Occurrence and fate of antibiotic resistant bacteria in sewage
1. Introduction and project background
This chapter describes the structure and objectives of the project, and provides the reader with the basic knowledge necessary to understand the problems addressed in the report. A description of the project is followed by a short introduction on antibiotics and antibiotic resistance, including an explanation of the problems occurring when antibiotic resistance is to be measured in bacterial populations. The public health and ecological concerns associated with the emergence of bacterial resistance are discussed. Finally, considerations are made concerning the importance of sewage in the spread of antibiotic resistance between different bacterial populations and environments.
1.1 Project structure and objectives
The project is composed of three parts, each part focusing on a particular feature concerning the occurrence and fate of resistant bacteria in sewage (Fig. 1). In the Part I, the effects caused by the discharge of waste effluent from a hospital and a pharmaceutical plant manufacturing antibiotics were investigated using Acinetobacter as a bacterial indicator. In the Part II, the effects of tertiary sewage treatment on total and relative numbers of resistant bacteria were monitored at two large-scale treatment plants for a period of six months. In the Part III, multiple-resistant strains isolated from treated sewage were analysed for their ability to survive in the aquatic environment. In addition, the impact of municipal sewage effluents on the spread of antibiotic resistance was further evaluated by studying the transfer of tetracycline resistance between Acinetobacter strains isolated from sewage and unpolluted freshwater, and by comparing the occurrence of resistant bacteria in shellfish exposed/non-exposed to sewage effluents.
Figure 1.1.
Schematic representation of the project
The general aim of the project was to identify factors influencing the occurrence of resistant bacteria in sewage and to study the fate of such bacteria along the sewage system. The assessment of the occupational risks associated with the occurrence of resistant bacteria in sewage treatment plants and the public health risks associated with the spread of multiple-resistant bacteria via municipal sewage effluents was not part of this study. The present work, however, represents a good basis for the development of future studies on risk assessment.
The following specific objectives were pursued as part of the project:
 | To assess the effects of waste effluent from a hospital and a pharmaceutical plant on the prevalence of resistant Acinetobacter in the recipient sewers (Part I). |
| To detect changes in the distribution of Acinetobacter strains/species associated with the discharge of waste effluent from these sources (Part I). |
| To determine to what extent sewage treatment reduces the total numbers of resistant bacteria depending on the antibiotic, the bacterial population and the treatment plant under study (Part II). |
| To evaluate the effects of sewage treatment on numbers of single and multiple-resistant bacteria (Part II). |
| To determine the ability of multiple-resistant bacteria originating from treated sewage to survive in laboratory marine microcosms and in membrane diffusion chambers immersed in a freshwater pond (Part III). |
| To demonstrate in vitro transfer of antibiotic resistance from bacteria isolated from sewage to related bacteria occurring in unpolluted aquatic environments (Part III). |
| To determine whether differences in the number of resistant bacteria exist in shellfish from sites exposed to treated sewage and in shellfish from unpolluted sites (Part III). |
1.2 What are antibiotics?
Antibiotics are substances produced by living organisms, which are able to kill or inhibit the growth of microorganisms. According to the literal sense of the word, substances produced synthetically (e.g. sulfonamides or quinolones) should not be termed antibiotics, and the use of a broader term (i.e. antimicrobial agent) would be more appropriate to indicate the complex of all substances having a harmful effect on microorganisms 1,1. However, the term antibiotic is used throughout the present report as a synonym of antimicrobial agent.
Antibiotics are selectively toxic substances as they affect pathogenic microorganisms more adversely than the host. The degree of selective toxicity depends on the specific mechanisms of action of the drug. The most selective agents are those affecting structures (e.g. cell wall) or functions (e.g. folic acid synthesis) present only in prokaryotic cells. The less selective antibiotics are those affecting protein (e.g. tetracyclines) or nucleic acid synthesis (e.g. quinolones), which are essential functions for both prokaryotic (bacteria) and eukaryotic cells (the host). Among the antibiotic agents produced synthetically, are some that are toxic for humans and animals, and their use is restricted to inanimate objects (i.e. disinfectants) or the surface of living tissues (i.e. antiseptics). These agents are generally termed biocides.
