Journal of Heredity 39: 327-335 (1948)
Induction of Morphological Changes in the Chromosomes and of
Abnormalities in Mitotic Divisions in the Root Meristem
Atomic Bomb Casualty Commission., Chugoku Military Gov't Region,
A.P.O. 317, Japan

WHILE conducting experiments with ribose nucleic acid and other compounds containing nucleotides at the Department of Botany, University of Wisconsin, during October‑November 1947, the author discovered that in onion root tips treated with solutions of the ribose nucleic acid of certain concentrations for certain lengths of time, the morphology of the chromosomes and their mitotic behavior are changed in many cells. He then undertook to investigate the induced changes under better controlled conditions. It is the purpose of this communication to report the method and results of this series of experiments and to discuss some of the important aspects of the results.

Bulbs of Allium cepa (2n=16) were set in tap water until a number of roots had grown to a length of a few millimeters. These bulbs were then set individually in vials filled with solutions described below in such a way that all roots and the basal parts of the bulbs were immersed in the solutions. Roots were then cut off at various time intervals and the meristems of these excised roots were fixed in 3‑1 Carnoy's fixative for at least 15 minutes and then stained by the Feulgen method, and smeared. Some preparations were made by smearing with aceto‑orcein. Both types of preparations were satisfactory for the present study. The photomicrographs presented in this paper were all taken from the Feulgen preparations.

Ribose nucleic acid is very slightly soluble in cold water. Therefore, in the later series of experiments sodium ribose nucleate was used, because of its greater solubility. The product of the Schwarz Laboratories, Inc., New York, N. Y. was used throughout the experiments. A series of eight solutions (in concentrations of 0.05, 0.1, 0.25, 0.5, 1, 2, 4, and 8 per cent) was made by dissolving appropriate amounts of the nucleic acid salt in cold distilled water. In one set of experiments, these solutions were used without being adjusted to the same pH. The pH of the distilled water was 5.8 and that of the 2% and 4% solutions was respectively 5 and 4.5. After 36 hours of culturing bulbs in these solutions, the pH of the solutions remained unchanged. In another set of experiments, 2%, 4% and 8% solutions were used after the pH of the solutions was raised to 7 by addition of a NaOH solution. Differences in the results between the two sets of experiments will be described presently.

The first externally visible effect of the treatment was that the growth of all treated roots was markedly suppressed. From the comparison of amounts of growth in several roots treated with 0.05%, 0.5% and 4% solutions for 24 hours, it was found that the greater the concentration of the solution, the less the growth. It was also found that in many of the treated roots retardation of growth was followed by death. The older roots showed, earlier in the treatment, a slight constriction in the region of elongation just behind the meristem. This became gradually more pronounced and at the same time, the older regions became progressively constricted until finally the entire root above the growing point became very thin. The meristematic region of these roots remained unchanged [328] in size, but the number of dividing cells in this region gradually decreased as the roots became more constricted. Finally the cells ceased dividing entirely. The constriction appeared earlier and the roots died more quickly in the more concentrated solutions. For instance, in an 8% solution, the constriction appeared in three or four hours and the roots died a few hours afterward. In a 1% solution, it took much longer before the constriction was recognized and several hours elapsed before the meristem ceased to show cell divisions.

In the neutralized solutions, the growth was suppressed as much as in the solutions of lower pH, but the constriction did not appear in any of the treated roots and none of them died even in the 8% solution which, if not neutralized, would have killed all within about 10 hours. Changes in the morphology and mitotic behavior of the chromosomes, induced by the neutralized solutions, were essentially the same as those produced by the more acid solutions. This indicates that the effects observed in treated roots are not due to an excessive concentration of hydrogen ions in the solutions. In a few preliminary experiments with Vicia faba, it was found that in 2% and 4% solutions the constrictions appear in a few hours and subsequently the roots die even in neutralized solutions. Evidently there are genetic variabilities in susceptibility to the action of sodium ribose nucleate.

Altogether about 350 roots were treated with solutions of various concentrations. Of these about 250 were excised and smeared directly from the solutions and the rest were left attached to bulbs and returned to tap water. Some of these roots were kept in water for eight hours and others for 24 hours or longer before they were cut off and smeared. The roots which took the full effects of the treatment did not regain a normal rate of growth in tap water and the chromosomal abnormalities persisted in these roots.

