R E V I E W S 37 Pedersen,L.C., Benning,M.M. and Holden,H.M. (1995) Biochemistry 34, 13305-13311 38 Holm,L. and Sander,C. (1995) Trends Biochem. Sci. 20, 345-347 39 Piepersberg,W. et al. (1988)Actinomycetologica 2, 83-98 40 Pang, Y. et al. (1994)Antimicrob. Agents Chemother. 38, 1408-1412 41 Udou,T., Mizuguchi,Y. and Wallace,R.J.,Jr (1989)FEMS Microbiol. Lett. 57, 227-230 42 Afnsa,J.A. et al. (1996)Antimicrob. Agents Chemother. 40, 2350-2355 43 Hall, R.M. and Stokes,H.W. (1993) Genetica 90, 115-132 44 Recchia,G.D. and Hall, R.M. (1995)Microbiology 141, 3015-3027 45 Martin, C. et al. (1990)Nature 345,739-743 46 Webb,V. and Davies,J. (1994)Trends Biotecbnol. 12, 74-75 47 Miller,G.H. et al. (1995)J. Chemother. 7, 31-44 Beyond vancomycin: new therapies to meet the challenge of glycopeptide resistance Thalia I. Nicas, Michael L. Zeckel and Daniel K. Braun ram-positive bacteria are and teicoplanin will continue, The incidence of infections caused by a major cause of infecor even increase, in the near resistant Gram-positive pathogens is tious diseases world- increasing, while emergence of vancomycin future. wide. With the widespread use The emergence of glycoresistance is reducing the number of of antimicrobials and other en- therapeutic options. New agents are being peptide resistance has focused vironmental pressures, the inattention on the need for alterrapidly evaluated as candidates to replace cidence of infections caused natives to vancomycin. Vancovancomycin; some of the most promising by antibiotic-resistant Grammycin-resistant Enterococcus include semisynthetic glycopeptides, positive bacteria has rapidly quinupristin-dalfopristin, oxazolidinones faecium (VRE) has become a increased in recent years, espe- and everninomycins. Alternative strategies, major infectious diseases probcially among nosocomial inlem in many hospitals in the including immunization and therapeutic fections. Vancomycin has been vaccines, may also have a role. USA, especially among debilia potent antibiotic against tated and immunocompromised T.I. Nicas*, M.L. Zeckel and D.K. Braun are in the multiresistant Gram-positive hosts ~,2. A Centers for Disease Lilly Research Laboratories, Eli Lilly and Company, bacteria for many years and, Control and Prevention (CDC) Indianapolis, IN 46285, USA. until recently, all major Gramstudy in the USA showed that *tel: +1 317 276 4236, fax: +1 ,317 277 0778, positive pathogens have been overall frequency of vancoe-mail: nicas_thalia_i@lilly.com uniformly susceptible. Use of mycin resistance in enterococci vancomycin in developed countries has increased over rose from <0.03% in 1988 to 7.9% in 1993, with 13.8% the past decade in response to the emergence and in intensive care units 3. This trend is continuing. Conspread of highly antibiotic-resistant microorganisms, cerns over glycopeptide resistance are not restricted especially methicillin-resistant Staphylococcus aureus to the enterococci; teicoplanin-resistant coagulasenegative staphylococci are well-documented and (MRSA) and methicillin-resistant Staphylococcus epidermidis (MRSE). In the USA, vancomycin is the there are increasing numbers of reports of coagulaseonly antibiotic approved for use against MRSA, negative staphylococci with intermediate or greater resistance to vancomycin 4-s. Species include both whereas in Europe, teicoplanin, a compound of the same chemical class (the glycopeptide antibiotics), is Staphylococcus haemolyticus and S. epiderrnidis. available. Although vancomycin resistance has not yet been reFor treatment of life-threatening infections, such as ported among clinical isolates of S. aureus, vancomycinendocarditis or sepsis caused by MRSA, and serious in- resistant S. aureus has been obtained in vitro 9 by plasfections caused by MRSE and other methicillin-resistant mid transfer from enterococci. Many scientists and coagulase-negative staphylococci, especially those inphysicians fear that it is only a matter of time before volving prosthetic devices, vancomycin is still a powvancomycin-resistant Staphylococcus aureus appears erful weapon. The increasing frequency with which in patients. Natural and laboratory transfer to other MRSA, MRSE and related bacteria are encountered in genera has been reported 1°,1~, and recent reports 12a3 hospitals makes it likely that high usage of vancomycin of isolates of Streptococcus mitis and Streptococcus G Copyright © 1997 Elsevier Science Ltd. All rights reserved. 0966 842X/97/$17.00 240 V,. 5 No. Pll: S0966-842X(97)01051-2 6 JUNI:" 1 9 9 7 R E V I E W S bovis that are resistant to vancomycin and teicoplanin have confirmed the ominous potential for spread of resistance to other pathogens. Concerns about the spread of vancomycin-resistant microorganisms have led the HICPAC (Hospital Infection Control Practice Advisory Committee) of the CDC to issue recommendations curtailing the use of vancomycin in many of the clinical situations where it has been routinely used]4; for example, empirical treatment of febrile neutropenic patients and initial therapy for some Clostridium difficile-associated diseases (CDAD). Although VRE is thought to be most problematic in hospitals in the USA, it is likely that such difficult-totreat pathogens will eventually be seen in many countries. VRE was first isolated from patients in Europe ]5,'6 and has subsequently been found in healthy volunteers ~7and animal TM and environmental sources 19. Effective prevention and treatment of infections caused by such highly resistant microorganisms will require new strategies and treatment methods. Some of the approaches under consideration to limit the spread of vancomycin resistance have been recently reviewed 2°. Even if spread can be delayed, new therapies will be needed. Combination therapies are an active area of investigation, but this approach has not been promising. In this article, we will discuss some possibilities for new agents, which are being sought