Stretching neuronsSmith and colleagues calculated howfast axons could grow and used theseparameters to find the limits of neurongrowth and the optimum speed ofgrowth for viability. Neurons were platedonto adjacent membranes and allowedto integrate across a 50 µm border between the membranes. The neuronswere then progressively stretched byusing a microstepper motor system toseparate the two membranes at a rate of3.5 µm per 5 min. The neurons not onlystretched to at least 1 cm in length after10 days, but also formed bundles com-prised of thousands of axons, thus gen-erating a nerve fibre structure similar tothat found in vivo (Fig. 1). This repre-sents an insight into the mechanism ofgrowth of white-matter tracts during development.

Smith says this was an unexpected find-ing: ‘The first thing I had in my mind wasthat we would have a long single cell, butinstead, they kept grouping more andmore together, coalescing into larger andlarger bundles that you can actually seewith the naked eye, about the size of ahuman hair.’ Smith thinks that this is partof what occurs in development: first, youhave a group of neurons signaling to eachother and as they grow, they cluster to-gether to form bundles. This was an im-portant goal for Smith and colleagues: to accomplish not only length, but number as well because, as with many cell types, itcan be speculated that strength in numberscan afford growth advantages to the cells.

Future studiesSmith and colleagues now intend tostudy the mechanisms of stretch-induced

growth and will be testing theirstretched-axon technology in animalmodels to show that these neuron graftsare connecting with the host tissue witha complete connection; this could bedemonstrated by flowing a dye acrossthe graft region and showing that anelectrical current is able to traverse thegraft. The ultimate aim is to developsuccessful transplant therapies for neurodegenerative diseases such asParkinson’s disease, CNS injury andoptic nerve damage, and also to expandtheir studies to include peripheral nerverepair.

Reference1 Smith, D.H. et al. (2001) A new strategy to

produce sustained growth of central nervoussystem axons: continuous mechanicaltension. Tissue Eng. 7, 131–139

DDT Vol. 6, No. 12 June 2001 updatenews

1359-6446/01/$ – see front matter ©2001 Elsevier Science Ltd. All rights reserved. PII: 1359-6446(01)01851-5 601

The end for diabetic kidney disease?Kathryn Senior, Freelance writer

Preclinical studies suggest that an in-hibitor of amadorase, the enzyme at thecentre of a newly discovered metabolicpathway that contributes significantly tothe development of diabetic nephrop-athy, could offer treatment hopes to dia-betic patients. The compound, DYN12,could help to delay, or even prevent, the onset of serious kidney problems.This approach is one of several currentlybeing developed to tackle the range oflife-shortening complications that arisein diabetics. ‘Antioxidants and inhibitorsof the β isoform of protein kinase C arebeing considered for cardiovascular andrenal complications, and specific anti-angiogenic drugs, such as integrinantagonists, are being developed forproliferative diabetic retinopathy,’ saysMichael Brownlee (Albert Einstein College

of Medicine, Bronx, NY, USA). However,Brownlee is impressed by the potentialof DYN12 and considers it to be ‘one ofthe more exciting new strategies’.

What is amadorase?Amadorase, a previously unknown fruc-tosamine kinase, was discovered, puri-fied and characterized by scientists wholater founded Dynamis Therapeutics(Wyndmoor, PA, USA). ‘We have shownthat amadorase is responsible for theproduction of 3-deoxyglucosone (3DG),a highly reactive dicarbonyl sugar that isa precursor to the advanced glycationend-products (AGEs),’ explains AnnetteTobia, President and CEO at Dynamis.AGEs were first shown by Brownlee toform on the surface of proteins, and to enable them to cross-link1. These

cross-linked proteins cause much of the damage to the glomerular basementmembrane that leads to problems in the kidneys of patients with diabetes.Previously, it was thought that 3DG re-sulted exclusively from non-enzymaticrearrangement, dehydration and frag-mentation of a fructoselysine-containingprotein (Fig. 1). Tobia and colleagueshave now shown that 3DG is also pro-duced as a by-product of a pathway that recovers lysine from fructoselysine(Fig. 1). ‘This enzymatically-controlledprocess is the major source of 3DG in the body,’ confirms Tobia.

