Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
21 being trans ferred by the process o f conj ugation . Shortly therea fter , other workers demonstrated that these conj ugative packages of genetic information , labeled R factors ( now cal l ed R plasmids ) , were prevalent not only in Shigella spp . , but also in Salmonella spp . and Escberichia QQli . S ince the time of their initial discovery , R plasmids have been shown-Â in complementary studies by epidemiologists , molecular geneticists , and bacterial physiologists--to be widely transmissible and speci fica l ly selectable when antimicrobial drugs are present in the environment . Increasing drug res istance among human isolates of cl inica l ly relevant bacteria during thi s period has been shown repeatedly to be due primarily to the prol i feration of R plasmids . MECHANISMS OF ACQUIRING AHTIMICRQBIAL BESISTAHCE IN BACTERIA CHRQMOSOHES Most drug resistance in cl inica l ly relevant bacteria is due to conj ugative trans fer of R plasmids and their clonal expans ion during exposure to antimicrobial drugs . But chromosomal inheritance and the trans fer of chromosomal mutant genes by trans formation also play a role . Bacteria can acquire new genetic information in three known ways : conj ugation , transduction , and trans formation . Coniuqation requi res that the donor bacterium possess both the means to dupl icate part of its genetic information and the means to attach itsel f to a recipient bacterium for DNA transfer . The donor must mate with a recipient bacterium that is phys iologica l ly capable of permitting the new DNA to enter and repl icate autonomously as a plasmid or permitting it to become incorporated by recombination into the recipient ' s chromosome . Transduction , a process much l ess important for the trans fer of drug res istance , depends on bacterial viruses ' known as bacteriophages to "package " pieces of the chromosome or plasmid of the donor organi sm and inj ect the package into the appropriate bacterium for uptake and incorporation of the foreign DNA . Transformation is the process by which DNA in solution is taken up directly by a bacterial cel l . Plasmid ( ci rcular) DNA is usua l ly taken up more efficiently than chromosomal ( l inear) DNA , in that many bacteria have enzymes , known as exonucleases , that attack only l inear fragments of DNA . Nevertheless , in some organisms , trans formation of chromosomal contents has been shown to be capable of trans ferring drug res istance . Chromosomal drug res istance , in most cases , i s due to mutation of pre-existing DNA . Although there are others , three ant imicrobial drugs stand out for their abi l ity to select chromosomal mutants that have acquired drug
22 res istance : streptomycin , nal idixic acid , and ri fampin . Their common characteristic is that they work by binding to protein targets within the bacterial cel l . Each o f the protein targets provides an important vegetat ive function for the cel l - -the streptomycin target is the 8 1 2 ribosomal protein , the nal idixic acid target is DNA gyrase , and the r i fampin target is RNA polymerase . Those targets perform essential functions for the cel l : protein synthes i s , the required winding of DNA , and the transcription of DNA into messenger RNA , respectively . The speci fic s ites of antibiotic binding in the targets can occasional ly be altered by mutation in such a way that the antibiotic no longer binds to the target and the target reta ins most of its function . Thus , the appropriate mutations in the genes for the 8 1 2 ribosomal protein , DNA gyrase , and RNA polymerase will lead to res istance to streptomycin , nal idixic acid , and ri fampin , respectively . The bacterium pays a price for the mutations , in that it acquires a less than opt imal "housekeeping" prote in in the process of evading the effect of the antibiotic ; thereafter , it i s typica l ly not as hardy as its nonmutant parents . Although chromosomal res istance usua l ly involves res istance that is specific for the selecting antimicrobial drug , it i s occas ional ly responsible for s imultaneous res istance to several antibiotics of different structures and s ites of act ion , e . g . , the mAX locus in ï¿½ 2211 affects uptake of tetracycl ine , ce foxitin , and chloramphenicol . Extens ive epidemiologic investigations of drug res istance in enteric bacteria have yielded l ittle evidence of the importance of chromosomal mutation in the acquis ition of drug res istance or of trans formation as a means by which drug res istance can be exchanged . Most enteric bacteria have been shown to have low efficiency in taking up DNA ( a property known as competence ) , unless they are treated so as to damage their permeabi l ity barriers temporarily . R PLA8MID8 AND TBANSPQSONS S ince their discovery about 3 0 years ago , R plasmids have been extens ively studied epidemiological ly and molecularly and have been shown to play a predominant role in drug res istance among bacteria . 7 , 2 2 - 2 4 , 2 7 , 2 9 , 3 4 , 5 5 , 8 5 , 9 1 , 9 2 , 9 8 Like other sel f-replicating nonchromosomal units of DNA , R plasmids carry "machinery" for efficient repl ication and genes for part icular drug-resistance phenotype . By definition , R plasmids carry genes that encode products that confer drug res istance in a bacterium . Often a single R plasmid contains multiple genes , each encoding a different kind of res istance . Some individual res istance genes encode res istance to multiple related antibacterial drugs . Almost
23 every drug-res istance determinant i s carried on a genetic unit ( usua l ly sma l l , occas ionally large ) , cal l ed a transposon , that can move from its location on the R plasmid to other l ocations--typ ical ly , but not exclus ively , other plasmids . That form of DNA rearrangement , or transposition , requires special genes and stretches of DNA that are parts o f the transposon . l 6 , 1 0 3 It i s useful t o cons ider transposons as freely movable genetic modules that can be assorted , reassorted , and added to and subtracted from evolving R plasmids as env ironmental pressures dictate . The abi l ity of most drug-resistance genes to transpose provides R plasmids with an extraordinary degree of genet ic plasticity . Although it i s not clear that the presence of ant ibiot ics in the environment has any influence on the extent of transposition of resi stance determinants , ant ib iotics exert a profound influence on the selection and pers istence of R plasmids with mult iple drug-resistance determinants . Studies of indigenous soi l bacteria in the preant ibiotic era showed not only that R plasmids were less prevalent , but al so that recovered R plasmids typically contained only one or two resi stance determinants each . 2 4 , 2 9 More recently , in sharp contrast , bacteria recovered in environments exposed to antibiotics have had a high prevalence of R plasmids with multiple drug-resistance determinants . 2 9 , 3 4 , 9 1 , 9 2 It i s important to note that these R plasmids are s imilar in many genetic respects to the preantibiot ic-era plasmids that did not have multiple drugÂ res istance determinants . Transposons , whose drug-res istance genes evolved as " protection" against the natural antibacterial substances in the soil , thus used plasmids al ready ava ilable . 7 , 2 9 Moreover , the same transposon can be found in an array of d i fferent plasmids ; in nature , drugÂ res istance elements are indeed promiscuous and can locate in a broad spectrum of genomes . Most bacteria , given the appropriate supplemental genes , are potentially capable of trans ferring DNA to other bacteria by conj ugation . Usual ly the suppl emental genes are carried on a plasmid . When present on an R plasmid , the suppl emental genes together make up what is cal l ed the res i stance trans fer factor , or RTF . Most R plasmids contain an RTF , which enables them to be conj ugatively trans ferred between bacteria . 2 9 Just as the transposon , at the level of the single drug-resistance determinant , is capable of movement to a di fferent R plasmid , the RTF-conta ining R plasmid i s capabl e o f movement t o other strains and other species o f bacteria . Although initial studies o n conj ugative R plasmids were l im ited to facultative gram-negat ive bacteria , 2 9 conj ugation clearly plays a maj or role i n the trans fer o f drug res i stance among facultat ive gram-positive bacteria 5 5 , 8 1 , 105 and among anaerob ic bacteria . l 5 , 9 6 , 1 1 4 Recently , i t has been found that some transposons i n some
24 qram-pos itive bacteria can bypass the need for RTF-conta ining R plasmids in conj uqation . 15 These " conj ugative transposons " conta in the equivalent of the RTF , as wel l as the R factor . Other studies involvinq anaerobic bacteria have shown that sublethal concentrations of tetracycl ine in v itro actual ly promote R-plasmid transfer , in addition to selecting bacteria that carry the druq-resistance determinant . 1 19 Of equal or qreater concern from the standpoint of druqÂ res istance prol i feration is the findinq that the tetracycl ine-res istance determinant ( Tc r ) can be conj uqatively transferred back and forth between bacteroides ( anaerobes ) and ï¿½ Â£2li ( facultative qram-neqative bacil l i ) . l l 9 Inasmuch as anaerobic bacteria , especial ly species of Bacteroides , are the predominant fl ora in the mammal ian qastrointestinal tract , the presence of back-andÂ forth trans fers suqgests that the reservoir for ma intenance , persistence , and spread of at least one druq-resistance determinant , Tc r , is enormous . The latter example of interspecies conj uqative trans fer , o f which there are many examples ( see Odel son et al . 9 6 for a review ) , brinqs up the issue of plasmid host ranqe . In addition to the qenetic attributes al ready discussed , R plasmids contain a veqetative origin of rep l i cation , or ï¿½ , which enables them to repl icate autonomously in host bacteria . Plasmids can be cataloqed into " incompatibi l ity" groups based on thei r ï¿½; those with the same ï¿½ cannot coexi st within a bacterium , because of competition for identical replication factors . 2 9 The spec i f ic oriV carried by a plasmid constitutes another element of its phenotype : its host ranqe . Narrow-host-range plasmids can repl icate in only a few species of bacteria , usual ly because only those bacteria provide additional factors requi red for plasmid repl ication . In contrast , broad-host-range plasmids can trans fer to and replicate in a larqe variety of bacteria . Results of trans fer studies in ï¿½ Â£2li and bacteroides ind icate that antibiot ic exposure--in this case to tetracycl ine--can increase the trans fer and sel ection of at least one form of broad-host-ranqe plasmid res istance to tetracycl ine . I n early studies of conj uqative trans fer of R plasmids amonq bacteria common in the qastrointestinal tracts of humans and farm animals , ï¿½ QQli was found to act as a qood donor and recipient of R plasmids ; Salmone lla spp . were not as qood donors and recipients . 2 9 In a more recent study involvinq ï¿½ typhimurium and ï¿½ QQli recovered from calves , stra ins of ï¿½ typhimurium were extremely proficient R plasmid donors--even better than stra ins of ï¿½ Â£2li . 1 2 4 Certain plasmids , belonging to a particular qroup of incompatible plasmids , Inc H2 , found in most strains of salmonel lae and pre ferentia l ly in ï¿½ typhimurium , showed a peak e ffic iency o f trans fer a t 3 0 Â° C and were conj uqatively trans ferred in
25 calves â¢ feces a fter excretion . 1 2 4 I nc H 2 plasmids have other crit ical features : they carry res istance to both penici l l in and the tetracycl ines , they transfer to ï¿½ ï¿½ , and they carry other non-druq-res istance determinants that increase the abi l ity of the bacteria to col on i z e the qastrointest inal tract . 1 2 4 I n every qeneral aspect of druq res istance studied ( i . e . , express ion of res istance qenes , maintenance of R plasmids , and trans fer of plasmids ) , Salmonella spp . and ï¿½ Â£2li have been found to be quite s imilar . 29 Althouqh Salmonella spp . miqht behave quite differently from ï¿½ QQli in the qastrointestinal tract , because of the former orqanism â¢ s abi l ity to invade entrocytes and thereby avoid ant ibiotics that cannot penetrate cel l s wel l ( i . e . , penic il l in and aminoqlycos ides , but not the tetracycl ines or chloramphenicol ) , evidence suqqests that R plasmid trans fer occurs with ease in Salmonel la spp . in vivo . For example , in an outbreak of qastroenteritis caused by ï¿½ typhimurium that a ffected 1 , 9 0 0 persons who ate contaminated turkey meat , the source stra in of bacteria isolated from the meat was ant ib iotic-susceptible , as were bacterial orqani sms isolated from persons who had not taken any antibiotics . In sharp contrast , a hiqh proport ion of the persons who were qiven chloramphenicol , ampic i l l in , or one of several other ant ibiotics had R plasmids bearinq ï¿½ typhimurium in the ir stool s ; that proves the abi l ity of salmonel lae to acquire R plasmids from the human qut . 2 a Thus , stra ins of salmonel lae , as wel l as ï¿½ QQli , can act as reservoirs for conj uqative R plasmids and thus qain an enormous selective advantaqe over other bacteria in the face of antibiotics in the environment . Epidemioloqic surveys of R plasmids in bacteria that are cl inica l ly important have shown that multiple-druq res i stance has increased proqressively since the beqinninq of the antibiotic era . 3 4 , 8 1 , 9 1 , 9 2 Amonq individual R plasmids , the number of druq-resistance qenes per R plasmid and the l ikel ihood that a qiven R plasmid is conj uqat ive have also been increasinq . 6 3 , 8 9 , 1 1 4 Althouqh early studies had indicated that the problem of druq res istance was most pronounced in the Enterobacteriaceae , the more recent studies show that the problem has spread . Druq res istance has now been found in bacteria of virtual ly a l l qenera that are i mp ortant . 1 5 , 2 2 , 2 7 , 3 4 , 5 3 , 5 5 , 8 1 , 8 5 , 9 1 , 9 2 , 9 6 , 9 8 , 1 0 2 , 1 0 5 , 1 1 4 , 1 19 , 1 2 2 I n summ a ry , R plasmids affect the microbial populations of humans in several ways : o R plasmids confer druq res istance on a qreat number of bacteria , includ inq pathoqens and commensal s . o R plasmids typica l ly confer druq res istance to several antimicrobial druqs s imultaneously .