The roots treated with 0.05% and 0.1% solutions were completely normal except for a slight reduction in growth rate even after 24 hours of treatment. The roots treated with higher concentrations showed a slight effect on the chromosome morphology even after a few hours, and the full effect on the morphology as well as on the mitotic behavior of the chromosomes in five to eight hours, depending upon the concentration of the solution. From the study of all treated roots it was found that the effects are qualitatively alike in all effective concentrations but are quantitatively different in different concentrations and times of treatment. The best concentration and time of treatment to obtain full effects in a large proportion of dividing cells in onion root tips are exposure to a 2%‑4% solution for between 8 and 16 hours.

In root tips which are fully affected by the treatment, chromosomes are to a greater or less extent shortened in length and increased in diameter at the metaphase (Figure 4A). The primary constriction at the kinetochore and the secondary constriction in the trabant‑bearing chromosomes of these condensed chromosomes appear as clear, narrow gaps (Figure 4A). Since the early prophase chromosomes are not noticeably shorter than the untreated ones of the same stage, the extreme condensation must occur during the late prophase. It is of interest to note in this connection that the coiling of the chromatids is clearly visible in most of the prophase chromosomes and in the metaphase chromosomes which are not excessively condensed (see Figure 5A). Furthermore, in a number of prophase chromosomes each of the sister chromatids was found to be already split. The details of these particular aspects of the effects will be described and discussed elsewhere by Professor Huskins.

Most of the prophase chromosomes have been found to bear numerous "hairs"; and to resemble closely in appearance the lampbrush chromosomes of the vertebrate ovocytes. Some of the metaphase chromosomes are also "hairy" (Figure 4B). Although very similar in shape and appearance the "hairs" in the present preparations may not be related in structure to those of the natural lampbrush chromosomes. The "hairs" in my preparations are Feulgen positive, whereas those in natural lampbrush chromosomes were shown by Duryee to be Feulgen negative. Feulgen positive lampbrush "hairs" have also been produced experimentally by treating the salivary chromosomes6 and the ordinary mitotic chromosomes10 with strong NaOH solutions. Although desoxyribose nucleic acid is present in both cases, as indicated by the positive Feulgen reaction, the "hairs" produced by the action of alkali and by the ribose nucleate solution are possibly not of the same nature, being in one case the product of direct chemical action of OH ions on the chromosomes; and in the other they are more likely to be the result of the physiological response of the nucleus to the disturbed cellular conditions induced by the ribose nucleate.

In some of the metaphase chromosomes, the arms of the sister chromosomes may be far apart while the kinetochores are not yet split (Figure 5A), and the "hairs" are invariably absent in these chromosomes. The "hairs" are visible, however, while the arms of the sister chromosomes are still closely associated with each other. It seems from these observations that the falling apart of the arms of the sister chromosomes is in some way related to the sudden disappearance of the "hairs," but what this relationship is, is not clear.

In many of the metaphase chromosomes, the splitting of the kinetochore is unusually delayed, and the X‑shaped figures (Figure 5A) similar to those found in colchicine‑treated root tips are found (see Levan,8 for similar figures in onion root tips treated with colchicine). Although delayed, the splitting does occur and the sister chromosomes may be found completely separate and lying closely side by side (Figure 5B) ;or they may be scattered in the cells (see Figure 5C) and also Nebel and Ruttle,9 for similar figures in colchicine‑treated material). In some of the treated roots, over 10% of the dividing cells have shown delayed splitting of the kinetochore and scattering of the anaphase chromosomes in the cell; but in most, if not all of these cells, the chromosomes move apart to the poles and cytokinesis follows. The distribution of the chromosomes in the anaphase is frequently irregular, however. Thus, one or more chromosomes may be slower in movement than the others. They may even be left on the metaphase plate, while others are already at the poles. In other cells the chromosomes may separate in unequal numbers. Cases of the latter type of irregularity that have been precisely analyzed show distribution of 12:20 and 13:19. A number of other anaphase figures have been found in which the two groups apparently differed in chromosome number but an accurate analysis was impossible because the chromosomes were too closely grouped together. Some of these might represent other possible unequal distributions. It is clearly evident from the accurately analyzed cases that unequal distribution, possibly at random, of the 32 chromosomes occurs, giving rise to cells with higher and lower numbers of chromosomes than the diploid. This the author considers to be due to random movements of sister chromosomes to the poles; that is, it is a matter of chance whether the sister chromosomes move together to the same pole or separate to go to opposite poles. In such a situation, it would be expected that all 32 chromosomes would move together to the same pole only with extreme rarity thus giving rise to a tetraploid cell. So far no tetraploid metaphase has been found even in material exposed to 72 hours of treatment in the solutions followed by return to tap water. This is probably due to the extreme rarity with which this would occur by chance. It may he added here that in all roots which took full effects of the treatment, metaphases are present in large numbers while anaphases are much less frequent than in untreated roots. Actual counts made in a few roots showed that among all the dividing cells 41% were at the metaphase and 5% at anaphase, the rest being mostly prophases.