from a variety of sources: the glycopeptide class itself, improvements in other well-accepted classes and entirely new chemical entities. We have not attempted to include all current investigational drugs but have highlighted some recent promising approaches. Molecular basis of glycopeptide resistance The emergence of resistance to vancomycin was first documented in E. faecium and Enterococcus faecalis in 1988, more than 30 years after the introduction of vancomycin. Two major forms occur: VanA isolates are resistant to both vancomycin and teicoplanin, and VanB isolates are usually resistant to vancomycin only21. The two forms have a similar mechanism of resistance but are genetically distinct. VanB isolates typically possess a regulator system that recognizes vancomycin, but not teicoplanin, as an inducer. VanB isolates with constitutive expression are resistant to teicoplanin as well22. An unrelated form of glycopeptide resistance can occur in staphylococci, especially coagulase-negative species4-8, and such isolates are often resistant to teicoplanin while remaining susceptible to vancomycin23. This form of resistance is chromosomally mediated and does not appear to be transferable, and its biochemical mechanism is poorly understood. Acquired resistance to glycopeptides in the enterococci occurs by a complex and unusual mechanism 2°'22'24. VanA enterococci have acquired nine new genes on a transposon. The molecular basis of resistance is well understood: of the three essential genes, the vanA gene product is related to the ligase that forms D-Ala-D-Ala, the terminal portion of the pentapeptide on the peptidoglycan precursor; the VanA protein forms D-AIa-D-Lac. D-Lac is synthesized by reduction of pyruvate by the vanH gene product. A dipeptidase that de- T R I - : ~ D S IN M , ( : R ( ) B , O I . O ( ; Y (a) OH HO , , ~ HO., I I~r12 OH3 /CH2OH ~ ~'O CI H H "°'"T "c, H~ T -,,.o. v o II I "'. o jL H0" T H H H ' -OH ". ', ,,, "N" - H , ! H CH31"[ I, V I,,,,H ,,,:, ',, .0 Ac.N. T H 7". "11/ ,' O-; ~ -0 O H3C H L-Lys-D-Ala-D-Ala NH2 (b) 0 H AcHNxv~ OH3 "" H O 0 O H3C H L-Lys-D-Ala-D-Lac NH2 Fig. 1. Binding of model cell wall precursor fragments to vancomycin. Vancomycin and other glycopeptides act by binding the cell wall precusor, lipidlinked N-acetylglucosamine-N-muramyl-pentapeptide, at the D-Ala-D-Ala terminus. This interaction prevents the precursor from being added to the growing cell wall polymer. In the interactions illustrated (a) N-acetyl L-Lys-D-Ala-D-Ala represents the terminal amino acids of vancomycinsusceptible bacteria. (b) N-acetyl L-LyS-D-Ala-D-Lacrepresents the terminal residues of vancomycin-resistant enterococci. Hydrogen binding sites are indicated by dotted lines. Note that the change from an amide to an ester in L-LyS-D-Ala-D-Lacremoves one major site of hydrogen bonding. grades D-AIa-D-Ala is encoded by the vanX gene and is also essential for resistance. Essentially, the bacteria have re-engineered their cell wall, making a new precursor in which the terminal D-AIa is replaced with D-Lac. This alteration, resulting in an ester rather than amide linkage of the terminal moiety, allows the loss of one of the key hydrogen bonds in complex formation with vancomycin 2s (see Fig. 1 ). Binding is at least 1000fold weaker, resulting in resistance. Both VanA and VanB resistance are transferable. The VanA transposon has been found on many different plasmids 26. At least one lab has transferred VanAtype resistance into S. aureus 9. The implications of this form of resistance in staphylococci are truly frightening, as vancomycin is the only consistently effective therapeutic option in most of the world. This spread 241 w,,..5 N~). 6 .IUNI~: 1 9 9 7 R E V I E W S tives explored by the scientists of the Lepetit Research Center=8,3°. The Lilly compounds address the problem of vancomycin resistance, whereas the Lepetit compounds remedy the relatively poor activity of teicoplanin against coagulasenegative staphylococci and improve overall potency. R BI [ NH OH % OH VRE active glycopeptides Lilly's most active compounds are H~ I II H II H J.,,,H exemplified by LY333328 (Fig. 2, O ~ ' " N/Jl"l°¢ N :"~ N " . Table 1). This compound is a semi°" N/J y~ N ~ N'~TH NHCH3 synthetic glycopeptide derived NH ~ O ~ H O ~'~'..~..-" from the naturally occurring compound A82846B (LY264826). The H O ~ H2N/ L ~ parent compound has the same peptide backbone as vancomycin, but differs in that 4-epivancosamine replaces the vancosamine sugar, and an additional epivancosamine is present on the molecule. A side chain is attached by reductive alkylation to the amino A B X R group of the epivancosamine that Vancomycin H OH H H is linked to glucose29'31. The original rationale behind the direction of the Lilly series came NH2 from comparison of vancomycin LY264826 OH H HHC - ~ - ~ . . ~ H and teicoplanin. Teicoplanin posO--5 sesses an acyl-linked aliphatic CH3 ] moiety that is believed to impart NH2 interesting pharmacokinetic propLY333328 OH H H¢.~v--....~....~ erties, such as longer serum halfH life; it also shows greater potency CH3 I than vancomycin against many Gram-positive species in vitro 27. Fig. 2. Structuresof vancomycin,LY264826and its semisyntheticderivativeLY333328. When alkyl- and acyl-linked groups were added to vancomycin at an analogous site, the new compounds may be inevitable, and the possibility underlines the did indeed exhibit longer half-lifes. At the time, none urgent need for new agents. of these improvements were sufficient to justify further pursuit, and new naturally occurring glycopeptides, such as LY264826, were more interesting in terms of Novel glycol~l~tides as vancomycin replacements The glycopeptide antibiotic class would not seem to be their potency and spectrum. After the appearance a promising place to look for compounds active against of glycopeptide-resistant enterococci, Lilly scientists vancomycin-resistant enterococci because the