By David N. Leff

Developers of a blood-thirsty, tumor-hungry anticancer reagent call their novel molecule ¿icon¿ - short for immunoconjugate. Its designers mixed and matched two natural components to concoct icon, which has a track record in mice of killing prostate and melanoma tumors. It owes part of its inspiration to the common camel.

Besides this Camelid concept, ¿icon¿s recipe whips up Factor VII and the Fc region of the antibody,¿ explained molecular geneticist Alan Garen, at Yale University in New Haven, Conn. ¿The other components that enter into the act when icon circulates in the blood are tissue factor as its target, plus two parts of the immune system that have Fc receptors ¿ the natural killer cells ¿ and complement. That mixture is responsible for destroying the blood vessels to which the icon has attached.

¿TIssue factor, our targeted molecule,¿ he continued, ¿is known for not being expressed on normal growing blood vessels ¿ the ones that respond to angiogenesis. The endothelial cells that line the inner surface of the blood vessels in the tumor vasculature do express tissue factor.

¿Endogenous Factor VII,¿ Garen continued, ¿is an essential protein in the clotting cascade of the blood. We¿ve converted Factor VII into an antibody-like molecule by attaching it to the Fc domain of the antibody. Now when it binds to tissue factor, instead of a blood clot, it causes the immune system to attack the cells that it is bound to. And this killing agent is the icon, which we produce in the laboratory.¿

Enter the camel¿s contribution:

¿There was a very startling discovery five or six years ago,¿ Garen recalled, ¿that camels, unlike the rest of the known mammalian world, make antibodies that are different from all the other mammals. We humans, and mice, make antibodies with both a heavy chain and a light chain. Camels have got rid of the light chain, and part of the heavy constant region. So they make an antibody that works very well in that way. We used it as a model for making our icons, which in some ways look like a camel¿s version of an antibody: It has only two chains, each one with the Fc region of the heavy chain, instead of what is normally the variable region that gives an antibody its specificity.¿

Camelid Antibody Structure Eased Icon Design

¿That¿s thrown away, and replaced by FVII, which now gives the icon its specificity,¿ Garen explained. ¿There¿s no camel protein in there at all. It just gave us the inspiration. It allows us to make the icon molecule easily. We don¿t have to bother with the light chains at all. So that producing the genes, then producing the molecule, was a much simpler process.¿

Garen is senior author of an early release paper in today¿s Proceedings of the National Academy of Sciences (PNAS), dated Oct. 9, 2001. Its title: ¿Targeting tissue factor on tumor vascular endothelial cells and tumor cells for immunotherapy in mouse models of prostatic cancer.¿

¿We grew prostatic tumor cells under the skin of immunodeficient SCID mice,¿ Garen told BioWorld Today. ¿They accepted these human cells, and allowed them to grow without rejecting them. Then when the tumors had grown quite large,¿ he recounted, ¿we injected them with the vector. We used a gene-delivery vector that could establish in vivo factories for synthesizing the icon. The one that seemed most promising was the adenovirus [AV], which had been widely tested as a gene therapy vector. The tumor cells that become infected acted as the factories producing the icon molecules and secreting them into the blood.

¿We chose to inject the vector directly into the tumors,¿ Garen recounted, ¿because tumor cells have receptors to the AV. When we did this, we avoided at least one potential complication ¿ that the virus might infect normal cells. So the infection was limited to the tumor cells. And those tumor cells, as long as they survived, were producing this icon molecule, and secreting it into the blood.

¿Because the icon cannot act within the cell, it had to get out,¿ Garen went on. ¿Because the TF was being expressed by the blood vessels, it had to get into the blood system. The icon then circulated in the blood, and in that way reached all tumors that might be vascularizing in the body, not just the one that we injected.

¿We stopped the experiment about six months after the last injection. At that point there had been no growth of tumors that on the surface appeared to have been eliminated. Since we were working with a prostate tumor, which produces PSA ¿ prostate-specific antigen ¿ that gave us a convenient measure of large tumor cells by their PSA titer. We also looked at the tumor site, and what we saw were some residual necrotic tumor cells, but no evidence of any live cells. So that was nice.¿

Next on the Yale team¿s agenda are clinical trials of icon¿s tumor-killing potential. ¿We¿ve been collaborating with two clinical colleagues,¿ Garen went on, ¿one former chief of gene therapy here at Yale until a few months ago, when he moved to San Diego as president of the Sidney Kimmel Cancer Center. He is arranging clinical trials using the same icon protocol, beginning with Stage 4 melanoma patients. If FDA approves the IND application, which will be requested a little before the end of this year, the trial would start early next year, in San Diego.¿

Melanoma Icon Trial In 2002, FDA Willing

¿We would inject the same AV vectors,¿ Garen said, ¿but using a human F VII, and an AV vector that encodes this human icon molecule, which should be ready some time in November.

¿In addition,¿ he continued, ¿we have another collaboration with a colleague at the University of Louisville, Kentucky, who is chairman of their ophthalmology department. He¿s tested icon in his animal models of macular degeneration, both mice and pigs. He has demonstrated that there is very impressive specificity of icon binding to the blood vessels that have grown into the retina in his models, and no binding to normal blood vessels.

¿Our icon has the potential, which we¿re actually testing,¿ Garen concluded, ¿as a molecule that can find blood vessels at a very early stage growing into a potential tumor. So diagnostically it has some hope in identifying early cancers.¿