R E S E A R C H H I G H L I G H T S

NATURE REVIEWS | MOLECULAR CELL BIOLOGY VOLUME 5 | AUGUST 2004 | 597

The correct disposal of waste is essen-tial for daily life, and this applies atthe cellular level too. Misfolded pro-teins in the endoplasmic reticulum(ER) are retro-translocated (or dislo-cated) to the cytosol for degradation,and blocking this pathway causes ERstress. Some components of the path-way are known — for example, thecytosolic ATPase p97 probably ‘pulls’misfolded proteins out of the ER.However, the channel that mediatesretro-translocation has been a matterof speculation. Two papers in Nature,though, now identify a protein thatmight form such a channel.

In the first paper, Lilley andPloegh studied the US11 protein ofthe human cytomegalovirus, whichtargets the major histocompatibilityclass-I heavy chain for retro-translo-cation. A Q192L mutant of US11cannot maintain this retro-transloca-tion, and the presence of this gluta-mine residue in the transmembraneregion of US11 indicated that itmight interact with host proteins.

The authors therefore developedan affinity-purification approach toexamine which cellular proteins bindto US11 versus US11Q192L, and theyidentified a protein — which wasnamed Derlin-1 — that specificallyassociated with US11. This trans-membrane protein is a homologue ofSaccharomyces cerevisiae Der1, whichis required for the degradation of asubset of misfolded ER proteins.

Using antibodies against Derlin-1in immunoprecipitation experi-ments, Lilley and Ploegh confirmedthe Derlin-1–US11 association andshowed that the class-I heavy chainwas also recovered with Derlin-1.This was not the case in cells express-ing US11Q192L, which indicates thatUS11 uses its transmembrane regionto recruit the class-I heavy chain intoa complex with Derlin-1.

Next, the authors created aDerlin-1 construct, in which greenfluorescent protein (GFP) wasattached to its cytosolic carboxyl ter-minus. This construct extended thehalf-life of the class-I heavy chain

from 5 to 30 minutes in US11-express-ing cells, which indicates that, withoutblocking the binding of the class-Iheavy chain to US11 or Derlin-1, theGFP moiety delays the exposure of theclass-I heavy chain to the cytosol.

In the second paper, Rapoportand co-workers used a differentapproach and identified the receptorthat links p97 (also known as VCP) tothe ER membrane. They actually iso-lated two proteins in near-stoichio-metric amounts — Derlin-1 and aprotein that was named VIMP (forVCP-interacting membrane protein).

As VIMP has a large cytosolicdomain, the authors proposed thatthis domain interacts with p97. Thisproved to be correct, and they showedthat this domain recruits p97 togetherwith its co-factors. Furthermore, usingimmunofluorescence microscopy,they showed that VIMP seems torecruit p97 to Derlin-1.

Next, Rapoport and colleaguesinvestigated whether Derlin-1–VIMPis involved in retro-translocation, andthey showed that Derlin-1–VIMPinteracts with the class-I heavy chain

in US11-expressing cells. In agree-ment with Lilley and Ploegh, theyalso noted that an interactionbetween Derlin-1 and US11 is func-tionally important for this process.

In the final part of their study,Rapoport and co-workers testedwhether the Derlin-1–VIMP complexhas a general function in the retro-translocation of misfolded proteinsand their results indicate that it canassociate with a large set of substrates.Furthermore, depleting Derlin-1levels in Caenorhabditis elegans usingRNA interference induced ER stressin many cell types.

Together, these groups have there-fore identified a putative waste-dis-posal chute in mammals calledDerlin-1. Derlin-1 mediates the retro-translocation of a subset of misfoldedproteins from the ER and, if it can beshown that Derlin-1 does form achannel, it seems only a matter oftime before other such retro-translo-cation channels are identified.

Rachel SmallridgeReferences and links

ORIGINAL RESEARCH PAPERS Lilley, B. N. &Ploegh, H. L. A membrane protein required fordislocation of misfolded proteins from the ER.Nature 429, 834–840 (2004) | Ye, Y. et al. Amembrane protein complex mediates retro-translocation from the ER lumen into the cytosol.Nature 429, 841–847 (2004)FURTHER READING Schekman, R. Cell biology:a channel for protein waste. Nature 429, 817–818(2004)

Waste-disposal chute

P R OT E I N T R A N S LO C AT I O N

Wired for cycling

• www.mpf.biol.vt.edu/research/budding_yeast_model/cyberyeast.htm

Hands up who’s managed toplough through a researchpaper on mathematicalmodelling. No? Then you areprobably an experimentalist...But, web sites might be moreaccessible when it comes tomathematical models.

The Budding Yeast CellCycle web site presents amathematical model of cell-cycle control in budding yeastand is an enhancement of arecent research paper by thegroups of John Tyson, BelaNovak and Frederick Cross.

The ‘Overview’ sectionpresents a complete wiringdiagram for the regulation ofcell-cycle progression basedon published data. Thediagram is clickable, allowingthe user to review theliterature sources.

The regulatory network hasbeen formulated into a set ofdifferential equations (see‘Mathematical model’section) and analysed bycomputer simulation. Theresearchers compared thephysiology of wild-type cellsand >100 mutant strains withthe simulated results thatwere derived from the wiringdiagram. The simulationsuccessfully predicted themutant phenotype in mostcases, but someinconsistencies were found(see ‘Problems’).

The ‘Conclusions’ section,under ‘Overview’, takes youto the ‘Simulate the model’page, which allows users tochange the value of anyparameter and re-computethe behaviour of the model.There is a short description ofeach parameter, its defaultvalue and the range thatallows a viable cell-cyclesimulation.

The web site has just beenupdated and is now evenmore user-friendly. Browsingthe site should giveexperimentalists a feel for thepotential of mathematicalmodelling and how itspredictive power can be usefulto design new experiments.

Arianne Heinrichs

WEB WATCH