James A. Reyniers

Use of the Germfree Animal Technic in the Study of Experimental Dental Caries I. Basic Observations on Rats Reared Free of All Microorganisms

pESPITE widespread interest in dental caries, and the consequent accumulJation of numerous well-planned experimental investigations in search of the causes of this disease, a clear understanding of its etiology and pathogenesis is definitely lacking. There is evidence, however, of many factors both systemic and local, modifying or contributory to the various process. Study of the interrelationships of these factors has given origin to a few attractive theories, but the validity of current theories still depends on the discovery of final proof of etiology. The bacterial flora of the mouth, and more specifically of the various lesion, is so complex that any potential cariogenic effects of microorganisms are constantly subject to modifying influences arising from the simultaneous action of some or all of the numerous species and varieties commonly present in this localized area. These modifying effects could be either additive or symbiotic, they could be neutralizing, or they could be either inhibitory or antibiotic. Even if the cause of dental caries is to be found in phenomena other than those associated with the metabolic activities of the local bacterial flora, as has been claimed by a few investigators, the objective demonstration of such phenomena is made exceedingly difficult, if not impossible, by the ubiquitous nature of oral microorganisms, some of which have the capacity, in vitro, to bring about a dissolution of tooth substance closely resembling in some respects certain stages in the dental caries process. In view of the foregoing considerations, we have undertaken a series of investigations with animals which are initially free of all bacteria, but to which single or multiple strains of known bacteria can be introduced. Only by this means does it seem possible to demonstrate clearly the bacterial effect on the teeth of animals fed a diet which in normal animals induces abundant tooth decay.

THE PURE-CULTURE CONCEPT AND GNOTOBIOTICS

Lightweight Stainless Steel Systems for Rearing Germfree Animals. B y BENGT E. GUSTA~SSON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Use of Plastics in the Design of Isolator Systems. By P. C. TREXLER. . . . . . . . . . . The Miyakawa Remote-Control Germfree Rearing Unit. B y MASAWMI MIYAKAWA. . Design and Operation of Apparatus for Rearing Germfree Animals. B y JAMES A. .. REYNIERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hcusing of Disease-Free Vertebrates. B y HENRY I,. FOSTER. . . . . . . . . . . .

DESIGN AND OPERATION OF APPARATUS FOR REARING GERMFREE ANIMALS

Rearing germfree animals and using them in experiments is necessarily a problem in biological engineering; without apparatus that will eliminate contamination and a t the same time permit animals to be maintained in it there is little point in attempting such experiments. The basic principles involved in the isolation of animals have been discussed elsewhere (Reyniers, 1943). What has not been discussed in any detail is the basic design, operation, and function of germfree apparatus. It is my purpose here to describe one such system, the Reyniers Germfree System 11, and the results obtainable from its use. Many installations for their production are in operation, and more are presently anticipated. Consequently, it seems proper a t this time to examine the steps involved in the operation of germfree equipment and the results obtained in order that the problems inherent in such work may be better understood. The fact that only one type of germfree apparatus is considered is satisfactory, apart from the limitations of space, for my purpose since the operational procedures are common to all such systems. Until recently there was little point in attempting a critical appraisal of germfree operations, simply because the techniques for rearing such animals were not sufficiently advanced to be reduced to routines, and the necessary degree of certainty relative to contamination had not been attained. As long as the germfree animal remained a laboratory curiosity and the apparatus for working with it was not standardized, there was little purpose in such an analysis. The production of a few germfree animals was merely experimental, and the costs involved were relatively less important than if germfree animals had been available in numbers and widely used. Highly skilled research personnel is no longer needed to operate germfree apparatus. Some species of animals can be reared germfree in numbers and maintained as colonies by breeding them through generations. Furthermore, this can now be done in a centralized establishment in much the same manner as in the commercial breeding of conventional animals. Shipment of germlree animals by public transportation to other centers is equally feasible (Reyniers and Sacksteder, 1958; Reyniers, 1957). Colonies of germfree animals can be tended by technicians, and the apparatus operated for experiments requires little more effort than does an autoclave. It is time, therefore, to discuss the operation of such equipment in terms of manpower, capacities, maintenance, and output of animals.

THE USE OF GERMFREE ANIMALS AND TECHNIQUES IN THE SEARCH FOR UNKNOWN ETIOLOGICAL AGENTS

The subject assigned to us for discussion must be treated, for the most part, theoretically. While the use of germfree techniques is rapidly expanding, the field is relatively new, and very little has been done with either the germfree animal or techniques in the search for unknown etiological agents. However, this is a matter of time, so that it is opportune now to discuss the problem. The thesis upon which germfree life studies rest is that of extending the pureculture concept to the animal by the use of methods that eliminate contamination (Reyniers, 1953; Reyniers and Trexler, 1953). The value of this concept to the development of bacteriology and infectious diseases is without serious question. I t would seem reasonable, therefore, that this concept should be extended to situations where an infectious agent is suspected, but unknown, by eliminating contaminants from the animal and the environment in which it lives. If these conditions can be achieved, then the animal is locked, in effect, into an isolated environment. Under these circumstances attention may be centered on the animal and not diverted by suspicion of contamination from the external environment or from microbes ordinarily contaminating a conventional animal. Theoretically, a germfree animal is one free from all living contaminants. Practically, the examination involved requires a function of available techniques for detecting contamination. Such techniques will encompass those of bacteriology, virology, parasitology, immunology, and pathology. If the levels of inquiry cannot be satisfied by such methods, then new methods must be devised. For example, if a disease occurs in a germfree animal, it may be reasonable to suspect a living contaminant even though conventional techniques do not reveal it. Further, if no disease is observed in the germfree animal during its life or in subsequent generations it may be necessary to cause a disease or, better, to bring stress to bear on the animal to bring out a contaminant. This would he the case where no spontaneous tumors occur but can be brought about experimentally through the use of a carcinogen. In the present state of knowledge, it is reasonable to assume, on a basis of experience, that microscopically visible microorganisms (for example, bacteria, yeasts, fungi, protozoa, endoparasites, and ectoparasites) can be eliminated from a mammal by cesarotomy or from oviparous animals by sterilizing the outer shell of the egg and hatching into a sterile environment. The techniques presently available are adequate for the purpose. Moreover, it should be remarked that the germfree animal is itself an excellent detection device in which contaminants can readily grow, probably because the animal is r*ithout previous microbial experience, and there is no antagonistic effect of one kind * The studies reported in this paper were aided hy Grant No. 48 from The Damon Runyon Memorial Fund for Cancer Kesearch, Inc., New York, N. Y., and Contract N.K. 131 067 from the Office of Naval Research, Washington, D. C.

