Open Book Publishers logo Open Access logo
  • button
  • button
  • button
GO TO...
book cover

17. Penicillin And Laboratory Animals: The Animal Rights Myth

The dramatic impact of penicillin on the treatment of patients with severe infections has rightfully been paraded by biomedical researchers as a vindication of the value of laboratory animals in research. Opponents of animal experimentation are equally emphatic that animal experiments not only played no part in the development of penicillin but also that reliance on such techniques might well have caused penicillin to be discarded (1,2). The basis of the animal rights argument is that Fleming did not use animals but the “humble culture dish” (1) and that penicillin would never have been used in patients if physicians had known of its supposed inordinate and unique toxicity to guinea pigs (2). Both of these claims are incorrect and based on a superficial assessment of the literature. Fleming, after his chance observation that contamination with spores of a strain of Penicillium inhibited the growth of a culture of Staphylococcus aureus, tested the toxicity of the mould culture on animals. He reported (3):

the toxicity to animals of powerfully antibacterial mould broth filtrate appears to be very low. Twenty c.c. injected intravenously into a rabbit were not more toxic than the same quantity of broth. Half a c.c. injected intraperitoneally to a mouse weighing about 20gm. induced no toxic symptoms.

Rather than animal experiments making no contribution to the recognition of the remarkable potential of penicillin, it has been argued that the development of this drug was in fact initially retarded by the omission of a simple and well-established animal test. The mouse protection test, whereby mice are injected intraperitoneally with pathogenic bacteria, and potential antibacterial substances administered to assess their ability to prevent death, had been described in 1911. From 1927 onwards Domagk and his team at Elberfeld used this in vivo technique routinely to assess the antibacterial efficacy of the azo dyes, this work culminating in the introduction of prontosil (which was inactive in vitro) and hence the sulphonamides (4). Had Fleming used this test during the studies that were the basis of his 1929 paper (3) he would have found, as did Florey and his colleagues at Oxford ten years later, that his crude broth contained sufficient active substance to protect mice artificially infected with susceptible organisms.

Penicillin and the Guinea Pig

The fact that penicillin, whilst being relatively innocuous for most species, appeared to be toxic to guinea pigs and hamsters, has served as a sheet anchor for the claims of antivivisectionists that species differences render animal experiments redundant and possibly dangerous. However, the apparent sensitivity of these species to penicillin, and indeed other antibiotics, is certainly not evidence of the futility of animal experimentation It is in fact a good example of the usefulness of the appropriate model in biomedical research.

The story began with the comprehensive and careful piece of work by Dorothy Hamre and her colleagues (5). These workers were interested in the effect of penicillin on gas gangrene. They therefore administered penicillin to guinea pigs that had been infected with Clostridium perfringens. The animals died 12-72 hours after receiving daily doses of penicillin for 3-4 days, yet showed no signs of gas gangrene.

Hamre therefore investigated the actual toxicity of penicillin to guinea pigs and found this species, unlike mice and rabbits, to be susceptible to repeated daily doses of the antibiotic. The samples of penicillin available at this time were very impure. It is clear from Hamre’s results with purer preparations that some of penicillin’s immediate and delayed effects were due to impurities. Nevertheless, daily doses 7-12 times those used in patients could result in lethargy, weight loss and often (but not invariably) death after 3-7 days. However, a dose approximate to that used clinically administered even for twenty days did not kill the guinea pigs.

Hamre thus stated:

The fact that present preparations are toxic for guinea pigs when given subcutaneously does not mean that penicillin is toxic for man. When treated with the same dose of penicillin per kg. as that given to man, guinea pigs did not die and, in fact, failed to show any signs of toxicity. However, it is suggested that chronic toxicity for man be borne in mind [my emphasis].

Hamre’s paper was carefully written and, as subsequent research showed, to some extent prophetic.