resist- screened naturally occurring and vancomycin-derived ance mechanism in VanA enterococci alters the ability semisynthetic glycopeptides against clinical isolates. Surof glycopeptides to bind their target, affording cross prisingly, the only compounds active against vancoresistance to all naturally occurring glycopeptides. The mycin-resistant enterococci were the N-alkyl vancoglycopeptide molecule is not particularly amenable to mycins27. Side chain addition was then pursued on chemical modification and, until recently, most changes other glycopeptides. Parallel modifications made on had very little positive effect on activity27,28. However, LY264826 resulted in a dramatic improvement over recent efforts to modulate the activity of this group of vancomycin. The best of these early compounds included antibiotics by chemical modification are now yielding an N-p-chlorobenzyl side chain on the disaccharide impressive results. The most interesting developments amino-sugar of LY264826 (Ref. 27). As LY264826 are probably the extension of the spectrum of glyco- possesses three reactive amines, selective chemistry peptide antibiotics to include vancomycin-resistant had to be developed for the specific synthesis of the enterococci, as seen in certain N-alkyl derivatives ex- desired mono-substituted compounds. Further modiplored by Lilly scientists 27,29, and the enhanced anti- fication of the alkyl group, including the use of bistaphylococcal potency by certain carboxamide deriva- phenyl and chloro-biphenyl side chains 29,31, led to the O i i 242 VO, 5 NO. 6 JUNE 1 9 9 7 REVIEWS Table 1. Activity of vancomycin, LY264826 and its semisynthetic derivative LY333328 ",b Vancomycin Microorganism Number tested MRSA MRSE Teicoplanin-resistant LY264826 LY333328 MICgo MIC range MIC~0 MIC range MICgo MIC range 40 2 30 3 4 0.5-4 1-4 4-8 1 2 0.13-2 0.25-4 4-32 0.5 0.5 _<0.06-0.5 0.0013-0.5 0.25-1 14,13 2 1-4 1 0.5-1 0.25 0.016-0.25 10,13 1024 128->1024 256 16->256 1 0.25-2 13,10 256 8-1024 256 1->128 0.13 0.008-0.13 Staphylococcus haemolyticus Vancomycin-susceptible Enterococcus faecalis and Enterococcus faecium VanA E. faecalis and E. faecium VanB E. faecalis and E. faecium Streptococcus pneumoniae 5 0.015-0.063 0.13-0.25 0.004-0.008 aln vitro activity against multiple strains expressed as the minimum inhibitory concentration (MlCgo)and range of MIC in gg ml ~. Data taken from Ref. 31. bAbbreviations: MRSA, methicillin-resistant Staphylococcus aureus; MRSE, methicillin-resistant Staphylococcus epidermidis. synthesis of compounds with high potency against the vancomycin-resistant strains of both VanA and VanB enterococci phenotypes3L These compounds have the unusual property of bactericidal activity against enterococci [minimum inhibitory concentrations (MICs) for VanA enterococci are typically 0.5-1 btg ml-1], good activity against MRSA and coagulase-negative staphylococci (MICs of 0.06-0.5 Bgm1-1) and remarkable potency against streptococci (MICs of 0.002-0.016 gg ml-I ). The compounds are also active against some teicoplanin-resistant coagulase-negative staphylococci. The semisynthetic glycopeptides act at the same site in peptidoglycan synthesis as vancomycin and teicoplanin 32. Their superior activity against vancomycinresistant enterococci does not appear to be based on their enhanced binding to the target peptides, at least as measured by binding to acetyl-Lys-D-Ala-D-Ala and acetyl-Lys-D-Ala-D-lactate. A model for the activity of these glycopeptide compounds has been proposed by Dudley Williams 33,34. At the cell surface where glycopeptides act in vivo, additional intermolecular interactions that increase affinity may also be important. LY333328 and related compounds readily form homodimers, a function that may also cooperatively enhance binding of the precursor at the cell surface32,34. They also interact with membranes in a manner that has been proposed to reduce their motion and further facilitate interaction with the precursor32,33.This ability to dimerize and interact with the membrane probably enhances binding to the point that it could account for the activity observed against VRE. However, a direct effect on bacterial membrane function cannot be excluded and may contribute to the enhanced bactericidal activity of the compounds. As might be predicted from this mechanism of activity, it has not been possible to select mutants resistant to LY333328 in either VRE or MRSA studies in vitro (T.I. Nicas, pers. commun.). LY333328 is now in phase I clinical trials in the USA. TREN,) 243 Teicoplanin analogs Attempts to modify teicoplanin and its analogs have been primarily directed towards the improvement of their activity against coagulase-negative staphylococci and retention of their safety and pharmacokinetic profiles. Scientists at Lepetit Research Laboratories have undertaken extensive studies on modifications of teicoplanin; some of the most interesting compounds emerging are from amide modifications of the terminal carboxyP °,~s. Two examples of these compounds, MDL 63,246 and MDL 63,042, are amide derivatives of the teicoplanin-related natural glycopeptide A-40,926 with the sugar on amino acid 6 removed (Fig. 3, Table 2). Compared with teicoplanin, these compounds show marked improvement in their activity against coagulasenegative staphylococci and moderate activity against VanA enterococci (Table 2). Recent studies with MDL 63,246 have confirmed its broad Gram-positive activity and efficacy in animal models of septicemia and endocarditis 36, but enterococcal activity is probably insufficient to predict efficacy against VanA isolates. Non-glycopeptide agents Non-glycopeptide antibiotics represent additional candidates for vancomycin replacements (Table 3). With the exception of the oxazolidinones, these drugs represent modifications of existing classes of antibiotics that have become more interesting with the increased need for new agents against infection caused by Grampositive bacteria. For many agents, lack of uniform bactericidal activity may be an issue in the effective treatment of enterococcal infection. The streptogramin antibiotic quinupristin-dalfopristin appears to have gone furthest in clinical development. The many recent reviews highlighting quinupristin-dalfopristin 37-39attest to the current interest in this area. There are also several antibiotic candidates in earlier stages of development, which may fulfill an important therapeutic role. VO, 5; N ' . ) . 