6) E. coli in the Human Gut.

(a) Acquisition of E. coli by neonates.

As was pointed out in the introduction, E. coli are present in the intestinal tract of all humans and most other warm blooded animals, generally in numbers around 108 to 109 per g of large intestinal contents. They comprise at maximum less than 10% of the bacterial flora of this ecosystem, which consists predominantly of anaerobic bacteria, which are bacteria which can only live in the absence of oxygen. E. coli being a facultative anaerobe can live both aerobically, in the presence of oxygen and anaerobically.

In order to colonise the intestinal tract, the E. coli has to overcome both the host defense mechanisms including such factors as the stomache acidity and also competition with other microorganisms including the E. coli already resident in the intestinal tract. None of these mechanisms apply to the neonate and so discussion on colonisation will begin with the new born baby. It should be realized that in the womb, the baby is in a bacteriologically sterile environment and thus at birth it is bacteriologically sterile. However, within a few days a baby is excreting bacteria in its faeces in numbers similar to adults and so obviously within the first few days of life it is being colonised by many bacteria including E. coli.

Early studies about half a century ago on the acquisition of E. coli by neonates were predominantly involved with trying to establish how the pathogenic E. coli are acquired. Although the pathogenic E. coli probably have mechanisms unique to them for colonisation, some lessons can nevertheless be learned from these studies. These clearly showed that these E. coli were derived from the mothers' genital tracts, which would in turn have acquired their E. coli from the intestinal tracts.

In a most extensive study on the acquisition of nonpathogenic E. coli, which was carried out in the 1974's, over 2500 strains of E. coli were collected from the faeces of 33 mothers, the faeces of their babies and the mucus extracted from the babies' mouths after delivery. Of these 33 babies, 28 babies were found to have E. coli in their faeces within the first few days of life before they left the maternity ward. Further it was found that 22 of the babies carried strains which were similar to those isolated from the mother. Unfortunately in not all cases could the mucus be obtained from the mouths of the babies, but, where it could be and E. coli cultured from it, then it tended to have an intermediate flora between the maternal and baby's faecal flora.

In a number of cases it was noted that the variety of types isolated from some but not all mother/baby pairs decreased from maternal faeces through baby's mucus to baby's faeces. On the basis of further investigations, it was confirmed that in at least in some cases the normal commensal E. coli flora found in the mother's faeces will be the source of at least some of her baby's E.coli and that transmission would most likely occur during the birth process. In some cases it was noted that one or two of the characteristics of the E. coli were either lost or gained during these transmissions, suggesting that there must be some environmental pressure on these organisms.

Under conditions, where the babies were virtually exclusively handled by their mothers, the babies were predominantly colonised by the E. coli from their mothers. It can thus be clearly stated that in normally delivered infants, where the mothers predominantly look after their own infants, the mothers' faecal E. coli will generally provide the major source of the commensal strains for her baby. However, this is not always the case and despite these extensive studies, no clear reason could be found why some babies did not acquire their mothers' faecal E. coli. The only general rule, which could be seen was that babies were more likely to be colonised by their mothers' faecal E. coli the longer the period of labour. Also in maternity wards, where there may be the introduction of competing E. coli from handlers dealing with other babies, then these factors may cause E. coli from other babies to be acquired through cross contamination.

In order to determine the role of non-maternal E. coli in the colonisation of human babies, a study on babies born by Caesarian section was instituted. Thereby, it was possible to exclude the effect on the colonisation of the neonate through the normal birth process. These neonates, though not in contact with their mother's faecal flora, nevertheless, were colonised by E. coli at the same time and to the same extent as the normally babies. Of the various routes by which it was shown that these babies were likely to have acquired their E. coli the most likely appeared to be via a feeding bottle from the nurse's hands; from the maternal faeces during an extended labour prior to the Caesarian section; from the maternal faeces via the mother's hand, perhaps contaminating the breast; from a gastric lavage tube via the nurse's hands; during oxygen administration and on three occasions via the nurses' hands. It should be noted that it was the practice in this ward at that time for the nurses to attend to most of the babies' needs. This would enhance the possibility of cross-infection.