26 o R plasmids are capable o f being transferred among strains of the same species and among d i f ferent species through the process of conj ugat ion . Trans fer can occur from commensal to pathogenic bacteria in the presence o f antibiotics . o R plasmids can carry additional genes , including virulence factors (discussed later ) . o Exposure to antimicrobial drugs causes an increase in the number and the spread of R plasmids by preferential ly selecting bacterial offspring that are drug-res istant and , in some cases , poss ibly by increasing the efficiency of conj ugative trans fer . ROLE OF ANTIMICRO BIAL DQSAGE IN SELECTION OF DRUG-RESISTANT BACTERIAL POPULATIONS Antimicrobials are used in three ways in agriculture . In the first , high ( therapeutic) doses are used for brief periods ( usua l ly no l onger than about a week to 1 0 days ) for the treatment o f infectious disease . In the second , l ow ( growth-enhancing ) doses are used for l ong periods in l ivestock to promote growth . In the third , l ow ( prophylact i c ) doses are used for a period up to 2 weeks to prevent di sease . Although the duration of antimicrobial use d i f fers between growth-promoting and prophyl actic purposes , the dosages for both " subtherapeutic" uses are typica l l y the same , i . e . , less than 2 0 0 grams/ton . The selection o f antimicrobial-resistant bacteria is a consequence of therapeutic use and of both types ( growth-enhancing and prophylactic) of subtherapeut ic use of antimicrobial agents . Dosages o f drugs that would be classi fied as subtherapeutic might wel l produce concentrations in the gastrointestinal tract that are sufficient to inhibit suscept ible species of bacteria . In Danish studies examining the inc idence of drug res istance in feces o f pigs given only intermittent therapeut ic courses of antimicrobial drugs ( none of the antibiot ics was fed ) , a high proportion ( 5 3 % ) of ï¿½ QQli strains were found to be res istant to at l east one commonly used antimicrobial drug . l l 5 Moreover , 5 3 % of the pigs carried tetracycl ine res istant ï¿½ QQli despite the fact that none of the anima l s had been exposed to tetracycl ine within the past year . l l 5 Using various methods of analys i s , Carpet recently found that continuous subtherapeutic doses o f ant imicrobial drugs caused a profound alteration in the fecal flora of mice , with a sign i ficantly increased proportion of res istant ï¿½ ï¿½ . 1 7 Although the data are overwhelming in showing that use of antimicrobial agents promotes the emergence of drug
27 res istance , the d i fferential effects o f the three kinds o f use i n selectinq druq res istance either i n anima l s or in vitro have not been wel l studied . Nevertheless , in a recently reported study , when three qroups o f p iqs were examined for the presence of druq res istance in their fecal col i form bacteria , a non-ant ibiotic-treated herd had a lower proportion of druq-resistant bacteria than a herd qiven ant ibiotics only intermittently and at a hiqh ( therapeutic ) dose : i solates from the latter herd had a l ower proport ion of druq-res istant col i form bacteria than those from a herd exposed continuously to antibiotics at a subtherapeutic dose . 58 Spec i f ical ly , the proport ion of the tetracycl ines res istance amonq the fecal isolates was found to be 2 6 % for the first herd , 7 6 % for the second herd , and 1 0 0 % for the third herd . s a An earl ier study by the same qroup o f investiqators yielded nearly identical result . 5 9 The feedinq of one of the tetracycl ines at subtherapeutic doses resulted in a l inear increase in tetracycl ine-res istant col i form bacteria , and the numbers of druq-resistant bacteria eventual ly equal ed or exceeded those of druq-res istant bacteria from anima l s treated therapeutical ly . These studies support the hypothes i s that any form of antibiotic exposure increases the prevalence of druq res istance . Moreover , they provide l imited evidence that druq res istance is at least as preval ent a fter continuous subtherapeutic use of antibiotics as a fter intermittent therapeutic use . EXPERIENCE WITH AUTIBIQTIC BESISTAHCE AfTER AHTI BIQTIC USE IN HUMANS The experience qa ined in the use o f antimicrobial druqs since the 1 9 4 0 s provides stronq evidence of the e f fects o f such aqents in the selection of antimicrobial -res istant commensa l s and pathoqens in humans . That experience emphas i zes the importance , constantly reiterated in c l inical teachinq , o f avoidinq unnecessary , prolonqed , inadequate ( subtherapeutic dos inq ) , or inappropriate ( for the et ioloqic bacteria ) treatment or prophylaxis with antimicrobial druqs . The current preval ence of antimicrobial -res i stant stra ins of bacteri a is a consequence of extens ive antimicrobial use in the last 4 0 - 5 0 years . The prevalence was much l ower in the pre-ant ibiotic era . Basel ine LeVels of Antibiotic Res istance--the Status in the Pre-antibiotic Era Insiqhts into the prevalence of ant ibiotic res istance in the pre-antibiot ic era come from examination of bacterial
28 i solates from primitive societies unexposed t o antimicrobial therapy and from study of stored i solates predating the introduction of penic i l l in . I n an examination of 2 1 human stool specimens and 19 soil specimens from an " antibiotic-virgin" population in the Solomon Islands in 1 9 6 8 , R plasmids (mediating streptomycin and tetracycl ine res istance ) were found in only two specimens ( 5 % ) . 3 5 I n a study o f human and animal communities in Rhodes ia , about 10% of 47 fecal specimens from Kalahari bushmen and 5 4 0 from animals conta ined gram-negative baci l l i res istant to one or more ant ibiotics . 7 4 I solates were frequently resistant to only a single drug ( often ampici l l in ) , and none contained R factors . However , the initial fecal cultures were streaked on media containing l ow concentrations o f ant ibiotics , wh ich may have been inhibitory to occas ional R-plasmid-conta ining stra ins and thus might have caused the ir prevalence to have been somewhat underestimated . I n a study o f another antibiotic-unexposed population , in North Borneo , 5 0 multiple-antibiotic-resistant stra ins were found among 1 , 0 1 7 fecal isolates ( more than hal f identi fied as ï¿½ Â£211 ) from 1 2 8 persons . Of the l atter , s i x strains ( al l ï¿½ Â£Qli) , conta ined R plasmids ; 2 4 those stra in s represented 0 . 6 % o f the origina l isolates . Skerman and Falkow in 1 9 6 9 studied the incidence of drug-res i stant fecal ï¿½ Â£Qli in members of a " pre-ant ibiotic" society in Austral i a and found that 1 6 % o f 2 4 7 ï¿½ QQli isolates were antimicrobial -res i stant and that less than 1% conta ined R factors . 3 0 , 1 1 0 I n a stored col lect ion of enterobacteri aceae from widely scattered parts o f the world ( Europe , the Middle East , and North America ) , origina l ly i solated between 19 17 and 1 9 5 4 , very l ittl e ant ib iotic resistance was found on suscepti b i l ity testing decades l ater . 4 5 Of 4 3 3 stra ins , 11 ( 2 . 5 % ) were res istant to ant ibiotics ( 9 to tetracycl ines , 2 to ampici l l in ) . No trans ferable ant ibiot ic-res istance plasmi d s were detected . None o f the 2 1 0 Salmonella spp . and none o f the 3 2 ï¿½ Â£Q!i stra ins in the study showed resistance t o any o f seven antibiot ics , to sul fonamides , or to trimethoprim . I n contrast , 2 4 % of stra ins were able to mob i l i z e a nonconj ugative plasmid in a rec ipient stra in , indicating the presence , in the host , of a conj ugat ive plasmid that was lacking ant imicrobia l -res i stance determinan ts . That preval ence is roughly comparabl e with that ( 17 % ) noted in J o g 6 fecal ï¿½ QQli stra ins isol ated from nonhospi ta l i z ed persons and that ( 3 3 % ) noted in 6 0 ï¿½ Â£Qli stra ins obta ined from J human , porc ine , and bovine sources . 1 1 Thus , it appear s that conj ugat ive plasmi ds were a s common among enterobacter iaceae before the introdu ction o f ant ibi otics a s they are i n the curren t antibi otic era . In contrast , under the select ive pressu re o f antibi otic use i n
29 humans and animal s , plasmid transfer and gene exchange have markedly expanded the population ( strains and species ) of res istant bacteria . Increases in Resistance After Wideapread Clinical Use of Antibiotics Experience over the last 4 decades of antimicrobial use in treatment and prophylaxis o f human infections indicates the profound effect of these practices on the preval ence of antimicrobial res istance in a variety of bacterial species . Experience t ime and again has emphasi z ed the importance of l imiting the use of antimicrobials to necessary and reasonable therapeutic indications and , indeed , even to holding some highly e ffective drugs in reserve . Be fore 1 9 4 6 , when penici l l in became general ly ava i l abl e , 8 5 % of cl inical isolates of Staphylococcus aureus at the Boston City Hospital were highly susceptible to penicil l in . 3 3 Within 3 years , most o f the strains there had become highly res istant to penicil l in . In most devel oped countries today , only about 1 0 % o f ï¿½ aureus isolates , whether of community or hospital origin , are susceptible to penicil l in . I solates of another gram-pos itive human bacterial pathogen , Streptococcus pneumoniae , col lected before 1 9 5 0 were tested years later and found t o b e uni formly susceptible to penicil l in . The situation is entirely d i fferent in Barcelona , Spa in , where recently about hal f o f a sma l l number of pneumococcal isolates were either moderately or highly res istant to penicil l in , and 6 2 % o f the same isolates were res istant to chloramphenicol . 6 0 This high degree of antibiotic res istance might wel l be related in part to widespread use o f those antibiotics , prevalent for many years in Spa in . 6 6 Problems of antimicrobial res istance i n bacterial species associated with widespread antimicrobial use in humans have been evident among gram-negative bac i l lary species notorious for caus ing infections in hosp itals . For example , whereas serratia strains isolated in the 1 9 5 0s were uni formly susceptible to kanamycin , nal idixic acid , and gentamicin and a minority of stra ins were res istant to streptomycin , a large proportion of strains were resistant to a l l but gentamicin by 1 9 6 7 . 3 3 Prophylactic use o f antimicrobial drugs , a s wel l a s use in therapeutic concentrations , might s imilarly exert a selective e f fect in humans , increasing the prevalence o f antimicrobial -resistant stra ins . The experience o n the Burn Service at Grady Hospital where prophyl act ic top ical use of gentamicin was extensive from 1 9 6 4 to 1 9 6 9 , offers a dramatic example . 1 0 8 I n 1 9 6 8 , more than 1 , 4 0 0 lb of that drug was used topica l ly on burns at that center . From 1 9 6 5 through
30 1 9 6 7 , 8 0-9 0 % o f a l l Pseudomonas aeruginosa i solates from patients in that burn unit were susceptible to gentamicin . By 1 9 6 9 , only nine percent were susceptible . Almost all the res istant isolates were almost a l l of a s ingl e pyocin type-Â a type that had in the past been seen only infrequently in the burn unit and had previously been susceptibl e to gentamicin . In mid- 1 9 6 9 routine top ical use of gentamicin was discontinued , and it was replaced with other topical drugs that would not have been used in systemic treatment of bacterial infections . In mid-19 7 0 , 9 5 % of ï¿½ aeruginosa isolates were susceptible to gentamicin--back to basel ine values that existed before the routine use of top ical gentamicin . 1 0 8 Effect of Reduced Use of Antibiot ics in Humans on Antimicrobial Besistance Quant itative aspects of overal l antimicrobial use in humans contribute to selection of res istant strains . Thus , rational choices of drugs for appropriate therapeutic and prophylactic indications are of continuing importance in treating humans . The same rational approach seems warranted in antimicrobial use for anima l s . I f one accepts the concept that populat ions of bacteria represent a common gene pool with cons iderable genetic fluidity and that some of the genetic information segregated for much of the t ime in individual components of the population are capable of be ing transferred among members of the population , then the effects of ant ibiotic use on one component ( e . g . , indigenous flora o f farm anima l s ) might b e relevant t o another component ( e . g . , commensal s and pathogens for humans ) . It is not known whether the current widespread preval ence of antimicrobial res istance genes among bacterial isolates from human and farm animal sources is too great to be reduced by any reasonable reduction o f overal l antimicrobial use : the issue warrants careful consideration . The extens ive reduction in gentamicin res istance in ï¿½ aeruginosa over a 1- to 2 -year period in the Grady Hospital Burn Unitl 08 cannot be general i z ed , because of the relatively short period o f use o f the drug and the confined location of its extens ive use . S ince the early 1 9 5 0 s , scattered studies have reported temporal relationships between decreased use of spec i fic antimicrobial agents in a given hosp ital or hospital ward and decreased preval ence o f nosocomial bacterial pathogens res i stant to those drugs . 7 8 However , caution must be exercised in interpreting such data , because many of the reported studies dealt with nosocomial outbreaks . Other epidemic-control measures that were instituted might have contributed as much as the changes in antibiotic use to the decreased prevalence of the res istant strain in some of the