6 Jumt'; 1 9 9 7 REVIEWS HO HO X" \0 -/ 0 . R3/tt'" ",4. tinamycin is a streptogramin combination that has been used in Europe but not in the USA. Streptogramin antibiotics are used as combinations of streptogramin A and streptogramin B or analogs; the two components bind at different sites on the ribosome, acting synergistically to inhibit protein synthesis. Quinupristin-dalfopristin (Fig. 4) is a fixed combination (30:70 ratio) of derivatives of the naturally occurring streptogramins pristinamycin Ia and Ib, with chemical modifications that increase solubility and allow use as an injectable agent 4°. Quinupristin, the streptogramin B component, appears to bind to the L24 protein of the 70S ribosomal subunit, whereas dalfopristin, the streptogramin A component, binds proteins L10 and L l l of the 50S ribosomal subunit. These two sites of action are likely to account for the synergistic activity of quinupristin and dalfopristin. In contrast to either component alone, the combination is bactericidal against some isolates. C [I. ," Y I L,,,,. " ,o o R2N ~ NO HO" v / "O HO C•OH CH2OH OH OH X Y Z R1 R2 HO Teicoplanin CH2OH CI H H MDL63,246 COOH H CI CH 3 R3 CH2OH . o-&-'-L A ~ Oo. ~ 5 . 5 " H I / N OH H N'~- ~ In vitro activity The quinupristindalfopristin combination demonOH H 2 N ~ ~ ~ NH2 MDL 63,042 COOH H Cl C H 3 strates synergistic activity41against a variety of Gram-positive bacteria, Fig. 3. Structures of teicoplanin and the related semisynthetic compounds MDL 63,246 and MDL including S. aureus, Streptococcus 63,042. Commercial teicoplanin is a complex of five structurally related natural products that differ pneumoniae and E. faecium 37,38 in the fatty acid acylatingthe glucosamine sugar. The most abundant component is shown. (Table 4). The most notable property of quinupristin-dalfopristin Quinupristin-dalfopristin in vitro is its activity against vancomycin-resistant GramChemistry and mechanism of action Quinupristinpositive cocci, including VanA and VanB E. faecium, dalfopristin (RP59500) is a new injectable antibiotic of Leuconostoc spp., Lactobacillus spp. and Pediococcus 42. the streptogramin class now being developed by Rh6ne Unfortunately, although active against E. faecium Poulenc under the name Synercid. The oral agent prisstrains in vitro, quinupristin-dalfopristin only has limited H N H N Table 2. Activity of teicoplanin, MDL 63,246 and MDL 63,042 Teicoplanin MDL 63,246 MDI. 63,042 Microorganism Number tested MICgo MIC range MICgo MIC range MICgo MIC range Staphylococcus aureus Staphylococcus epidermidis Staphylococcus haemolyticus Enterococcus faecalis Enterococcus faecium E. faecium (VanB) E. faecium (VanA) Streptococcus pneumoniae 20 25 34 31 13 25 20 9 0.5 8 32 2 0.5 0.5 >128 0.06 0.25-0.5 0.25-32 0.25-64 0.06-0.5 0.06-0.5 0.125-0.5 0.25 0.25 1 1 0.5 0.5 32 <0.03 0.125-0.5 <0.03-0.5 0.125-2 0.125-1 0.06-1 0.125-0.5 4-64 0.125 0.25 0.5 0.25 0.25 0.5 16 <0.03 <0.03-0.125 <0.03-0.25 0.06-1 0.03-0.25 0.06-0.5 0.125-0.5 0.5-32 0.06-0.25 In vitro activity against multiple strains expressed as the minimum inhibitory concentration (MlCeo)and range of MIC in pg m1-1. Data taken from Refs 35,36. TR.N,, ,N M,(:.,,.,,,L,,(,Y 244 V,,, _5 N'.>. 6 JUNV: 1 9 9 7 REVIEWS Table 3. Comparison of new agents to replace vancomycin Compound Class Projected stage of development, 1997 LY333328 Glycopeptide Phase II MDL 63,246 MDL 63,042 Glycopeptide Preclinical Poor in vitro activity against most vancomycin-resistant Enterococcus faecium (VRE) Exceptional potency against coagulasenegative staphylococci Quinupristin-dalfopristin (Synercid) Streptogramin combination Phase III Poor activity against Early clinical experience with efficacy against VRE Limitations suggested by preclinical studies Outstanding features Most uniformly bactericidal of new agents Enterococcus faecalis, variable bactericidal activity, resistance development U-100592 Oxazolidinone Phase II Not bactericidal Oral agent, new chemical class SCH 27899 Everninomycin Phase II Not bactericidal New chemical class activity against E. faecalis (MICg0 of _>4gg m1-1). This may prove to be a limitation in clinical use because the majority of nosocomial enterococcal isolates are E. faecalis. Furthermore, superinfection with E. faecalis has been reported during quinupristin-dalfopristin treatment of vancomycin-resistant E. faecium 43. In one report [P. Linden et al. (1996) 36th Int. Conf. Antimicrob. Agents Chemother., New Orleans, LA, USA, Abstr. LM32], E. faecalis was identified in 15 of 68 patients receiving quinupristin-dalfopristin therapy. Although quinupristin-dalfopristin demonstrates a bactericidal effect against S. aureus and S. pneumoniae, it is only bacteriostatic against E. faecium; as many E. faecium strains are currently resistant to aminoglycosides, this absence of bactericidal activity may be a significant shortcoming in the treatment of enterococcal endocarditis. However, unlike the majority of other possible vancomycin replacements (and vancomycin itself) quinupristin-dalfopristin is also active in vitro against the important respiratory pathogens Moraxella catarrhalis, Legionella spp. and Mycoplasma pneumoniae; Haemophilus influenzae is only moderately susceptible (MICg0 of 4-8 pg m1-1)41. Resistance Resistance to streptogramins may occur by at least three different mechanisms. The most important mechanism is plasmid-mediated target modification that confers