In a maternity ward situation where there was the practice of separating the infants from their mothers for 8 to 72 hours post-natally and caring for them by nursery staff there was an increase in the transmission of E. coli among the neonates. In fact in those conditions, the spread was much more significant than the acquisition from the mother. But in those cases where there was direct transmission from mothers to their babies, there was also greater contact between them.

Breast-fed babies were also found to have a narrower range of different E. coli types than in bottle-fed bed babies. Human milk has been shown to be inhibitory to strains of E. coli but the degree of inhibition varied from strain to strain. The faecal E. coli from breast-fed infants were found to be more sensitive to the bacteriocidal effects of human serum and more often spontaneously agglutinating. These studies suggested that breast milk favours the proliferation of mutant strains of E. coli, which have a reduced virulence.

It has been shown that the strains initially colonising neonates are able to persist for up to 18 months and it has been suggested that such colonisation should not be left to chance. It has even been suggested that babies should be deliberately colonised in this early state by known non-pathogenic E. coli which would thus competitively inhibit colonisation by the pathogenic types. Some studies done in these aspects have shown that these artificially acquired E. coli will also persist in the babies faecal flora.

In one study 14 newborn babies in an intensive unit and who were not receiving antibiotics, were given 1 ml of culture broth containing 108 to 109 bacteria, of a strain of E. coli marked to be resistant to sodium azide. In three of the babies the strain became the only E. coli to be excreted in the faeces for a week or more. In four others it was part of the enterobacterial flora for a week, while in six it was excreted for less than a week and from one baby it could never be cultivated. These results show the great variability which must exist in the interaction between E. coli and neonates.

An investigation was carried out into the spread of E. coli in child day-care centres. In general a contamination rate with faecal E.coli of between 11% and 21% was noted for most of the animate as well as inanimate items. However, the fist size toy balls, which were supplied bacteriologically sterile reached a contamination rate of 46%. By comparing two types of diaper use, it was demonstrated unequivocally that use of overclothing over the diaper, significantly reduced the faecal E. coli contamination of the environment. Use of paper diapers further significantly reduced this contamination compared to use of cloth ones. By comparing the antibiotic resistance patterns of the E. coli which were isolated from the environment, it was found that all characteristics unique to the centre from which they were isolated. This showed that the children tend to act as foci for the spread of E. coli and that the faecal-oral route of transmission must be considered the most likely.

(b) Adults also acquire E. coli.

Although adults will have an established microbial flora in their intestinal tract, including E. coli, they will still be colonised by new types. Adults are generally considered to have a resident and transient E. coli flora. Thus certain types, the resident types, will be found regularly, whenever sampling is performed but there will also be found a variety of types, the transient types, which may be isolated on a few successive days and then never again. Occasionally for reasons not fully understood a transient type can become a resident type, replacing the previous resident type.

This is an extremely simplified picture, because the actual situation in the intestinal tract of an adult is far more complex, even with respect to only the one species of E. coli. In a study of the complete stools from different healthy adults, ten samples were taken along the length of the stool about 2 cm apart. From each sample at least ten E. coli were characterised and in a number of samples up to 100 E. coli were. Thus, from each individual stool over 200 E. coli were characterised. In general the main E. coli of the stool was present in most of the samples, but other types were apparently randomly scattered around. In addition certain rare types such as antibiotic resistant strains were only found when specifically saught, they were not among the randomly selected few hundren strains obtained from the samples.

This study confirmed that there is a resident or predominant E. coli flora, but there are a large variety of types which are also present, in varying numbers. When one considers that there may be at least 100,000,000 E. coli in a faeces and all that could reasonably be sampled and characterised were a few hundred, or at most one in a million it should be realised what a complex flora is present in the human intestinal tract. It should further be noted that while this study was performed over 20 years ago and has not been repeated, indicating also how daunting such an investigation is.

A group of eight women (aged 42-56 years) and the 61 year old husband of one was observed for periods ranging from two to five months and a small number of their faecal E. coli regularly collected and characterised. They all lived in a suburb of North London and predominantly ate at home, four went on holiday during the time of study and one also ate at a canteen.

