Lasting biological effects of early environmental influences. II. Lasting depression of weight caused by neonatal contamination.

Certain specific-pathogen-free (SPF) mice bred and maintained under semiprotected conditions have an intestinal flora which is qualitatively simpler (although not quantitatively smaller) than that of mice of the same genetic stock produced under ordinary conditions. They are also heavier at weaning time, grow at a faster rate, and reach a greater adult weight than ordinary mice. When SPF mice are contaminated per os shortly after birth with certain bacterial cultures isolated from the intestinal contents of adult ordinary mice, these bacteria multiply extensively throughout the gastrointestinal tract and persist at extremely high levels until weaning time. Such bacterial infections do not affect significantly either weaning weight, growth rate, or maximum adult weight. In contrast, weight depression could be consistently brought about by contaminating newborn SPF mice per os with bacteria-free filtrates of homogenates of intestines from ordinary mice. The weight-depressing agent passed through Millipore discs of 0.45 and 0.22 micro porosity, but was held back at 0.10 micro porosity. The depression of weight caused by either intestine homogenate or filtrates thereof could be detected within a few days after contamination (of 2 day old mice) and persisted throughout the adult life of the contaminated animals. When intestine homogenate of the SPF mice used in this study were introduced per os into newborn SPF mice, they did not affect their growth rate or adult weight. On several occasions, but not consistently, bacteria-free filtrates capable of depressing the weight curve of SPF mice produced alterations in the appearance of tissue cultures of BHK-21 and mouse embryo cells. When tissue cultures so infected were introduced into newborn SPF mice, the weight of these animals was depressed early and lastingly. An agent exhibiting weight-depressing activity has been transferred from mouse to mouse over many passages by contaminating newborn SPF animals per os. Weight depression was achieved with extremely small doses of material (10(-5) ml of intestine homogenate of 10(-4) of mouse embryo culture). Under these conditions, none of the animals showed obvious signs of disease except reduced weight. Only very young SPF mice (preferably less than 3 days old) proved susceptible to the weight-depressing effect of the filtrates of intestine homogenates or of infected tissue cultures prepared therefrom. After oral contamination, it took approximately 1 wk before the intestinal homogenate obtained from contaminated animals exhibited a high level of weight-depressing activity. The growth-depressing effect could be transmitted from one generation to the next by mating SPF mice that had been contaminated shortly after birth and that were consequently smaller than control SPF animals.

Lasting biological effects of early environmental influence. 3. Metabolic responses of mice to neonatal infection with a filterable weight-depressing agent.

A lasting depression of body weight was consistently produced in SPF mice by infecting them orally 2 days after birth with a nonlethal, bacteria-free filtrate, prepared from the intestine of young SPF mice previously infected with an unidentified agent. Neonatal infection caused a decrease of muscle DNA and of muscle and brain protein in the adults. No other effect was detected in the chemical composition of various organs. Incorporation of (14)C-amino acid into the acid precipitable fractions of liver, kidney, muscle, and brain was lower in infected than in control animals. No difference in incorporation was recognized in the thymus and spleen. The free amino acid pool of adults, measured as blood levels of free amino nitrogen, was decreased by neonatal infection. Surprisingly, the food intake of young animals infected neonatally was higher than that of the controls, measured on the basis of body weight. Their fecal excretion of nitrogen was also higher. The comparative responses of infected and control adults to a stressful situation was measured by giving them intravenously the antituberculous vaccine BCG. Under these conditions, the mice infected neonatally excreted some 20% more nitrogen in their urine and 40% more in their feces than did the controls. The mechanisms through which neonatal infection caused a lasting weight depression are discussed in the light of these metabolic findings.

Alterations in the mouse cecum and its flora produced by antibacterial drugs.

Addition of penicillin, Terramycin, or kanamycin to the drinking water of adult mice rapidly induced in them an enlargement of the cecum. In all animals, this occurred within 12 hr after the beginning of drug administration-the effect being most pronounced with penicillin. The cecums remained enlarged and generally continued to increase in size as long as the antibacterial drugs were administered. The increase in wet weight of the cecums was due primarily to an accumulation of water in the lumens during the first 24-48 hr of drug administration. At that time, there were no detectable histological changes in any case, but the bacteriological picture differed from drug to drug. The cecums were free of bacteria in animals receiving penicillin, fusiform-shaped bacteria and bacteroides were present in those receiving Terramycin, and lactobacilli and bacteroides in those receiving kanamycin. After the initial 48 hr, an abundant and complex secondary microflora developed in all treated animals, its composition being characteristic for each type of antibacterial drug. When penicillin was administered for 2 wk, the cecal weights and microbial populations did not return to normal levels for over 14 days after discontinuance of the drug. This recovery period could be shortened to 10 days by giving the mice food contaminated with cecal homogenates prepared from normal animals. A period of 7 or 8 days was required for the cecal weights and microflora to reach normal levels when the administration of penicillin lasted only 24 hr; this period could not be shortened by giving the animals contaminated food. The effects of drugs on the size and bacterial contents of the cecum have been discussed in the light of earlier findings concerning the characteristics of the huge cecums uniformly found in germfree mice. Taken together, these observations support the hypothesis that certain elements of the intestinal microflora-not yet completely identified-play an essential role in maintaining the integrity of the water-transport mechanism in the intestinal epithelium.

