180 | MARCH 2004 | VOLUME 2 www.nature.com/reviews/micro

H I G H L I G H T S

Researchers working on the bacterialpredator Bdellovibrio have madewhat they hope will be the next bigleap forward in their research withthe publication in Science of thecomplete genome sequence ofBdellovibrio bacteriovorus HD100.

Members of the genus Bdellovibrioare Gram-negative δ-proteobacteriathat are obligate predators of otherGram-negative bacteria and arefound in a wide range of aquatic andterrestrial habitats, including thehuman gut. The basic steps in thelife cycle of the bdellovibrios havebeen known since the 1970s, but thepaucity of available genetic systemsfor studying this unique genus has meant that progress in eluci-dating the molecular genetics and biochemistry of each step has beenrelatively slow.

The publication of the B. bacteri-ovorus genome sequence has thereforebeen keenly awaited by the small com-munity of Bdellovibrio researchers.The single, circular chromosome issurprisingly large — ~3.78 Mb inlength — and is predicted to encode3,584 proteins. 55% of the codingsequences have been assigned aputative function. Perhaps surpris-ingly, there is no evidence for thehorizontal transfer of genes fromprey into B. bacteriovorus.

The Bdellovibrio life cycle has twomain phases: attack and growth. Thedetection of prey involves chemotaxis— the genome contains 20 methyl-accepting chemotaxis protein (MCP)genes. There is no evidence in thegenome that bdellovibrios caneither secrete or detect the signallingmolecules homo-serine lactones,ruling out a role for quorum sensing.

After they have detected theirprey, bdellovibrios begin the growthphase of their life cycle by firstattaching to the prey cell then pene-trating the outer membrane and

Researchers working on Candida albicanshave identified a key phosphatase that theybelieve is important for the yeast-to-hyphatransition in this fungus, according to arecent report in Molecular Microbiology.

In Saccharomyces cerevisiae, the SIT4gene encodes the catalytic subunit of a type2A protein phosphatase that regulatesseveral cellular processes, includingpseudohyphal development and cell-cycleprogression. Using data available from theC. albicans genome sequencing project,Lee et al. retrieved the nucleotide sequenceof the C. albicans SIT4 gene — the encodedprotein, C. albicans Sit4p, has 89% amino-acid identity to S. cerevisiae Sit4p.

To investigate the role of Sit4p in C. albicans, heterozygous and homozygousSIT4 mutant strains were generated.Although both strains are viable, thegrowth rate of the yeast form was reducedand hyphal outgrowth was inhibited, bothon plates and in liquid culture.Reintroduction of SIT4 restored normalgrowth and hyphal formation. Furtherinvestigation using time-lapse microscopy

revealed that SIT4 is necessary for truehyphal growth to be maintained. A murinemodel of systemic candidiasis was used toconfirm that SIT4 is involved in thetransition between yeast and hyphal formsin vivo, and also that this switch is requiredfor C. albicans virulence.

Lee et al. went on to take a more detailedlook at the role of SIT4 by using DNAmicroarrays to investigate the differencesin transcription across the whole genomebetween sit4 mutant and wild-type cellsunder different growth conditions.Compared with wild-type cells, thetranscription of genes encodingmodulators of protein biosynthesis wassignificantly reduced in the sit4 mutantcells, indicating that Sit4p can regulatetranslation. A comparison of thetranscription profiles under hypha-inducing conditions showed that thetranscription of a total of 198 genes wassignificantly altered. Gene expression wasboth up- and downregulated; the set ofSIT4-dependent genes that are upregulatedduring hyphal formation includes XOG1

and YNR67, two hyphae-induced genesthat have not been identified previously.As these genes both encode glucanases,Lee et al. propose that SIT4 might beinvolved in controlling growth of the cellwall in C. albicans. The microarray dataalso proved that disrupting SIT4expression affects the expression of HOG1,which encodes a mitogen-activated proteinkinase (MAPK) that is thought to beinvolved in cell wall remodelling in C. albicans during osmotic stress.

The morphological switch from one formof growth to another is clearly a majorcellular event that requires remodelling ofthe cell wall. With this latest research,Lee et al. seem to have identified a keyregulator of this process.

Sheilagh Clarkson

References and linksORIGINAL RESEARCH PAPER Lee, C.–M. et al. The serine/threonine protein phosphatase SIT4 modulatesyeast-to-hypha morphogenesis and virulence in Candidaalbicans. Mol. Microbiol. 51, 691–709 (2004)FURTHER READING Johnson, A. The biology of mating in Candida albicans. Nature Rev. Microbiol. 1, 106–116(2003)

Candida remodelling

F U N G A L P H Y S I O LO G Y

First predatory bacterium sequenced

G E N O M I C S

The life cycle of B. bacteriovorus. B. bacteriovorus is shown in yellow, the bacterial prey cell isshown in blue. Image courtesy of Stephan C. Schuster, Snjezana Rendulic and Jürgen Berger.