There is considerable clinical and experimental evidence that 3DG is amajor factor in the development of dia-betic nephropathy. For example, elevatedlevels of 3DG and 3-deoxyfructose (its

detoxification product) are found in theplasma and urine of diabetics but not innon-diabetics2. 3DG is particularly ele-vated in patients with signs of diabeticnephropathy. Also, Pimagedine (Alteon;New Jersey, NJ, USA), an inhibitor of AGEcross-linking, has been shown to reduceAGE associated renal pathology in ani-mal models. ‘Phase III clinical trials ofPimagedine showed that AGE inhibitionled to a significant reduction in progres-sion of diabetic nephropathy, comparedwith conventional treatment using angiotensin-converting enzyme (ACE)inhibitors,’ explains Brownlee. The randomized, double-blind ACTION 1 (A Clinical Trial In Overt Nephropathy)trial involved 690 patients at 56 clinicalsites in the USA and demonstrated a28% reduction in nephropathy, suggest-ing that Pimagedine also slowed the progression of retinopathy3.

Preclinical studiesAfter isolating and purifying amadorase,scientists at Dynamis showed that DYN12,

an effective in vitro inhibitor of amado-rase, could reduce 3DG levels in theplasma of diabetic rats. In the treatedgroup, the rats showed a 53% reductionof 3DG levels compared with a controlgroup that received only saline. ‘Theability of DYN12 to reduce systemic 3DGlevels suggests that other diabetic com-plications could also be potentially treat-able. We are already starting to look atatherosclerosis and nerve complicationsbut don’t currently have the resources toinvestigate retinopathy,’ reports Tobia.‘Phase III trials indicate that blocking theeffects of AGEs has a significant effect ondiabetic renal complications; becausethe role of DYN12 will be to prevent theformation of AGE precursors, rather thanto block them once they have been produced, the Dynamis approach couldhave even greater efficacy,’ predictsBrownlee. However, the competitionfrom Alteon is hotting up. In January itannounced preclinical results for its ownAGE-formation inhibitor, ALT946, show-ing that this drug was more effective

than Pimagedine, both in vitro, and in vivo in the diabetic rat model4.

Future studiesBoth companies are well aware that theclinical impact of an agent that couldprevent diabetic nephropathy and couldpotentially reduce the incidence of otherdiabetes-associated complications wouldbe enormous. ‘Diabetes is the most com-mon cause of renal failure, blindness andlower-limb amputation in the workingpopulation – it is also a major risk-factorfor heart attacks, with about one-third ofall heart attacks occurring in diabetic pa-tients,’ says Brownlee. Dynamis is, there-fore, urgently searching for a corporatepartner to extend its development ofDYN12. ‘Although preclinical work iscontinuing, we are not yet able to em-bark on the two-year toxicology studyrequired before clinical trials couldbegin. As soon as we form a partnershipto fund the next stage of the work, thefirst stage of clinical investigations couldbegin within two-and-a-half years,’ saysTobia. If this is done successfully, Tobiaanticipates that patients will be testedfor elevated levels of 3DG as a marker fornephropathological risk. ‘Only patientswho are likely to develop kidney com-plications as a result of their diabeteswould need to receive treatment, but a truly preventative therapy could benefit millions of people worldwide,’she predicts.

References1 Brownlee, M. et al. (1986) Aminoguanidine

prevents diabetes-induced arterial wallprotein crosslinking. Science 23, 1629–1632

2 Wells-Knect, K.J. et al. (1994) Deoxyfructoseconcentrations are increased in humanplasma and urine in diabetes. Diabetes 43,1152–1156

3 Appel, G. et al. (1999) Pimagedine lowerstotal urinary protein and slows progressionof overt diabetic nehprothapy in patientswith Type 1 Diabetes Mellitus. AmericanSociety of Nephrology 32nd Annual Meeting,Miami Beach, FL, USA, 7–8 November 1999

4 Forbes, J.M. et al. (2001) Renoprotectiveeffects of a novel inhibitor of advancedglycation. Diabetalogia 44, 108–114

update news DDT Vol. 6, No. 12 June 2001

602

Figure 1. 3-Deoxyglucosone (3DG) pathways. A reversible reaction between glucose andthe amino groups of lysine-containing proteins initiates the formation of the stableketoamine, fructoselysine. Significant amounts of 3DG are produced enzymatically fromfructoselysine by amadorase. 3DG is either detoxified or goes on to form harmful AGEs.Pimagedine converts 3DG to harmless triazines.

Drug Discovery Today

Glucose + lysine

Fructoselysine

Fructoselysine-3-phosphate

Amadorase (a fructosamine kinase)

3-Deoxyglucosone (3DG)+ lysine + Pi

Detoxification by aldehyde

reductase

3-Deoxyfructose (3DF)

Advanced glycationend-products (AGEs)

Diabetic nephropathyand other complications

Pharmacologicaldetoxification by

Pimagedine

Triazines

Less importantnon-enzymatic conversion