There were occasional attempts to explain the apparent toxic effects of penicillin over the following decade, but no progress was made until a group of Belgian workers suggested that a change in the composition of the bacteria in the gut was responsible. De Somer and his colleagues showed that the Gram-positive microorganisms that inhabit guinea pig intestine are completely removed by the penicillin and are replaced by insensitive bacteria. Injection into normal guinea pigs of a sterile filtrate of a culture of these bacteria produced a condition identical to penicillin poisoning (6). It was thus supposed these new bacteria colonising the gut produced toxins that are absorbed by the guinea pigs causing the illness and perhaps death of the animals. The theory that the guinea pigs thus die from an enterocolitis, rather than from a direct toxic effect of the penicillin was supported by the observation that penicillin is not toxic to germ-free guinea pigs (7). Such animals were unaffected by penicillin 240 mg/kg/day but conventionally reared guinea pigs given this dose died within 7 days of severe infection of the colon (Fig. 17.1).

Fig. 17.1 Effect of penicillin in normal and germ-free guinea pigs. Figures are number of animals per group. Data from S. B. Formal, G. D. Abrams, H. Schneider, and R. Laundy (1963), ‘Penicillin in germ-free guinea pigs.’ Nature, 198:712.

The fact that penicillin itself is therefore not toxic is unlikely to impress the implacable opponent of animal experimentation, who requires only a simple quote for argument. However, this reaction, noted in the guinea pig in 1943, was in fact a prediction of what became, within ten years, a common iatrogenic condition in patients.

Antibiotic Associated Colitis

Pseudomembranous enterocolitis was first described in 1893 (8). It was commonly observed in post-operative patients. The characteristic of the condition is the appearance of a pseudomembrane over the gut mucosa. Microscopically the membrane can be seen to be composed of mucin, fibrin, white blood cells and sloughed mucosal epithelial cells. Occasionally the complete membrane is sloughed off and passed out as a “cast of the colon” (8).

Although most early cases of pseudomembranous colitis occurred post-operatively, antibiotics also became implicated as causative factors soon after their introduction in the 1940s (9). The location of the lesion following antibiotic use was invariably the colon rather than the small intestine, thus this particular condition was named “antimicrobial-induced pseudomembranous colitis.”

The cause of the condition was the removal of sensitive bacteria from the gut by the antibiotic. This then results in profound alterations in the composition of the gut microflora due to recolonisation of the gut by other micro-organisms. This phenomenon was noted to follow administration of penicillin given by mouth or by injection. Generally the change in intestinal flora produced by penicillin (and other antibiotics) was of no significance, since the normal intestinal flora became re-established after therapy was stopped. In some patients however, after prolonged administration of the antibiotic, a superinfection occurred in the gut resulting in life-threatening pseudomembranous colitis (10).

Guinea Pigs and Humans: Remarkably Similar

The particular micro-organism recolonising the gut and causing pseudomembranous colitis eluded identification for many years. This, sometimes fatal, condition was originally attributed to mucosal ischaemia or a viral infection. Ultimately stools from affected patients were shown to contain a toxin that damaged cultured cells (11). Shortly after, this toxin was shown to be derived from Clostridium difficile (12, 13). It thus appeared that C. difficile was the pathogen that caused pseudomembranous colitis. This was not easy for everyone to accept. C. difficile had been described in 1935 (14). It was difficult to isolate and slow to grow in culture (thus it was designated the “difficult clostridium”). As it was found in the stools of quite healthy infants C. difficile was considered to be innocuous and thus of little interest. When it became clear that infants are resistant to the effects of C. difficile toxin only until about 12 months of age (possibly due to the absence of toxin receptors on the immature cells of their gut (11) it became generally accepted that C. difficile was a pathogen and indeed the cause of pseudomembranous colitis.

Returning to the animal studies, veterinary research has strengthened the apparent similarity between guinea pigs and humans. Penicillin-induced inflammation of the large intestine of guinea pigs has been shown to be caused by clostridial toxins, and C. difficile has been isolated from the gut contents of these animals (16). Newborn guinea pigs, like germ-free animals, have been shown not to be susceptible to penicillin (17). Perhaps neonatal guinea pigs, like babies under 12 months of age, are insensitive to clostridial toxins.

