sequences. The resulting tree, generatedusing a maximum likelihood method, di-vided the enzymes into two clear groupswith the E. coli and A. aquifex (such asHelicobacter pylori) sequences in differentclasses3. When other enzymes were characterized, those that were grouped with A. aeolicus were found to requiremetal for activity, whereas those (likeSalmonella typhi) that were grouped withE. coli, were not. There was no othercommon feature between the enzymesin either class3. Furthermore, PD404182was discovered to be a potent inhibitorof all enzymes in both classes apart fromthe hyperthermophiles. It is thereforehoped that analogues of this compoundcould prove to be active against a widerange of gram-negative bacteria.

Future prospectsRobert Kretsinger (University of Virginia,VA, USA) says ‘Specific KDO 8-P synthase

inhibitors could well be viable drugs, butit is important to remember that onlyone in a hundred promising drugs everreaches the market.’ Woodard agreeswith this cautious assessment: ‘Our in-hibitor does not kill bacterial cells as wellas we would like.’ However, it will be dif-ficult to improve its binding throughchemical modification until a crystalstructure of the enzyme–inhibitor com-plex has been solved. Woodard’s groupis initially concentrating on trying to improve the solubility of PD404182 andits ability to cross the cell wall but, so far,they have not had much success.

However, if these problems can beovercome, PD404182 would have twoimportant advantages as a lead com-pound. First, it has previously beenpatented as a non-steroidal anti-inflam-matory drug, so some data on its toxicity and metabolic properties is avail-able in the literature. Also, compounds

targeting the LPS pathway could be useful as drugs even if they are not cyto-toxic. The lipopolysaccharide layer pre-vents many antibiotics that are activeagainst gram-positive bacteria from en-tering gram-negative cells. It is possiblethat a compound like PD404182, givenas adjuvant therapy with one of theseantibiotics, could break down this layerenough to allow the antibiotic entry intothe cells.

References01 Birck, M.R. et al. (2000) Identification of a

slow tight-binding inhibitor of 3-deoxy-D-manno-octulosonic acid 8-phosphatesynthase. J. Am. Chem. Soc. 122, 9334–9335

02 Duewel, H.S. et al. (2001) Substrate and metalcomplexes of 3-deoxy-D-manno-octulosonate-8-phosphate synthase from Aquifex aeolicus at1.9A resolution. J. Biol. Chem. 276,8393–8402

03 Birck, M.R. and Woodard, R.W. (2001)Aquifex aeolicus 3-deoxy-D-manno-2-octulosonic acid 8-phosphate synthase: anew class of KDO 8-P synthase. J. Mol. Evol.52, 205–214

update news DDT Vol. 6, No. 10 May 2001

1359-6446/01/$ – see front matter ©2001 Elsevier Science Ltd. All rights reserved. PII: S1359-6446(01)01806-2500

Novel flow-based assay technologies de-signed to study the interaction betweenpathogens and host cells could lead directly to new drug discovery targets.LigoCyte Pharmaceuticals (Bozeman,MT, USA) has developed two promisinganti-infective technology platforms: aProteoflow™ system to develop new tar-gets for anti-infective therapeutics andvaccines, and nanoparticle technologythat can be used to provoke an immuneresponse without the risk of infection.

ProteoFlow™ assaysThe flow-based assays originally devel-oped by LigoCyte were inspired by thecell–cell interactions involved in inflam-mation, where white blood cells are

recruited and transported by shearforces in the blood and are required tostop instantaneously at the site of inflam-mation. This ability to attach to inflamedtissue has been subsequently studiedusing a flow-based assay consisting of aglass tube lined with endothelial cellsthrough which blood components areflowed, and observing and quantifyingthe characteristics of the fluorescently labelled white blood cells by video mi-croscopy. By altering the expression ofproteins, such as receptors, on the cellsand adding various purified proteins, antibodies or inhibitors, the interactionsof the immune cells with the endothelialcells can be characterized, and com-pounds that inhibit inflammation can be

studied (Fig. 1). Similar models usingepithelial cells simulate the lining of theintestines and airways.

LigoCyte has now applied this tech-nology to the development of assays tostudy the interaction of pathogens suchas Escherichia coli and Candida albicans,both with host cells and homotypicbinding between the pathogens them-selves. For example, E. coli form coloniesby adhering to other E. coli cells thathave already attached to mucosal epi-thelium, which results in the upregula-tion of virulence factors. Inhibition of theinitial attachment to the host cell would,therefore, reduce the likelihood of colo-nization and further infection at existingsites of inflammation, such as a surgical

A sticky end for pathogens?Joanna Owens, Joanna.Owens@current-trends.com

wound. Using the ProteoFlow assay,pathogens can be passed through artifi-cial blood vessels or along epithelial cellsin the presence of various inhibitors, toreveal the mechanisms used to adhere to the host cells. Robert Bargatze, VicePresident of R&D, LigoCyte, believesthat the identification of adhesins is avery exciting area of research: ‘Adhesinsrepresent virulence factors in terms ofthe organism or its toxin’s ability to adhere and colonize and are the keyplayers in pathogenicity.’