Some of these women only yielded very few E. coli types. From one subject, throughout the three months period of study, only one E. coli type was isolated. Three other subjects only yielded 3, 4 or 7 different types each. From one of these subjects seven of the 12 specimens examined contained only one unique E. coli type and the only other types isolated from this person were closely related types, which may have been derived from this resident type.

A great variety of types was noted in other subjects, but all showed that one type was probably the persistant type because it was isolated repeatedly over a period of months. The person, who was found to carry the greatest variety of types also uniquely ate in canteens. The husband and wife carried the same type, but the variety of types carried by the wife was only a third of those carried by the husband.

The preliminary conclusions, which can be drawn from these observations are that there is a persistant E. coli flora in healthy adults, but it can be modified, probably through food intake. A special study was designed to investigate this further. It involved six nurses, who were working in a central London hospital and was carried out over an eight-week period. During the first and last two weeks, the nurses ate a normal diet. In the middle four weeks, the nurses were exclusively on a diet of tinned food and ultraheat-treated milk with vitamin supplements. To ensure that the introduction of E. coli was as reduced to a minimum, all cutlery was sterilised by heating and only disposable plates and cups were used. Even the tooth brushes were soaked in hydrogen peroxyde and sterilised water was used for cleaning the teeth. Specimens from all faeces passed by the six nurses were cultured for E. coli and 10 colonies representing as many different colonial variants as possible were examined. The mean number of types found per specimen in each nurse before, during and after the sterile diet was calculated. This revealed a clear trend from an initial rich variety, which dropped to low variety during the period of the sterile diet and rose to a greater variety then in the first phase after the sterile diet. Despite the precautions to avoid the nurses acquiring new E. coli during the sterile diet phase, some new serotypes were observed for the first time during this period. These may have been present before, but missed due to the difficulty of sampling. They may also have been acquired while the nurse may have on occasion not adhered strictly to the sterile diet. Also within the ecological pressures in the intestines to which the E. coli flora is subjected there may have been some variation among the types causing a new type to emerge. However, this study clearly showed that food is probably the major source of the intestinal E. coli of a healthy adult person.

This was was also confirmed in a different study, involving no intervention at all by the experimenters who only acted as careful observers. This time a group of students travelling from many parts of the world to Dublin, Ireland were involved. Faecal specimens had been collected prior to their departure and then weekly for the first four months of their stay in Dublin. The most noteworthy observation was that very little overlap of E. coli types from one specimen to the next could be observed. There was only a suggestion of a persistant flora, although representatives of it could not always be isolated from a specimen, but may then be found weeks later. Also a number of transient types made single appearances or occasionally multiple appearances.

Similar continuous changes in types of E. coli carried were also shown in a group of peace-corps volunteers who travelled to Morocco. In this study the emphasis was on determining whether the volunteers developed travellers' diarrhoea. Just like the students arriving in Dublin, the flora changed after the volunteers arrived and continued to change irrespective of whether they developed diarrhoea or not.

(c) Microbiological Factors involved in acquiring E. coli.

For some time it has been known that a number of bacteria produce substances, which have a lethal effect on different types of the same or related species. These substances are given the general name of bacteriocines, i.e. sunstances which kill bacteria. Colicines are a specific subset of these bacteriocins, being substances produced by some strains of E. coli which will kill other strains of E. coli.

Following the discovery of colicines, it has been considered likely, that they would play a role in the ecology of the intestine giving the colicinogenic strain a selective advantage over sensitive strains. In an experiment in which human beings consumed mixtures of pairs of cultures of an E. coli strain, differing only in that one was a colicine producer while the other was not. It was found that only some colicines promoted survival in the intestinal tract, while others had no observable effect.

The intestinal tract is in constant motion, predominantly due to peristalsis causing the food particles to be digested and the main nutrients to be obtained from the food. Thus it would be of considerable advantage to an organism in that tract to be able to adhere to the intestinal wall rather than be in continuous movement downwards. Thus, in order to colonise a human, E. coli have to be able adhere to some extent to the surface cells of the intestinal tract. An important early finding was that adherence of E. coli to human mucosal cells is mediated by receptors specific for the carbohydrate mannose, which is commonly found on the surface of mucosal cells.