It is clear that the now common, clinical problem of antibiotic-induced colitis (15) is analogous to the condition observed in guinea pigs by Hamre and her colleagues 50 years ago. An intensive investigation of the effects of penicillin on the intestinal flora of the guinea pig might have forewarned clinicians of the life-threatening superinfections that can occur with prolonged use of some antimicrobials. However it would be uncharitable to suggest the early experimental workers in the field were remiss. On the contrary, the admonition by Hamre that “chronic toxicity (of penicillin) for man be borne in mind” was an accurate and perceptive observation.

There is no basis for citing the effect of penicillin on guinea pigs as a prime example of species difference; in fact, it is an example of the exact opposite, namely of strikingly similar effects on the guinea pig and on people. The misrepresentation of the history of penicillin is but one example of the distortion of scientific fact perpetrated by opponents of animal research. It is the responsibility of scientists to correct these errors before they gain unwarranted credibility with the general public.

An earlier version of this chapter was published as: Burying the penicillin myth. RDS News July 1995 6-7.


  1. Sharpe R (1988) The Cruel Deception: The Use of Animals in Medical Research. London: Thorsons.
  2. Ruesch, H (1982) Naked Empress. Klosters: Civis Publications.
  3. Fleming, A (1929) On the antibacterial action of cultures of a Penicillium, with special reference to their use in the isolation of B. influenzae. Br. J. Exp. Pathol, 10 226-36.
  4. Florey, H (1953) The advance of chemotherapy by animal experiment. Conquest 41 4-14.
  5. Hamre, D M, Rake, G, McKee, C M and MacPhillamy, H B (1943) The toxicity of penicillin as prepared for clinical use. Am. J. Med. Sci. 206 642-52.
  6. De Somer, P, Van De Voorde, H, Eyssen, H and Van Dijck, P (1955) A study on penicillin toxicity in guinea pigs. Ant. Chemother 5 463-69.
  7. Formal, S B, Abrams, G D Schneider, H and Laundy, R (1963) Penicillin in germ-free guinea pigs. Nature 198 712.
  8. Finney, J M T (1893) Gastroenterostomy for cicatrizing ulcer of the pylorus. Bull. Johns. Hopkins Hosp. 4 53.
  9. Bartlett, J G and Gorbach, S L (1977) Pseudomembranous enterocolitis (antibiotic-related colitis). Adv. Int. Med. 22 455-76.
  10. Mandell, G L and Sande, M A (1990) Penicillins, cephalosporins and other beta-lactam antibiotics, in Gilman, A G, Rall, T W, Nies, A S. and Taylor, P. eds. Goodman & Gilman’s The Pharmacological Basis of Therapeutics. 8th edition. New York: Pergamon Press, pp. 1065-97.
  11. Kelly, C P, Pothoulakis, C. and LaMont, J T (1994) Clostridium difficile colitis. New Eng. J. Med. 330 257-62.
  12. Bartlett, J G, Chang, T W, Gurwith, M, Gorbach, S L and Onderdonk, A B (1978) Antibiotic-associated pseudomembranous colitis due to toxin-producing clostridia. New. Eng. J. Med. 298 531-34.
  13. Larson, H E, Price, A B, Honour, P and Borriello, S P (1978) Clostridium difficile and the aetiology of pseudomembranous colitis. The Lancet 1 1063-66.
  14. Hall, I C and O’Toole, E (1935) Intestinal flora in new-born infants with a description of a new pathogenic anaerobe. Bacillus difficilis. Am. J. Dis. Child. 49 390-402.
  15. Anon. (1995) Antibiotic-induced diarrhoea. DTB 33 23-24.
  16. Lowe, B R, Fox, J G and Bartlett, J G (1980) Clostridium difficile-associated cecitis in guinea pigs exposed to penicillin. Am. J. Vet. Res. 41 1277-1279.
  17. Manning, P J Wagner, J F and Harkness, J E (1984) Biology and diseases of guinea pigs, in Fox, J, Cohen, B and Loew, F eds. Laboratory Animal Medicine. New York: Academic Press, p. 173.