TRENDS in Biotechnology Vol.19 No.10 October 2001

http://tibtech.trends.com 0167-7799/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0167-7799(01)01745-0

375Research Update

Meeting Report

The Protein Expression meeting was held

in McLean, Virginia, USA, 5–6 April 2001.

Genome projects, despite theirmagnitude, focus on the characterizationof a single type of molecule and userelatively few optimized technologies toefficiently generate data. The field ofproteomics is different. Proteomicsfocuses on an enormously diversecollection of smaller protein moleculesencoded by seemingly endless genomicDNA – the precise properties of eachprotein can only be determined after ithas been expressed and purified. Thiscrucial goal formed the focus of a recentconference on heterogeneous proteinexpression and purification.

Problems in protein expression

A variety of systems has been developedfor the synthesis and purification ofheterologous proteins. Decisions onwhich system to try for a recentlydiscovered gene are typically based on a compromise between economics,reliability and the putative processingneeds of the target protein.Unfortunately, because of complexityarising from differences in amino acidcompositions, folding pathways, cofactorrequirements, subunit configurationsand post-translational modifications, it isextremely difficult to predict how geneswill behave in these systems or whetherthe resulting proteins will show theirnative activity. Work by Peter LeMotte(Pfizer Global Research andDevelopment, Groton, CT, USA)demonstrates that although existingsystems can be cleverly adapted to solvespecific expression problems, there areno guaranteed approaches, thuselucidating the entire human proteome isan unrealistic target even for the largestresearch organizations.

Narrowing the field

The number of gene products to bestudied can be reduced by identifyingthose few that are somehow related to

disease. These peptides are often smalland present in trace amounts, causingconventional detection methods to missthem completely (Alex Karavanov,Ciphergen Biosystems, Freemont, CA,USA). This problem is furthercompounded by the requirement forquantitative data from large populationsto assure statistical significance of asuspected target. New technologies in thisarea generally attempt to increase systemspeed and sensitivity (John Wiktorowicz,Lynx Therapeutics, Hayward, CA, USA),or capacity and data analysis (DavidLubman, University of Michigan,AnnArbor, MI, USA). Despite someprogress, problems with protein labeling,resolution and identification still presentsignificant limitations in this area.

The shotgun approach

Once target proteins are selected, theymust be expressed and purified beforethey can be characterized. A clear trendin optimizing this process is the parallelevaluation of expression systems usingplasmids that allow genes to be expressedin a variety of cells from a single PCRreaction. One vector, described byBarbara Morris (Novagen, Madison, WI,USA), uses multiple nested promoters toallow a gene to be expressed inEscherichia coli, insect and mammaliancells from a single plasmid. MichaelBrasch (Invitrogen Corporation,Rockville, MD, USA) described a secondstrategy, the Gateway system, in whichgenes are rapidly moved into variousexpression vectors by phage mediatedsite-specific recombination. Because thetransfer takes place without additionalPCR amplification, only the sequence ofthe donor plasmid must be verified,allowing validated gene libraries to becreated. Brasch listed several genomeproject collaborations that are generatingdonor plasmids for use in this system.

Optimizing expression systems

Despite the development of technologiesto rapidly evaluate existing systems,

there are some genes that seem to defyefficient expression in any availablesystem. New systems with greaterexpression, folding and secretioncapabilities are therefore crucial.Several presentations described thegenetic modification of bacterial cells for efficient expression and foldingthrough the addition of new genes (RudiGlockshuber, Eidgenoessische TechnischeHochschule, Zurich, Switzerland), or bycombining rational modification withgenetic selection for resistance to toxicproducts (Fergal Hill, Avidis SA,Saint Beauzire, France). Althoughsignificant problems remain, transgenicplants are also becoming increasinglyused for the production of complexproteins (Brandon Price, GoodwinTechnologies, Plantation, FL, USA).

Several protein-fusion technologies werealso described, including a systematicapproach for identifying highly-solublefusion partners developed by RogerHarrison (University of Oklahoma,Norman, OK, USA), as well as reversiblyinsoluble partners that allow simplepurification by precipitation (AshutoshChilkoti, Duke University, Durham, NC,USA). Our group has developed a methodby which the fusion partner can be madeto self-cleave from the target proteinfollowing purification. A clear message isthat the complexity of the proteome isreflected in the myriad of approachesbeing used for the synthesis of itscomponents. Even the seminal fluid oftransgenic pigs is being considered for commercial scale expression(Michael Dyck, Universite Laval,Quebec, Canada).