Some strains of E. coli can carry S-fimbriae, which are specific for sialic acid. This is a component of many glycoproteins, present on mucosal cell surfaces. They also bind to sialic acid-containing structures on human buccal epithelial cells. Age differences were found among the abilities of human buccal epithelial cells to bind S-fimbriate E. coli. The mean number of bacteria bound/cell rose slightly with the age of the individuals. Such adherence may reflect the ability of the organism carrying these fimbriae to displace a resident strain, which does not.

A major problem resulting from the indiscrimate use of antibiotics, has been the emergence of antibiotic resistance among both human and animal pathogens. It has been known for many years that humans, who are not currently taking or who have not recently taken antibiotics, will be carrying in their intestines strains of E. coli which are resistant to one or a number of antibiotics. These would generally be present in comparatively low numbers and only able to be identified if special antibiotic containing media are used, which will let them grow, while the majority of sensitive E. coli will not.

In many cases these resistances will be on separate pieces of DNA unlinked to the main E. coli genome. These plasmids, as such independent DNA pieces are known, can be passed from one E. coli to another. Thus, if antibiotic is taken by the host for whatever reason then the majority E. coli flora can rapidly become resistant. This resistance may not just be restricted to the antibiotic being taken but to a whole battery of antibiotics. Such antibiotic resistant E. coli are also no different from other E. coli and can be transmitted from person to person, passed on to babies in exactly the same way that non-resistant strains are passed around. A further feature of these plasmids is that they do not just harbour genes conferring antibiotic resistance for the strain but also genes coding for some virulence factors or toxins. Thus, these pathogenic E. coli would be selected.

A strain of E. coli, which has been in artificial culture for many years has been shown that it can certainly survive in the human intestine and transfer hybrid plasmids to the resident E. coli flora. A number of studies have also demonstrated that E. coli carrying antibiotic resistance conferring plasmids will be found in the faeces of patients in a hospital environment and that these plasmids as well as their E. coli can be transmitted from one patient to another.

In general for plasmids to be maintained in E. coli in the natural environment they must confer some selective advantage to their host bacterium. However, when the carriage of plamids conferring resistance to the penicillin group of antibiotics were studied among hospital isolates of E. coli, they were found to comprise a great diversity of different types, conferring resistance in different ways. This further suggested that there must be a wide distribution of such strains in the community.

A study was carried out on the carriage of E. coli resistant to the antibiotic trimethoprim by children in day care centres and their family members. This revealed that such children are frequently colonised by such resistant E. coli and that about a quarter of their family members were similarly colonised. Mothers were the most likely to be carrying such trimethoprim resistant E. coli, followed by siblings and fathers were least likely.

7) E. coli in Domestic Animals.

(a) Pigs

During the first day of life of piglets, their intestinal tract is flooded with large numbers of E. coli as well as other bacteria. At that time the pH of the stomache is high, i.e. it is not producing the acid which is produced in later life. When the pH drops shortly thereafter, and the stomache becomes more acid, the numbers of all bacteria except lactobacilli decrease.

The significance of pili for the colonization by E. coli of the porcine intestinal tract was demonstrated in a series of specific colonization experiments, in which the colonisation of piliated and non-piliated forms by piglets was investigated. The results obtained from these studies confirmed the hypothesis that the pili facilitate intestinal adhesion and colonization by E. coli. It was shown that the receptor for the E. coli K88 antigen was genetically determined in the pig population and studies since then have shown that the different variants of the K88 antigen have significantly different receptor-binding properties.

The faecal E. coli of two pigs throughout their 210 day life were examined and these studies indicated that even in the absence of antibiotic selection pressure some types of E. coli persist for long periods, while others appear and disappear. The introduction of antibiotics causes certain resistant types to be selected but only some of these will persist. Extensive studies on the ecology of E. coli of weaned and unweaned pigs demonstrated generally a highly complex flora with many different types being present. Thus, it should be considered that there are strong similarities between the normal E. coli flora of humans and pigs.

(b) Cattle.

Again it was found that the alimentary tract of calves is flooded at birth with large numbers of bacteria. But this occurred to a lesser extent than in piglets. It has been suggested that this may be due to calves being less intimately contaminated with faecal matter than piglets. Bovine colostrum, the first milk produced by the cow, has been shown to be bactericidal to E. coli. This can be seen as a defense mechanism against infection and needs to be considered when discussing colonization of the calf.