1.2.1 Classification
Antibiotics are classified based on their chemical structure. Each class of antibiotics is characterised by a typical core structure and the various members of the class are differentiated by the addition or subtraction of secondary chemical structures from the core structure. The main classes of antibiotics currently used in clinical practice include penicillins, cephalosporins, tetracyclines, aminoglycosides, fluoroquinolones, potentiated sulfonamides, macrolides and glycopeptides.
Antibiotics can also be classified as broad, intermediate or narrow spectrum, depending on the range of bacterial species against which they are active 1. Broad-spectrum antibiotics include compounds active against both Gram-positive and Gram-negative bacteria like quinolones, tetracyclines, and third generation cephalosporins. Intermediate spectrum antibiotics generally include substances with reduced activity against some Gram-negative bacterial species (e.g. ampicillin, amoxicillin, first and second generation cephalosporins). Narrow spectrum antibiotics are only effective against restricted groups of bacteria. For example, penicillin is only active against Gram-positive bacteria, whereas, aminoglycosides, sulfonamides and trimethoprim are solely active against aerobic bacteria 1.
1.2.2 Mechanisms of action
Antibiotics constitute quite a heterogeneous group of chemicals. Depending on the chemical structure, antibiotics exert an effect on different structures or functions of the bacterial cell (Fig. 1.2). The major mechanisms of action are inhibition of the cell wall synthesis (e.g. penicillins and vancomycin), damage of the cell membrane function (e.g. polymixins), inhibition of protein synthesis (e.g. aminoglycosides, tetracyclines, chloramphenicol, lincosamides and macrolides), inhibition of the nucleic acid synthesis (e.g. quinolones and rifampicin), and metabolic antagonism (e.g. sulfonamides and trimethoprim)1.
Figure 1.2.
Sites of action for selected antibiotics. PABA, para-aminobenzoic acid; DHFA,
diydrofolic acid; THFA, tetrahydrofolic acid. Modified from Prescott and Baggot 1.
1.3 What is antibiotic resistance?
Antibiotic resistance is a relative term. A bacterial strain can be defined resistant if it survives in the presence of higher antibiotic concentrations in comparison with phylogenetically related strains 2. Thus, antibiotic resistance is not a bacterial property that can be determined by studying a single strain, but only by comparison under identical conditions of two or more strains belonging to the same genus or species.
The above-mentioned definition of antibiotic resistance refers to in vitro conditions. Under in vivo conditions, antibiotic resistance is a context-dependent term as it depends on the location of the bacterium and the bioavailability of the drug. For example, bacteria are less susceptible to antibiotics when assembled in biofilms (complex communities of microorganisms embedded in a matrix of extracellular material) compared with the same organisms living separately 3. In aquatic environments, binding of the antibiotic molecule with ions or substances present in sediment strongly reduces both the activity of the drug and its absorption in the fish intestine 4.
1.3.1 Molecular mechanisms
Bacterial resistance to antibiotics can be caused by different molecular mechanisms 5. The most common mechanisms include: reduced drug uptake (e.g. membrane impermeability to cephalosporins); active drug efflux (e.g. tetracycline efflux pumps); drug deactivation (e.g. hydrolysis of penicillins by beta-lactamases), modification of the drug target (e.g. mutations of the DNA gyrase leading to quinolones resistance); increased concentration of the drug target (e.g. increased folic acid production that counteracts the inhibition of such production by sulfonamides), or alternative pathways to elude the drug effect (e.g. synthesis of folic acid using an enzyme which is not affected by sulfonamides)(Fig. 1.3).
Figure 1.3.
Molecular mechanisms of antibiotic resistance. Modified from Hayes and Wolf 5.