resistance to macrolides, lincosamides and streptogramin B (MLS) by methylation of their common ribosomal-binding site. As this mechanism does not confer resistance to dalfopristin or other streptogramin A compounds, quinupristin-dalfopristin is active against such resistant strains, even when this resistance is constitutively expressed. Resistance may also be mediated by drug-modifying enzymes or efflux; however, the frequency of these kinds of resistance is currently low. Resistance to quinupristin-dalfopristin has been selected in MRSA in vitro, but the mechanism involved in this example has not been elucidated 44. Although resistance to only one of the components of quinupristin-dalfopristin may not be apparent in in vitro susceptibility testing, such resistance may result in diminished efficacy, as was noted in one study of rabbit endocarditis caused by S. aureus resistant only to the quinupristin component 4s. In such isolates quinupristin-dalfopristin is generally not bactericidal. \ o N HN" ~Y D N I O~'-H ~.N~ 9 .JP-...S~A"-.../N N Quinupristin o ~ ' " o N / ' ~ ~ ~ "".../ i .H o 7 /-~N~ OH ",W o <o7 Dalfopristin Fig. 4. Structure of the streptogramins quinupristin and dalfopristin. R E V I E W S patients received concomitant agents with in vitro activity against the vancomycin-resistant enterococci and MICso (range from R e ~ (3) patients dying within the first 5 Microorganism several sudles) MIC range days of therapy were excluded from 37,38 efficacy determinations. Staphylococcus aureus 0.25-1 0.03-4 37,38 Overall, these results still suggest Coagulase-negative staphylococci 0.25-2 0.03-4 37,38 that quinupristin-dalfopristin may Streptococcus pneurnoniae 0.25-2 <0.125-2 Vancomycin-susceptible 1-4 0.25-8 38 prove a useful antibiotic, especially for Enterococcus faecium VRE infections caused by E. faecium. Vancomycin-resistant 4 0.06-32 40 Its use against other infections reE. faecium mains uncertain. Given the reported Enterococcus faecalis 4-32 0.25-32 38 emergence of resistance during therapy and its inconsistent bactericidal In vitro activity against multiple strains of Gram-positive bacteria expressed as the minimum activity, especially with regard to coninhibitory concentration (MlCgo) and range of MIC in pgm1-1. stitutive MLS isolates, it is unlikely to become the therapy of choice for Differences in distribution resulting from the dif- staphylococcal infections. Other factors that might ferent pharmacokinetics of the two components, with limit its overall utility as a vancomycin replacement suboptimal dalfopristin concentrations at the site of include its poor activity against the most common infection, have been suggested to account for the lack nosocomial enterococcal species, E. faecalis, superof efficacy observed in treatment of experimental endoinfection with the resistant E. faecalis, and decarditis in one study using an MLS-resistant S. aureus velopment of resistance in E. faecium arising during isolate 46. Resistance to quinupristin-dalfopristin has therapy. been reported in clinical trials; in one study [C.A. Wood Other agents in development etal. (1996) 36th Int. Conf. Antimicrob. Agents ChemoSeveral additional agents demonstrate activity against ther., New Orleans, LA, USA, Abstr. LB13] of 24 pavancomycin resistant Gram-positive pathogens in vitro; tients receiving quinupristin-dalfopristin for the treathowever, most of these are still at a relatively early stage ment of vancomycin-resistant E. faecium infection, of the drug development process. Furthermore, many three patients (12.5%) had isolates of E. faecium resistant to the drug. Emergence of resistance has also been of these drug candidates are bacteriostatic, suggesting limited potential in treating serious Gram-positive encountered in a patient with S. aureus bacteremia bacterial infections such as endocarditis. during quinupristin-dalfopristin therapy 47. Table 4. Activity of quinupristin--dalfopristin Clinical trials Phase III clinical trials are also under way with Synercid in more than 310 centers in the USA and Europe. In one Phase Ill trial, more than 200 patients have received the agent for the treatment of documented VRE. Additional Phase III trials are under way for the treatment of community-acquired pneumonia, nosocomial pneumonia and complicated skin and skin-structure infections. More than 700 patients with severe Gram-positive infections have received Synercid through an emergency-use program. Among patients with bacteremia caused by vancomycin-resistant enterococci, 67% showed improvement and eradication of the bacteria 39. These results, however, may be misleading because (1) no control group was available, (2) many 0 Everninomycins The oligosaccharide antibiotics of the everninomycin class have been known since 1964 (Ref. 51) but were not developed for medical use because of nephrotoxicity. Recently, interest in the class has been revived, and a more-potent and less-toxic analog, SCH 27899, has been brought forward by Schering as a candidate for development s2. MICg0 values for clinically significant Gram-positive bacteria F Flg. 5. Structure of the oxazolidinone U-100592. TRENDS 'N Oxazolidinones The oxazolidinones are novel synthetic agents that are active against both Gram-positive bacteria and mycobacteria in vitro 48. If the drugs currently being studied are developed, oxazolidinones would represent the first new class to be developed in over a decade. The oxazolidinone class possess bacteriostatic activity in vitro against resistant Gram-positive pathogens. These drugs inhibit protein synthesis, acting at the step preceding the interaction between N-formylmethionine-tRNA and the 30S ribosomal subunits with the initiator codon, probably inhibiting recognition of the 3' upstream ribosome-binding sequence of mRNA (Ref. 49). U-100592 (see Fig. 5) is an oxazolidinone that is active against MRSA and enterococci (MICg0 _<4~tgm1-1) in vitro s° (Table 