Embryonic Responses to Structurally Related Inhibitors

MANY of the so-called nucleic acid antagonists inhibit amphibian development (Bieber, 1954; Bieber, Nigrelli, & Hitchings, 1952; Waddington, Feldman, & Perry, 1955); a number of them affect embryological processes differentially (Liedke, Engelman, & Graff, 1954, 1955), some of them interfering primarily with early cleavage stages where synthesis is not appreciable, others affecting most drastically only late development in which differentiation and synthesis are marked. These compounds, which bear some structural resemblance to the naturally occurring purines, affect other biological systems also, among them Tetrahymena gelii, Escherichia coli, and tissue cultures of mammalian cells (Gillespie, Engelman, & Graff, 1954).

This paper reports the study of the effects of 35 compounds, substituted benzothiadiazoles, benzimidazoles, quinoxalines, and other related compounds. While 8 of these compounds had been used previously (Liedke, Engelman, & Graff, 1954), 27 are additional new compounds, some of them only newly synthesized. The structures of the parent compounds are illustrated in Text-fig. 1. Since it had been observed in the previously reported compounds as well as in those 27 newly reported (Table 1) that a nitro substituent appeared to enhance inhibitory activity, a number of compounds were selected from both groups for more intensive study and comparison. It appeared provisionally that the nature and position of the substituents rather than the structure of the nucleus, whether benzimidazole, quinoxaline, benzotriazole, or benzothiadiazole, would determine, not only the intensity of inhibition, but also the nature of the inhibition and the morphological response of the embryo.

The methods are the same as those described previously (Liedke, Engelman, & Graff, 1954, 1955). Table 1 shows that 17 of the 27 newly tested compounds were inhibitory toward Rana pipiens embryos on continuous exposure starting at Shumway (1940) stages 3,8,14, or 18 at concentrations of 0·1 mg. /ml. excepting where solubility was a limiting factor. Table 2 affords a comparison of the selected compounds at a much lower concentration, 0·01 mg./ml.; the conditions were otherwise the same. The compounds were selected so that they had similar group substitutions on the respective benzothiadiazole, benzimidazole, quinoxaline, or benzotriazole nucleus. Eight of these selected 16 compounds are also listed in Table 1, i.e. they were new ones, while the other 8 had been used in previous experiments (Liedke, Engelman, & Graff, 1954, 1955). Table 2 shows that the substituted compounds are more effective inhibitors than the unsubstituted ones; benzothiadiazole and benzimidazole at low concentrations permit of normal development, but the 5-nitro-substituted compound in each case is strongly inhibitory. Both Tables 1 and 2 show that the benzotriazoles affect late tail-bud stage 18 embryos more severely than they do younger embryos. Two of the benzothiadiazoles, on the other hand, 4-M-6-NBD and 5-NBD, arrest both early and tail-bud stage embryos impartially.

By varying the length of exposure and using a concentration of 0·1 mg. / ml. it was found that the time required to stop exposed 2-cell stages in early cleavage was one hour or less for all four benzothiadiazoles and for all four quinoxalines, but it took 3 hours for 4-M-6-NBZ, and indeed 6 hours for 5-NBT, and even 20 hours for 5-NBZ. The other benzimidazoles (BZ, 6-E-4-NBZ) and the benzotriazoles (4-M-6-NBT and 6-E-4-NBT) arrested the embryos only in later stages on exposures up to 20 hours, starting at the 2-cell stage. To stop blastulae (stage 8) in the early gastrula stage, using the same concentration of 0·4 mg./ml., required only 1-2 hours for 4-M-6-NBD and for 5-NBD, but 3-5 hours for all quinoxalines and for 6-M-4-NBD, and even 10 hours for 5-NBT and 5-NBZ. All other compounds required longer exposures, and even then, after 20 hours, arrested the embryos at later neurula or tail-bud stages (BD, 4-M-6-NBT, BZ). Sometimes development was normal even in spite of 20 hours’ exposure (6-E-4-NBZ, 5-M-4-NBT). To stop all further development and cause death of embryos first exposed in tail-bud stage 18, using the same concentration of 0·1 mg./ml., required only 5-6 hours for 5-NBT, for 4-M-6-NBD, and for 5-NBD, but it took 12-15 hours for 4-M-6-NBT, and at least 30 hours for 5-NBZ and the quinoxalines. Exposure at stage 18 also showed again clearly that 5-NBZ is the most effective and the unsubstituted BZ the least effective in the benzimidazole series. Thus 24 hours of exposure to BZ permitted almost normal development to stage 24, while 24 hours of exposure to 4-M-6-NBZ and only 8 hours to 5-NBZ already caused arrest at stage 22/23 as abnormal hydrops embryos.