An important distinction should be made between natural and acquired resistance. Bacteria are termed naturally, intrinsically or constitutively resistant when resistance is due to characteristic features typical of the species. For example, Pseudomonas aeruginosa is naturally resistant to penicillins, due partly to the inability of the drug to diffuse through the outer membrane 6 and partly to the deactivation of the drug by chromosomally encoded enzymes (i.e. beta-lactamases)7.
In contrast, acquired resistance emerges in a bacterial population that was previously susceptible, because of modifications of the bacterial DNA caused by either chromosomal mutation or horizontal gene transfer. Natural resistance results from a long process of genetic evolution, whereas, acquired resistance can arise within a short time (one or few generations)5.
1.3.3 Acquisition by chromosomal mutations
Mutation is a heritable change in the sequence of the DNA occurring due to errors during DNA replication 8. The frequencies of chromosomal mutations leading to antibiotic resistance depend on the specific antibiotic. For example, mutation frequencies are high for compounds like nalidixic acid, rifampicin and streptomycin (10-8 to 10-10 cells per generation), low for erythromycin and are not known to occur for vancomycin and polymixin B 9. For antibiotics like streptomycin, a single mutation can determine a 1000-fold increase in the resistance levels 9. In contrast, for other drugs (e.g. quinolones) the acquisition of resistance is a gradual, step-wise process in which different mutations are involved 10.
1.3.4 Acquisition by horizontal gene transfer
Horizontal gene transfer is the relocation of genetic material from one bacterial cell (donor) to another (recipient). Such a transfer may occur directly by physical contact (conjugation) or indirectly, using the surrounding medium (transformation) or bacteriophage (transduction) as vectors 11(Fig. 1.4). Bacterial transfer of antibiotic resistance has been demonstrated to occur in various natural habitats, including water, sediment, soil, plants and animals 12,13. The DNA transferred from the donor to the recipient may be contained in mobile genetic elements called plasmids, structures of circular DNA that reproduce independently from the chromosome 11. Unlike chromosomes, plasmids generally do not encode functions essential to bacterial growth, but functions that are of importance under particular conditions, such as antibiotic resistance, heavy metal resistance, metabolic functions, or production of antibiotics, toxins and virulence factors 14.
1.3.5 Intracellular migration of resistance genes
Antibiotic resistance genes can migrate from one site to another on the bacterial genome using small vectors called transposons 15 and integrons 16. These genetic elements containing antibiotic resistance genes are able to move between different sites of the bacterial genome without any requirement of DNA homology. This process is known as non-homologous recombination (to a site that does not match with the gene) and differs from the normal process of genetic recombination, which requires a high degree of DNA homology (a near perfect match) 11,17. Both transposons and integrons make it possible for new antibiotic resistance genes to be acquired by plasmids and subsequently spread in the bacterial population by mechanisms of horizontal gene transfer, as suggested by the frequent recovery of these genetic elements as part of broad host plasmids 18,19.
Figure 1.4.
Mechanisms of bacterial genetic transfer (Levy 20).
1.3.6 Measurement of resistance in bacterial populations
The value of the term "measurement of antibiotic resistance" in environmental microbiology generally differs from that in clinical studies. The main concern for environmental microbiologists is to investigate the distribution of antibiotic resistance in bacterial populations rather than the level of resistance in individual strains. Unfortunately, culture methods are not efficient enough to determine the actual prevalence of antibiotic resistance in a bacterial population. In fact, only a small proportion of the aquatic bacterial flora (<1%) can be cultured on laboratory media 21.
The method traditionally used for the measurement of antibiotic resistance at the population level consists in standard bacteriological counts on media containing specific concentrations of antibiotics. The main drawback of this method is the use of a single breakpoint for the determination of antibiotic resistance. In fact, the use of a single breakpoint, corresponding to the amount of antibiotic agent added to the medium, does not take into account the variability in the levels of antibiotic resistance existing among different bacterial species. Consequently, bacteria characterised by intermediate levels of resistance can be classified either as resistant or susceptible depending on the concentration of antibiotic added to the medium.