Besides the unequal distribution of the type just described, there has been found another type of abnormality in the chromosome distribution in the same root tips. This is illustrated in the elongated cell to the right in Figure 3D, which contains a large and a very small nucleus. The larger resting nucleus is approximately equal in size to the normal diploid nucleus in the neighborhood. The micronucleus at the opposite pole in this cell contains a very small number of chromosomes, probably only a few. If 32 chromosomes were involved in the unequal distribution in this cell, the number of chromosomes in the larger nucleus would be close to 32 and its size should be much larger, instead of being approximately the same as the normal diploid nucleus, as it actually is. It is evident, therefore, that 16 chromosomes instead of 32 were distributed unequally to the poles in this cell. A number of cells at the anaphase stage showing unequal distribution of 16 chromosouies have actually been found. In Figure 6A‑D are shown 0‑16, 2‑14, 3‑13 and 5‑11 distributions. Other distributions have been found also. In the telophase, the smaller groups of chromosomes form micronuclei, when the chromosome numbers are very small, and in the resting stage following cytokinesis, these nuclei become heteropycnotic.

Careful observations of these 16 chromosomes which are undergoing the second type of unequal anaphase distribution have revealed that these chromosomes consist of strands which are the same in number and size as those of the anaphase chromosomes in the untreated root tips. This inevitably means that these chromosomes had not reduplicated since the anaphase of the previous cycle. Unreduplicated metaphase chromosomes, 16 in number per cell, have been also found. Evidently being unable to split into sister chromosomes, these 16 chromosomes have moved to the poles at random.

In many anaphase cells, particularly in those with unequally distributed chromosomes, one or more chromosomes are delayed in movement. This is illustrated in Figure 4C and 6B. In addition to lagging chromosomes, there have been found in the same cells and in different cells one or more akinetic fragments (Figure 4F). These fragments are of different lengths, but are composed always of the same number of strands as the other chromosomes in the same cell. In Figure 4F the main body of the chromosome from which the fragment is separated can be partly seen as a protrusion from the telophase nucleus at the left. The cause of the chromatid breakage in the formation of these fragments will be discussed later.

In many of the fully affected roots, a number of late anaphase cells have been found showing one or more chromosome bridges between the two groups of chromosomes (Figure 4D). These bridges are formed by the sister chromosomes which have moved apart to the poles but stuck together at the very tips of their free ends. In most of the cases, these chromosomes separate before the anaphase stage terminates, but in some cases they remain attached until the cell plate breaks them apart. In these cases, the breakage might not be exactly at the points of attachment: then one chromosome with a terminal duplication and another with a deletion might arise.

In the roots treated with 8% solution at pH 7 for 20 hours, a number of metaphases were found in which the chromosomes are contracted far more than in those treated with less concentrated solutions (Figure 4E). In some instances many of the chromosomes are almost spherical and the primary and secondary constrictions are not visible (note these features in Figure 4E). These chromosomes also show the lampbrush‑like structure. Furthermore, the nuclear membrane seems to be still present in these cells and the chromosomes are all lying closely against it. The membrane disappears later and the chromosomes, after splitting, move apart to the poles as in normal mitosis. In addition to these abnormalities, there have been found a large number of cells in which the chromosomes are in a more or less compact mass due apparently to their extreme stickiness. A similar abnormality was found also in the roots treated with less concentrated solutions, but in a much lower frequency.

