The study of the reactions of the embryonic organism to chemical substances is one of the main approaches to the analysis of embryonic development. Well known is the use of the lithium ion which produces in embryos of diverse species a variety of developmental abnormalities. However, in the case of a substance like lithium an analysis of the particular sequence of chemical reactions which leads eventually to the manifestation of a morphological abnormality encounters a peculiar difficulty. Since lithium reacts with numerous proteins, a wide range of reactions has to be considered as the initial step in the teratogenic mechanism.

For this reason it seemed advantageous to consider substances as potential teratogenic agents which are known to react only with one or a few definite chemical components of the cell. A class of substances which seems to fulfil these requirements is known in biological chemistry as ‘analogues’ (Woolley, 1950, 1952; Martin, 1951). Because of their close resemblance in chemical structure to vitamins, metabolic intermediaries, or other cell constituents these analogues interfere specifically with the proper function of the corresponding natural substances. Thus at least the starting-point of the disturbed mechanism is indicated and its further analysis greatly facilitated. Malformations in the chick embryo have been produced with a series of vitamin analogues. This work has been recently summarized by Cravens (1952). From previous studies by other authors very little has become known about the role of a disturbed protein metabolism in abnormal morphogenesis. Since amino-acid analogues offer the possibility of a more direct interference with protein metabolism, the use of such analogues seemed of interest in this connexion. The following outline is a preliminary report concerning disturbances in the development of chick embryos which were brought about by some amino-acid analogues.

In the present series of experiments tests were carried out on explanted chick embryos. The in vitro technique provides simplified conditions under which the developing embryo can be subjected to a variety of experimental factors and the results can be more readily observed and analysed.1 Chick embryos were removed from the egg after about 24 hours of incubation. In most experiments embryos from the late headfold stage up to the 6-somite stage were selected. Controls and experimental embryos of exactly matching developmental stages were used for comparison. The embryos were transferred to the surface of a mixture of agar and an extract of egg. A total volume of 2 ml. of this nutrient was used. Except for small modifications the procedure for explantation and maintenance follows a technique described by Spratt (1947). The explanted embryos were kept in the incubator for about 18 hours. During this period the control embryos develop the primitive organs which correspond to those of a 15–20-somite embryo developing in the egg; the primitive central nervous system, optic vesicle and otocyst, a beating heart, a regular string of somites, &c., can be observed. Thus, while the period of explantation is short, it includes very fundamental steps in organogenesis.

In the present experiments two groups of analogues were used. The one comprised analogues of the aromatic amino-acid phenylalanine, the other group included analogues of the aliphatic amino-acids leucine and valine.2 The structural formulae of the analogues and of the corresponding natural amino-acids are found in Table I. With the analogues of phenylalanine, i.e. thienylalanine, and p-chlorophenylalanine, abnormal forms of development were obtained at concentrations of the analogue in the culture medium of 2· 5 and 1· 5 mg./ml. respectively. Ortho-chlorophenylalanine was without any effect at a concentration of 1·5 mg./ml. One of the main defects which was observed under these conditions was a retardation of growth. As an example, data for four controls explanted at the headfold stage gave an average length of 4· 0 mm. at the end of the incubation period, while five embryos developing on a thienylalanine medium reached only a length of 2· 9 mm. If the natural amino-acid phenylalanine and the analogue were added to the medium simultaneously and in equal concentrations, the inhibition of growth was reversed at least in part (5 embryos with an average length of 3 ·4 mm.). In addition to the inhibition of growth, a variety of abnormal forms of development were observed. Very prominent were a zigzag neural tube, an irregularly formed brain, and the somites were often smaller and irregular in shape. However, the addition of phenylalanine did not reverse the abnormal forms of development.

TABLE 1

Natural amino acids and their analogues used in experiments

Natural amino acids and their analogues used in experiments
Natural amino acids and their analogues used in experiments

The analogues of leucine and valine3 interfere with normal development at much lower concentrations than the analogues of phenylalanine. Effects of the chlorine derivatives were observed regularly with 0·25 mg./ml. and of bromoallylglycine with 0· 15 mg. / ml. At these concentrations the inhibition of growth was slightly less than the inhibition found with thienylalanine. However, the leucine analogues seem to affect mainly the increase in length and less the increase in the other dimensions, while thienylalanine affects more evenly the growth in all dimensions.

The most characteristic abnormality obtained with the leucine analogues consists in an impairment of the segmentation of somites, that is, in a failure of the somites to separate in the direction of the longitudinal axis of the embryo. The chlorine analogues exert this effect only during a short time. Two to four somites which normally separate 3–5 hours after explantation remain fused and thus form what we call a ‘somite block’. Somites which form later separate again normally. In recent experiments we obtained evidence that the transient nature of this effect is due to the lability of the chlorinated analogues. In corroboration of these experiments we found that with a stable analogue like bromoallylglycine the failure of the somites to separate is much more extensive and usually involves all somites which enter the state of separation after the analogue is added.