The work ahead

Although scientists have routinelyexpressed genes and characterized theirproducts for decades, the completed

Conquering the proteome: complexity, cooperation

and commerce

David W. Wood

‘...the complexity of the proteome is

reflected in the myriad of approaches being

used for the synthesis of its components.’

human genome sequence has increasedthe scope of this research. It has becomeapparent that those who will reap themajor rewards of the genome must be ableto find and study important genes rapidly,effectively finding needles in a ‘genomichaystack’at a frantic pace. To address thischallenge, Edward Eisenstein (NationalInstitute for Standards and Technology,Rockville, MD, USA) closed the meetingby proposing a framework for a proteinexpression consortium. This group wouldbe comprised of representativelaboratories from government, academiaand industry, and would ultimatelyproduce a public database in which

expression and purification techniquesare systematically described usingestablished standards.

Despite the benefit of such acooperative effort, it is clear that manyhurdles would have to be overcome. Thefirst is simply defining the standards forwhat constitutes an acceptable expressionsystem. Approaches for expressingproteins are often target specific and moreintuitive than rigorous, whereas resultsare typically qualitative and often judgedsubjectively. It also seems unlikely thatresearchers will take time to test methodson well-known proteins that are notparticularly interesting, and some

scientists may be even more unlikely toshare unexpected results, which mightprove valuable. Finally, the greatestchallenge might simply be the complexweb of commercial interests andintellectual property that biotechnologyhas become. It is this aspect of the postgenomic era, driven by relentlesseconomic pressures, that might pose thegreatest threat to characterizing thehuman proteome.

David W. Wood

Dept of Chemical Engineering, PrincetonUniversity, Princeton, NJ, USA.e-mail: dwood@princeton.edu

TRENDS in Biotechnology Vol.19 No.10 October 2001

http://tibtech.trends.com 0167-7799/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0167-7799(01)01744-9

376 Research Update

Relieving biologics’ bottleneck

Jon Williams

BioLOGIC 2001 was held at the Noga Hilton,

Geneva, Switzerland from 4–6 June.

With 150–200 protein products in clinicaldevelopment, the clamour for increasedmanufacturing capacity forbiopharmaceuticals has reached anunprecedented level. This is particularlyapparent in Europe where theinfrastructure of the young biotechnologyindustry is still undergoing rapiddevelopment. Given that a new facilitytakes between 3 and 5 years to build, itseems unlikely that current expansionplans will provide sufficient capacity and,consequently, demand will continue tooutstrip supply for some time.

Delegates at this conference weretherefore keen to hear the latestsolutions for averting the crisis and,although there were no quick fixes onoffer, they were not disappointed by thesound advice presented.

Alternative expression systems

One option attendees were keen to explorewas the possibility of using alternatives tothe microbial and mammalian cellsystems traditionally employed. RichardFrancis (GlaxoSmithKline, Greenford,UK) and Frederic Bader (Centocor, Inc.,Malvern, PA, USA) were among thosewho discussed the relative pros and consof adopting a transgenic system.

Centocor, for instance, are currentlyevaluating the merits of transgenictechnology, particularly with regard to the

production of monoclonal antibodies. Thereduced capital outlay and the relative easeof altering production capacity countedheavily in the favour of using a transgenicsystem. However, this was not blinkeredendorsement. Any savings garnered fromadopting a transgenic system would takea minimum of three years to materialiseand, moreover, significant downstreamprocessing would still be required beforethe product was ready for market.

Several concerns were raised during an involved panel discussion.Many delegates feared that publicopinion, although currently amenabletowards the use of transgenic systemsfor medicinal purposes, could swing towards a more hostile position,jeopardising manufacturing programmes.The risk of transmissible spongiformencelaphopies spreading from transgenicanimals, though minimal, was also mooted.

The general consensus, therefore, wasthat transgenic expression systems wouldultimately provide an attractive additionto current biopharmaceutical productionmethods but that they would not providea panacea to the crisis.

Manufacturing strategy

One of the key themes of the conferenceprogramme was the need for biotechcompanies to develop a clearmanufacturing strategy as early aspossible in product development.Joseph Santangelo (Microscience,Reading, UK) stressed that the

manufacturing process is as importantas the product itself. Time spentcharacterising the process before clinicaltrial is imperative if expensive changesat a later stage, particularly post-licence,are to be avoided.

Similarly, Chris Holloway (ERAConsulting, London, UK) stated that theregulatory authorities are more likely toaccept that processes are comparable ifalterations are anticipated and madeearly during development. Unforeseenpost-licence changes could incursignificant delays awaiting regulatoryapproval, leaving the market open forcompetitors to exploit.

The importance of developing a robust process was reinforced by ReginaHodits (Apax Ventures GmbH, Munich,Germany), who said that biotechcompanies would be looked upon morefavourably by investors if they coulddemonstrate a workable manufacturingstrategy.

Selection criteria for contract

manufacturing organisations

Several delegates drew on their ownexperience to offer pertinent advice to thosebiotech companies considering the use ofContract Manufacturing Organisations

‘...transgenic expression systems would

ultimately provide an attractive addition to

current biopharmaceutical production

methods...’