An alternative approach is to use a group of phylogenetically related organisms as bacterial indicators of antibiotic resistance. This method is based on the principle that spatial and temporal differences observed in the levels of antibiotic resistance of the bacterial indicator are indicative of the selective pressure to which the entire bacterial population is exposed. Thus, this method does not aim to determine the exact prevalence of antibiotic resistance in the bacterial population under study, but rather to detect the effect of potential sources of antibiotic resistance on the bacterial population.
The use of bacterial indicators offers various advantages compared with bacteriological counts on antibiotic selective media. The isolation and identification of bacteria makes possible the use of standardised methods for antibiotic susceptibility testing, namely the disc-diffusion method and the dilution method. When antibiotic susceptibility testing is performed on a large number of bacterial isolates, results can be used to understand the distribution of antibiotic resistance within the target bacterial population and consequently to define appropriate breakpoints for the classification of resistant and susceptible strains.
Today, the nucleotide sequences of many genes encoding for antibiotic resistance are available. DNA-DNA hybridisation and PCR methods are currently used to investigate the presence of resistance genes in environmental bacteria. In comparison with phenotypic methods, genetic methods offer the great advantage to investigate also non-culturable bacterial species. However, the currently available genetic methods are only able to quantify, to a limited degree, the presence of a gene in a bacterial population. Furthermore, since a number of different genes can encode resistance to the same antibiotic agent, genetic methods cannot be used for quantitative assessment of resistance.
1.4 The microbial threat
In the last decades, bacterial resistance to antibiotics has assumed an increasing importance with regard to its impact on both public health and ecology. Obviously, the primary problem is represented by the emergence of antibiotic resistance among bacteria pathogenic to humans and animals, which makes difficult the treatment of some life-threatening infections. However, independent from the risks for human health, is the spread of antibiotic resistance and the problems raised in ecological nature. In fact, the introduction and selection of resistant bacteria in the environment can lead to structural changes in the composition of microbial communities, with possible deleterious effects on the balance of natural ecosystems.
1.4.1 The emergence of resistance in human pathogenic bacteria
In the past, bacteria were the most important cause of disease and mortality among humans. The introduction of antibiotics in human medicine has markedly reduced the impact of bacterial diseases on human mortality. Nevertheless, the extraordinary capacity for adaptation of bacteria soon allowed these organisms to develop mechanisms of resistance enabling them to overcome the toxic effects of antibiotics.
A survey on enterobacterial isolates collected between 1917 and 1954 has demonstrated that bacteria were generally susceptible to antibiotics before these drugs became commonly available in human medicine 22. However, other studies indicate that resistant bacteria were present at the time, although they were not prevalent in bacterial populations 23. Thus, it appears that the indiscriminate use of antibiotics has played a major role in the emergence of antibiotic resistance by exerting a selection in favour of resistant bacteria.
The first case of penicillin resistance in E. coli was reported in the 1950’s. Since then, things have taken a turn for the worse. Today, antibiotic resistance represents an important problem in the therapy of various human pathogenic bacteria. Three bacterial species causing life-threatening infections (Pseudomonas aeruginosa, Mycobacterium tubercolosis and Enterococcus faecalis) can demonstrate resistance to any available antibiotic 20. Vancomycin is the only effective drug for treatment of infections caused by methicillin-resistant Staphylococcus aureus (MRSA), but the occurrence of strains with reduced susceptibility to this antibiotic has already been reported 24. Problems may also occur in the therapy of hospital infections caused by Acinetobacter baumannii, Haemophilus influenzae, Klebsiella pneumoniae, Neisseria meningitidis and Streptococcus pneumoniae 20.
