnant. Fig. 3. Split chromosomes with short, thick chromatids still attached at the centromere. Fig. 4. Restitution nuclei, one from a metaphase (lower right) showing densely packed chromosomes just reverting to interphase, and two from anaphases, showing bridges of different widths.
Figs. 5, 6. Onion, coumarin, . exposure—45 hr. recovery. Fig. 5. Hypertrophied cell with giant nucleus and karyomere. Fig. 6. Polyploid anaphase.
Figs. 7-10. Lily, coumarin, 3 hr. exposure. Fig. 7. n tangle of chromosomes probably arising from an interrupted anaphase. Fig. 8. Thick, split, clumped chromosomes of the blocked metaphase. Fig. 9. Blocked metaphase (below): chromosomes in the optical plane show the split but little contraction. Above is a blocked metaphase with the chromosomes tightly packed, just prior to reversion to interphase. Fig. 10. Double nucleus from a blocked anaphase (lower left). Lobed nucleus from a blocked metaphase in which the chromosomes were scattered.
The course of transformations in the blocked metaphases of both lily and onion is exemplified by the sequence : 9 lower, 2, 3 and 8, 9 upper, 4 lower right, 10 upper right, 5 (or more typically a uninucleate giant cell consonant with less scattering of the chromosomes), 6. Magnification photography by Lew Sunny.