The physiological mechanisms underlying the various effects of the ribose nucleate solutions on the meristem cells cannot be discussed in detail at this stage of the investigation. But it does seem appropriate to consider the physiological nature of some of the outstanding effects. It was noted above that the growth is markedly suppressed in all effective concentrations of the ribose nucleate solutions. This means that the rate of cell division in the meristem is considerably reduced. However, from the examination of root tips cultured for various lengths of time (8, 12 and 16 hours) it was found that the proportion of dividing cells to all cells in the root tips is approximately the same regardless of the time of treatment. Furthermore, this proportion is also about the same as that in the normal root tips growing in tap water. This must mean that the duration of the mitotic cycle is enormously prolonged by the treatment. In tap water it is about four hours4; in the present experimental solution it must be considerably longer than that. It was also observed that the proportion of cells in metaphase is increased with respect to the cells in other stages as the treatments are prolonged. For instance, in root tips treated with 8% solution for six hours, about 25% of all dividing cells are in metaphase, while in roots treated with the same solution for 19 hours, the percentage is nearly doubled. It is apparent, therefore, that the metaphase stage is prolonged considerably more than the other stages. Moreover, since the proportion of the dividing cells to the total number of cells in the roots does not appreciably increase, the number of cells that enter into the divisional phase must be very small.

Among the various effects of ribose nucleate solutions on the chromosomes, probably the most interesting are the inhibition of chromosome reduplication and the formation of akinetic fragments. These two effects seem to be closely re­ated to each other. A breakage that causes fragmentation of the chromosomes is probably initiated some time before the fragment is observed as such, probably when the chromosome reproduces. At this time a replica of the chromosome is formed along its entire length, except at one point which will appear later as the point of breakage. Fragmentation would then be due to a partial inhibition of the chromosome reduplication. For a similar type of fragments found in the pollen grain of Tradescantia treated with mustard‑gas vapor, Koller7 offers a similar explanation for the cause of the breakage. He believes that the inhibition of reproduction is due to inactivation of a locus as the result of chemical combination of the locus with the mustard gas molecules, and a consequent inhibition of reduplication or polymerization of the desoxyribose nucleic acid. We will discuss later whether or not this hypothesis can adequately explain the present case.

A consideration of the cause of the reproductive failure in the treated chromosomes raises a question as to the nature of the component in the ribose nucleate solution that exerts this and other effects on the chromosomes. Zittle12 studied the substrate activities of five different commercial products of sodium ribose nucleate (including that of the Schwarz Laboratory). He found that these products invariably contain a large quantity of polynucleotides but very little of the tetra‑nucleotide or mono‑nucleotide. From the consistently low content of mono‑ and tetra‑nucleotide in all the different products, it can be assumed that the ribose nucleate used in the present experiment is very closely similar to the samples analyzed by Zittle.

