NATURE 212(5063): 697-699 (Nov. 12, 1966)
Department of Agricultural Botany, University College of Wales, Aberystwyth

HERITABLE changes are induced in flax plants of the variety 'Stormont Cirrus' grown in different environments1. Plants of the original variety (plastic genetroph) are changed into a large stable genotroph L or into a small stable genotroph S according to the nutrients supplied; plants treated similarly yield similar uniform progeny which breed true. L is six or more times the size of S and both varieties have remained unchanged for ten generations irrespective of the nutrients subsequently supplied. They behave as two distinct genetic types in giving equilinear inheritance when they are reciprocally crossed and there is no transmission through reciprocal grafts, so that it is likely that nuclear changes have occurred in the course of induction. The varieties differ from normal genotypes, however, in the comparatively large increase of the variation in the F1 of their reciprocal crosses; this variation is inherited by the F2 and succeeding generations with the occurrence of a range of types (epitrophs) varying in size from L to S (refs. 2 and 3). Presumably when the nuclear factors responsible for the difference between L and S are brought together in the F1 they become unstable, the degree of instability being influenced by the external and cytoplasmic environments. Genotrophs varying in weight between L and S can also be induced directly from the plastic genotrophs by varying the environments in which the plastic plants are grown.

Induction and inheritance. The balances of environmental factors necessary for induction of L and S were not all known so that induction experiments were not always successful. All recent investigations show, however, that induced changes will occur if the plastic plants are grown in conditions which promote rapid growth—fertile soil, high temperature and a day about 14 h long. Nitrogen (ammonium sulphate fertilizer) is added to induce L, and phosphorus (triple superphosphate fertilizer) is added to soil with a lower pH to induce S. Certain other combinations are also effective.

Large and small genotrophs have been induced from another variety of flax, 'Lyral Prince'. The differences of plant weight, height and general appearance between L, S and the original plastic form are similar to those between the genotrophs of 'Stormont Cirrus'. A diallel cress between L and S of both varieties gives no complementation. and no pronounced reciprocal differences. There is increased F1 variation with F1 instability in crosses of L and S both between varieties and within them, but not between large genotrophs of the two varieties or between the two small genotrophs. The implication is that similar induced changes occur in both varieties, that they probably have a nuclear basis, and that the genotrophic differences override the genotypic differences between the varieties at least with respect to plant weight.

The differences of plant weight between genotrophs and between epitrophs are essentially quantitative, and the range of variation observed would normally be attributed to several genes or to polygenes. The variation may alternatively arise from changes at one locus, existing in many states. Because of the unstable nature of the F1 there is no certainty that linkage investigations to determine whether chromosomal factors are responsible for the difference between L and S would give a positive answer, and such investigations have se far been inconclusive. In the F2 of crosses between L and S of 'Stormont Cirrus' with the mauve flowered variety 'Mandarin', Tyson found normal segregation of flower colour and no evidence of linkage with plant weight. One other useful marker, a white flowered mutant, has been obtained by irradiating L of 'Stormont Cirrus' with gamma rays from a cobalt-60 source. In the F2 and F3 of crosses of this mutant (denoted LW) with S, flower colour segregated normally and there was a significant association with plant weight in the opposite direction from that expected from linkage. In another experiment with plants grown under different conditions, no association was obtained in the F2. Examination of the chromosomes with the ordinary light microscope, like the linkage experiments, failed to confirm the suspected nuclear basis for the induced changes, but other methods have been more successful.

Genotrophs and varieties. Estimates of nuclear DNA were made in 4C resting nuclei in apical meristems by Feulgen photometry of the L, S and plastic genotrophs of 'Stormont Cirrus' and 'Lyral Prince' and of other flax and linseed varieties which for the past seven or eight generations had been grown under ordinary field conditions not likely to induce inheritable changes. The data are given in Table 1. It will be seen that, according to this method of analysis, there is 16 per cent more nuclear DNA in L of 'Stormont Cirrus' than in S, and that the plastic form has an intermediate amount. Similarly, in 'Lyral Prince' L has 10 per cent more DNA than S and the plastic form again has an intermediate amount. These differences are statistically highly significant and imply that the environments which first induced the permanent inheritable changes in the weights of plants have also induced permanent changes in the amount of nuclear DNA.

Although these estimates were made under uniform conditions of fixing and staining, and with extensive replication, it is possible that differences in the amounts of DNA could be due to differences in the penetration of fixatives or stains in the morphologically distinct tissues of the different genotrophs. To rule out this interpretation, the nuclei were extracted from the cells and stained and analysed in isolation. The results were in complete agreement with those given in Table 1.


