For treating brain disorders, it is not only necessary to develop a therapeutic. That therapeutic then has to get where it's needed.

The brain is not alone in posing delivery challenges, as a generation of RNAi researchers trying to get their therapeutics to anywhere other than the liver can attest. But the brain is unique in having its own fence, in form of the blood-brain barrier (BBB), a layer of cells with extremely tight junctions that keeps many things – including toxins and infectious agents, but also including therapeutics – away from the brain.

No brain is an island, of course, and so there are transport molecules whose job it is to get things across the blood-brain barrier via a mechanism called transcytosis.

Genentech Inc. has a research program devoted to identifying promising transport molecules. In the Dec. 10, 2015, online issue of Neuron, company researchers reported several new receptors that could fit the bill.

The research was motivated by several factors, corresponding author Ben Chih told BioWorld Today.

Earlier studies on one such transport system, the transferrin receptor, unexpectedly revealed temporary toxicity because antibodies targeting the transferrin receptor also affected red blood cells, leading to transient anemia in mice treated with that antibody. (See BioWorld Today, May 6, 2013.)

There was also basic scientific curiosity about which other such pathways might exist in the brain, and the feeling that Genentech was well situated to explore the question. "Antibody engineering is one of Genentech's strengths," Chih said, and the company culture, as well as its size, allows for collaborations that would be difficult, if not impossible, at most academic institutions.

In the studies now published in Neuron, Chih and his colleagues began by looking at two proteins, lipoprotein receptor-related protein 1 and the insulin receptor. Both receptors are well studied and are widely considered to be promising candidates for brain transport. But in their experiments, the researchers found that when they raised antibodies against those receptors, those antibodies were not delivered into the brain in clinically relevant amounts, because the receptors' expression levels in the cells that line the BBB were not high enough.

"They just don't have the capacity," Chih said.

The team next looked specifically for proteins that were expressed at extremely high levels within the BBB. An mRNA-based approach was likewise unsuccessful. But when the team applied proteomics to the problem, it was able to identify three transmembrane proteins that were present in high levels – an extracellular matrix metalloproteinase inducer, the glucose transporter Glut1 and the solute carrier CD98 heavy chain (CD98hc).

Tracing studies showed that of the three antibodies, CD98hc delivered the highest levels of antibodies to the brain after a peripheral injection into animals, and so the team focused efforts on developing a bispecific antibody that targeted CD98hc and beta-secretase (BACE), the rate-limiting enzyme in the processing of amyloid precursor protein to beta-amyloid.

A bispecific CD98hc/BACE-targeting antibody developed by the Genentech scientists was also taken up into the brains of mice, and it did not alter the expression levels or function of CD98hc.

Chih noted that the current form of the antibody is "not meant to be the final therapeutic drug," In principle, BBB-crossing antibodies could take several forms, including that of antibody-drug conjugates.

BACE is a drug target for Alzheimer's disease. Cortellis Clinical Trials Intelligence lists two small-molecule inhibitors of BACE in phase III trials, Merck & Co Inc.'s verubecestat (MK-893) and Astrazeneca plc's AZD3292. Two more, JNJ-5486911 (Janssen Pharmaceuticals Inc.) and E-2609 (Biogen Inc./Eisai Co. Ltd.) are in phase II.

The team is now looking for antibodies with the optimal affinity to deliver antibodies to the brain via CD98hc. From earlier work on the transferrin receptor, it is clear that antibodies that bind too weakly don't get transferred into the brain – but antibodies that bind too strongly are no better from a therapeutic standpoint, because they get transferred into the brain and then transferred straight back out. (See BioWorld Insight, June 27, 2011.)

"It's like a Goldilocks problem," Chih said.