31 outbreaks . Nonetheless , the data suggest the potential for reduction of antimicrobial res istance by l imiting the use o f some drugs . When antimicrobial use in humans has been decreased , usually in a hospital setting , res istant strains did not necessarily disappear quickly . Res istant bacteria can sometimes persist despite the absence of antimicrobial selective pressures . In a study o f 56 infants who were known to have been colonized with R-plasmid-containing kanamycinÂ res istant enterobacteriaceae during a stay in an intens iveÂ care nursery in which kanamycin was extens ively used , 8 9 % were sti l l colon i z ed with kanamycin-resistant strains 2 months after their return to the community . 2 0 Intestinal carriage o f such R-plasmid-containing kanamycin-resistant strains gradually decreased , but a fter 12 months or more , 4 6% o f the infants sti l l harbored the res istant organisms . EFFECT OF ELIMINATION OF SUBTHEBAPEQTI C AHTIBIQTIC USE ON LEVELS OF AHTIBIQTIC BESISTAHCE I N FARM ANIMALS What i s the counterpart in animal production o f the above-described circumstance of the withdrawal or shi ft o f antibiotic use in humans , and what changes have such shi fts made in the prevalence o f antibiotic-resistant col i form bacteria in their fecal bacterial population? Results o f long-term studies o f the responses of the intestinal col i form bacteria of swine to cessation o f antibiotic exposure suggest that changes occur slowly ( V . Hays , Univers ity of Kentucky 4 ' personal communication , 1 9 8 8 ; also see Langlois et a l . S 6 , 5 ) . In a separated herd o f p igs ( see additional detai l below ) mainta ined o n subtherapeutic concentrations o f tetracycl ine ( 50 - 1 0 0 grams/ton of feed ) for 13 years ( s ince 1 9 7 2 ) , tetracycl ine res istance averaged over 9 0% in the fecal col i form population . Such a high level o f res i stance to tetracycl ine , almost exclusively R-plasmid-mediated in enterobacteriaceae in this setting , was accompan ied by res istance to one or more other antibiotics . Does that high level o f res istance indicate such extens ive permeation o f R plasmids and transposons throughout the intestinal population o f ï¿½ QQli and other related col i form species so extens ive as to stab i l i z e their predominance and preclude diminution o f their maj or pos ition even i f exposure t o antimicrob ials i n feed ceased? O n the bas is o f experience with the use and withdrawal of specific antibiotics in relat ively closed populations of hospital ized patients , one might expect a protracted period to elapse before res istance returned to control levels in a comparable popul ation that had not been exposed to antibiotics . The e ffects o f antibiotic withdrawal on antimicrob ial res istance in intestinal col i form organ isms in the swine herd
32 mentioned above has been extensively documented over a 1 3 - year period ( V . Hays , 1 9 8 8 , personal communication ) . Thi s herd o f p igs , establ ished i n 1 9 6 3 , that received ant ib iotics routinely as feed additives and in inj ectable form ( when needed for treating sows and p igs ) , no s ingl e ant ibiotic was used continuously . After 1 9 7 2 , the herd was kept free of any antibacterial agents as feed additives or for therapy . The l evel o f tetracycl ine res istance among fecal col i form bacteria was over 9 0 % in 1 9 7 2 . The level o f resistance subsequently decl ined , but very s l owly . E ight years a fter cessation of exposure to the tetracycl ines and other antimicrobial s , the l evel of tetracycl ine res istance was sti l l 5 7 % ; by 19 8 5 , it had dropped to 3 0% . Those results certainly indicate that an early postwithdrawal " snapshot" should not be the basis for evaluation o f the e ffects of discontinuation o f antibiotic use . Why did reversion to l ower levels o f ant ibiotic resistance in the col i form flora take so l ong in the experiments in swine j ust described? First , some stra ins ( 0- serotypes ) o f ï¿½ QQli appear better constituted than others to persist and multiply in the colon , perhaps by virtue o f speci fic plasmid carriage and particular surface antigens . 4 2 I n p iglets , the presence o f the plasmid K8 8 confers adhesive properties on some stra ins of ï¿½ ï¿½ ' faci l itating attachment to brush borders o f intestinal cel l s . 4 9 S imilarly , i n calves a comparabl e pl asmid , K9 9 , i s important 1 in enteric disease . 1 3 In a situat ion where over 9 0 % o f col i form organisms start out with R-plasmid-mediated res istance to the tetracycl ines ( or other antimicrobial s ) , stra ins with a selective advantage , such as the presence o f coloni zation factors o r nutrient sequestrat ion , are preferential ly retained . Thus , once R plasmids and transposons are extens ively establ ished in the col i form flora , they can be retained by selective advantages other than those provided by antibiotic use . Second , persi stence o f R-plasmid-containing res istant stra ins might be fac i l itated by environmenta l factors . In the p igs that showed only very slow loss o f antib ioticÂ res istant intestinal col i form bacteria over 1 3 years , the isolat ion o f the experimental anima l s from other anima l s could have p layed a role in the slowness o f repl acement o f such bacteria . The quarters o f the animal s were undoubtedly fouled initia l ly with their own excreta , which contained res istant col i form bacteria . Re introduction o f the same bacterial flora from a constantly soiled environment must have occurred repeatedly over the years . Exposure o f the animals to untreated animals whose intestinal flora contains col i form bacteria that colonize efficiently but are suscept ible to ant ibiotics , might be required i f the ant ibiotic-res istant stra ins are to be "diluted out " more rap idly .
33 Chickens fed subtherapeutic doses o f oxytetracycl ine rapidly began to excrete an intestinal population in which over 9 0 % o f col i form bacteria were tetracycl ine-res istant : despite frequent cleaning of the cages , the chickens continued to excrete high concentrations of tetracycl ineÂ res istant organisms . 64 The resistance determinants might have been plasmids that had establ ished stable relationships with the host bacteria . Insofar as conj ugative plasmids help to mediate chromosomal gene recombination and genetic transfer among members of a population , population fitness , rather than individual bacterial fitness might be enhanced by these plasmids . 6 2 The proportion of resistant col i form bacteria decreased only when the chickens were mixed with other chickens excreting antibiotic-susceptible col i form bacteria or when they were moved to different cages . Plasmids might acquire multiple res istance determinants when the host bacteria are exposed to a s ingle drug in an animal ' s intestinal tract . The long-term use of tylos in , a macrol ide antibiotic , as a feed additive in p igs resulted in the evolution of an intest inal streptococcal population that was multiple-drug-resistant . l 4 Such multiple-drug-resistant plasmids appeared to differ from streptococcal plasmids present in control antibiotic- free p igs only in the presence of added res istance determinants . The acquis ition by conj ugative plasmids of individual or mult iple res istance transposons from sma l ler , nonconj ugat ive plasmids under the influence of antibiotic exposure might faci l itate subsequent dissemination of the res istance determinants . Whether pers istence of ant ib iot ic res istance in intestinal col i form bacteria a fter cessation of ant ib iotic exposure indicates the presence of res istance determinants cannot be known , but such pers istence could be important in determining the results of cessation . I f determinants are present , whether they are on conj ugative or non-conj ugative plasmids might affect the ult imate rate of spread o f res istance . Although the level o f res istance t o tetracycl ine rema ined relatively high a fter discontinuation of the tetracycl ines as feed additives in Great Brita in , there was an observable decl ine . 1 1 1 The prevalence of antimicrobial res istance t o one or more drugs in human isolates of salmonellae is 16-3 1 % ( 8 - 2 2 % res i stance to tetracycl ine , 5 - 1 9 % to amp ic i l l in ) ( see Chapter V) , and prevalence of res istance to ampicill in and tetracycl ine in human isolates of ï¿½ Â£Qli is 2 3 -3 2 % and 2 5 - 2 9 % , respectively . 3 These data suggest that the dispers ion o f res istance genes might not yet have gone so far as to be irrevers ible .
34 SPECIFIC MECHANISMS OF BES ISTAHCE TO ABTIMICBOBIAL DRUGS IN PATHOGENS FROM AHIMAL SQURCES RATIONALE AND EFFICACY OF MOLECULAR GENETIC TECJINIOUES IN RELATING HUMAN BACTERIAL ISOLATES TO FARM QRIGINS A p ivotal issue concerning the potential spread of animal -borne pathogens to humans is the abil ity to determine that a given pathogen is " the same" in both l ocations . Because bacteria divide asexually , each parent gives rise to two daughter cel l s that , assuming no maj or acquis ition or loss of genetic material ( e . g . , an R plasmid) , would be genetica l ly and biochemically identical with the parent . The daughter cel l s are "clonal " with respect to each other and to the parent . By definit ion , clonal bacteria have the same recent origin and are therefore identical or nearly ident ica l genetical ly and biochemical ly . Clonal ity cannot be absolutely ensured in the field or in the laboratory , but its existence can be shown with statistical analytic procedures . In thi s analysis , two bacteria are labeled operational ly clonal i f the test comparison shows so many s imilarities that the probab i l ity that the bacteria examined are d i fferent approaches zero . The demonstration of clonal ity , with some degree o f statistical certa inty , i s necessary but not sufficient for proving that a given pathogen has been transferred from an animal source to humans ( either through the food cha in or by some other means ) . Also required is evidence that the pathogen in question is not normally a part of the human flora . With cons istently pathogenic bacteria , that confounding factor , which would otherwise present substant ial background noise , i s not usual ly a problem . Modern epidemiologic techniques for invest igating human outbreaks of infectious disease usua l ly make it possible to determine whether a pathogen has entered the human environment from an outs ide source . The same techniques can be appl ied to each prior source in the l inear chain of contagion , so that , assuming the rarity of a given pathogen in a given setting , the primary source can be determined with reasonable certa inty . The power of the ep idemiologic approach rests squarely on the abi l ity to prove clonal ity . A number of recent deta iled reviews have examined the various methods ava i lable and have analyzed the ir resolving strength . l , 3 2 , 4 1 , 1 0 7 , 1 3 1 Other reviews have examined the evolutionary stabil ity of bacterial chromosomes and pl asmids , with the greatest attent ion focused on E Â· QQli and Salmonella spp . 1 2 , 4 3 , 8 2 , 1 0 0 , 1 0 5 , 1 17 , 1 2 1 , 12 5 , 12 6 The consensus i s that , for large class i f ication proj ects involving hundreds of stra ins of a given species , s imple , inexpens ive techniques , such as isoenzyme analys is , are capable of yielding adequate
35 discrimination o f strains . l This biochemical approach has been useful in fol lowing the evolution and dissemination o f clones o f a s ingle species , such as B Â· QQ11 . 1 0 7 Many of the s imple biochemical approaches are not useful for identi fication , because their discriminatory power is insufficient . When sma l l numbers of isolates are to be identi fied the use of mol ecular genetic techniques to " fingerprint" plasmids with DNA-probe technology 1 2 2 or restriction- fragment-length polymorphism is extremely useful for exactly the reasons that make these techniques worthless for the large numbers of isolates from population studies gathered over several years . Most plasmids--especial ly R plasmids , because of their genetic plastic ity--are unl ikely to be invariant over time . Moreover , because of their conj ugative nature , plasmids are spread horizonta l ly between strains ( and species ) of different clonal origin . Those e f fects are minimal in short-term outbreaks , so identity of plasmid profiles , particularly in the background of substantial plasmid divers ity , is strong evidence of clonal ity . In some instances , the use of molecular genetic techniques to analyze plasmids has achieved great certa inty of stra in ident i fication . 8 , 4 3 , 1 00 , 1 0 4 , 1 17 In s ituations in which either the bacteria have no plasmids or the plasmids are particularly stable ( e . g . , those needed to preserve stra in virulence ) , the use of cloned , random chromosomal sequences as probes to identi fy clones o f Salmonella spp . has been valuable in research settings . 12 5 R PLA$MIDS OF VEGETABLE QRIGIN AND ANIMAL ORIGIN AND THE EMERGENCE OF DRUG BESISTAHCE In evaluations of the origin of drug-res istant bacteria in humans , investigators have analyzed fecal bacterial isolates from meat-eaters and of vegetarians and have compared the flora for res istance . The frequency of drugÂ res istant bacteria in vegetarians i s at least as high as , i f not higher than , that i n omnivores . No studies have used the molecular epidemiologic techniques described above to demonstrate the origin of the drug-resistant bacteria in vegetarians , but cross-contamination of vegetables which 1 have been shown to carry high numbers of bacteria , 8 , 9 9 presumably occurs in the environment . On the basis of the preceding discussion , increased drug res istance in the environment may be said to reflect the presence of increased amounts of substances with ant ibacterial activity , probably from any source .