3), and no cross resistance to other antibiotic classes has, as yet, been noted. U-100592 is well tolerated in single oral doses of <1000 mg, and Phase II trials were expected to begin in 1996. MICROBIOLOGY 246 vo, 3 No. 6 JUNE 1997 R E V I E W S are comparable or lower than those of vancomycin, and SCH 27899 is active against VRE (Table 3). Its spectrum includes MRSA and methicillin-resistant coagulase-negative staphylococci (MICg0 of 0.5 gg m1-1) and VRE (MICg0 of 0.25 gg ml-l) 53. The compound is bacteriostatic, and its mechanism of action is unknown. Glycylcyclines Two new semisynthetic tetracyclinelike antibiotics containing the N,N-dimethylglycylamino substituent at the 9-position of minocycline (DMG-MINO) and 6-demethyl 6-deoxytetracycline (DMG-DMDOT) appear to have activity against MRSA (MICg0 of <_2gg m1-1)54 and vancomycin-resistant strains of E. faecium and E. faecalis (MICg0 of_<0.5 ggm1-1) in vitro ss. These compounds are active against isolates resistant to tetracycline and minocyclines6. Resistance to tetracyclines is mediated by two major mechanisms: active efflux and ribosomal protection. The glycylcyclines are able to evade both of these mechanisms; they are poor substrates for the transporters mediating efflux and may avoid the ribosomal protection mechanism by their greater affinity for ribosomesST,SL The potential for development of resistance to the glycylcyclines is unknown. Like tetracyclines, they are bacteriostatic. Fluoroquinolones A large number of quinolones with enhanced activity against vancomycin-resistant enterococci are undergoing development sg. Agents in Phase III development include trovafloxacin, clinafloxacin and DU-6859a (in Japan). Compared to ciprofloxacin, these compounds are much improved in the potency of their activity against Gram-positive bacteria, especially S. pneumoniae. Resistance to ciprofloxacin emerged with unexpected speed in staphylococci; only time will tell if these new agents will do better in retaining their activity. However, resistance to ciprofloxacin may portend resistance to the newer fluoroquinolones 63. N o v e l ~-lactam antibiotics A novel penicillin-binding protein 2a (PBP-2a) accounts for [3-1actam resistance in MRSA and coagulase-negative staphylococci. Efforts to develop [3-1actams capable of inhibiting PBP-2a activity to achieve MRSA activity are progressing in several research labs but, although there are examples of interesting activity in vitro, as yet clinical studies have not been reported. Vaccines instead of vancomycin? Alternatives to antibiotics for the prevention and treatment of infections caused by Gram-positive pathogens can also be in the form of vaccines. For S. aureus, both active and passive immunization are being investigated for disease prevention. Active immunization would be likely to include individuals at high risk for S. aureus bacteremia and systemic infection, such as those with internal catheters, prosthetic heart valves and joints and vascular grafts, and patients receiving long-term hemoor peritoneal dialysis. Candidates for short-term passive immunization with polyclonal antibodies raised in normal hosts include trauma and burn victims and people undergoing elective surgery. TRENDS IN MI(ZR()B|()I.()(;Y 247 The capsule of S. aureus is under study as a possible immunogen. Many strains of S. aureus possess a surface polysaccharide capsule, which is thought to contribute to the pathogenesis of disease. This capsule may be one of 11 different serotypes, but -80% of clinical blood isolates of S. aureus possess either type 5 or type 8 capsular polysaccharide61-63.Encapsulated strains of S. aureus have been shown to resist opsonophagocytosis64,65.In the presence of type-specific antibodies, however, opsonophagocytosis and killing of S. aureus are enhanced 64-66. Work is in progress to develop a bivalent type 5 and type 8 S. aureus vaccine composed of processed capsular polysaccharides (CPs) conjugated to carrier protein: specifically, a nontoxic recombinant form of Pseudom o n a s aeruginosa exotoxin A (Refs 66,67). The carrier protein may be needed for vaccine development because the relatively small molecular size of the S. aureus polysaccharides means that they are unlikely to be sufficiently immunogenic alone. This contrasts with the vaccines developed using the capsule polysaccharides of S. p n e u m o n i a e , Neisseria meningitidis and H. influenzae type b. Studies of immunization with type 5 and type 8 conjugates in both mice and humans have shown that each conjugate can elicit anti-CP antibodies 67. Human subjects had pre-immune immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies to type 5 and type 8 CPs before immunization. Two weeks after immunization with type 5 conjugate, subjects showed an increase in anti-type 5 CP IgG, and six weeks postimmunization a >30-fold rise in IgG and 10-15-fold rise in IgM was noted. For a type 8 CP conjugate, an -eightfold increase in anti-type 8 CP IgG and fivefold rise in corresponding IgM antibodies were observed at both two and six weeks postimmunization. Similar results were observed with a high-molecular-weight non-conjugated form of type 8 CP. For both types, a second immunization did not not result in significant further increases in anti-CP antibody levels. Small declines in anti-CP IgG levels for both types occurred six months postimmunization, and studies suggest that this decline may be somewhat greater if the two conjugates are given together6L In assessing the results of these studies, the authors hypothesize that no booster effect was observed in humans after the second immunization because the first immunization acted as a booster in subjects with low levels of pre-existing anti-S, aureus CP antibodies induced by colonization with these strains of S. aureus or microorganisms with antigenically similar polysaccharides. They also suggest that concurrent administration of type 5 and type 8 