In an attempt to determine what component of the sodium ribose nucleate solution is responsible for the various effects, four different mononucleotides of ribose nucleic acid are now being used in tests with onion roots. Complete results will be reported later. However, a preliminary result of the experiment with adenylic acid can be presented here. [334] In the onion roots treated for 18 hours in a 1% solution of this nucleotide, in the manner similar to that of the previous experiments, it was found that the growth is suppressed, a constriction is formed in the region above the meristem. The morphology and mitotic behavior of the chromosomes in the meristem is also changed in somewhat the same way as in the ribose nucleate solutions. Detailed examinations of the chromosomes in the meristem of these roots have shown the following effects: unusually high condensation, lamphrush structure, anaphase bridges, earlier splitting of the sister chromatids in prophase and scattering over the entire cells during anaphase. Delayed splitting of the kinetochore, fragment formation and the random distribution of the chromosomes have not been found. Apparently the kinetochores always split in time and the chromosomes always reduplicate. In contrast to the complete or partial inhibitory action on chromosome reproduction of the sodium ribose nucleate solutions, the adenylic acid is entirely devoid of such action. The delayed splitting of the kinetochore and the inhibition of chromosome reduplication are evidently due to some other substance. The possibility that this substance is a monoor tetra‑nucleotide seems to be excluded for the following reason. It was found by Zittle12 that the commercial products of free ribose nucleic acid contain a large amount of acetic acid‑soluble mono‑ and tetra‑nucleotides. The proportion of the nucleotides in the purified product of Schwarz Laboratory was 28%. the rest being mostly polynucleotides. If the effects of the ribose nucleate solution were entirely due to tetranucleotides or to a certain mononucleotide, it would be expected that the root tips treated with a solution of free ribose nucleic acid would duplicate completely the effects of the ribose nucleate solution. Onion root tips treated with a saturated solution of free nucleic acid for 6, 12 and 18 hours did not completely duplicate the ribose nucleate effects but did duplicate those of the adenylic acid. This result indicates that neither the mononucleotides nor the tetranucleotides in the sodium ribose nucleate are able to delay splitting of the kinetochore nor to inhibit completely or partially chromosome reproduction.

The content of mononucleotides in the sodium ribose nucleate is very small, if there is any at all, yet it produces the effects that could be attributed to these low nucleotides. Either the amount of these nucleotides present is large enough to produce the effects or else they are formed from the tetra‑ and polynucleotides which the cells might have taken up from the solution. This latter possibility would depend upon whether these nucleotides are able to enter into cells of the root meristem.

According to Zittle, both the sodium ribose nucleate and the free nucleic acid contain large amounts of polynucleotides. It might, therefore, be that the effects produced by the sodium ribose nucleate which are not attributable to low nucleotides are due to polynucleotides. The fact that these few effects are not produced by the ribose nucleic acid could be explained by the insolubility of this compound in water.

A possibility that the effects of the ribose nucleate solution are due to an impurity seems to be excluded, because it was found recently by Mr. Irving Galinsky that a certain component of mononucleotides induces many of the same effects.

Tests for other components of the mononucleotide molecule are now tinder way and results will be reported elsewhere. However, the result of experiments with uracil, a pyrimidine base of ribose nucleic acid, may he reported here. This compound is only very slightly soluble in water; therefore, a saturated solution was used. Onion roots treated with this solution showed after seven hours of treatment a relatively high contraction of the chromosomes, lampbrush structure and a large number of chromosome bridges during anaphase. After 18 hours of treatment no further change in the morphology of the chromosomes was found. The growth of the root was not appreciably suppressed. [335] Whether other purine and pyrimidine bases can induce partly or entirely the same changes as the solutions of ribose nucleate and mononucleotide remains to be seen.

It is of interest to note that the chromosomes in many tumor cells are increased in number and show stickiness. clumping, etc.1,2,8 Many of these abnormal features of the chromosomes in tumor tissues parallel those observed in the meristem cells cultured in ribose nucleate solutions. The close similarities in the chromosome abnormalities between these tissues and the fact that the nucleic acid content of the tumor cells is often greater than in normal tissue (see Stowell11) suggest that these abnormal features in neoplastic tissues may possibly be due to an excessive amount of nucleic acid.

I am gratefully indebted to Professor C. L. Huskins for his invaluable suggestions and criticism.

Literature Cited

  1. BIESELE, J. J. Cancer Res. 4:540. 1944.
  2. BIESELE, J. J., and E. V. COWORY. J. Nat. Cancer Inst. 4:373. 1944.
  3. BIESELE, J. J., H. POYNER, and T. S. PAINTER. Univ. of Texas Publ. No. 4243. 1942.
  4. DERMAN, H. Bot. Rev. 6:599. 1940.
  5. DURYEE, W. R. Cytology, Genetics and Evolution. Univ. of Penn. Press. 1941.
  6. KODANI, M. J. Hered. 33:115. 1942.
  7. KOLLER, P. C. Symp. of Soci. for Exper. Biol., No. 1, Cambridge Univ. Press. 1947.
  8. LEVAN, A., Hereditas. 24:471, 1938.
  9. NEBEL, B. R., and M. L. RUTTLE, Jour. Hered. 29:3. 1938.
  10. PAINTER, T. S., Anat. Rec. 87:34. 1943.
  11. STOWELL, R. E. Symp. of Soci. For Exper. Biol., No. 1, p. 190. 1947.
  12. ZITTLE, C. A. J. Biol. Chem. 160:527. 1945.