Lasting biological effects of early environmental influences. IV. Notes on the physicochemical and immunological characteristics of an enterovirus that depresses the growth of mice.

Physicochemical and immunological techniques have been used in an attempt to characterize a filterable agent, separated from the intestines of mice raised under ordinary conditions of husbandry, which produces a lasting depression of weight in specific pathogen-free (SPF) mice when administered to them orally shortly after birth. Although this agent has not yet been identified, it will be tentatively designated here as enterovirus. The mouse enterovirus can be readily sedimented by ultracentrifugation and by precipitation at pH 4.3; it does not pass through cellophane membranes. Its infective power is completely destroyed by ultraviolet radiation, but is resistant to heating at 56 degrees C, exposure to ether, treatment with trypsin, ribonuclease, and deoxyribonuclease. Dialysis and treatment with ether and nucleases greatly increase the infective activity of the intestinal filtrates containing the enterovirus, a finding which suggests that these procedures eliminate or destroy some inhibitory substance(s). The mouse enterovirus causes hemagglutination of mouse red blood cells. When injected into rabbits, it elicits in them an immune response that renders their serum capable of neutralizing its weight-depressing activity. As measured by inhibition of hemagglutination or complement fixation, the sera of infected mice do not exhibit any significant activity against usual mouse viruses. Centrifugation of the mouse enterovirus in 50%-20% sucrose gradient gave almost complete recovery of the infectivity and of hemagglutinating activity in the same fraction. In contrast, the protein content of the material was distributed through the various fractions. Consequently, this procedure resulted in a marked increase of specific activity.

The anaerobic bacterial flora of the mouse cecum.

Refinements of the anaerobic technique such as the utilization of prereduced media and the removal of traces of oxygen from the gases used by passage through a hot copper oven resulted in quantitative and qualitative improvements in the recovery of anaerobic bacteria from the cecum of SPF mice. The commonest morphologic types of bacteria in the mouse cecum were tapered rods. These characteristically gave "speckled" colonies on agar plates. Those identified were species of the genera Fusobacterium, Eubacterium, and Clostridium. Several of the morphologic types seen with phase microscopy still could not be cultivated. The implications of these findings for the study of the intestinal flora are discussed.

Lasting biological effects of early environmental influences. 8. Effects of neonatal infection, perinatal malnutrition, and crowding on catecholamine metabolism of brain.

The effects of neonatal infection, perinatal malnutrition, and crowding on the metabolism of brain catecholamine were studied in specific pathogen-free mice. Metabolic turnover of catecholamine was determined by measuring the incorporation of precursor tyrosine-(14)C into brain tissue, catabolic activity of norepinephrine-(3)H at various times after intracisternal injection, and tissue levels of dopamine and norepinephrine. The rate of tyrosine incorporation was decreased by neonatal infection but was increased by perinatal malnutrition and crowding. There was no difference in catabolic activity of norepinephrine between infected, crowded, and control groups. In the malnourished group, however, the total radioactivity from norepinephrine was significantly higher than in the control group (1/2) and 2 hr after injection. The brain contents of dopamine and norepinephrine were depressed in the malnourished group. There was no significant difference in catecholamine levels between infected, crowded, and control groups. In the malnourished group, treatment of the mothers with growth hormone prevented almost completely weight loss during lactation, and also resulted in higher fetal weight. Hormone treatment restored to normal the levels of brain catecholamine and the enzymatic activity of brain tyrosine hydroxylase in progeny of malnourished mothers.

Lasting biological effects of early environmental influences. VII. Metabolism of adenosine in mice exposed to early environmental stress.

The metabolism of adenosine 3',5'-monophosphate (cyclic AMP) was studied in specific pathogen-free mice exposed to neonatal infection with mouse enterovirus or to malnutrition during early life. Metabolic activity was determined by measuring the turnover of cyclic AMP-8-(14)C to respiratory (14)CO(2), its incorporation into various organs and plasma, and the binding activity of synaptosome for cyclic AMP. Early malnutrition increased the catabolism of cyclic AMP as measured by expiration in respiratory CO(2). The level of cyclic AMP was lower in plasma and its incorporation into various tissues was decreased in infected and malnourished animals. Metabolic products of cyclic AMP were isolated from plasma by ion exchange chromatography. Cyclic AMP-8-(14)C had completely disappeared 9 hr after injection. Fewer metabolites of cyclic AMP were detected in infected or malnourished groups than in controls and the metabolic reaction from 5'-AMP to adenosine seemed to be slow in these animals. The ability to incorporate cyclic AMP to synaptosome was also impaired in the experimental groups. The concentrations of brain cyclic AMP were lower in infected or malnourished animals than in controls. Depression of accumulation of cyclic AMP probably resulted from excessive activity of phosphodiesterase, rather than from impairment of adenyl cyclase. Intraperitoneal administration of theophylline brought the activity level of phosphodiesterase to normal in infected or malnourished mice; this fact probably accounted for enhanced accumulation of brain cyclic AMP.