vaccines may result in moderately decreased immunogenicity and humoral response. Further efforts are in progress to develop a bivalent vaccine for type 5 and type 8 strains of S. aureus. Alternatives to antibiotics for C. difficile The major use of oral vancomycin has been for topical treatment of CDAD in the colon. C. difficile has been implicated as a causative agent in 10-20% of all cases of antibiotic-associated diarrhea and in almost all cases W'.. S No. 6 Ju~: 1997 R E V I E W S Questions for future research * Will VanA or VanB glycopeptide resistance appear in methicillin- • • • • • resistant staphylococci? There is some question as to whether the two forms of resistance mechanism, both of which alter peptidoglycan synthesis, are compatible. Can infection control measures and limited use of glycopeptide make an impact on the spread of vancomycin resistance? Recent concern over the use of glycopeptide in animal feed suggests faith in this approach. Are the new agents in development adequate as replacements for vancomycin in treating methicillin-resistant Staphylococcus aureus and coagulase-negativestaphylococci? How important is bactericidal activity in therapeutic outcome? Is the proposed mechanism of action of glycopeptides active against vancomycin-resistant Enterococcus faecium a complete explanation? Will new agents be available fast enough to respond to growing resistance? Can vaccines or passive antibody provide an alternative to antibiotics in some infections? of pseudomomembranous colitis, a potentially lifethreatening disease. CDAD is generally seen in hospitalized patients who receive antibiotics. In one study at a large hospital in the USA, 21% of 399 patients acquired C. difficile and, of these, 37% developed diarrhea 68. Concern over the rapid spread of VRE and HICPAC guidelines recommending alternative treatment TMhave led to a decline in vancomycin therapy of CDAD in the USA. Oral metronidazole has been used as a replacement for vancomycin but is not a totally satisfactory alternative: metronidazole-resistant C. difficile has been observed (R. Fekety, pets. commun.), and use of metronidazole (as well as use of vancomycin) has been associated with VRE bacteremia in seriously ill patients 1,69. New therapies are clearly needed. The pathogenesis of CDAD is mediated by synergistic effects of two C. difficile-produced toxins, A and B, on the colon. One preventative measure for CDAD that is currently under investigation uses either bovine or avian polyclonal antibodies raised against toxins A and B. Syrian hamsters given intragastric clindamycin and challenged with toxigenic C. difficile provide a useful model for CDAD. In one approach, bovine antibodies were synthesized in cattle vaccinated with a toxoid made from a C. difficile culture filtrate TM. Antibodies to the toxins were obtained from the coiostrum of cows at parturition. Bovine IgG with activity against toxins A and B was used to passively immunize hamsters and was shown to protect animals from diarrhea and death after challenge with C. difficile. Animals were protected if given the hyperimmune antibody preparation before the onset of diarrhea, but attempts to treat hamsters who developed diarrhea after termination of prophylaxis failed. Recently, C. difficile toxinneutralizing antibodies have been detected in the stools of normal humans after oral ingestion of bovine immunoglobulin preparations 71. These experiments indicate that polyclonal antibodies with high activity to C. difficile toxins A and B can protect sensitive animals from the effects of CDAD and suggest that such an approach may be beneficial in humans. TRENDS IN M I ( ' R O B I O L O ( ; ¥ Similarly, another research group has hyperimmunized laying hens with recombinant fragments of toxins A and B to obtain antibodies that neutralize both toxins. These avian antibodies are obtained from the yolks of the eggs. When delivered orally, the antibodies are effective in preventing or treating CDAD in Syrian hamsters administered clindamycin and challenged with toxigenic C. difficile. The infected animals survive after the termination of treatment and there has been no incidence of relapse in the hundreds of animals treated so far (Ophidian Pharmaceuticals, pers. commun.). Several groups of investigators have examined the feasibility of active immunization 19.v2-74.Recently, using a toxoid prepared from culture filtrates of toxigenic C. difficile, Syrian hamsters immunized by the subcutaneous or intraperitoneal route were protected from death when challenged with clindamycin and C. difficile, although most developed diarrhea 19. However, all animals given intranasal mucosal immunization and later boosted intraperitoneally survived and were protected from diarrhea. When hamsters immunized intranasally with intraperitoneal boosting were rechallenged with clindamycin and C. difficile 140 days after the initial challenge, some were protected from both death and diarrhea, suggesting that immunity does persist over time. Further studies are in progress using different preparations of toxoid and different modes of administration, with an interest in eventual human C. difficile vaccine development. Conclusions Vancomycin resistance has increased the urgency of the search for new antibacterial agents. Apart from activity against glycopeptide-resistant strains, the compounds currently under study have few advantages over vancomycin; the enhanced bactericidal activity of the new glycopeptides may be one advantage. Use of vaccines or anti-toxins may have the potential to reduce some of the need for antibiotic therapy, but studies are still in early phases. Until new agents currently in development prove their use, combination antibiotic therapy will also have to be carefully evaluated. Compounds from both old and new chemical classes await further clinical trials to establish if any will prove to be sufficiently well-tolerated and efficacious to serve as a vancomycin replacement. References 1 Edmond,M.B.et al. (1995) Clin. Infect. Dis. 20, 1126-1133 2 Murray,B.E. (1995)Clin. Infect. Dis. 20, 1134-1136 3 Centersfor DiseaseControl (1993)Morb. Mortal. Weekly Rep. 45, 597-599 4 Sanyal,D. etal. (1991) Lancet 337, 54 5 Schwalbe,R.S.,Stapleton,T. and Gilligan, P.H. (1987)New Engl. ]. Med. 316, 927-931 6 Veach,L.A.et al. (1990)]. Clin. Microbiol. 