The problem of antibiotic resistance is of particular concern for immunosuppressed patients, suchas those affected by HIV, cancer or chronic diseases, as antibiotic therapy represents the only way to overcome bacterial infections for these people. Serious problems may also occur in developing countries where the use of new and expensive drugs is limited by their cost and availability. In addition to the risks for human health, this situation incurs a worldwide increase in the cost of hospital care, including the use of new expensive drugs, increased costs for bacteriological analysis and prolonged hospitalisation 25.
1.4.2 The spread of resistance among environmental bacteria
Antibiotic resistance is not only found in pathogenic bacteria but also in environmental organisms inhabiting terrestrial and aquatic habitats. The occurrence of resistant bacteria in nature may have originated from antibiotic-producing organisms, as suggested by the evidence that in some cases the mechanisms and genes protecting these organisms from the antibiotics they produce are similar to those responsible for resistance in clinical isolates 26. However, higher numbers of resistant bacteria occur in polluted habitats compared with unpolluted habitats 27,28, indicating that humans have contributed substantially to the increased proportion of resistant bacteria occurring in the environment.
Possible mechanisms by which humans enhance the spread of antibiotic resistance among environmental bacteria include the deliberate or accidental introduction of antibiotics, resistant bacteria and resistance genes into the environment. Antibiotics exert a selection in favour of resistant bacteria by killing or inhibiting growth of susceptible bacteria (see section 1.5.1); resistant bacteria can adapt to environmental conditions and serve as vectors for the spread of antibiotic resistance; resistance genes can be taken up by indigenous bacteria and spread by mechanisms of genetic transfer (see section 1.3.4).
The main risk for public health is that resistance genes are transferred from environmental bacteria to human pathogens. The ability of resistant bacteria and resistance genes to move from one ecosystem to another is documented by the various cases in which transmission of resistant bacteria has been demonstrated between animals and humans 29,30. The inclusion of certain growth promoters in animal feed has been recognised as a cause for the selection of resistance genes in the commensal microflora of animals and their transmission to humans via the food chain 29,30. Similarly, drinking and bathing water could represent a source for the acquisition of resistant bacteria in humans. However, further studies are necessary to validate this hypothesis.
The ecological consequences associated with the dissemination of resistant bacteria in the environment have been scarcely investigated. However, it appears evident that environmental contamination with antibiotics, resistant bacteria and resistance genes affects the biodiversity of natural ecosystems. Antibiotics are likely to determine a reduction in the levels of microbial diversity by the suppression of susceptible organisms, including bacteria, fungi, protozoa and algae. Resistant bacteria and genetic elements could find favourable conditions to become predominant in habitats contaminated by antibiotics, thereby, altering the original composition (balance) of natural microbial communities.
1.5 Spread of antibiotic resistance in sewage
Sewage is waste matter resulting from the discharge into the sewers of human excreta and wastewater originating from the community and its industries. Sewage contains a high content of both organic and inorganic matter, as well as high densities of living organisms, including pathogenic, commensal and environmental bacteria. This chacteristic composition makes sewage a particularly suitable ecological niche for the growth and spread of antibiotic resistance.
1.5.1 Antibiotic selective pressure
The acquisition of antibiotic resistance genes is generally independent of the presence of antibiotics. However, the exposure of bacteria to antibiotics confers an ecological advantage to resistant strains on susceptible strains, allowing them to become predominant in the bacterial population. This situation is commonly termed as antibiotic selective pressure and can occur in either the host in vivo (eg. human or animal body) as a consequence of chemotherapy or in the environment, for example when antibiotic residues are introduced in sewage.
Residues of antibiotics administered to humans and animals reach the sewage systems in urine or faeces, in the form of either parent compound or degraded metabolites depending on the pharmacology of the specific antibiotic. Furthermore, an unknown amount of antibiotics enter the sewers by waste derived from antibiotic production and disposal of a surplus of drugs. Indeed, various antibiotics have been found in municipal sewage, including fluoroquinolones, sulfonamides and erythromicin metabolites 31-34. The antibiotic concentrations found in sewage vary between 1 and 100 m g per liter. Such concentrations are 100- to 1000fold lower compared with those necessary to inhibit resistant bacteria, but are sufficient to affect susceptible bacteria 35,36. Therefore, the occurrence of such antibiotic concentrations in sewage has the potential to select for antibiotic resistance.