The effects observed with the chlorine analogues of leucine are irreversible on addition of leucine. This irreversibility may also be a consequence of the labile nature of the chlorine analogues. During their breakdown these compounds may form covalent bonds with the tissue proteins which are not affected by the addition of the natural amino-acid. Such a possibility is borne out again by the use of the stable analogue bromoallylglycine, the effects of which are reversible. Whether there exists a dependence of the reversibility of the effects of bromoallylglycine and thienylalanine upon the stage of development of the embryos is now being investigated.

Histologically the process of the separation of somites consists in the rearrangement of the cells in the mesodermal plate and the formation of mesenchymal tissue which forms fibrous septa between the somites. Which phase of this process is actually suppressed is now being investigated. In considering the morphological mechanism of the separation the question arises whether the decrease of the growth rate of the embryo by leucine analogues may prevent the somites from separating. From our experimental material this possibility seems unlikely. We found equally extensive blocks of somites in embryos in which no retardation of longitudinal growth was observed after addition of leucine analogues, and the retardation of growth by thienylalanine leads only in very rare instances to the formation of ‘somite blocks’.

So far we have found no evidence indicating that the suppression of the separation affects the differentiation of the somite into a dermomyotome and a sclerotome. In sections the development of these tissue components is found to have progressed to about the same extent in ‘somite blocks’ as in normal somites.

The chlorinated leucine analogues very rarely cause disturbances of the central nervous system. There may be a somewhat higher incidence with bromoallylglycine, but a greater number of experiments is required to make these figures more definite.

Summarizing the description of our results the striking differences in the effects brought about by the two groups of analogues may be pointed out. Phenylalanine analogues produce malformations only in concentrations 10 times higher than those found to give malformations with leucine analogues. Phenylalanine analogues suppress somite separation only as a rare exception, while this anomaly can be found regularly with leucine analogues. On the other hand, abnormally small and irregularly shaped somites are characteristic for thienylalanine and appear only rarely with leucine analogues. Abnormalities of the central nervous system are brought about very frequently by phenylalanine analogues but very rarely by the chlorine analogues of leucine.

An attempt at a more specific interpretation of the observed effects seems premature. Even the more extensive studies of the effects of amino-acid analogues on the growth of bacteria and other lower organisms have, apparently, not yet led to more definite suggestions about the mechanism of their action (Dittmer, 1950). However, it is likely that in many instances the analogues interfere more or less directly with the formation of new proteins. An inhibition of the oxidation of the amino-acids, leading to a diminished energy production, and hence to a reduced synthesis of proteins, seems unlikely in view of the recent results of Frieden, Hsu, and Dittmer (1951).

For the problem of abnormal development there remains the question as to how a diminished rate of protein synthesis would give rise to such malformations as were observed in our experiments. Since the tested analogues are antagonists of amino-acids which occur in practically all cytoplasmic proteins, the impairment of the utilization of these amino-acids should give rise to suppression of protein formation in general. In this case different analogues should bring about, at least qualitatively, the same type of malformation. Since the results of our experiments show qualitatively different effects of different analogues, this assumption cannot hold. In this case we have to postulate that different analogues inhibit the formation of different proteins to a different extent. This postulate can be tested experimentally by the measurement of the rate of incorporation of labelled amino-acids into different proteins in the absence and presence of amino-acid analogues. If different analogues inhibit the synthesis of proteins to a different extent, the rate of incorporation of labelled amino-acids into different proteins should correspondingly be reduced by different degrees. Experiments to test this hypothesis are in progress in our laboratory.

Cravens
,
W. W.
(
1952
).
Ann. N. Y. Acad. Sci
.
55
,
188
.
Dittmer
,
Karl
(
1950
).
Ann. N. Y. Acad. Sci. 52, 1274
.
Frieden
,
HSU
,
Lien
Tien
, &
Dittmer
,
Karl
(
1951
).
Jour. biol. Chem
.
192
,
425
.
Martin Gustav
,
J.
(
1951
).
Biological Antagonism
.
New York
:
Blakiston Co
.
Spratt
,
N. T.
, Jr
. (
1947
).
J. exp. Zool
.
106
,
345
.
Woolley
,
D. W.
(
1950
).
Antimetabolites
.
Ann. N. Y. Acad. Sci
.
52
,
1235
.
Woolley
,
D. W.
(
1952
).
A Study of Antimetabolites
.
New York
:
John Wiley & Sons, Inc
.
1

The work using this technique was carried out at our laboratory by U. Rothfels and M. Curry.

2

We are indebted for the supply of chlorine analogues of phenylalanine, leucine, and valine to Dr. H. J. Klooster (U. Colo.), and for the bromoallylglycine to Dr. K. Dittmer (U. Florida).

3

In the remaining part of the paper reference is made only to results obtained with leucine analogues. Parallel experiments with the corresponding chlorine analogues of valine gave essentially the same results.