Varying the length of exposure confirmed the observation that the benzotriazole compounds have a greater effect on later embryonic stages and that the nitro-substituted compounds are the most effective inhibitors in each of the groups with the possible exception of the quinoxalines.

Histological examination of the embryos confirmed the morphological observations, and showed that the benzothiadiazoles, benzimidazoles, and quinoxalines cause similar typical abnormalities and developmental arrests varying only in degree and depending, of course, on stage exposed. In contrast, however, the benzotriazoles cause a different type of arrest and hypomorphism. In the former there is always a selective cellular response, necrosis of early differentiating cells or of highly active mitotic cells, and enlargement and arrest of sensitive cells in prophase stages. In the benzotriazoles, on the other hand, the whole cell population is arrested equally, all nuclei take on a paler stain, and at the same time there is marked delay in development and often there is swelling of the embryo, particularly in the neurula stage. After exposure to benzotriazoles and return to normal medium (10 per cent. Ringer solution) the embryos, unless they die suddenly, continue with delayed normal or hypomorphic development and a tendency toward cyclopia if exposure took place in the blastula stage. Such drastic delay in development, always characteristic of the benzotriazoles, was, however, also found in a benzimidazole, 5-NBZ, where, on the other hand, it was coupled with even more extensive necrosis of sensitive tissues. Exposure to 5-NBZ at stage 18 destroyed not only early differentiating cells, but practically all cells of the nervous system as well. Even the growing posterior end was severely affected; one day after exposure to 5-NBZ the budding tail had become a swollen edematous ball filled with necrotic elements (Plate, fig. A). The chorda cells failed to segregate and to differentiate from the lateral somite cells (Plate, figs. A, B; compare control embryo, fig. C). Later this tail end turned into a wrinkled epidermal projection containing but a few mesenchyme cells. No regeneration of nerve-cord, chorda, or muscle took place. Instead of a head and a tail, only anterior and posterior epidermal knobs developed in which further growth and differentiation ceased. Despite this, and in the absence of a nervous system except for a few regenerating peripheral cells enclosing the necrotic mass (Plate, fig. E) the embryo survived. All of the benzothiadiazoles, benzimidazoles, and quinoxalines brought about a moderate degree of necrosis of sensitive structures; only a few early differentiating cells were so affected while other cells ceased dividing and remained enlarged. The nitro-substituted compound, 5-NBZ, however, was highly necrotic. 5-NBZ also delayed development of the whole embryo. On the other hand, 5-NBT, which produces similar delay in development, exhibits very little necrotic activity. Figs. B, D, and E (Plate) illustrate this remarkable difference. It is difficult to see how such an abnormal embryo as that produced by 5-NBZ (Plate, figs. A and E) could survive at all, whereas embryos exposed to the benzotriazoles which caused the same developmental delay unaccompanied by necrosis so often die suddenly.

These experiments demonstrate that the inhibitory effects of certain heterocyclic compounds are influenced by substituents; enhancement of inhibition is brought about by nitro-substitution in benzothiadiazoles, benzimidazoles, and benzotriazoles. It is of comparative interest that syndactylism, the teratogenic response of the chick embryo to the naturally occurring imidazoles, pilocarpine, and pilocarpidine, appears to be a qualitative function of the parent structure, and that steric configuration affects activity only quantitatively. In the ‘trans’ configuration (iso) compounds the teratogenic activity is always more intense than in the ‘cis’ compounds (Landauer, 1956). Similarly, cleavage of the seaurchin egg is also inhibited more by ‘para’ compounds than by ‘ortho’ or ‘meta’ compounds (Druckrey, Dannenberg, & Schmähl, 1953).