In a particular study, it was possible to sample over a 10 month period a large number of healthy calves ranging in age from 0 to 8 weeks from individual farms over a wide area of England and South Wales. A wide variety of types were found, but generally they were different to the ones normally associated with healthy humans. It was further shown that the types found on the carcass surfaces were similar to those found in the rectum.

Calves, like humans, acquire their E. coli from their mothers, but as the environments of these neonatal animals are more likely to be contaminated with faeces from the other animals in the herd and/or other farm animals these are also likely to be acquired by the young animals. With progressing age the diversity of the E. coli of farm animals decreases.

(c) Cats and Dogs.

Most of the very few studies undertaken have centred on seeing whether these pets harbour human-pathogenic E. coli and they have indeed shown that they can be present in household pets, such as dogs and cats. Certain pathogenic types were found in 12% of dogs and 8% of cats examined. Similar results were obtained in kittens. A study of the carriage by healthy cats and dogs in Australia of antibiotic resistant E. coli, revealed that 60% of dogs and 26% of cats carried such E. coli. Many of these E. coli were multiply resistant. In addition 60% of these strains could transfer their resistances to suitable recipient strains. Thus it is clear that these pets can be vehicles for the transmission of pathogenic E. coli to humans, but the role these and other E. coli play in the intestinal tract of these animals normally awaits elucidation.

(d) Rats and Mice.

E. coli have been isolateded in very small numbers in the stomaches of mice less than 12 days old. The numbers then rose and fell again at the age of 21 months. At all times the numbers in the small intestine were very small while in the large intestine they were large. Again the numbers rose up to the 18th day and then fell. E. coli can generally be isolated from the faeces of mice at all times.

When E. coli strains of differing genetic composition were individually introduced into germfree mice they would colonize, however, when two or more strains were in competition, then a colonizing order of the different phenotypes could be demonstrated, with some strains more successful than others. An extensive study on antagonisms between strains of E. coli, which were genetically very similar, within the intestinal tracts of germfree mice, showed that strains became host "adapted". Such "adapted" strains then became the dominant population. If these strains were then cultured on artificial media it made them lose this "adaptation". These observations may explain why certain strains in the human intestine will remain dominant for considerable times, preventing any incoming strains from supplanting them. It is possible that such "adapted" strains can even prevent the colonisation by the pathogenic E. coli and thus protect the host.

(e) Domestic poultry.

It was found that some types of E. coli will persist throughout the life of the birds, while others appeared early but disappeared later. Certain others only appeared late but then persisted till slaughter. These types probably have a particular ability to colonize the chicken intestinal tract. Also there may be differences in the characteristics by a strain of E. coli required to colonise a young bird, compared to an adult one.

(f) Birds.

In a study on recreational lakes and bathing areas, the E. coli flora of sea-bird droppings, mainly gulls, was examined. These studies suggested that these birds may carry different kinds of E. coli than humans. However, more recently it has been found that sea birds can carry E. coli, which are pathogenic for humans.

(g) Fish

There has been very little information available about the presence of E. coli in fish. It has generally been considered that the presence of E. coli in fish is an indicator for faecal pollution of the waters in which the fish are swimming, and their presence is considered a human health hazard. It is generally assumed that E. coli may not normally colonise fish.

In experiments recently performed on Rainbow Trout, it was found that E. coli did not become established at 6°C but did become established in the intestines of the fish at 15°C. In fish aquaculture it has definitely been shown that the E. coli will multiply and survive in the fish intestinal tract. Again the growth of the E. coli was only noticeable in the warmer summer months, when water temperatures of 26-29°C were attained. E. coli did not grow in winter when water temperatures were around 10°C. Most of these studies have also noted that the transfer of antibiotic resistances can occur between strains of E. coli in the fish.

8) The Effect of Antibiotics on Intestinal E. coli.

(a) The development and transfer of resistance.