(arbitrary units)
  S. Cirrus Plastic 85.8
  L 93.7
  S 80.5
  L. Prince Plastic 90.5
  L 95.3
  S 86.3
Flax varieties  
  Percello   93.3
  L. Monarch   88.5
  S. Gossamer   85.2
  Hollandia   87.8
  S. Motley   87.7
  Rembrandt   87.0
  N. Princess   86.5
  Mandarin   85.3
Linseed varieties  
  Dakota   94.3
  Royal   87.6

Other measurements of nuclei confirm the nuclear change inferred from photometric estimates of DNA. First, the nuclear volumes are larger in L than in S. Second, total dry mass of nuclei as measured by interference microscopy differs significantly between genotrophs and runs parallel to the variation in the amount of DNA. Third, a chemical analysis of nuclear DNA in germinating seedlings by Holdgate4 showed 10 per cent more DNA in L than S. Although this last result agrees reasonably well with expectation, it must be regarded as an approximation because the method used did not fully exclude the effects of the size of cells or differences of mitotic rate between the tissues of the different genotrophs. Tests were also carried out to determine whether the differences might be due to a faster metabolic turnover of L compared with S; we measured the nuclear DNA in plants grown in environments conducive to different growth rates, but the results were again the same as in Table 1.

There are also differences of nuclear DNA between the flax varieties in Table 1. They are smaller than those between the genotrophs and are not significant. Even so, 'Mandarin' and 'Percello', which have the smallest and largest amounts of DNA respectively, are also characterized by the smallest and largest plant weights; therefore it is possible that inherited differences between some of these varieties are in part genotrophic, or environmentally induced. The two linseed varieties are larger and taken together have significantly more nuclear DNA than the varieties of flax, although 'Royal' has a larger plant weight than 'Dakota'.

Further confirmation of the correlation between nuclear change and phenotype was obtained from an analysis of the derivatives of a cross between LW and S. Large and small plants (epitrophs) were selected in the F1 and in subsequent generations up to the F4 where there were obvious differences of size between the large and small epitrophs even though these were still somewhat less than the difference between LW and S. Nuclear DNA estimates for F5 plants with blue and white flowers of the selection lines and of L, LW and S are given in Table 2. L has about 16 per cent more DNA than S as before, and LW is almost identical with L. The differences between the large and small epitrophs are significant and in the expected directions; that is, large epitrophs have more nuclear DNA than the small, but, as with the weights of the plants, the differences are not as great as those between LW and S. Variation in DNA between F5 plants descended from different F4 plants was significantly greater than between plants descended from the same F4 plants and also greater than the variation between plants within any one genotroph descended from different parent plants, so that it is probable that further divergence can be expected between the epitrophic lines, with the possibility that it may eventually be possible to detect epitrophs with the same DNA content and intact weight as the L and S genotrophs.


  Nuclear DNA
(arbitrary units)
L 84.5
LW 85.5
S 75.2
F5 of the cross S x LW
 Selection for high plant weight  
    Blue flowered 77.2
    White flowered 82.3
 Selection for low plant weight  
    Blue flowered 75.7
    White flowered 76.6

Induction of nuclear change. Several observations have shown that the first five weeks of growth is the critical period during which the environments induce heritable changes. The question arises as to whether the nuclear changes occur at this time and, to answer this, plastic plants of 'Stormont Cirrus' were grown in the contrasting environments, with high nitrogen for the induction of L and high phosphorus for the induction of S. Estimates of nuclear DNA were then made on plants sampled at weekly intervals. The estimates, in Fig. 1, are expressed as percentages of the amounts in plastic plants grown in a normal, non-inducing environment and sampled at the same time. This ratio shows gradual changes of nuclear DNA in the expected directions during the five weeks. At the end of this period the difference in nuclear DNA is practically identical with that already described between L and S genotrophs. Not only do nuclear changes take place during this period of five weeks but they are also apparently completed in that time.

Fig. 1. Changes in DNA in the nuclei of plastic plants grown in inducing environments during the first five weeks of growth. The two points show the percentage values of large and small genotrophs. As judged by changes in nuclear DNA the induction is complete by the end of the first five weeks.

Nature of nuclear change. The differences of nuclear DNA demonstrated by Feulgen photometry are probably in the main located in the chromosomes. The best evidence for this is that the DNA differences apply not only to 4C interphase nuclei but to metaphase chromosomes as well. Any differences associated with nucleolar or other extra-chromosomal material in the nucleus must therefore be small or non-existent because such matter is entirely, or almost entirely, absent at metaphase. It is therefore pertinent to consider how the differences could arise in the chromosomes. Flax has 15 pairs of chromosomes of approximately equal length. If only one particular pair of chromosomes were involved in the change, then, with an overall 16 per cent DNA difference between L and S, the F1 of L x S would contain one chromosome twice the size of its homologue. At metaphase 1 of meiosis, this feature would be obvious in the form of a grossly asymmetrical bivalent. No such difference has been observed and it is reasonable to suppose that the DNA changes are spread over several, perhaps all, of the chromosomes in the complement.

It is not known how the DNA changes are brought about. They may involve chromosome material which is not part of the permanent DNA helix or helices or they may, alternatively, result from an altered polynemy of whole chromosomes or segments of them. Our evidence does suggest, however, that they are induced at a particular and specific stage of development and that the quantitative changes in nuclear DNA are associated with the quantitative changes in plant weight.

  1. Durrant, A., Heredity, 17, 27 (1962).
  2. Durrant, A., Nature, 196, 1302 (1962).
  3. Durrant, A., and Tyson, H., Heredity, 19, 207 (1964).
  4. Holdgate, D. P., thesis, University College of Wales, Aberystwyth (1964).