36 EFFECTS OF BESISTAHCE TO ABTIKICBOBIAL DRUGS ON BACTERIAL VIRULENCE Two mechanisms enable bacteria to change from a drug-susceptible to a drug-res istant phenotype : mutation of chromosomal genes or acquisition of new genetic information , typically in the form of R plasmids . The former often involves the alteration of a target of the antibiotic ( e . g . , the S 1 2 ribosomal protein and streptomycin or the RNA polymerase and ri fampin ) . The antibiotic target i s usually required to perform a vital vegetative function o f the cel l , so the e f fect of the mutation must be l imited speci f ical ly to the antibiotic-binding domain . Occas iona l ly , point mutations have pleiotropic effects and can subtly or profoundly alter the abil ity of the organisms to exist in their environment . Insofar as those vegetative functions are required for the infective stage of the pathogen , the virulence of the pathogen wil l natural ly be attenuated . Perhaps in part for that reason , nearly a l l drug res istance in bacteria , particularly in pathogenic spec ies , is due to R-plasmid acquis ition , as reviewed above . Most drug-resistance genes act either by inactivating an antibiotic or by excluding it from its target s ite . 2 2 , 2 3 , 2 9 , 8 7 As opposed to the mutation of chromosomal genes , the acquis ition of new genes ( on the R plasmid ) might be expected to have l ittle i f any effect on bacterial virulence . Presumably , no vegetative function has been a ltered . Any pleiotropic consequences to the bacterial cel l might be expected to be l imited to the effect of the sma l l additional burden of repl icating new genes and synthes i z ing new gene products . To balance that theoretical deleterious e f fect on cel l physiology , it should be pointed out that R plasmids can carry genes in addition to those essential for drug res istance . Thus , the selection of R plasmids that also contain toxin genes or other virulence factors would indirectly select for the acqyisition of new virulence factor s . 2 5 , 2 8 , J l , J 7 , 3 9 , 4 0 , 7 0 , 7 1 , 7 3 , 7 7 , 8 5 , 1 3 0 Two approaches have been used to analyze the effects of R-plasmid acquisition on the virulence of a bacterial stra in . Some invest igators have begun with a strain known to be virulent in an experimental infection model , genetica l ly trans ferred a given R plasmid into the stra in , and then compared the parent and the recipient for virulence in the model . 5 , 1 0 , 19 , 5 0 , 54 , 8 5 , 1 12 , 1 2 3 Others have surveyed isolates in the field ( e . g . , a hosp ital setting ) , determined which stra ins were drug-suscept ible and which drug-resistant , and then evaluated the amount of disease that the two classes of bacterial strains were capable of producing . 4 4 , 4 7 studies comparing bacterial stra ins that differ only in whether they contain an R plasmid have genera l ly reveal ed l ittle difference in virulence , although a few exceptions are
37 notable . Two early reports of the e ffects of an R plasmid on ï¿½. typhimurium1 1 2 , 1 2 3 showed decreased virulence in the strain that had acquired the R plasmid by conj ugative transfer . The molecular basis of those results was not pursued ; but they might be explained by more recent studies that have shown many cl inical isolates to be rel atively poor recipients in conj ugat ive crosses . The effic iency of mating , though , can be overcome with a mutation in the smooth l ipopolysaccharide ( LPS ) , which concomitantly reduces bacterial virulence . 7 2 A s ingle mutation is sufficient for that effect , so a single back-mutation to the virulent wild type readily appears in an appropriate setting , such as the animal gastro intestinal tract . The consequences of R-plasmid acquisition can be dramatic , in contrast , when the R plasmid contains , in addition to drug-res istance determinants and RTFï¿½ a virulence gene , such as enterotoxin 2 5 , 3 7 , 4 0 or hemolys in , l ï¿½ O or when the R plasmid is capable of mobil i z ing 8 5 or recombining with 7 7 a virulence-encoding plasmid . Thus , in special cases in which R plasmids are l inked with virulence genes , selection by antibacterial agents might promote spread of virulent strains . For example , aerobactin plasmids in ï¿½ QQli stra ins isolated are commonly associated with res istance to one or more antimicrobials , such as the tetracycl ines and ampici l l in . 4 8 The broad epidemiologic studies suggest that R plasmids do not interfere with bacterial pathogenicity and might s ignificantly increase the severity of disease . The best examples include cases of plasmid coinheritance of drug-res istance factors and various enterotoxins , as in cl inical isolates of Staphylococcus aureus7 3 and B Â· coli . S , 2 5 , 3 7 , 6 9 Even without speci fic evidence of a coinherited virulence factor , it is clear that drug-res istant bacteria are culprits in maj or outbreaks of disease . 4 7 A recent review examined 1 7 5 publ ished and unpubl i shed reports evaluating the effects of res istance to spec i fic antibiotics on the outcome of bacterial infections in both community and nosocomial settings . 4 4 Regardless of the setting , the drug-res istant bacteria , compared with drugÂ susceptible bacteria , were found to be associated with i l lness that was s ign i ficantly more severe ; in particular , the mortality , the l ikel ihood of hospital i zation , and the length o f hosp ital i z ation were at least twice as great . Further analys is determined that the underlying causes of the worse outcome in hospital i zed patients infected with a variety of drug-resistant bacteria were twofold : drug res istance led to a high incidence of antibiotic fa ilure , and drug-resistant pathogens emerged in the superinfection that fol l owed prior antib iotic treatment for a different disease . In summa ry , R pl asmids have been shown repeatedly to increase the virulence of bacteria , both when spec i fic mechanisms can be elucidated 2 8 and when they cannot . 4 4 The
38 few experimental s ituations in which reduced virulence a fter R-plasmid acquisit ion has been demonstrated are unl ikely to re flect the importance of selective pressures in nature . Any environmental factors that select for spread of R plasmids in bacterial populations ( such as use of antimicrobial drugs ) are l ikely to coselect for spread of coinherited virulence factors . THE SPBEAD OF ANtiMICRQBIAL-BESISTAHCE GENES THROUGH BACTERIAL POPULATIONS IN STAGES Bacteria l ive in the world not in pure culture , but as mixed stra ins and species competing in ecosystems , many of which are carried by animals or humans . 1 2 9 Bacteria isol ated from humans a hal f-century ago had res istance plasmids , but not the res istance genes found in comparable isolates today . 2 1 The use of each new antibacterial agent since than has commonly led eventual ly to the emergence and spread on plasmids of genes that encode res istance to the agent . 9 3 The recent appearance of new res istance genes and of older ones in pathogens that had been free of them indicates that the process of emergence and spread is continuing . 2 6 , 8 0 , 8 3 , 8 6 We review here evidence that the use of antimicrobial agents advances the emergence and spread of antimicrobialÂ res istance genes through bacterial populations in discrete stages triggered by speci fic events than are increas ingly ident i f iable . A previously unknown res istance gene can emerge by becoming mob i l i zed from an obscure stra in or by evolving from an ancestral gene . Use of an ant imicrobial agent could mob i l i ze a res istance gene from an obscure stra in by selecting for the stra in ' s overgrowth and thus increasing its contact with other stra ins . Use would s imilarly favor evolut ion by selecting for increas ingly res istant mutants of the ancestral gene . Mobil i zation and evolution would each be l ikely to progress through stages , and some res istance genes might progress through stages of both . 3 6 , 1 2 7 Broad-spect rum P-lactamases able t o hydrolyze newer P Â lactam ant ibiotics , for example , have recently emerged apparently by two-step mutations of 8 -lactamases ( SHV- 1 or TEM-2 ) that had emerged earl ier . 5 2 , 1 1 6 Use of the older pÂ l actams had presumably mob i l i zed the older res istance genes to stages of suf ficient prevalence for their very rare mutants that were res istant to the new P-lactams to occur often for selection by use of the new agents and thus the beginn ing of the new stage . Ant ibacterial agents can affect bacterial populat ions directly only by inhibiting susceptibl e stra ins . A bacterial ecosystem so depleted by an ant imicrobial agent can be
39 repopulated by the progeny o f a s ingl e bacterium that has a gene that encodes res istance to the agent . overgrowth of a clone in the face of such selection can generate overnight a b i l l ionfold ampl i fication of the number of copies of the res istance gene . That ampl i fication might be transitory , however , because the isogeneic bacteria that carry these cop ies might prove less fit for the diverse niches of the ecosystem than would their heterogenic predecessors or successors . SIGNIFICANCE OF PLA8MIPS Once inserted on a plasmid , a resistance gene can repl icate and pers ist in the niches of a l l the strains to which the pl asmid can be transferred . Recombination by various mechanisms , such as transposition or s ite-spec i fic recombination , moveover , can move the res i stance gene to other plasmids capabl e of trans fer to additional strains and thus to additional niches , beyond the host range of the first plasmid . 6 , 7 5 LINKAGE AHD DISSEKINATION OF BESISTAHCE GENES Recombination of a res istance gene with additional plasmids and cores idence of the recombinant plasmids with additional plasmids and chromosomes in new stra ins would a l so tend to l ink or associate the res istance gene with other genes in the presence of different selection . Such other genes would include other res istance genes in the presence of selection by other antimicrobials , other adaptive genes on chromosomes , and genes that were maintained on plasmids and had unknown survival values before antibacterials began to be used . 2 1 , 1 0 9 Selection for any of these would preserve or ampl i fy the res istance gene in the absence of the agents to which it encodes res istance . Spread of plaque and cholera across continents showed long ago that the world ' s bacterial ecosystems interconnect . 3 8 Accordingly , dissemination of an emerged resistance gene through them would await only sufficient amp l i f ication , recombination and trans fer of the gene to get it into intercoloniz ing stra ins . Studies of res i stance-gene phenotypes and more recently of the ir nucleotide sequences are , in fact , reveal ing few examples of parochial res istance genes confined to one area , and those examples may prove to have been premature . 4 6 , 7 9 In recent years , moreover , new technology has made poss ible the detection , not j ust of dist inctive res istance genes , but of distinctive plasmids carrying them that have
40 become widely distributed among bacterial genera , animal and human hosts , and widely separated geographic areas . Two small multicopy plasmids that encode res istance to sul fonamide and streptomycin and that share homology over hal f their length were found in escherichia and salmonel lae from animals and humans in many parts of the world . 4 , 1 2 8 Three distinctive multiple-res istant plasmids were found in col lections of salmonella isolates from both anima l s and humans in the United States ; one of them was endemic in cattle in 2 0 states and sporadic in humans in at least two other states . 9 5 A plasmid in an isolate from Venezuela was found also to carry one of the earl iest gentamicin res istance genes in seven genera of enterobacteriaceae from e ight widely separated medical centers in the United States . 9 4 A plasmid that encodes res istance to trimethoprim and other agents was in isolates of ï¿½ QQli from p igs on a farm and from patients in a medical center 4 0 miles away . 1 3 It has become possible t o recognize identical o r closely rel ated speci fic transposons in plasmids from bacteria isolated in different parts of the world . 1 0 1 , 1 1 8 One large ( 2 0-kilobase ) transposon , Tn 2 1 , has been found in plasmids from some of the earl iest multiple-res istance shigel lae isol ated in Japan and in a remarkable number of d i fferentÂ looking plasmids from other genera in d i fferent parts of the world . They include nearly a l l of the varied plasmids that first brought gentamicin res istance to 2 0 medical centers in Germany and a number of plasmids from various locations that carry d i fferent P-lactamase genes . 6 1 , 1 0 6 A l ineage has been proposed for the evolution of T n 2 1 and its g rogress ive acquis ition o f di fferent res istance genes . 1 2 Tn 2 1 appears to have spec i fic recomb inat ion s ites that permit exchange of new res istance genes between its progeny . 1 1 I f Tn 2 1 continues to be found in additional res istance plasmids , it wi l l ra ise the possib i l i ty that resistance might have progressed in the world quite d i fferently if it had not evolved . SPBEAD OF RESISTANCE GENES IN STAGES The preceding observations indicate that res istance genes emerge and spread in success ive stages . 9 0 The initial mob i l i zation or evolution of the gene , insertion into a plasmid , trans fer to strains within the plasmid host range , recombination with other plasmids , and insertion into other transposons and consequent l inkage with other genes in the presence of d i fferent antibiotic selection , as wel l as the paral lel evolution of the plasmids and transposons themse lves ( e . g . , Tn 2 1 ) toward fitness in more niches , are all discrete steps in the dissemination of the gene .