TUMORIGENISIS AND THE GERMFREE CHICKEN: A PRELIMINARY REPORT*

This communication considers briefly some of the possibilities and problems inherent in such an approach. Obviously, this depends upon how a contaminant is defined and upon the techniques available for detecting it. The fact that the definition has these practical limits does not change it. A germfree animal must, by definition, be free from living contaminants; if a contaminant can be detected, the animal is not germfree. It may be considered a gnotobiofe (gnos-known; bi~s-life)~ since it is associated with a known contaminant. As such, its value may be of equal or greater importance than if it were germfree. Absolute freedom from contamination for an animal may be a theoretical goal, but nevertheless it is an important one to investigate. At present chickens can be obtained free from bacteria, yeasts, fungi, protozoa, endoparasites, ectoparasites, and many disease-causing v i r u s e ~ . ~ They can be reared germfree to maturity and a t least into one generat i~n.~ Whether germfree chickens can also be obtained free from latent or cytopathogenic viruses is not yet known; first, because the subject has not been investigated in germfree animals and, second, because such tests are difficult to make on a living animal and are presently not available for inclusion in routine testing programs. However, it is the investigation of these not-too-well-defined areas that offer so much to cancer research. In the final analysis, what constitutes a living being is a matter of definition. At the moment it does not matter whether cell particulates are or are not called viruses as long as they can be isolated, examined, and defined; neither does it matter whether they are considered as contaminants. These are matters that can be decided a t a later date when more information is available. In considering germfree animals for cancer research we were motivated by two main objectives : first, the possibility that subcellular particles might be related to tumorigenesis and, second, the desire to study immunity to tumor formation. The germfree animal, being free from contamination, constitutes a “pure” animaL6 Because it has never had experience with microbes, it may be considered as LLimmunologically immature” in the sense that certain defensive systems with which it is endowed by nature to protect it from microbial assault are not activated at birth. On this basis it is possible to apply the pureculture concept to the search for etiological entities, secure in the knowledge that no contaminants will be added from the environment and that the animal

Fractionation studies of tumor tissues from germfree chickens.

The reduction in “take” time with the repeated passage of homogenates of the original chemically induced tumor in germfree birds, described elsewhere by Keyniersl suggests that, during the course of the repeated passages, an increase in the pathogenicity or in the quantity of an infectious agent in the tumors occurs. This thinking is, of course, based upon the concept that such neoplastic growths and viral infections are caused by the interaction of the host and an independent infectious agent. For the present let us ignore the possible source and the initiation of the process and examine the tumor itself. The first question that arises seems to be: “DO these tumors contain what we consider to be virus or viruslike particles?’’ We undertook ultracentrifugal fractionation and electron microscopic studies of the fractions for such orientation purposes, and we present some results of these studies correlated, as far as possible, with the biological work described elsewhere by Reyniers.’ It must be pointed out that no strictly quantitative comparative analyses have been carried out; instead, our object has been to obtain some fundamental information regarding the submicroscopic morphologic characteristics of these tumors. For the initial experiments, frozen tumor tissue was finely minced, frozen and thawed 3 times, and then homogenized in a Waring Blendor to yield a 25 to 30 per cent suspension of tissue in phosphate-buffered salinc solution. The homogenate was then diluted to a 7.5 to 10 per cent suspension, after which it was angle-centrifuged at 4000 g for 15 min. to obtain an opalescent extract of the tumor tissue for further fractionation. No filtration or other preliminary clarification was used in these early experiments. Later, after learning something of the gross morphologic composition of the crudc extracts, the 30 per cent blender homogenate was ground in a mortar with carborundum and then filtered through nylon with No. 503 Celite as a 10 per cent suspension. Still later a modification of the fluorocarbon emulsification technique first described by Gessler et d2 was used to eliminate lipid and extraneous protein material from the extracts prior to the centrifugal fractionation. An ultracentrifugation sequence was used (20,000 rpm for 45 min. and 40,000 rpm for 90 min. in a Spinco No. 40 rotor) similar to the cycle used in earlier work to separate equine encephalomyelitis virus3 from a normal or microsomal component4 of chick embryos. It was found that there was a progressive increase in the total quantity of the particulate (microsomal) fraction sedimented a t 40,000 rpm from the crude saline extracts of the tumors obtained from successive passages. All preparations for electron-microscopic examination were prepared by allowing a small drop of thirtyfold concentrated pellet solution (from the original 7.5 or 10 per cent tissue extract) to filter through a collodion