Since the introduction of antibiotics, E. coli were seen to play a particularly crucial role in the development of resistance. It was the discovery of antibiotic-resistance transfer (R) factors in Japan in the 1960's, that provided the explanation for many of these resistance developments. These studies showed that E. coli and related enteric organisms including the pathogenic Shigella and Salmonella species may carry plasmids, which are extrachromosomal pieces of DNA, which code for the resistance to a number of antibiotics. These plasmids can be transferred from one strain to another and even from one species such as E. coli to a Salmonella.

Resistance can also be developed within the bacterium's chromosome, either through various forms of mutational events or by incorporating some of the plasmid's DNA into the chromosome. Currently both plasmid-mediated and chromosomally mediated antibiotic resistance is being regularly observed.

A particular problem with respect to the penicillins and related antibiotics are E. coli which produce enzymes called Extended-spectrum b-lactamases (ESbs). These cause the bacteria which produce them to be resistant to most penicillins, cephalosporins and aztreonam. The genes coding for the production of these ESbs are plasmid-mediated and such strains are often resistant even to currently available b-lactamase inhibitor-b-lactam drug combinations as well, which contain an inhibitor to some of the older enzymes which split penicillins and are known as b-lactamases.

E. coli does not produce just one b-lactamase. Many chromosomally mediated b-lactamases of E. coli have been distinguished. E. coli can also increase the level of b-lactamase production, by increasing the number of b-lactamase gene copies. By altering the regions on the chromosome affecting gene transcription or by changing the complex regulatory apparatus of inducible b-lactamase expression E. coli can also increase production of b-lactamases.

E. coli has developed a different strategy to combat the quinolone drugs. Here the organism produces DNA affected by the drugs, with an altered structure, which is now unaffected by them. In addition there may also be decreased permeability through the outer membrane for the drugs. Zidovidine (AZT), a thimine analogue, which is active against the human immunodeficiency virus, has been increasing used and E. coli can readily become resistant to it by the loss of a specific enzyme, which converts AZT to the biologically active triphosphate form. Thus AZT is inactive on E. coli.

(b) The Behavior of resistant E.coli in the Human Intestine.

With the realization in the 1960's that enteric organisms can transmit antibiotic resistance determinants in packets to each other including human pathogens, it led to increased efforts to examine the importance of such resistant organisms in the intestine. It must be stressed that in an environment, where a given characteristic is not required, such as antibiotic resistance in an antibiotic-free situation, the production of the factors involved in the resistance is a waste in resources and energy for the organism. Such resistance markers are therefore readily lost, when not required. However, for the population as a whole it is worthwhile maintaining a few bacteria carrying these resistances in a tranferrable form so that the characteristics can become readily available to the whole population when the antibiotic stress is encountered.

In an experiment specifically addressing the problem of the survival of Resistance(R)-factor carrying E. coli in the human intestine, cultures of these organisms were ingested by adult healthy individuals in gelatin capsules containing 1010 to 1011 viable organisms. Following this a rapid disappearance of the R-factor bearing strains was observed. It was suggested that this was largely due to the ecological disadvantage which the R-factor carrying bacteria have in an environment, where there is no pressure to maintain the R-factors.

In the natural state no significant difference in the residence times in humans between resistant and sensitive strains was observed. A low level of R-factor plasmid carrying strains was maintained. The R-factors were less widespread in those serologically distiguishable groups, known to persist well in the human intestinal tract and were more common in rarer ones. R-plasmid carrying E. coli are regularly found in the faeces of patients in hospitals and these plasmids as well as their E. coli can be transmitted.

More recently it was shown that for plasmids to be maintained in E. coli in the natural environment they must confer selective advantage to their host bacterium. However, the great diversity in the carriage of plamids conferring b-lactam resistance, which was noted earlier, among hospital isolates of E. coli suggests a wide distribution of such strains in the community. An extensive study in the Netherlands has clearly shown that there certainly is significant carriage of resistant E. coli in the healthy population in the developed world. The source of these organisms is most probably the food chain.

© Copyright Microbionet.

All literary matter in Microbionet is covered by copyright, and must not be reproduced, stored in a retrieval system, or transmitted in any form by electronic or mechanical means, photocopying, or recording, without written permission. This page is to be read in conjunction with the Disclaimer.

Use of this site signifes your agreement to the Legal Notices. Legal Notices © Copyright Microbionet. All rights reserved.