41 Each o f the stages sketched above can be seen to have the qual ities that characteri ze the steps in an enzymatic cascade : amp l i f ication and irrevers ibi l ity . At any stage , a given amount of ant ibacterial use would select more cop ies of the res istance gene than it would have at the preceding stage . S imilarly , the use needed to ma inta in a new stage would be less than the threshold amount needed to init iate 1t ' â¢ 76 Each new stage begins with a mutation , a chance encounter of rare strains in vast populations , a conj ugat ive trans fer , or a recombinational event . Each such event has a low probabil ity of occurrence , and the ultimate prevalence o f a res istance gene depends o n the product of the probab i l ities o f the events that inaugurate each of its stages of spread . We know , for example , that recombination of DNA segments that encode the TEM- 1 P-lactamase with plasmids res ident in stra ins of Hemophilus infl uenzae and Neisseria gonorrheae was the final step in the disastrous emergence of res istance to penici l l ins in these two pathogens 15 years ago . We can only quess , however , at the cha in of preceding events that required 3 0 years of penici l l in use before those final stages could occur . 9 , 1 3 3 A res istance gene i s an essential participant in each of the improbable events cited above , there fore , the chance that any of them will occur is a function of the preval ence of the res istance gene in the preceding stage . EFFECT OF ANTIMICRQBIAL USE ON SPBEAD OF RESISTANCE GENES Because use of antimicrobial agents can greatly ampl i fy the prevalence of genes that encode res istance to them at each stage , such use can be seen as the ma in force driving the progress ion of res istance genes through their stages o f spread . I f res istant bacteria arose only through frequent s ingle-stage mutation of susceptible stra ins , as was once thought the genesis of a res istant stra in isol ated from any patient might be due entirely to antimicrobial use in that patient or a neighbor . I f , however , the res istance in each such i solate is encoded on a complex genetic element that has been assembled in sequent ial stages , each triggered by a rare event , the l ineage of that res istance might be traced back to almost anywhere . Antimicrob ial use by the patient might have produced overgrowth and mani festation of the res istant stra in in that patient , but the evolution of the stra in ' s res istance genome and its spread to the pat ient required much greater use elsewhere . The emerg ing connection between the molecular evolution of the genomes that carry a res istance gene and its stage of
42 spread suqqests that even very l ittle interchanqe between bacterial populations miqht be sufficient for wide dispers ion throuqh them of a res istance qene that had been selected to an advanced staqe by intens ive antimicrobial use in the first population . The analoqy would be to a virus that evolves to hiqh contaqiousness on one continent and becomes epidemic on a second continent when taken there by a s inqle tourist . The qrowinq information reviewed above is enouqh to show that there is a qlobal epidemioloqy of antimicrobialÂ res istance qenes and qenomes , but not enouqh to measure them . A ranqe of model s can be considered . At one extreme , a model would propose that antimicrobial use in bacterial populations in one reqion would have no effect on l ater resistance-qene prevalence in distant bacterial populations . At the opposite extreme is a model that proposes that the prevalence of res istance qenes in a bacterial population is a function of total antibacterial use in a l l bacterial populations , however remote . Neither of those extreme model s appears to be appropriate . Al l the above evidence indicates some effect o f use in one population o n prevalence of res istance in others . For a more spec i fic example , we have qood reason to bel ieve that current prevalence of penici l l inase-produc inq Neisseria gonorrbeae ( PPNG ) in a number of American cities is a consequence of antibiotic use in areas of the Far East 1 5 years aqo . 2 9 There appears to be ( see Chapter V) qradients of preval ence o f resistance from hiqh to moderate to l ow as one moves from populat ions of intense to moderate to no use of antimicrobials . The critical question is : How much qreater is the prevalence of res istance ( and how much sooner is it attained ) in populations of moderate or no use than would be the case if there had been no ( or fewer) populations of intense use . IMPLICATIONS FOR ANTIMICROBIAL USE The model s indicate the potential effect of the use of ant imicrobial aqents in animal bacterial populations on the prevalence of res istance qenes in human bacterial populations . Nearly hal f the antimicrobial use in the United States is in animals . The pool of res istance qenes in animal flora in the United States may be est imated ( see Chapter V) to be 1 0 times that in the total human flora . Animal bacterial popul at ions are not distant from human bacterial popul ations , in as much as cont inuous samples of the animal populat ions flow on sl auqhtered carcasses throuqh food distribut ion cha ins to most households in the United States . Althouqh recent work with new technoloqy has elaborated many of the observat ions presented above , the qlobal ef fect
43 o f antibacterial use on bacterial resistance has l ong been grasped intuitively by workers studying res istance in the l aborato ï¿½ and by cl inicians coping with resistance in their patients . 9 , 1 3 4 Many have urged that antimicrobial use be minimi z ed whenever poss ible and have identi fied use o f antimicrobials in animal feed as the largest use category that could be reduced in the United states . 6 7 Their concerns and recommendations , however , were based on indirect evidence . The complexity of the processes outl ined above and the hugeness of the bacterial populations in which they operate had precluded reconstructing the entire chain of use that led to the presence of a particular res istance gene in the isolate from a particular patient . 68 DUAL ROI.ES OF SALMQNEI.I,AE AS PATBOGENS AND TBACERS The first direct evidence that antimicrobial use in animal s led to res istance in bacteria in humans came from an epidemic of multiple-res istant salmone l l ae among calves in Britain , apparently augmented by the antimicrobial use , that spread to infect humans . It became the occas ion to enact legislation banning routine addition of antibiotics to animal feed in Brita in and countries of the European Common Market ( see Chapter V) . 2 Some observers felt that the epidemic in Britain had been exceptional , ascribable to pecul iar husbandï¿½ practices there , and not pertinent for the United states . 5 1 When the National Research Counci l Committee to study the Human Health Effects of Subtherapeut ic Ant ibiotic Use in Animal Feeds met in 19 8 0 , it could sti l l be arqued that in the United States animal and human bacterial res istance genes were in separate pools that did not interchange . 8 8 The committee j udged that hazard to human health associated with the subtherapeutic use of ant imicrobials in animal feeds was neither proved nor d isproved . s a A report publ ished two years later found three d i fferent examples in which isolates of salmonellae from anima l s and humans in the United States carried res istance plasmids that had the same distinctive restrict ion endonuclease fragments . Several l ater studies used the same method to trace spread of res istant stra ins of salmonel lae from animals to humans in recognized food-borne outbreaks in the Un ited States ( see Chapter V) . Al l the direct evidence comes from salmone l l ae , because they are pecul iarly traceable . Relatively rare in human flora , they s ignal their presence by produc ing a conspicuous i l lness . Laboratories routinely find them with selective media even when they are sparse in stool flora and send them to reference l aboratories , where their serotypes are discriminated from more than a 1 , 0 0 0 poss ible types . More
44 than 40, 000 strains are winnowed out of human flora in the United States by that elaborate system each year , and their serotypes are recorded . A parallel system exists for salmonel lae isolates from animals . Althouqh the ir relative rarity in human flora helps to make salmonel l ae traceable , it also leaves them a very smal l part o f the antimicrobial -res istance problem . I t is probably reasonable to estimate that , of all the courses of antimicrobial therapy administered to humans in the United States , includinq al l the expens ive and somet imes toxic second and third qeneration aqents used to c ircumvent res istance to older aqents , much less than 1 % are directed aqa inst infection by salmonel lae . DUAL LINES OF EVIDENCE OF FEED-ADDITIVE HAZARD Two l ines of evidence of hazard associated with antimicrobial use in animal feed additives have been devel opinq in the last several years . One l ine o f evidence , as outl ined here , arises from the qrowinq understandinq that the res istance qenomes in bacterial isolates of humans are the products of extens ive evolutionary development that required l onq exposure of vase bacterial populat ions to antibacterial aqents . Where the populations miqht not be critical , qiven the qrowinq evidence of their interconnect ion . Which antibacterials were used miqht also not be critical , qiven the cl ose l inkaqes observed between res istance qenes and the ease with which some plasmids , once made preval ent by use of an antimicrobial , appear to acquire new res istance qenes by s ite-spec i f ic recombination . 9 7 The second l ine of evidence arises from epidemiol og ic observations , supported by molecular f inqerprint inq o f plasmids , of outbreaks and other surveys of salmone l l a infection in the United States i n recent years . They provide valuable examples for the first l ine of evidence , but a l so direct evidence of the ef fect of antibacterial use in animals on the outcome of spec ific types of human infect ion . The first l ine of evidence addresses a l l res istance in human bacterial fl ora , but sti l l cannot yield enouqh direct evidence for development of a quant itat ive risk assessment model . The second l ine of evidence addresses a subset o f human bacterial infections , but can yield direct evidence for a risk assessment model , as developed in Chapter VI I . SUMMARY OF THE BIOLOGIC IMPACT OF DRUG RES ISTANCE Althouqh present at low incidence in the preant ibiotic era , druq res istance has burqeoned s ince the wide use of
45 ant imicrobial drugs began 4 decades ago . Most drug res istance is due to the presence of transposable genet ic elements , cal led transposons , that are found on R plasmids , which are usual ly conj ugative . Because of the flexib i l ity of this mechanism of genetic express ion , bacteria can acquire res istance to a given antimicrobial agent quite eas ily and at minimal cost in terms of growth rate and general hardiness . Because of the abi l ity of R plasmids to add additional transposons in modular fashion , the conj ugative acquis ition of an R plasmid typical ly confers drug resi stance phenotypes - in addition to that spec i f ical ly selected by the presence of antibiotic in the environment . R factors make a direct contribution to the abi l ity of bacteria to cause disease . Unfortunately , however , R factors occas ional ly carry virulence factors as wel l . The mere presence of the drug-resistance phenotype has been shown , in epidemiologic surveys , to increase severity of i l l ness . The mechanism of the latter is twofold : antimicrobial res istance leads to delays in the selection of an appropriate therapeutic agent and permits the different ial outgrowth of the res istant organism during antimicrobial therapy or some other infection or presumed infection . With greater understanding of the molecular biology of drug res istance , R-pl asmid passage and strain dissemination have become better understood as wel l . R plasmids are genetical ly plastic and in many cases are transmissible to a wide variety of stra ins and species . Therefore , the outbreak of disease due to a clona l ly expanded strain can often be fol l owed with great precis ion by fol lowing its plasmid molecular profile , as deduced from restriction fragmentÂ length polymorphism . With this sophisticated technology , RÂ plasmid-containing stra ins of salmone l l a have been traced unequivocal ly from the farm through the food cha in to clusters of human patients . Outbreaks have provided strong evidence of the deleterious role that ant imicrobial use in the environment has pl ayed in the selection and propagation of antimicrobial res i stant bacterial pathogens of humans . BEFEBENCES 1. Achtman , M . , and G . Pluschke . Cl onal analys is o f descent and virul ence among selected Escherichia Â£Qli . Ann . Rev . Microbial . 4 0 : 1 8 5 - 2 1 0 , 19 8 6 . 2. Anderson , E . s . Drug resistance in Salmone l l a typhimurium and its impl icat ions . Br . Med . J . 3 : 3 3 3 , 1968 .