Somatic reduction division also has been induced by caffeine and uracil.

Acta Hort. (ISHS), 300:377-380, 1992
Cytological characterization of cell suspension cultures of fruit trees.
Blando, F., Giorgetti, L., Tonelli, M.G. and Nuti Ronchi, V.

Abstract: Using a new suitable method to initiate a cell suspension culture in apple (Malus x domestica Borkh.) and quince (Cydonia oblonga Mill.), it has been possible to detect some cytological mechanisms of chromosome reduction, firstly reported in carrot cell cultures (Nuti Ronchi, 1990; Nuti Ronchi et al., 1990). These segregational events, namely somatic-meiosis and prophase-reduction, are present in all analysed cultures, haploid metaphases (n=17) being 3.2% in apple lines. The importance and role of these phenomena in cell cultures of different non-embryogenic species are discussed in comparison with the carrot cell culture model.


Fundamental and Molecular Mechanisms of Mutagenesis, 452 (1): 67-72, 2000
Inducing somatic meiosis-like reduction at high frequency by caffeine in root-tip cells of Vicia faba
Yihua Chen, Lihua Zhang, Yihua Zhou, Yuxuan Geng and Zhenghua Chen
Institute of Genetics, Chinese Academy of Sciences, Beijing 100101, People's Republic of China

Abstract: Germinated seeds of Vicia faba were treated in caffeine solutions of different concentration for different durations to establish the inducing system of somatic meiosis-like reduction. The highest frequency of somatic meiosis-like reduction could reach up to 54.0% by treating the root tips in 70 mmol/l caffeine solution for 2 h and restoring for 24 h. Two types of somatic meiosis-like reduction were observed. One was reductional grouping, in which the chromosomes in a cell usually separated into two groups, and the role of spindle fibers did not show. The other type was somatic meiosis, which was analogous to meiosis presenting in gametogenesis, and chromosome paring and chiasmata were visualized.


Journal of Agricultural Biothechnology, 8(2): 147-150, 2000
Somatic Meiosis-like Reduction Induced at High Frequency by Nucleotide Analogues in Vicia faba
Zhang Lihua, Chen Yihua, Chen Zhenghua
Institute of Genetics, Chinese Academy of Sciences, Beijing 100101, People's Republic of China

Abstract: Root tip cells of Vicia faba were treated by caffeine and uracil to induce high frequency of somatic meiosis-like reduction. Optimum treatments including the concentration of the chemicals and the time of restoration were screened out. The highest frequency of somatic meiosis-like reduction approached up to 54.0% and 35.7% for caffeine and uracil treatment respectively. The process of somatic meiosis-like reduction was observed, which existed in two types. One was chromosome reductional grouping, in which chromosomes in a cell usually separated into two groups without the role of spindle fibers. The other type was somatic meiosis, which was analogous to meiosis in gamatogenesis and characterized by pairing and chiasmata of chromosomes. Establishment of this inducing system has a profound meaning in the study of producing haploid cells as well as in cell engineering breeding.


In Vitro Cellular and Development Biology - Plant, 37(5): 654-657(4), September 2001
Meiosis-like reduction during somatic embryogenesis of Arabidopsis thaliana
Yihua C.; Lihua Z.; Yuxuan G.; Zhenghua C

Abstract: Somatic meiosis-like reduction was observed in some cells of the embryogenic callus of Arabidopsis thaliana. Two types were identified. One type was somatic chromosome reductional grouping, in which the chromosomes in a cell were separated directly at either prophase or metaphase. Chromosome reductional grouping happened more frequently in polyploid cells, and the morphology of the chromosomes did not show the role of the spindle fibers. The other type was somatic meiosis which was analogous to the process of gametogenesis, characterized by the pairing and synapsis of homologous chromosomes. The roles of somatic meiosis-like reduction in somatic embryogenesis and somaclonal variations are discussed.