The fate of antibiotics in sewage depends on their chemical properties. Lipophilic and non-readily degradable substances are likely to be retained in the sludge, whereas, hydrophilic substances may be able to pass through treatment plants and end up in the natural recipients receiving treated sewage 37. It also appears that the solubility in water of drug metabolites is generally higher compared with the parent compounds 37. Thus, it is likely that a large proportion of the antibiotic residues introduced into the sewage system can reach surface waters through municipal sewage effluents.
1.5.2 Non-antibiotic selective pressure
Among the multitude of substances occurring in sewage, there are some that have the potential to select for antibiotic resistance, even though they are not antibiotics. Heavy metals and biocides are two important groups of non-antibiotic substances showing this property. Heavy metals are widespread in sewage as a consequence of industrial pollution. Biocides are introduced into sewage by hospitals, farms, slaughterhouses and food-processing establishments; where these agents are used for the disinfection of environments and utensils, or by the community, due to the presence of these agents in house-hold products, such as soaps and dishwashing detergents.
There are two possible ways by which heavy metals and biocides can select for antibiotic resistance. The genes encoding resistance to heavy metals and biocides can be located together with antibiotic resistance genes on either the same genetic structure (e.g. plasmid), or different genetic structures within the same bacterial strain. Alternatively, bacteria can have unspecific mechanisms of resistance to different substances, including heavy metals, biocides and antibiotics. In both cases, exposure to one substance results in the selection of bacterial strains able also able to resist the other substance (co-selection).
Genes encoding resistance due to heavy metals and antibiotics often co-exist on plasmids 38. In addition, unspecific mechanisms conferring resistance to both heavy metals and antibiotics are known to exist in some bacterial species (e.g. active pump-efflux system encoded by the marA gene in E. coli). The co-selective property of heavy metals is confirmed by the indirect evidence that bacteria isolated from heavy metal-polluted marine sediment are significantly more resistant to antibiotics compared with bacteria isolated from unpolluted sites 39.
Although genes encoding resistance to biocides have been found on plasmids and integrons 40, these substances are more likely to select for antibiotic resistance by induction of unspecific mechanisms of multiple resistance. Laboratory experiments have shown that biocides such as triclosan and pine oil can select for resistance to different antibiotics when bacteria are exposed to low concentrations of biocide 41,42. Accordingly, the co-selective effect of biocides for antibiotic resistance could be particularly marked when these substances are dispersed in the environment, because of dilution and formation of concentration gradients.
1.5.3 Optimal conditions for horizontal gene transfer
Sewage is a suitable habitat for the transfer of resistance genes across different groups of bacteria. In this habitat, environmental bacteria meet resistant bacteria selected by use of antibiotics in human and veterinary medicine. Consequently, resistance genes occurring in bacteria of human and animal origin can be transferred to environmental bacteria, contributing to the formation of an environmental pool of resistant bacteria and resistance genes.
The high concentrations of bacteria, nutrients and suspended solids in sewage are all factors enhancing horizontal gene transfer 43-45. High bacterial concentrations increase the chance that donor and recipient cells come in contact. Nutrients are more likely to have an indirect influence on the occurrence of gene transfer by increasing the concentration and the metabolic activity of bacteria. Suspended solids provide ideal surfaces on which the various components contributing to the process of horizontal gene transfer (bacteria, free DNA and bacteriophages) are concentrated.
Plasmids and transposons harbouring antibiotic resistance genes are widespread in the bacterial flora of sewage 46,47. Multiple-resistant bacteria isolated from sewage can transfer plasmid-mediated antibiotic resistance at high frequencies in the laboratory 48,49. Experiments performed using membrane chambers immersed in sewage have shown that high frequencies of transfer may also occur under real conditions 50,51.