46 2a . Aserkoff , B . , and J . B . Bennett . Effect on antibiot ic therapy in acute salmonellosis on the fecal excretion of salmonel lae . N . Engl . J . Med . 2 8 1 ( 12 ) : 6 3 6-64 0 , 1 9 6 9 . 3. Atkinson , B . A . , and v . Lorian . Antimicrobial agent susceptibil ity patterns o f bacteria in hosp itals from 1 9 7 1 to 1 9 8 2 . J . Clin . Microbial . 2 0 : 7 9 1-7 9 6 , 198 4 . 4. Bagdasarian , M . R . , R . Lurz , B . Ruckert , F . c . H . Frankl in , M . M . Bagdasarian , J . Frey , and K . N . Timmis . Spec i fic-purpose plasmid cloning vectors . I I . Broad host range , high copy number RSF1010-derived vectors , and a host-vector system for gene cl oning in Pseudomonas . Gene 16 : 2 3 7 , 19 8 1 . 5. Balows , A . An overview of recent experiences with plasmid-mediated antibiotic res istance or induced virulence in bacterial diseases . J . Antimicrob . Chemother . 3 : 3 - 6 , 1 9 7 7 . 6. Bennett , P . , J . Heritage , A . Perry , J . Harvey , and c. Zagagl ia . Transposition and plasmid evolution : variations on a theme , p . 3 5 5 . In S . B . Levy and R . P . Novick , Eds . Antibiotic Res istance Genes : Ecol ogy , Trans fer , and Express ion . Banbury Report # 2 4 . Cold Spring Harbor , New York : Cold Spring Harbor Laboratory , 19 8 6 . 7. Bennett , P . M . , and M . H . Richmond . Plasmids and their influence on bacterial evolution . In L . N . Ornston and J . R . Sokatch , Eds . The Bacteria . Vol VI . New York : Academic Press , 1 9 7 8 . a. Brunner , F . , et al . The plasmid pattern as an epidemiologic tool for Salmonella typhimurium epidemics : comparison with the lysotype . J . Infect . Dis . 14 8 : 7 - 1 1 , 1983 . 9. Brunton , J . L . , D . Cl are , and M . A . Meier . Molecular epidemiology of antibiotic res istance plasmids of Haemophi lus spec ies and Ne isseria gonorrhoeae . Rev . Infect . Dis . 8 : 7 1 3 , 19 8 6 . 10 . Bryan , L . E . , A . J . Godfrey , and T . Schollardt . Virulence o f Pseudomonas aeruainosa stra ins with mechanisms of microbial pers istence for P-lactam and aminoglycos ide antibiotics in a mouse infection model . Can . J . Microbial . 3 1 ( 4 ) : 3 7 7 - 3 8 0 , 19 8 5 .
47 11 . Cameron , F . H . , et al . Nucleotide sequence o f the AAD ( 2 " ) aminoglycos ide adenyltransferase determinant aadB . Evolut ionary relationship of this region with those surrounding aadA in R53 8 - 1 and dhfri i in R3 8 8 . Nuc . Acids Res . 14 : 8 6 2 5 , 19 8 6 . 12 . Campbel l , A . Evolutionary s ign i ficance o f accessory DNA elements in bacteria . Ann . Rev . Microbiol . 3 5 : 5 5 -8 3 , 19 8 1 . 13 . Campbel l , I . G . , B . J . Mee , and s. M . Nikolett i . Evolution and spread of IncFIV plasmids conferring res istance to trimethoprim . Antimicrob . Agents Chemother . 2 9 : 8 07 , 19 8 6 . 14 . Christie , P . J . , and G . M . Dunny . Antibiot ic selection pressure resulting in multiple antibiotic res istance and l ocal i z ation of res istance determinants to conj ugative plasmids in streptococci . J . Infect . Dis . 14 9 : 7 4 - 8 2 , 1984 . 15 . Clewel l , D . B . , and c. Gawron Burke . Conj ugative transposons and the dissemination of ant ibiotic resistance in streptococc i . Ann . Rev . Microbial . 4 0 : 6 3 5 - 6 5 9 , 19 8 6 . 16 . Cohen , s. Transposable genetic elements and plasmid evolut ion . Nature 2 6 3 : 7 3 1-7 3 8 , 1 9 7 6 . 17 . Corpet , D . E . Ant ibiotic res idues and drug res istance in human intestinal flora . Antimicrob . Agents Chemother . 3 1 : 5 8 7 - 59 3 , 1 9 8 7 . 18 . Corpet , D . E . Antibiotic res istance from food . N . Engl . J . Med . 3 18 : 1 2 0 6- 1 2 0 7 , 1 9 8 8 . 19 . cutler , R . R . Relationship between antibiotic res istance , the production of "virulence factors " and virulence for experimental animals in Staphylococcus aureus . J . Med . Microbial . 12 : 5 5 - 6 2 , 1 9 7 8 . 20. Damato , J . J . , D . v . Eitzman , and H . Baer . Pers istence and disseminat ion in the community of R- factors o f nosocomial origin . J . Infect . Dis . 12 9 : 2 0 5-2 0 9 , 1 9 7 4 . 21. Datta , N . , and V . Hughes . Conj ugative plasmids in bacteria of the " pre-ant ibiotic" era . Nature 3 02 : 7 2 5 , 1983 . 22 . Davies , J . General mechanisms of antimicrobial res istance . Rev . Infect . Dis . 1 : 2 3 - 2 9 , 1 9 7 9 .
48 23 . Dav ies , J . , and D . L . Smith . Plasmid-determined res istance to antimicrobial agents . Ann . Rev . Microbial . 3 2 : 4 69 -5 1 8 , 1 9 7 8 . 24 . Davis , c . E . , and J . Anandan . The evolution o f R factors : a study of preantibiotic community in Borneo . N . Engl . J . Med . 2 8 2 : 1 1 7 - 12 2 , 19 7 0 . 25. de Lopez , A . , s. Kadis , and E . B . Shotts , Jr . Trans fer of drug-res istance and enterotoxin product ion in porc ine Escherichia ï¿½ strains and relationship between K8 8 antigen and raffinose (mel itose ) fermentat ion . Am . J . Vet . Res . 4 3 : 4 9 9 - 5 0 1 , 1 9 8 2 . 26 . Dickgiesser , N . , P . M . Bennett , and M . H . Richmond . Penic i l l inase-producing Neisseria gonorrhoeae : a mol ecular comparison of 5 . 3 . -kb-1actamase plasmid . J . Bacterial . 152 : 1 17 1 , 1 9 8 2 . 27 . E isenstein , B . I . Bacterial variation and antibiotic action . Adv . Intern . Med . 2 6 : 3 9 3 -4 2 6 , 19 8 0 . 28 . Elwel l , L . P . , and P . L . Shipley . Plasmid-mediated factors associated with virulence of bacteria to animals . Ann . Rev . Microbial . 3 4 : 4 6 5-4 9 6 , 1 9 8 0 . 29 . Falkow , s. Infectious Multiple Drug Res istance . London , England : Pion Ltd . , 197 5 . 30 . Falkow , s. The prevalence and ecology of R- factors , p . 70 . In s . Falkow , Ed . Infectious Multiple Drug Res istance . London , England : Pion Ltd . , 1 9 7 5 . 31. Falkow , s. , and D . Portnoy . Bacterial plasmids-an overview . Cl in . Invest ig . Med . 6 : 2 07 - 2 12 , 1 9 8 3 . 32 . Farrar , w . E . , Jr . Molecular analys is o f plasmids in epidemiologic investigation . J . Infec . Dis . 1 4 8 : 1- 6 , 1983 . 33 . Finland , M . Changing patterns of susceptibi l ity o f common bacterial pathogens t o ant imicrob ial agents . Ann . Intern . Med . 7 6 : 1 0 0 9 - 1 0 3 6 , 1 9 7 2 . 34 . Finland , M . Emergence of antibiotic res istance in hospitals , 1 9 3 5 - 1 97 5 . Rev . Infect . Dis . 1 : 4 -2 1 , 1 9 7 9 . 35. Gardner , P . , D . H . Smith , H . Beer , and R . c. Moel lering , Jr . Recovery o f res istance (R) factor from a drug- free community . Lancet 2 : 7 7 4 -7 7 6 , 1 9 6 9 .
49 36. Ghuysen , J . M . Bacterial active-s ite i n penic i l l inÂ interactive proteins and domains : mechanism , structure , and evolut ion . Rev . Infect . Dis . 1 0 : 7 2 6 , 1 9 8 8 . 37 . Gowal , D . , s . Saxena , M . Mago , and L . Rao Bhau . Plasmid mediated enterotoxin production and drug res istance amongst Eschericbia 2211 from cases o f infantile diarrhea . Ind . J . Pad . 5 2 : 5 7 - 5 9 , 1 9 8 5 . 38 . Greenough , w . B . Gram-negative bacil l i . Virbio cholerae , p . 12 08 . In G. L . Mandel l , R. G. Douglas , Jr . , and J . E . Bennett , Eds . Principles and Practice of Infectious Diseases . Second Edition . New York : John Wiley and Sons , 1 9 8 5 . 39 . Hamood , A . N . , R . D . Subletr , and c . D . Parker . Plasmid-mediated changes in virulence of vibrio cholerae . Infect . Immun . 5 2 : 4 7 6-4 8 3 , 19 8 6 . 40 . Harnett , N . M . and c . L . Gyles . Linkage o f genes for heat-stable enterotoxin , drug res istance , K9 9 antigen , and col icin in bovine and porc ine stra ins o f enterotoxigenic Escherichia col i . Am . J . Vet . Res . 4 6 : 4 2 8 -4 3 3 , 1 9 8 5 . 41. Hartl , D . L . , and D . E . Dykhuizen . The population genetics of Esqberichia 2211 Â· Ann . Rev . Genet . 18 : 3 1- 6 8 , 1 9 8 4 . 42 . Hart l ey , c . L . , H . M . Clements , and K . B . Linton . Escberich ia 2211 in the faecal flora o f man . J . Appl . Bact . 4 3 : 2 6 1-2 6 9 , 1 9 7 7 . 43 . Holmberg , s . D . , et al . Drug-resistant Salmonel l a from animals fed antimicrobial s . N . Engl . J . Mad . 311 : 617-62 2 , 1984 . 44 . Holmberg , s . D . , s. L . Solomon , and P . A . Blake . Health and economic impacts of antimicrobial res istance . Rev . Infect . Dis . 9 : 1 0 6 5 - 1 07 8 , 1 9 8 7 . 45. Hughes , v . M . , and N . Datta . Conj ugat ive plasmids in bacteria of the "pre-antibiotic" era . Nature ( London ) 3 02 : 7 2 5 -7 2 6 , 1 9 8 3 . 46 . Hummel , R . , H . Tschape , and w . Whitte . Spread of Plasmid-mediated nourseothricin res istance due to ant ibiotic use in animal husbandry . J . Basic Microbiol . 2 6 : 4 6 1-4 6 6 , 1 9 8 6 .
