clonal distribution? To answer this question,researchers at the University of California,San Francisco constructed two plamids:pYK20, containing the mecA gene alone,and pYK644, containing mecA and its repres-sors mecR1 and mecI. Analysis of PBP2aexpression revealed that in all ‘experienced’staphylococcal host strains — defined asmethicillin-sensitive S. aureus (MSSA) strainscreated by the excision of SCCmec from thechromosome of MRSA strains — pYK20 wasstably maintained. In naive staphylococcalhost strains — MSSA strains that have neverhosted mecA — pYK20 was unstable and PBP2aexpression was heterogeneous and variedwith the host genotype.

pYK644, by contrast, was stably maintainedin naive hosts, implying that the mecA-associatedregulatory genes can influence the clonalrestriction of mecA acquisition. In most clinicalMRSA isolates, the mecA regulatory genes aredeleted and mecA is regulated by blaR1 andblaI, genes that are homologous to mecR1

A new study has shown that the genetic back-ground of recipient staphylococcal strains canhave an important influence on whether or notthe mobile genetic element responsible formethicillin resistance can jump between strains.

Methicillin-resistant Staphylococcus aureus(MRSA) are not only the most commoncause of nosocomial infections, they are alsosignificant community-acquired pathogens,particularly in children. S. aureus resistance tomethicillin is conferred by the presence of amodified penicillin-binding transpeptidase,PBP2a, which is encoded by the mecA gene.mecA is carried on a mobile genetic elementknown as staphylococcal cassette chromosomemec (SCCmec), which is integrated into the S. aureus chromosome. There are 4 (and perhapsmore) main classes of SCCmec and, despite thefact that SCCmec is mobile, its distribution is narrow and is principally restricted to just 5 clonal complexes worldwide.

What is the role of the genetic backgroundof the staphylococcal host in this restricted

90 | NOVEMBER 2003 | VOLUME 1 www.nature.com/reviews/micro

H I G H L I G H T S

An important proof-of-concept study hasshown that is possible to inhibit HIVinfectivity by expressing a specific receptorfor HIV in a Lactobacillus species that iscommonly found in the vaginal microflora.

Of the 40 million individuals presentlyinfected with HIV-1, an estimated 80% wereinfected by heterosexual transmission, witha higher rate of transmission to females thanto males. With the continuing absence of aneffective vaccine, researchers are turning tonovel anti-viral therapies that can empowerwomen to actively protect themselvesagainst infection.

The bacterial microflora present in thevagina of healthy, non-menopausal women isdominated by the genus Lactobacillus. Leeand colleagues obtained naturally occurringlactobacilli isolates from vaginal swabs takenfrom healthy volunteers and, after evaluationfor favourable growth and colonizationability, Lactobacillus jensenii was selected asthe experimental strain.

In this initial study, Lee et al. decided toexpress the HIV-specific CD4 receptor in L. jensenii. The coding sequences for the firsttwo extracellular domains of CD4 (2D CD4),which are sufficient for high-affinity bindingto the gp120 extracellular glycoprotein ofHIV, were modified by PCR to conform to

the codon usage in lactobacilli, then clonedinto a Lactobacillus expression plasmid andprotein expression was optimized. A CD4capture ELISA and HIV-1 gp120-bindingassay were used to confirm that therecombinant 2D CD4 was in the correctconformation and was biologically active.

Could the 2D CD4 expressed by L. jenseniiblock the infectivity of HIV? In an HIV entry

assay using a reporter virus, 2D CD4purified from L. jensenii and conditionedculture media containing secreted 2D CD4were both shown to inhibit HIV-1 infectionof HeLa cells in a dose-dependent manner. Infollow-up assays, Lee et al. looked at theeffects of co-incubating intact recombinantL. jensenii with both a laboratory strain ofHIV-1 and a natural strain. The recombinantlactobacilli were shown to inhibit the entryof the laboratory strain by 95% and thenatural strain by 55%. As natural strains ofHIV-1 are known to be less sensitive to CD4inhibition, this result is modest yet stillstatistically significant.

Vaginal microflora as HIV inhibitors

H I V

Look beforeyou leap

E V O L U T I O N

NATURE REVIEWS | MICROBIOLOGY VOLUME 1 | NOVEMBER 2003 | 91

New research published in Genes andDevelopment from the Cormack lab shows thatCandida glabrata has at least two groups ofadhesin virulence genes located in subtelomericregions that are regulated by a definedmechanism of transcriptional silencing.

Candida is responsible for 8% of seriousnosocomial infections in the United States. Incommon with other Candida species, C. glabratausually populates mucosal surfaces, but canbreach the mucosal barrier when host defencesare lowered and cause life threatening systemicinfections. Unlike the diploid opportunisticpathogen Candida albicans, C. glabrata ishaploid, which makes genetic manipulations farmore straightforward. Previous work identifiedan adhesin protein — Epa1 — which enablesC. glabrata to stick to epithelial cells. Epa1 is amember of a large class of cell wall proteinsthat are found in many fungi, includingSaccharomyces and Aspergillus species.