28, 2064-2068 7 Degener,J.E. etal. (1994)]. Clin. Microbiol. 32, 2260-2265 8 Froggatt,J.W. et al. (1989)Antimicrob. Agents Chemother. 33, 460-466 9 Noble,W.C., Virani, Z. and Cree, R.G. (1992)FEMS Microbiol. Lett. 72, 195-198 10 Biavasco,F. et al. (1996)Eur.J. Clin. Microbiol. Infect. Dis. 248 w,L. 5 No. 6 JIJNE 1997 REVIEWS 43 Chow, J.W. et al. (1997) Clin. Infect. Dis. 24, 91-92 44 Kang, S.L. and Rybak, M.J. (1995)Antimicrob. Agents Chemother. 39, 1505-1511 45 Fantin, B. et al. (1995)Antimicrob. Agents Chemother. 39, 400-405 46 Entenza, J.M. et al. (1995) Antimicrob. Agents Chemotber. 39, 1419-1424 47 Kostman, J.R. et al. (1995) Clin. Infect. Dis. 21,746 48 Brickner,8.J. (1996) Curr. Pharm. Des. 2, 175-194 49 Eustice, D.C. et al. (1988) Antimicrob. Agents Chemother. 32, 1218-1222 50 Jones, R.N., Johnson, D.M. and Erwin, M.E. (1996) Antimicrob. Agents Chemother. 40, 720-726 51 Weinstein, M.J. et al. (1964) Antimicrob. Agents Chemother. 4, 24 -32 52 Ganguly, A.K. et al. (1989) Heterocycles 28, 83-88 53 Urban, C. et al. (1996)]. Antimicrob. Chemother. 37, 361-364 54 Weiss, W.J. et al. (1995)]. Antimicrob. Chemother. 36,225-230 55 Fraise, A.P. et al. (1995)J. Antimicrob. Chemother. 35, 877-881 56 Testa, R.T. et al. (1993) Antimicrob. Agents Chemother. 37, 2270-2277 57 Bergernon,J. et al. (1996)Antimicrob. Agents Chemotber. 40, 2226-2228 58 Someya,Y., Yamaguchi, A. and Sawai, T. il995) Antimicrob. Agents Chemother. 39, 247-249 59 Gootz, T.D. and Bright),,K.E. (1996) Med. Res. Rev. 16, 433-486 60 Harrington, G.D. et al. (1995) Diagn. Microbiol. Infect. Dis. 21, 27-31 61 Fournier, J.M. et al. (1987)]. Clin. Microbiol. 25, 1932-1933 62 Arbeit, R.D. et al. (1984) Diagn. Microbiol. Infect. Dis. 2, 85-91 63 Sompolinsky,D. et al. (1985)J. Clin. Microbiol. 22, 828-834 64 Karakawa, W.W. et al. {1988) Infect. Immun. 56, 1090-1095 65 Xu, S., Arheit, R.D. and Lee,J.C. (1992) Infect. lmmun. 60, 1358-1362 66 Fattom, A. et al. (1990) Infect. lmmun. 58, 2367-2374 67 Fattom, A. et al. (1993) Infect. lmmun. 61, 1023-1032 68 McFarland, L.V. et al. (1989) New Engl. J. Med. 29, 204-210 69 Shay, D.K. et al. (1995}J. Infect. Dis. 172, 993-1000 70 Lyerly,D.M. et al. (1991) Infect. Immun. 59, 2215-2218 71 Kelly,C.P. et al. (1997) Antimicrob. Agents Chemother. 41, 236-241 72 Libby, J.M., Jointer, B.S. and Wilkins, T.D. (1982) Infect. Immun. 36, 822-829 73 Kim, P-H., Iaconis, J.P. and Rolfe, R.D. (1987) Infect. Immun. 55, 2984-2992 74 Fernie, D.S. et al. (1983) Dev. Biol. Stand. 53, 325-332 54, 15-24 11 Power, E.G. et al. (1995)J. Antimicrob. Chemother. 36,595-606 12 Krcmery,V., Jr, Spanik, S. and Trupl, J. (1997)]. Chemother. 8, 325-326 13 Poyart, C. et al. (1997) Antimicrob. Agents Chemother. 41, 24 -29 14 Hospital Infection Control Practice AdvisoryCommittee (1995) Infect. Control Hosp. Epidemiol. 16, 105-113 15 Uttley, A.H.C. et al. (1988) Lancet 1, 57-58 16 Leclerq,R. et al. (1988) New Engl. ]. Med. 319, 157-161 17 Van Der Auwera, P. et al. (1996)J. Infect. Dis. 173, 1129-1136 18 Bates,J., Jordens, J.Z. and Griffiths, D.T. (1994)J. Antimicrob. Chemother. 34, 507-516 19 Torres, C. et al. (1994)J. Antimicrob. Chemother. 33, 553-561 20 Arthur, M., Reynolds, P. and Courvalin, P. (1996) Trends Microbiol. 4, 401-407 21 Arthur, M. and Courvalin, C. (1993) Antimicrob. Agents Chemother. 37, 1563-1571 22 Arthur, M. etal. (1996) J. Infect. 32, 11-16 23 Woodford, N. etal. (1995) Clin. Microbiol. Rev. 8,585-615 24 Walsh, C.T. et al. (1996) Chem. Biol. 3, 21-28 25 Bugg, T.H. et al. (l 991) Biochemistry 30, 10408-10415 26 Dutka-Malen, S. et al. (1990) Antimicrob. Agents Chemother. 34, 1875-1879 27 Nagarajan, R. (1993) ]. Antibiot. 46, 1181-1195 28 Malabarba, A., Nicas, T.I. and Thompson, R.C. ¢1997) Med. Res. Rev. 17, 69-137 29 Cooper, R.D.G. et al. (1996) ]. Antibiot. 49, 575-581 30 Ciabatti, R. and Malabarba, A. (1994) Chim. Ind. 76, 300-304 31 Nicas, T.I. et al. (1996) Antimicrob. Agents Chemother. 40, 2194-2199 32 Allen, N.E., Hobbs, J.N., Jr and Nicas, T.I. (1996)Antimicrob. Agents Chemother. 40, 2356-2362 33 Williams, D.H. (1996) Nat. Prod. Rep. 1996, 469-477 34 Beauregard,D.A. et al. (1995) Antimicrob. Agents Chemother. 39, 781-785 35 Malabarba, A. et al. (1995)J. Antihiot. 48,869-883 36 Goldstein, B.P. et al. (1995) Antimicrob. Agents Chemother. 39, 1580-1588 37 Bryson, H.M. and Spencer, C.M. (1996) Drug 53,406-415 38 Low, D.E. (1996) Microb. Drug Resist. 1,223-234 39 Griswold, M.W., Lomaestro, B.M. and Briceland, L.L. (1996i Am. J. Health-Syst. Pharm. 53, 2045-2053 40 Bouhallab, S. et al. (1988) Eur. J. Med. Chem. 23, 143-148 41 Neu, H., Chin, N-X. and Gu, J-W. (1992)J. Antimicrob. Chemother. 30A, 83-94 42 Collins, L. et al. (1993) Antimicrob. Agents Chemother. 37, 598-601 In t h e o t h e r Trends j o u r n a l s A selection of recently published articles of interest to TIM readers. *Virulence and cAMP in smuts, blasts and blights, by J.W. Kronstad - Trends in Plant Science 2, 1 9 3 - 1 9 9 *Multi-gene vaccination against malaria: a multistage, multi,immune response approach, by D.L. Doolan and S.L. Hoffman - Parasitology Today 13, 1 7 1 - 1 7 7 *Cytokine regulation of endothelial cell function: from molecular level to the bedside, by A. Mantovani, F. Bussolino and M. Introna - Immunology Today 18, 2 3 1 - 2 4 0 *HIV integrase: a target for AIDS therapeutics, by M. Thomas and L. Brady- Trends in Biotechnology15, 1 6 7 - 1 7 2 .RNA editing: getting U into RNA, by M.L. Kable, S. Heidmann and K.D. Stewart - Trends in Biochemical Sciences 22, 1 6 2 - 1 6 6 249 V,,,-5 N,. 6 99V