Histologically, none of the benzotriazoles, not even the highly inhibitory 5-nitro compound, elicit selective cellular responses in the embryo as do the benzothiadiazoles, benzimidazoles, and quinoxalines. 5-NBT resembles 5-NBZ in activity except for the lack of selectivity evinced by 5-NBT. While nitrosubstitution increases activity and other substitutions decrease it, the type of the response elicited seems to be determined by the nucleus, not by the substituent. The benzotriazoles all cause one characteristic response, while the other series of compounds, benzothiadiazoles, benzimidazoles, and quinoxalines, cause another response which is perhaps best characterized by selective cellular necrosis.

There is some suggestive evidence that the compounds used in this study, particularly the benzimidazoles, could possibly act as purine and pyrimidine antagonists. Studies of the embryological effects of this type of inhibitory agent have been made by Brachet (1946, 1950), Bieber (1954), Liedke, Engelman, & Graff (1954,1955), Waddington, Feldman, & Perry (1955); Analysis of specific embryogenic responses has shown that some embryogenic processes can be inhibited in varying degree by a number of the agents employed, while other similar agents have no effect. It has also been shown that some agents, azaguanine and benzimidazole, for example, may have inverse effects on different embryos, i.e. azaguanine is highly inhibitory to chick but not to amphibian embryos, while the situation is reversed for benzimidazole.

While the responses of the embryo are often similar, or differ only in degree, and do not illuminate the underlying biochemical processes, it is of interest that one group of compounds, the benzotriazoles, are more or less uniquely different from the other three groups used in this study. It was shown that the former affect neurula and older stages more than younger ones; they may cause mesodermalization of the notochord or cyclopia (Liedke, Engelman, & Graff, 1954, 1955), while the latter cause selective cellular necrosis of early differentiating cells and sensitive structures. Such differences in response are not found with most agents, i.e. Trypan blue (Waddington & Perry, 1956).

It is, of course, possible that this is not a real difference, but only a matter of degree, similar to that described by Waddington, Feldman, & Perry (1955) as the relatively mild effect of ethionine in comparison to the far more drastic inhibition by azaguanine in the chick embryo. They concluded that ‘there is some over-all parallelism between the pattern of sensitivity of the chick embryo to azaguanine and that of the presumed protein synthesis, indicated by the incorporation of methionine-S³⁵ and sensitivity to ethionine’. However, no such parallelism was found by them for amphibian embryos with glycine-C¹⁴ and methionine-S³⁵ (Waddington & Sirlin, 1954; Sirlin, 1955; Sirlin & Waddington, 1956). If it were only a matter of degree, one might assume that the embryo could respond first with limitation of metabolism to maintenance without abnormalization, secondly, with mesodermalization and cyclopia indicating disturbance of inductive gradients and processes, and thirdly, with selective necrosis of sensitive regions. Such sensitive regions vary with the stage of development, and are usually composed of highly active cells concerned with morphogenic movements, mitosis, or early differentiation. One could assume then that the benzotriazoles are relatively mild inhibitors which produce only the first and second response, while the other compounds call forth the third or most severe response. This explanation is illustrated in the Plate, figs. A and B. In both cases the embryos are arrested and delayed in development (first response), but in the 5-NBZ embryo (fig. A) the sensitive dorsal structures are already necrotic (third response), while they are still normal in the 5-NBT embryo (fig. B). The only difficulty is that the explanation fits only these two compounds with the highly active nitro group substituents, but not the others, where selective necrosis of sensitive areas occurs (third response) without delay in development (first response). Hence, sensitive structures are affected first with all compounds other than the benzotriazoles, while less sensitive structures still manage to develop more or less normally.

Presumably, all the compounds used in this study could affect nucleic acid metabolism, but the unique activity of the benzotriazoles is somewhat puzzling. Only with the benzotriazoles does swelling occur with lighter pigmentation and loss of cells before death, particularly in neurula arrests, all signs of surface layer disturbances. It is possible, therefore, that the benzotriazoles may exert some action on the surface coat of the embryo (Holtfreter, 1943) like that of surface chelating or polymerizing agents. Such action has been reported by Gustafson & Horstadius (1955) for some antimetabolites in studies on echinoderm eggs.

This work was supported in part by grants from the Damon Runyon Memorial Fund and the American Cancer Society.

Grateful acknowledgement is made to Mr. Edward R. Hajjar, of the Department of Medical Photography, Francis Delafield Hospital, for preparation of the photomicrographs.