When the EPA1 gene was deleted from thechromosome, the resulting mutant lost virtuallyall ability to attach to epithelial cells in the lab.So, EPA1 was an excellent candidate for a viru-lence gene. However, when the same mutantwas tested for pathogenicity in mouse modelsof candidiasis it was still virulent. Pathogenslike C. albicans, Trypansoma brucei andPlasmodium species commonly encode familiesof virulence genes. So, one possible solution tothis conundrum was that C. glabrata mighthave a family of adhesins. The hunt was on forEPA1 homologues and now De Las Peñas et al.report the characterisation of 5 homologues ofEPA1 in C. glabrata.

The 5 EPA1 homologues are located in twoseparate subtelomeric clusters. AlthoughSaccharomyces cerevisiae and C. glabrata havelargely similar genetic organisation (with syn-tenic gene arrangement) the subtelomeric

regions, including the EPA genes, are completelydifferent. Knocking out both subtelomericEPA clusters reduced virulence of the resultantmutant strain in a mouse model. Genomicsequencing studies have subsequently uncov-ered additional potential EPA1 homologues, sothe family is growing.

Curiously, De Las Peñas et al. found thatonly EPA1 was expressed in vitro. Not onlywere the other EPA genes repressed, but amarker gene — URA3 — was repressed wheninserted at either subtelomeric locus. So, bothsubtelomeric EPA clusters are regionallysilenced. This study also showed that a keycomponent of the transcriptional silencingmachinery in S. cerevisiae — SIR3 — regulatestranscriptional silencing of the EPA genes in C. glabrata. Even though the telomere genecomplements of these two fungi are different,the gene silencing mechanism is conservedbetween these species.

The subtelomeric location of the adhesingenes, coupled with a defined silencing mech-anism, means that C. glabrata is the firstpathogen for which the mechanism of tran-scriptional silencing of virulence factors hasbeen defined. Intriguingly, regulation of sub-telomeric silencing in S. cerevisiae has beenlinked to cellular stresses. Since cellular stress islikely to be common during infection, thisraises the exciting possibility that C. glabratavirulence factors could be switched on and offduring infections.

Susan Jones

References and linksORIGINAL RESEARCH PAPER De Las Peñas, A. et al.Virulence-related surface glycoproteins in the yeast pathogenCandida glabrata are encoded in subtelomeric clusters andsubject to RAP1- and SIR-dependent transcriptional silencing.Genes Dev. 17, 2245–2258 (2003)WEB SITEBrendan Cormack’s laboratory: http://www.mbg.jhmi.edu/FacultyDetails.asp?PersonID=361

Stuck on you

F U N G A L V I R U L E N C E

This in vitro demonstration of theinhibition of HIV-1 infectivity by arecombinant constituent of the vaginalmicroflora is an important proof-of-principle study in the search for novelmethods to halt the HIV epidemic. As yet,the engineered lactobacilli are a long wayfrom clinical development. However, Leeand colleagues have already gone on toachieve stable integration of the 2D CD4coding sequence into the L. jenseniichromosome, thereby circumventing theproblem that the antibiotic-resistancegenes present on the expression plasmidwould pose for clinical trials. The fact thatin a recent issue of PNAS another teamreport on BMS-378806, a new small-molecule inhibitor that blocks the dockingof HIV-1 to CD4, reflects the fact that theefforts of HIV researchers to thwart HIV-1are not wholly devoted to the search for avaccine, and all aspects of the infection andtransmission cycle are being investigatedfor potential new targets.

Sheilagh Clarkson

References and linksORIGINAL RESEARCH PAPERS Chang, T. L.-Y. et al.Inhibition of HIV infectivity by a natural human isolate ofLactobacillus jensenii engineered to express functionaltwo-domain CD4. Proc. Natl Acad. Sci. USA 100,11672–11677 (2003) | Lin, P.-F. et al. A small molecule HIV-1inhibitor that targets the HIV-1 envelope and inhibits CD4receptor binding. Proc. Natl Acad. Sci. USA 100,11013–11018 (2003)FURTHER READING Shattock, R. J. & Moore, J. P.Inhibiting sexual transmission of HIV-1 infection. Nature Rev.Microbiol. 1, 25–34 (2003)

and mecI, respectively and which also regulateblaZ, the gene encoding β-lactamase.Katayama and colleagues found that the blaregulatory regions did have a permissiveeffect on mecA acquisition in naive strains,although this effect was not as strong as thatof the mec genes.

Horizontal transfer of the mec element haslong been recognised as a major contributorto the evolution of MRSA strains, yet theclonal restriction of mecA distribution hasalways been puzzling. This work highlightsthe important contribution of the geneticbackground of the host staphylococcal strainto this process.

Sheilagh Clarkson

References and linksORIGINAL RESEARCH PAPER Katayama, Y. et al. Jumpingthe barrier to β-lactam resistance in Staphylococcus aureus.J. Bacteriol. 185, 5465–5472 (2003)FURTHER READING Enright, M. C. et al. The evolutionaryhistory of methicillin-resistant Staphylococcus aureus (MRSA).Proc. Natl Acad. Sci. USA 99, 7687–7692 (2002) | Chambers,H. F. Solving staphylococcal resistance to β lactams. TrendsMicrobiol. 11, 145–148 (2003)