Novel targetsThe adhesion of pathogens via proteinand carbohydrate ligands expressed onthe cell surface is difficult to study accu-rately in a static assay, such as a petridish, because some cells cannot bind inthis environment, and yet are able tobind when studied in a flow-based assay.In fact, some cell-binding interactionsoccur only under flow conditions. Theidentification of ligands involved inpathogen–cell adhesion present novelgroups of targets for therapeutics andvaccines and has great potential for usein drug and vaccine targeting. For exam-ple, if a protein responsible for pathogenattachment to epithelial cells in the gas-trointestinal mucosa is identified, thisprotein could theoretically be used totarget drugs to the epithelial mucosa.Indeed, LigoCyte has used an adhesinfrom reovirus, which infects immune-uptake sites known as M-cells in mucosaltissue, to target a vaccine (orally ornasally administered) to these cells, andthey have achieved a high mucosal IgAimmune response. This is of particular interest because it has previously beendifficult to raise IgA responses from intra-muscular injections. Further, Bargatze revealed that LigoCyte’s partnering of atherapeutic antibody and vaccine has resulted from the identification of a hydrophobic protein involved in thepathogenesis of C. albicans. ‘In an in vitrohuman cell model, an excellent correla-tion between the adhesion of C. albicans

to cells and invasion of the vascular endothelial was observed, which wassubsequently blocked by a monoclonalantibody raised to the protein.’

Nanoparticle technologyLigoCyte has also developed a syntheti-cally engineered virus-sized particle,consisting of a polymerized liposome onwhich antigens can be presented andoriented and used for adhesion, tar-geting and to provoke an immune response. Complex antigens can, there-fore, be mimicked by using a combi-nation of simple ligands presented at different heights and densities, whichcould result in a similar immune responseto that observed after exposure to thepathogen, but without the virulence.The nanoparticles can also be used totarget large amounts of antigen to, forexample, an M-cell, to increase the ‘pay-load’ of antigen delivered to an immuneuptake site. LigoCyte is currently carry-ing out research funded by the National

Institutes of Health (Bethesda, MD, USA)to develop this technology for the inhi-bition of eosinophil recruitment to thelungs of asthma patients.

DDT Vol. 6, No. 10 May 2001 updatenews

501

Schematic representation ofnanoparticle technology.

Figure 1. ProteoFlow™ flow-based assay showing (a) fluorescently labelled neutrophilrecruitment in the presence of a cytokine, tumour necrosis factor-α (TNF-α). (b) Theaddition of a nanoparticle inhibitor (developed by LigoCyte) after 19 min results in arapid decrease in neutrophil attachment to the vascular endothelium.

02468

10121416182022242628303234363840

0 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29Time (min)

Num

ber

of b

ound

neu

trop

hils Infuse nanoparticle inhibitor

Drug Discovery Today

(a)

(b)

Ongoing R&DLigoCyte is involved in several partner-ships with other companies that areusing these technologies in drug andvaccine development. Corixa (Seattle,WA, USA) has licensed both an antibody

and a vaccine based on LigoCyte’s work, and Merck (Rahway, NJ, USA) is evaluating compounds developed forinflammatory bowel disease, differen-tiating lead candidates using LigoCyte’sflow-based assays. Bargatze says that

the next step for LigoCyte is: ‘to take advantage of the targeting to the M-cellsite using both antigens and DNA constructs that encode antigens, for the development of highly effective vaccines.’

update news DDT Vol. 6, No. 10 May 2001

1359-6446/01/$ – see front matter ©2001 Elsevier Science Ltd. All rights reserved. PII: S1359-6446(01)01805-0502

Researchers might have found the mech-anism by which oral contraceptives(OCs) help prevent ovarian cancers.They think it is mainly a direct effect ofthe hormone progestin on the ovarianepithelium, an effect that is unrelated toovulation inhibition. This knowledge ishoped to enable the development ofOCs that provide enhanced ovarian cancer prevention benefits.

Ovarian cancer is the sixth most com-mon cancer among women worldwideand causes more than 100,000 deathsper year1. However, routine OC use hasbeen shown to protect against the dis-ease. Data from large cohort and case-control studies show that the risk ofovarian cancer is decreased by 40% inpast/present users of OCs and by >50%in long-term users (>5 years)2.

Mechanism of OC protectionWhy OC use reduces ovarian cancer riskis not known with certainty, although thegeneral assumption has been that thesedrugs reduce the lifetime number of ovulations. Repeated cycles of ovulationcan cause recurrent damage to ovarian epithelium, and can eventually result ingenetic mutations, triggering cancer.

However, this explanation is not con-clusive. Little information is available on

progestin-only contraceptives, but it ap-pears that this type of contraceptive isalso associated with a reduced risk ofovarian cancer, while often having no effect on ovulatory cycles. These pro-gestins (i.e. synthetic formulations of thefemale hormone progesterone) are in-cluded in OCs because they thicken thecervical mucus, thus making it difficultfor the sperm to reach the uterus or fallopian tube.

Furthermore, long-term OC use hasbeen shown to have a disproportionatelygreater protective effect than can be at-tributed solely to ovulation suppression.This realization prompted researchers at the Duke University Medical Center(Durham, NC, USA) to look for otherbiological effects of OCs that could beresponsible for their protective properties.

Role of progestin?Gustavo Rodriguez and his team exam-ined the ovaries of monkeys that hadbeen treated for three years with OCs.Four groups of cynomolgus macaques(75 in total) were randomized to receiveno hormones, the OC Triphasil (a combi-nation of an oestrogen and a progestincomponent), the oestrogen componentof Triphasil, or the progestin componentof Triphasil. Results showed that treat-ment with Triphasil or its progestin com-ponent, unlike the other two groupstested, resulted in an increased rate ofapoptosis in the ovarian epithelium3

(Table 1).This finding that the hormone could

activate one of the key in vivo mecha-nisms to eliminate cells that are prone tomalignant transformation, was supported

New opportunities for enhancedovarian cancer preventionMartina Habeck, Freelance writer

Table 1. Apoptotic effect of hormone treatment on the ovarianepithelium of cynomolgus monkeys

Hormone treatment Apoptotic cell counts (median percentage)

No hormones (control) 3.9%Oestrogen component of Triphasil 1.8%(ethinyl estradiol)

Triphasil 14.5%Progestin component of Triphasil 24.9%(levonorgestrel)