50 47 . John , J . F . , Jr . , K . T . McKee , Jr . , J . A . Twitty , and w. Schaffner . Molecular epidemiology of sequential nursery epidemics caused by multires istant Klebsiella pnegmoniae . J . Ped . 1 0 2 : 8 2 5-8 3 0 , 1 9 8 3 . 48 . Johnson , J . R . , et al . Aerobactin and other virulence factor genes among strains of Eschericbia ï¿½ caus ing urosepsis : Association with patient characteristics . Inf . Immun . 5 6 : 4 0 5-4 12 , 1 9 8 8 . 49 . Jones , G . w . , and J . M . Rutter . Role of the K8 8 antigen in the pathogenes i s of neonatal diarrhea caused by Escherichia ï¿½ in piglets . Infect . Immun . 6 : 9 18 , 1972 . so . Kinsman , o . s . , J . Naidoo , and w . c . Noble . Some effects of plasmids coding for ant ibiotic res istance on the virulence of Staphylococcus aureus . Br . J . Exp . Pathol . 6 6 ( 3 ) : 3 2 5-3 3 2 , 198 5 . 51 . Kiser , J . s . A perspective on the use of ant ibiotics in animal feeds . J . Anim . Sci . 4 2 : 1 0 5 8 , 1 9 7 6 . 52 . Kl iebe , c . , et al . Evolution of plasmid-coded res istance to broad-spectrum cephalosporins . Antimicrob . Agents Chemother . 2 8 : 3 0 2 , 19 8 5 . 53 . Kotarski , s . F . , T . L . Merriwether , G . T . Tkalcevic , and P . Gemski . Genetic studies of kanamycin res istance in Campylobacter ieiuni . Antimicrob . Agents Chemother . . 3 0 : 2 2 5 -2 3 0 , 1 9 8 6 . 54 . Kras inski , K . M . Virulence versus res istance . Bul l . N . Y . Acad . Med . 6 3 : 2 3 7 -2 5 2 , 1 9 8 7 . 55 . Lacey , R . w . 19 7 5 . Antibiotic res istance plasmids of Staphylococcus aureus and their c l inical importance . Bacteriol . Rev . 3 9 : 1-3 2 , 1 9 7 5 . 56 . Langol is , B . E . , G . L . Cromwe l l , and v . w . Hays . Influence of chlortetracycl ine in swine feed on reproductive performance and on incidence and pers istence of ant ibiotic-resistance enteric bacteria . J . Anim . Sci . 4 6 ( 5 ) : 1 3 6 9 - 1 3 8 2 , 1 9 7 8 . 57 . Langlois , B . E . , G . L . Cromwe l l , and V . w . Hays . Influence of type of ant ibiotic and length of antibiotic feeding period on performance and pers istence o f antibiotic-res istant enteric bacteria i n growing finishing swine . J . Anim . Sci . 4 6 ( 5 ) : 1 3 8 3 - 1 3 9 6 , 1 9 7 8 .
51 58 . Langlois , B . E . , et al . Antibiotic resistance i n pigs fol l owing a 1 3 year ban . J . Anim . Sci . 6 2 ( Suppl . 3 ) : 1 8 -3 2 , 19 8 6 . 59 . Langlois , B . E . , et al . Antibiotic res istance o f fecal col i forms from swine fed subtherapeutic and therapeutic level s of chl ortetracycl ine . J . Anim . Sci . 58 : 6 6 6 - 67 3 , 1984 . 60 . Latorre , c . , T . Juncosa , and I . Sanfel iu . Antibiotic resistance and serotypes of Streptococcus pneumoniae strains isolated in a chi ldren ' s hospital in Barcelona , Spain . Antimicrob . Agents Chemother . 2 8 : 3 57 - 3 5 9 , 1 9 8 5 . 61 . Levesque , R . c . , and G . A . Jacoby . Molecular structure and interrelationships of multires istance P-lactamase transposons . Plasmid 19 : 2 1 , 1 9 8 8 . 62 . Levin , B . R . The maintenance of plasmids and transposons in natural populations of bacteria , pp . 5 7 - 70 . In s . B . Levy and R . P . Novick , Eds . Antibiotic Res istance Genes : Ecology , Trans fer , and Expression . Cold Spring Harbor , New York : Cold Spring Harbor Laboratory , 19 8 6 . 63 . Levy , s. B . Emergence of antibiotic-res istant bacteria in the intestinal flora o f farm inhabitants . J . Infect . Dis . 1 3 7 : 6 8 8 -6 9 0 , 1 9 7 8 . 64 . Levy , s. B . , G . B . Fitzgerald , and A . B . Macone . E ffect o f an ant ibiotic supplemented feed on the ecology of .1..t.. 2211 on a farm , pp . 4 3 - 6 1 . In J . Drews and G . Hoqenauer , Eds . Topics in Infectious Diseases . II . R Factors : Their properties and Poss ible Control . Berl in : Springer-Verlag , 1 9 7 7 . 65 . Lewis , M . J . Trans ferable drug res istance and other trans ferable agents in strains of Escherichia Q2li from two human populat ions . Lancet 1 : 1 3 8 9 - 1 3 9 3 , 1 9 6 8 . 66 . Linares , J . , J . Garau , c . Dominquez , and J . L . Perez . Antibiotic res istance and serotypes of Streptococcus pneumon iae from patients with community-acquired pneumococca l disease . Antimicrob . Agents Chemother . 2 3 : 5 4 5-54 7 , 1 9 8 3 . 67 . Linton , A . H . Antibiotic-res istant bacteria in animal husbandry . Br . Med . Bul l . 4 0 : 9 1 , 1 9 8 4 .
52 68 . Linton , A . H . Antibiotics , animals and man-an appraisal o f a contentious subj ect , p . 2 6 . I n M. Woodbine , Ed . Antibiotics and Ant ibios is in Aqriculture . Proceedinqs of the 9th Easter School in Aqricultural Sciences , Nottinqham ( Enqland ) Univers ity . London , Enqland : Butterworth , 1 9 6 2 . 69 . Liven , E . Relationships between production o f enterotoxin and various druq res istance patterns i n strains of Escberichia Q2li isolated from p iqlets sufferinq from col ienterotoxaemia . Acta Vet . Scand . 2 0 : 3 9 6 -4 0 3 , 1 9 7 9 . 70 . Luria , s . E . , and M . Delbruck . Mutations o f Bacteria from Viruses Sens itive to Virus Res istance . Genetics 2 8 : 4 9 1- 5 1 1 , 19 4 3 . 71. Macrina , F . L . Molecular cloninq of bacterial antiqens and virulence determinants . Ann . Rev . Microbiol . 3 8 : 19 3 -2 19 , 1 9 8 4 . 72 . Manninq , P . and M . Achtman . Cel l -to-cel l interactions in conj uqatinq Escherichia QQli : The involvement of the cel l envel ope , pp . 4 0 9 -4 4 7 . In M . Inouye , Ed . Bacterial Outer Membranes . New York : Wi ley , 1 9 7 9 . 73 . Mauff , G . J . , I . Rohriq , u . Ernzer , w . Lenz , M . Berqdol l , and G . Pulverer . Enterotoxiqenicity o f StaphYlococcus aureus stra ins from cl inical isolates . Eur . J . Clin . Microbiol . 2 : 3 2 1- 3 2 6 , 1 9 8 3 . 74 . Mare , I . J . Incidence of R factors amonq qram neqative bacteria in druq- free human and animal communities . Nature ( London ) 2 2 0 : 1 0 4 6 - 1 04 7 , 1 9 6 8 . 75 . Marinnez , E . , and F . de la Cruz . Transposon Tn2 1 encodes a RecA- independent s ite-speci fic inteqration system . Mol . Gen . Genet . 2 2 1 : 3 2 0 , 1 9 8 8 . 76 . Mayer , K . H . , et al . Molecular evolution , spec ies distribution and cl inical consequences o f an endemic aminoqlycos ide res istance plasmid . Antimicrob . Aqents Chemother . 2 9 : 62 8 , 1 9 8 6 . 77 . McConne l l , M . M . , G . A . Will shaw , H . R . Smith , S . M . Scotland , and B . Rowe . Transpos ition o f ampici l l in res istance to an enterotoxin plasmid in an Escherichia Q2l1 stra in of human oriqin . J . Bacteriol . 1 3 9 : 3 4 6 - 3 5 5 , 1979 .
53 78 . McGowan , J . E . , Jr . Antimicrobial resistance i n hospital organisms and its relation t o antibiotic use . Rev . Infect . Dis . 5 : 1 0 3 3 -1 04 8 , 1 9 8 3 . 79 . Medeiros , A . A . , R . Levesque , G . A . Jacoby . An animal source for the ROB-1 ï¿½ -lactamase of Haemophi lus influenzae type b . Antimicrob . Agents Chemother . 2 9 : 2 12 , 19 8 6 . 80 . Medeiros , A . A . , and T . F . O ' Brien . Ampici l l in resistant Haemophi lus influenzae type B possessing a TEM-type ï¿½-lactamase but l ittle permeabil ity barrier to ampic i l l in . Lancet 1 : 7 1 6 , 197 5 . 81. Mederski-samoraj , B . D . , and B . E . Murray . High-level res istance to gentamicin in cl inical isolates of enterococci . J . Infect . Dis . 1 4 7 : 7 5 1 -757 , 19 8 3 . 82 . Mercer , A . A . , et al . Conservation o f plasmids among Escherichia Â£Qli o f diverse origins . Infect . Immun . 4 6 : 649-657 , 1984 . 83 . Meyer , J . F . , B . A . Nies , and B . Wiedemann . Amikacin res istance mediated by multires istant transposon Tn2 4 2 4 . J . Bacterial . 1 5 5 : 7 57 , 1 9 8 3 . 84 . Michel , J . , J . Mi l l er , and T . Sacks . E f fect o f an R plasmid on the virulence of a hospital strain of Escherichia QQli . Chemotherapy 2 6 : 4 2 7 -4 3 0 , 1 9 8 0 . 85 . Murray , B . , D . Evans , M . Penaranda , and D . Evans . CFA/ I -ST plasmids : Comparison of enterotoxigenic Escherichia Â£Qli ( ETEC ) o f seroq roups 0 2 5 , 0 6 3 , 0 7 8 , and 0 1 2 8 and mob i l i z ation from an R factor containing epidemic ( ETEC ) isolate . J . Bacterial . 1 5 3 : 5 6 6-57 0 , 1983 . 86 . Murray , B . E . , B . Mederski-Samaroj . Trans ferable ï¿½Â lactamase : a new mechanism for in vitro penici l l in res istance in streptococcus faeca l i s . J . Cl in . Invest . 72 : 1168 , 1983 . 87 . Murray , B . E . , and R . c . Moel lering , Jr . Patterns and mechanisms o f antibiotic resistance . Med . Cl in . North Am . 6 2 : 8 9 9 - 9 2 3 , 1 9 7 8 . 88 . National Research Council . Committee on The E ffects on Human Health o f Subtherapeutic Use o f Antimicrobials in Animal Feeds . E ffects on Human Health of Subtherapeutic Use of Antimicrobials in Animal Feeds . Washington , D . C . : National Academy Press , 19 8 0 .
54 89 . Neu , H . c . , c . E . Cherubin , E . D . Lonqo , B . Flouton , and Â· J. Winter . Antimicrobial res istance and R- factor trans fer amonq isolates o f Salmonella in northeastern United States : A comparison of human and animal i solates . J. Infect . Dis . 1 3 2 : 6 1 7 - 62 2 , 19 7 5 . 90 . Nies , B . A . , et al . R17 67 , an example of the evolution o f res istance plasmids . Plasmid 1 3 : 1 6 3 , 1 9 8 5 . 91 . O ' Brien , T . F . , J . F . Acar , A . A . Mede iros , R . A . Norton , F . Goldste in , and R . L . Kent . Internat ional comparison o f prevalence of res istance to ant ibiot ics . J. Am . Med . Assoc . 2 3 9 : 1 5 1 8 - 1 5 2 3 , 1 9 7 8 . 92 . O ' Brien , T . F . , R . A . Norton , R . L . Kent , and A . A . Medeiros . International surveil lance o f preval ence of ant ibiotic res istance . J. Antimicrob . Chemother . 3 ( Suppl . C ) : 5 9 - 6 6 , 1 9 7 7 . 93 . O ' Brien , T . F . and the Members of Task Force 2 . Res istance of bacteria to ant ibacterial aqents : Report of Task Force 2 . Rev . Infect . Dis . 9 ( Suppl . 3 ) : S 2 4 4 , 1987 . 94 . O ' Brien , T . F . , et al . Intercont inenta l spread of a new ant ibiotic res istance qene on an epidemic plasmid . Science 2 3 0 : 8 7 , 19 8 5 . 95 . O ' Brien , T . F . , et al . Molecular epidemiol oqy o f antibiotic res istance i n sa lmonella from anima l s and humans in the United States . N . Enql . J . Med . 3 0 7 : 1 , 1982 . 96 . Odel son , D . A . , J . L . Rasmussen , c . J. Smith , and F . L . Macrina . Extrachromosoma l systems and qene transmiss ion in anaerobic bacteria . Plasmid 17 : 8 7 - 109 , 1 9 8 7 . 97 . Ouelette , M . , L . Bissonnette , and P . H . Roy . Precise insertion of antibiotic res istance determinants into tn2 1-l ike transposons : nucleotide sequence of the OAX- 1 P-l actamase qene . Proc . Natl . Acad . Sci . 8 4 : 7 3 7 8 , 1 9 8 7 . 98 . Pratt , w . B . and R . Fekety . The Antimicrob ial Druqs . New York : Oxford Univers ity Press , 19 8 6 . 99 . Reminqton , J . s. , and s. c . Schimpff . Please don ' t eat the salads . N . Enql . J . Med . 3 0 4 : 4 3 3 -4 3 5 , 19 8 1 .