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | MARCH 2004 | 181

Sometimes, practising self-restraint can have itsadvantages. A recent study by GenoveffaFranchini, Christophe Nicot and colleaguesprovides an example of this, in which human T-cell lymphotrophic virus type 1 (HTLV-1)inhibits the expression of its own proteins — astrategy that could help it to avoid detection bythe immune system.

p30II — an HTLV-1 protein of previouslyunknown function — localizes to nuclei, and theauthors therefore suspected that it might functionin regulating viral gene expression. To investigatethis, they co-expressed an HTLV-1 plasmid with ap30II complementary DNA, and found that p30II

expression inhibited production of viral proteinsin a dose-dependent manner.

The HTLV-1 Tax protein stimulates viral geneexpression, so the authors investigated whether the effects of p30II are mediated through inhibitionof Tax-induced transcription. When p30II was co-expressed with a Tax cDNA construct, there wasno change in the transcription of Tax-dependentreporter genes. However, when Tax was expressedfrom a plasmid containing the full-length HTLV-1sequence, p30II inhibited transcription by Tax,indicating that p30II mightprevent the expressionof Tax at the post-transcriptional level.

Using semiquantitative or real-time PCR onRNA from cells co-transfected with an HTLV-1provirus and a p30II construct, the authors foundthat p30II expression had little effect on the levelsof viral mRNAs in total cell extracts. However,when cytoplasmic RNA was analysed, there was a significant decrease in the level of the Tax/RexmRNA, which contains sequences for both Taxand Rex — another positive regulator of viralgene expression. The Tax/Rex transcript, incontrast to other viral mRNAs, was detected innuclear extracts, indicating that p30II might bindspecifically to this transcript and cause itsretention in the nucleus.

The Tax/Rex and p21Rex transcripts areproduced by differential splicing of viral genomicRNA, and therefore contain different splicejunctions. To confirm the specificity of theinteraction between p30II and Tax/Rex mRNA,Franchini, Nicot and colleagues made reporterconstructs containing the splice junctions ofeither p21Rex or Tax/Rex at their 3′ ends.Co-expressing p30II inhibited expression from theconstruct containing the Tax/Rex splice junctions,but had no effect on the p21Rex construct.

The authors then tested whether p30II is able tomove in and out of the nucleus. They transfectedone set of cells with a plasmid expressing a redfluorescent protein (RFP) and another with agreen-fluorescent-protein-tagged p30II construct.When the two cell lines were fused, p30II was unableto move into the nuclei of the RFP-expressing cells,unlike a control protein that shuttled between thenucleus and cytoplasm. This indicates how p30II

binding toTax/Rex mRNA might cause its nuclearretention, and therefore prevent its translation.

To confirm the inhibitory function of p30II in abiologically relevant situation, the authors testedits ability to suppress the production of viralproteins in human T cells chronically infectedwith HTLV-1. Expression of p30II from arecombinant lentivirus vector reduced the levelsof viral proteins in supernatants from these cells.

By inhibiting the production of viral proteins,p30II could enable HTLV-1 to avoid immunesurveillance at key stages of infection — a strategyused by several other viruses. Whether this is thecase, and if so, how the timing of p30II activity isregulated will be important questions for futureinvestigation.

Louisa Flintoft

References and linksORIGINAL RESEARCH PAPER Nicot, C. et al.HTLV-1-encoded p30II is a post-transcriptional negative regulator of viral replication. Nature Med. (18 Jan 2004) doi:10.1038/nm984

HTLV-1 — showing some self-control

V I R O LO G Y

peptidoglycan to gain access to theperiplasm. Once there, the Bdellovibrioalters the shape of the prey cell,causing it to round up. Both of theseprocesses clearly require hydrolyticenzymes and the importance ofhydrolysis to the Bdellovibrio lifecycle is underlined by the abundanceof hydrolytic enzymes — a total of293 are encoded, including 150 pro-teases and peptidases — and trans-porters — 244 open reading frames— in the genome.

Aside from being interesting asthe ‘world’s smallest hunter’, scien-tists are also interested in Bdellovibrioin the hope that this genus might bea source of novel enzyme-basedantimicrobial compounds. The pub-lication of the genome sequencecould therefore potentially be thefirst step towards novel antibacterialstrategies.

Sheilagh Clarkson

References and linksORIGINAL RESEARCH PAPER Rendulic, S. et al. A predator unmasked: life cycle of Bdellovibriobacteriovorus from a genomic perspective.Science 303, 689–692 (2004)