Comparison of the Different Teratogenic Effects of Three Commercial Samples of Trypan Blue

The teratogenic action of the azo dye trypan blue (Gillman et al., 1948) has been widely accepted. Commercially sold preparations of the dyestuff injected into pregnant rats at days of gestation usually cause a great increase in foetal mortality with subsequent intra-uterine resorption of the dead foetuses; at the same time a substantial proportion of the surviving foetuses are deformed. The efficacy of trypan blue preparations in causing malformations and foetal resorptions is dependent to some extent upon the genetic make-up of the animals used (Gunberg, 1958; Tuchmann-Duplessis&Mercier-Parot, 1959; Beck et al., 1960). It is, therefore, necessary to use a susceptible pure-bred strain in experiments of this kind.

Other factors have also been shown to influence the teratogenic action of trypan blue. It was recently shown (Beck et al., 1960) that the action of preparations of Grübler’s trypan blue could be modified by recrystallization from hot water so that the resultant product caused significantly fewer foetal resorptions than did the parent compound. Preparations of the mother liquor remaining after the recrystallization were found to be as active as the parent compound in producing foetal resorption, and a preparation of the main coloured component of the mother liquor was found to produce a resorption rate which lay between the crude commercial product and the control group. No conclusion regarding the occurrence of malformations was advanced, since the M.R.C. hooded rats used in this experiment responded mainly by a raised resorption rate and produced very few malformed foetuses.

Furthermore, different commercially sold preparations of trypan blue have been found to produce quantitatively varied responses (Tuchmann-Duplessis & Mercier-Parot, 1959), but there must be some uncertainty as to the actual dose of trypan blue administered in these and similar experiments, since commercial samples of trypan blue contain highly variable amounts of sodium chloride, varying from T5 to 75 per cent.

The present investigation deals with the effects of three commercial preparations of trypan blue—standardized by titration to contain the same amount of azo dye—on a colony of Wistar rats which respond by producing both malformations and resorptions. Two preparations were found to be highly active but the third almost totally inactive. In addition, estimations of the Ld₅₀ of one of the teratogenic and the non-teratogenic preparations were carried out on adult non-pregnant female rats in order to see whether there was any correlation between teratogenic activity and acute toxicity.

The animals were divided into four groups. One group was injected with a preparation of trypan blue sold by G. T. Gurr, another with a preparation offered by Flatters & Garnett, and the third with an alternative preparation of Flatters & Garnett. The fourth was a control group. The dyes were given subcutaneously at days of pregnancy in a dosage of 0·05 mol. of azo linkage/kg.; they were dissolved in normal saline to a concentration of 0·01 mol./ml. Controls received 5 ml. of normal saline/kg. The animals were killed at days of pregnancy, the uteri were removed, resorption sites counted, and surviving foetuses examined for evidence of external malformation.

The concentration of azo grouping in each sample was estimated by volumetric titration against titanous chloride (British Pharmaceutical Codex, 1954) and this was repeated at the conclusion of the experiment to ensure that the sample had not deteriorated.

The acute toxicity (Ld₅₀) was calculated for the Gurr preparation and for batch 2 of the Flatters & Garnett preparation; Kärber’s formula, which is designed for dealing with small series, was used (Kärber, 1931). Sixteen rats divided into 4 dosage groups of 4 rats per group were used in the Gurr series and 20 rats in 5 dosage groups of 4 rats per group in the Flatters & Garnett series. A single dose of 2 per cent, trypan blue in saline was given subcutaneously and death within 48 hours was considered to be a positive reaction. When a single injection was too bulky smaller quantities were injected simultaneously at various sites.

Table 1 is a summary of the results obtained. There is little difference between the response to injections of Gurr’s trypan blue and that to injections of normal saline. No abnormal embryos were found and the slightly raised resorption rate following the injections of Gurr’s trypan blue is not statistically significant (0·20 > P > 0·10).

Following the injection of the first sample of Flatters & Garnett trypan blue there were only 9 survivors from 33 implantation sites. Of these only 3 were normal and 6 were malformed. These values are significantly different from those obtained with the Gurr sample and also from the controls (P < 0·01 in each case for resorptions, as well as for malformations when these are expressed as a proportion of the total survivors).

The second sample obtained from Flatters & Garnett was also very active but less so than the first batch. However, the resorption and malformation rates are still significantly different from those of the Gurr and control groups (P < 0·01 in each case). Out of 67 implantation sites there were 17 normal survivors only, 13 abnormal foetuses, and 37 resorptions.

Toxicity results are shown in the final column of Table 1. It is seen that the Ld₅₀ of the Gurr sample was 475 mgm./kg. while that of batch 2 of Flatters & Garnett was only 360 mgm./kg. The significance of this difference is difficult to evaluate. If the standard error of each individual value is taken to be its square root, then the values of the Ld₅₀ become 475±44 mgm./kg. for the Gurr sample and 360 mgm.±38 mgm./kg. for the Flatters & Garnett sample (i.e. each value ± twice the standard error). The highest likely Ld₅₀ for the Flatters & Garnett preparation is therefore 398 mgm./kg., and the lowest likely for the Gurr product 431 mgm./kg.; consequently there seems to be a real difference between the two figures. However, the calculation of the standard error in this case is open to criticism, and it is therefore impossible to state categorically that the difference in the toxicity of the two preparations was not due to chance.

The Gurr sample of trypan blue when tested for its teratogenicity is found to behave very differently from either of the Flatters & Garnett samples. Provided, therefore, that each sample is, in fact, trypan blue as the makers claim, the conclusion seems inescapable that commercially marketed trypan blue is not a pure dye. Indeed, preliminary chromatographic experiments at present being carried out in this laboratory show that all commercially marketed samples of trypan, blue tested contain more than one chromogenic impurity. Commercial trypan blue is therefore a mixture of substances, and one, or a combination, of the components of the mixture may be teratogenic. It is, therefore, likely that the differing potencies of the commercial samples used in this experiment depended upon the amount of teratogen(s) which each contained, and it is interesting to note that the non-teratogenic product appears to have been less toxic than the teratogenic sample tested. (The Ld₅₀ of the third product could not be estimated since insufficient quantities were available.)

Whatever may be said about the purity of commercial samples of trypan blue, there is no doubt that the pure dye is theoretically a definite chemical entity to which a structural formula has been given (Conn, 1946). A method of preparation has been described (Hartwell & Fieser, 1936) and suggestions made for eliminating impurities which are bound to occur during the process of preparation of the dye. It seems unlikely that these methods of purification are being used in the commercial manufacture of the dye, and devices for preparation of the pure substance are at present under investigation in this laboratory. Whether trypan blue is itself a teratogenic agent must therefore remain an open question until it is tested in a completely pure form and can be distinguished from any possible closely related and similarly coloured contaminants.

1. Trois préparations commerciales de bleu trypan ont été injectées à des rattes Wistar, à 8 j. $12$ de gestation.

2. La quantité de colorant azoïque administrée (exprimée en liaisons azo titrables) a été la même pour chaque série.

3. Deux des préparations avaient des propriétés tératogènes bien marquées, tandis que la troisième était presque totalement inactive.

4. Une solution physiologique normale a été injectée à des rats témoins.

5. La dose létale 50 d’un des colorants tératogènes était plus faible que celle du colorant non tératogène.

I should like to express my thanks to Dr. F. Jacoby for his most helpful advice and criticism during the writing of this paper, as well as to Drs. D. B. Moffat, K. S. Dodgson, and S. L. Stone for their encouragement.