55 1 0 0 . Ri ley , L . w . e t al . Evaluation of isolated cases of salmone l l osis by plasmid profile analysis : Introduction and transmission of a bacterial clone by precooked roast bee f . J . Infect . Dis . 1 4 8 : 1 2 - 1 6 , 19 8 3 . 1 0 1 . Rol l ins , L . D . , L . N . Lee , and D . J . LeBlanc . Evidence for a disseminated erythromycin res istance determinant mediated by Tn9 1 7 - l ike sequences amonq qroup D Streptococci i solated from piqs , chickens , and humans . Antimicrob . Aqents Chemother . 2 7 : 4 3 9 , 1 9 8 5 . 1 0 2 . Saqara , H . , A . Mochi zuki , N . Okamura , and R . Nakaya . Antimicrobial res istance of Campyl9bacter 1e1uni and Campyl9bacter ï¿½ with special re ference to plasmid pro f i l es of Japanese cl in ical isolates . Antimicrob . Aqents Chemother . 3 1 : 7 1 3 -7 1 9 , 1 9 8 7 . 1 0 3 . Saunders , J . R . Transposable res istance qenes . Nature 2 5 8 : 3 8 4 , 19 7 5 . 1 0 4 . Schaberq , D . R . , L . s. Tompkins , and s. Falkow . Use of aqarose qel electrophores is of plasmid deoxyribonucleic ac id to finqerprint qramneqat ive bacteria . J . Cl in . Microb . 1 3 : 1 1 0 5 - 1 1 08 , 19 8 1 . 1 0 5 . Schaberq , D . R . , and M . J . Zervos . Interqeneric and interspec ies qene exchanqe in qram-pos itive cocci . Antimicrob . Aqents Chemother . 3 0 : 8 17 -8 2 2 , 19 8 6 . 1 0 6 . Schmidt , F . , and I . Kl opfer-Kaul . Evolut ionary relationship between TN2 1-l ike elements and pBP2 0 1 , a pl asmid from Kl ebs iella pneumoniae mediatinq resistance to qentamicin and e iqht other druqs . Mol . Gen . Genet . 1 9 7 : 1 09 , 1 9 8 4 . 1 0 7 . Selander , R . K . , D . A . Cauqant , and T . s. Whittam . Genetic structure and variation in natural populations of Escherichia QQli , pp . 1 6 2 5 - 1 64 8 . In F . c . Neidhardt et al . , Eds . Escherichia ï¿½ and Salmonella typhimurium : Cel lular and Mol ecular Biol oqy . Washinqton , D . C . : American Society for Microbiol oqy , 1987 . 1 0 8 . Shulman , J . A . , P . M . Terry , and c . E . Houqh . Col on i z ation with qentamicin-res istant Ps eudomona s aeruqinosa , pyoc ine type 5 , in a burn unit . J . Infect . Dis . 1 2 4 : S 1 8 -S2 3 , 19 7 1 .
56 109 . S i lver , s . Mechanisms of plasmid-determined heavy metal res istances , p . 1 7 9 . In s . B. Levy , R. c . Clowes , and E . L . Koeniq , Eds . Mol ecular Biology , Pathoqenicity , and Ecology of Bacterial Plasmids . Internationa l Plasmid Conference on Molecular Biology , Pathoqenicity , and Ecology o f Bacterial Plasmid& , 19 8 1 , Santo Dominqo , Dominican Repub l ic . New York : Plenum Press , 1 9 8 1 . 1 1 0 . Skerman , F . J . , and s . Falkow . Unpubl ished data as cited by s . Falkow . In s . Falkow , Ed . Infectious Multiple Druq Res istance . London , Enqland : Pion , Ltd . , 1969 . 1 1 1 . Smith , H . W . Pers istence o f tetracycl ine res istance in piq L. Â£Qli . Nature 2 5 8 : 6 2 8 - 6 3 0 , 19 7 5 . 1 1 2 . Smith , H . w . The effect on virulence o f trans ferrinq R factors to Salmonella typhimurium in vivo . J . Med . Microbial . 5 : 4 5 1-4 5 8 , 1 9 7 2 . 1 1 3 . Smith , H . w . , and M . A . Linqqood . Trans fer factor in Escherichia Â£Qli with particular reqard to the ir inc idence in enteropathoqenic stra ins . J . Gen . Microbial . 6 2 : 2 8 7 - 2 9 9 , 19 7 0 . 1 1 4 . Smith , J . c . , s . M . Markowitz , and F . L . Macrina . Trans ferable tetracycl ine res istance in Clostridium difficile . Ant imicrob . Aqents Chemother . 19 : 9 9 7 - 1 0 0 3 , 19 8 1 . 1 1 5 . Spqaard , H . Incidence of druq res istance and transmiss ible R factors in stra ins of L. 2211 from faeces of healthy piqs . Acta Vet . Scan . 14 : 3 8 1 - 3 9 1 , 197 3 . 1 1 6 . souqakoff , w . , s . Goussary , G . Gerbaus , and P . Courval in . Plasmid-mediated res istance to thirdÂ qeneration cephalosporin& caused by point mutat ions in TEM-type peni c i l l inase qenes . Rev . Infect . Dis . 1 0 : 8 7 9 , 1988 . 1 1 7 . Sp ika , J . s , s . H . Waterman , G . w . Soo Hoo , et al . Chl oramphenicol -res istant Salmonella newport traced throuqh hamburqer to da iry farms . N . Enql . J . Med . 3 1 6 : 565-57 0 , 1987 . 1 1 8 . Steen , R . , and o . Skold . Plasmid-borne or chromosomal ly mediated res istance to Tn7 is the most common response to ub iquitous use of trimethoprim . Ant imicrob . Aqents Chemother . 2 7 : 9 3 3 , 19 8 5 .
57 1 19 . Ta l ly , F . P . , and M. H . Malamy . Res i stance factors in anaerobic bacteria . Scand . J . Infect . Dis . 4 9 ( Suppl ) : 5 6 - 6 3 , 19 8 6 . 1 2 0 . Tanaka , M . , T . Yamamoto , and T . Sawa i . Evolution o f complex res istance transposons from a n ancestra l mercury transposons . J . Bacterial . 1 5 3 : 1 4 3 2 , 1 9 8 3 . 1 2 1 . Taylor , D . N . , K . Wachsmuth , and Y . H . Shanghuan . Sa lmone l l a associated with marij uana : a multistate outbreak traced by plasmid fingerprinting . N . Engl . J . Mad . 3 0 6 : 12 4 9 - 12 5 3 , 1 9 8 2 . 1 2 2 . Tenover , F . C . Minireview : Studies of antimicrobial res istance genes us ing DNA probes . Ant imicrob . Agents Chemother . 2 9 : 7 2 1 - 7 2 5 , 19 8 6 . 1 2 3 . Th iel e , E . Virulence of R- factor-bearing Salmonella typhimuriurn . Infect . Immun . 2 : 5 1 6 - 5 1 8 , 19 7 0 . 1 2 4 . Timoney , J . F . R plasmids in pathogenic enterobacteriaceae from calves , pp . 5 4 7 - 5 5 5 . In s. B . Levy et a l . , Eds . Molecular Biology , Pathogenicity , and Ecology of Bacterial Plasmids . New York : Plenum , 19 8 1 . 1 2 5 . Tompkins , L . s. , et al . Cloned , random chromosoma l sequences as probes to identi fy Sa lmone l l a species . J . Infect . Dis . 1 5 4 : 1 5 6 - 1 6 2 , 19 8 6 . 1 2 6 . Tompkins , L . s. , et al . Molecular epidemiology of Legione l l a spec ies by restriction endonuclease and a l l oenzyme analys is . J . Cl in . Microb . 2 5 : 18 7 5 - 18 8 0 , 1987 . 1 2 7 . Tschape , H . , E . Tietze , R . Prager , w . Vo igt , E . Wolter , and G . Seltmann . Plasmid-borne streptothricin res istance in gram-negat ive bacteria . Plasmid 12 : 1 8 9 , 1984 . 1 2 8 . van Treeck , u. , F . Schmidt , and B . Wiedemann . Molecular nature of a streptomycin and sul fonamide res istance plasmid ( pBP1 ) prevalent in cl inical Escherichia col i stra ins and integration of an ampic i l l in res istance transposon ( TnA ) . Antimicrob . Agents Chemother . 19 : 3 7 1 , 19 8 1 .
58 1 2 9 . Veldkanp H . , H . van Gemerden , w. Harder , and H . J . Laanbroek . Microbial Competition . See Competition Amonq Bacteria : An overview , p . 2 7 9 . In M . J . Kluq and c . A . Reddy , Eds . current Perspectives in Microbial Ecoloqy . Washinqton , D . C . : American Society for Microbioloqy , 1 9 8 4 . 1 3 0 . Waalwij k , c . , J . Van Den Bosch , D . MacLaren , and J . de Graa f f . Hemolys in plasmid codinq for the virulence o f a nephropathoqenic ï¿½ 2211 stra in . Infect . Immun . 3 5 : 3 2 - 3 7 , 1982 . 1 3 1 . Wachsmuth , K . Genotypic approaches to the diaqnos is o f bacterial infections : plasmid analyses and qene probes . Infect . Control 6 : 1 0 0 - 1 0 9 , 19 8 5 . 1 3 2 . Wi l l iams , R . E . o . Chanqes in the virulence and antibiotic res istance of staphyl ococcus aureus , p . 9 9 . In M . Finland , w . Marqet , and K . Bartmann , Eds . Bacterial Infect ions : Chanqes in the ir causat ive Aqents , Trends , and Poss ible Bas is . New York : Spri nqer verlaq , - 19 7 1 . 1 3 3 . Yeunq , K . H . , et al . A novel 4 . 9 -kilobase plasmid associated with an outbreak of peni c i l l inase-producinq Neisseria qonorrhoeae . Conc ise Communic . p . 1 1 6 2 .
IV QUANTITATION OF ANTIBACTERIAL AGENTS USED IN LIVESTOCK AND POULTRY FEEDS Expos inq larqe numbers of bacteria to an antibacterial aqent at concentrations that inhibit qrowth is the most e f fective means of selectinq antimicrobial -resistant stra ins ( or spec ies , if the initial population comprises a mixture o f spec i es ) . The concentration o f the antimicrobial chemical to which bacteria are exposed determines both the number of res istant orqanisms that may be isolated initia lly and the maqnitude of res istance to that antimicrobial chemical . The durat ion of exposure to an antimicrob ial compound has an important role in the el imination of suscept ibl e stra ins or spec ies in mixed populations , e . q . , in the l ower qastrointestinal tracts of humans and anima l s . Ant ibacterials are used in l ivestock and poultry in several dosaqes : at hiqh ( therapeuti c ) concentrat ions to treat establ ished infectious diseases and at l ow ( subtherapeutic) concentrations for enhancement o f qrowth and for disease prevent ion ( prophylaxis ) . The Food and Druq Administration ( FDA) approves separately each l abel claim for e f ficacy for each dose and animal species . The Feed Addit ive Compendium summ a rizes a l l FDA-approved feed additives , label c l a ims , doses , animal types , and federa l requlations . It is estimated that approximately 7 5 % o f da iry calves , 6 0 % of bee f cattle , 7 5 % o f swine , and 8 0 % o f the poultry marketed have rece ived one or more ant imicrobial druqs in their feed at s ome t ime . 1 , 2 Twenty antimicrobial druqs have been approved by FDA as feed additives ( Table IV- 1 ) . Penic i l l in and chlortetracycl ine were approved in 1 9 5 1 and oxytetracycl ine in 1 9 5 3 . Nonantibiotic ant imicrob ial druqs , such as nitro furans and sul fonamides , are used in the same fashion . I n add ition , ionophores , such as monens in , which have l ittl e ant ibacterial act ivity , are extens ively used as cocc idiostats in the poultry industry . Combinat ions o f ant imicrob i a l s such as chlortetracycl ine and hyqromyc in or chl ortetracycl ine , penici l l in , and sul fathiazole--are approved as feed add itives . POPULATION OF LIVESTOCK AND POULTRY It is useful to know the populat ion of each spec ies o f animal for calcul atinq the amounts o f penic i l l in o r the 59