Lasting biological effects of early environmental influences. VI. Effects of early environmental stresses on metabolic activity and organ weights.

Specific pathogen-free mice were exposed to three different kinds of environmental stress during early life: (a) by infecting them with a mouse enterovirus on the second day after birth; (b) by placing the mother during pregnancy and lactation on a mildly deficient diet containing wheat gluten supplemented with See PDF for Structure small amounts of lysine and threonine; (c) by combining a (neonatal infection) and b (early malnutrition). All animals survived the three types of stresses, but all exhibited marked depressions of metabolic activity, and of body weights and organ weights. These depressions lasted throughout the experimental period even though all animals were placed under optimum conditions of nutrition and husbandry after weaning, and maintained under these same conditions thereafter. Metabolic activity was determined by measuring the turnover of (14)C-acetate and (14)C-glucose in respiratory CO(2), and their incorporation in total lipids of liver and brain. The utilization of (14)C-acetate was profoundly depressed in all experimental groups with regard to both elimination in respiratory CO(2) and their incorporation in total lipids of liver and brain. In contrast, the utilization of (14)C-glucose was much less affected; its incorporation into lipids was not decreased and its elimination in respiratory CO(2) was depressed only in animals having experienced both neonatal infection and early malnutrition. The extent of weight depression per 100 g of body weight differed according to the organ and the type of stress. Irrespective of the organ, however, depression of weight was largest in animals having experienced both neonatal infection and early malnutrition. And irrespective of the type of stress, the brain exhibited the smallest depression of weight relative to total body weight.

Coliform bacteria in the intestine of mice.

Mice of the NCS and NCS-D colonies, bred at the Rockefeller University, harbored in their intestine an endemic strain of slow lactose-fermenting Escherichia coli 081+/-: :H21 serotype. In addition, NCS mice have recently acquired E. coli 0109+/-:K48:H14. Both strains persisted during the period of observation, whereas they were not encountered in the feces of mice from two other colonies. Other coliform strains encountered were more transient in their occurrence. Since strains of E. coli 081+/-: :H21 and 0109+/-:K48:H14 are extremely uncommon in human beings, it seems probable that they possess specificity for the mouse host.

The gastrointestinal epithelium and its autochthonous bacterial flora.

Colonization of the gastrointestinal tract by bacteria of the normal flora was followed by bacteriological and special histological techniques in mice from several colonies. These histological techniques were designed to preserve the intimate associations that become established between particular strains of microorganisms and the epithelium of the mucosa of certain areas of the gut. The findings were as follows: 1. The various strains of bacteria of the normal flora became established in the different areas of the guts of infant mice according to a definite time sequence. 2. The first types of bacteria that could be cultured from the gut were lactobacilli and Group N streptococci. Within the first day after birth, these bacteria colonized the entire digestive tract and formed layers on the stratified squamous epithelium of the nonsecreting portion of the stomach and of the distal esophagus. 3. The bacterial types that appeared next were coliforms and enterococci. From about the 9th to the 18th day after birth, these bacteria could be cultured in extremely high numbers from the cecum and the colon. Histological sections of those organs taken during the first 2 or 3 days of that interval revealed microcolonies of Gram-positive cocci in pairs and tiny Gram-negative rods embedded in the mucous layer of the epithelium. The microcolonies were well separated from the mixture of digesta and bacteria that occupied the center of the lumen; they may have consisted of the coliforms and enterococci mentioned above; but this possibility remains to be proved. 4. Histological sections also revealed that, at about the 12th day after birth, long, thin Gram-variable rods with tapering ends were present, side by side, with the small Gram-negative rods and Gram-positive cocci in the mucous layer. By the 15th day after birth, the fusiform bacteria formed thick layers in the mucus, and seemed to be the only bacteria remaining in that location. It has not yet been possible to enumerate these tapered rods by culture methods, but as judged by visual appearances in the histological sections, they seemed to outnumber all other bacteria in the cecum and the colon by a factor of as much as 1000. It must be stressed that these bacterial layers are readily disrupted and even washed away by conventional histological techniques; their discovery was largely due to the use of the special histological techniques described in the text. The bacteriological and histological findings described here constitute further evidence for the hypothesis that symbiotic associations exist between microorganisms and animals, and that a very large percentage of the bacteria in the gastrointestinal tract constitutes a true autochthonous flora. The constant occurrence of several distinct associations of bacteria with the special histological structures of the animal host renders obsolete the notion that the intestine constitutes a chemostat in which the bacterial populations are randomly mixed. For a full understanding of the ecology of the normal microflora, it is necessary to think of body surfaces as distinct microenvironments in which virtually pure cultures of a few species of microorganisms interact with their host and the adjacent microbial populations. Experiments based on this hypothesis are admittedly difficult to design, but on the other hand studies based on the assumption that microorganisms exist as mixtures in the gastrointestinal tract will be only of limited value and may often be misleading.