15 November 2009. It’s hard enough to make a drug that does the right thing. Designing compounds to fight neurodegenerative disease comes with the additional challenge of making sure they actually reach the brain. At the Society for Neuroscience annual meeting held 17-21 October 2009 in Chicago, a handful of groups tackled this “problem behind the problem” of drug development by presenting different tactics for overcoming the blood-brain barrier.
Thanks to an intricate, coordinated system of tight endothelial junctions and active transport enzymes, the brain enjoys remarkable protection from pathogens and most blood-borne substances greater than 400 daltons. Unfortunately, this physiological barrier also keeps out the vast majority of compounds that could help sustain and restore the critical organ when it succumbs to Alzheimer disease or any number of central nervous system scourges.
AngioChem Inc., a biotech company in Montreal, Canada, has addressed this problem by harnessing proteins naturally expressed at high levels on brain capillaries. Low-density lipoprotein receptor-related proteins (LRPs) are endocytic and signaling receptors; they help shuttle into the brain large, essential molecules, such as glucose and insulin, which would ordinarily be excluded because of size.
Taking advantage of this transport pathway, researchers at AngioChem have designed a set of peptides that can be attached to candidate drugs, enabling them to hitch a ride with LRP across the blood-brain barrier. These peptides, cleverly called Angiopeps, bind to LRPs and can range in size from six to more than 30 amino acids. In the case of small molecules, the peptides are hooked to the compound of interest using linkers that enable them to be cleaved off once the drug crosses the blood-brain barrier. However, larger biologics—such as enzymes, nucleic acids, and monoclonal antibodies—do not require this release step.
The system is adaptable. “We can have different types of linkers, different length peptides, different numbers of peptides incorporated. There are numerous possibilities,” Betty Lawrence, vice president of development, told ARF. “It’s something that needs to be examined for each project.” The technology is called the Engineered Peptide Compound (EPiC) platform.
Thus far, AngioChem has applied the EPiC system with small anticancer drugs, larger peptides, as well as with small interfering RNAs (siRNAs) and monoclonal antibodies. At SfN, the company presented data showing brain uptake of various EPiC compounds and early clinical trials in brain cancer patients of its lead candidate, ANG1005 (paclitaxel, a mitotic inhibitor used in cancer chemotherapy). With the EPiC platform, three anticancer drugs (paclitaxel, doxorubicin, and etoposide) given intravenously were able to reach the brain five to 10 times faster than they would have without EPiC.
The researchers showed similar enhancement with monoclonal antibodies and siRNAs—large molecules that are far worse at crossing the blood-brain barrier. Ordinarily, these biologics enter the brain >100 times more slowly than glucose, i.e., their transport rates are less than 1 percent of glucose’s. With EPiC, they were getting into the brain at rates ranging from five to 10 percent of glucose’s, CEO Jean-Paul Castaigne told ARF. In Alzheimer disease immunotherapy, how much antibody actually crosses the BBB remains a big, unsolved question even as expensive large-scale trials are under way.
In a Phase 1 trial of 55 people with malignant gliomas, the EPiC-paclitaxel drug had no CNS toxicity and did not trigger an immune reaction when given to patients as two intravenous infusions three weeks apart. The drug was able to reach the brain tumors at therapeutic concentrations (see Drappatz et al. SfN poster abstract). And in a separate Phase 1b/2 study of patients with more advanced disease (i.e., advanced solid tumors and brain metastases), the drug reduced tumor sizes in multiple organs in 15 of 21 participants (see Kurzrock et al. SfN poster abstract).
AngioChem is collaborating with two different companies to develop EPiC compounds for neurodegenerative disease—one a monoclonal antibody (ANG3101), the other a siRNA (ANG3201). Castaigne said he cannot at this point disclose which companies and therapeutics are involved. However, one can make educated guesses. At an AngioChem-sponsored SfN breakfast symposium, Karoly Nikolich, chairman of AngioChem’s scientific advisory board and a consulting professor at Stanford University, Palo Alto, California, specifically mentioned Alzheimer and Parkinson diseases in a talk on how EPiC can be used to deliver substances besides paclitaxel. The next speaker at the symposium was Dinah Sah of Alnylam Pharmaceuticals in Cambridge, Massachusetts. Alnylam, which specializes in therapeutics based on RNA interference technology, has a grant from the Michael J. Fox Foundation to develop novel PD therapies (see company news release).
The monoclonal antibody and siRNA programs are in the early preclinical stage, with only animal data. The company has successfully incorporated the Angiopep backbone onto the target molecules, and has shown that the new products efficiently cross the blood-brain barrier, reach the brain parenchyma, and get to their target cells in mice, Castaigne told ARF. “This is really early research. We have signals that it is going in the right direction,” he said. “We are now improving and optimizing.”
Meanwhile, other researchers are taking a different tack to help therapeutics reach the brain. These methods use high-frequency sound waves to open the blood-brain barrier in a temporary, localized manner. In a SfN slide talk, James Choi (a Ph.D. student in the lab of Elisa Konofagou at Columbia University, New York) described his latest work using focused ultrasound to deliver BBB-impermeable molecules into specific brain regions. After demonstrating they could use the technique to open the blood-brain barrier in the hippocampus of wild-type mice (Choi et al., 2007), Choi and colleagues showed they could do this in an AD mouse model (APP/PS1) without affecting the timing of BBB opening and molecular delivery (Choi et al., 2008). For the new study, presented at SfN and published last week in the journal Ultrasound in Medicine & Biology (Choi et al., 2009), the focus was safety and efficacy (i.e., how large the compounds can be, and how precisely they are delivered). First, the researchers injected into the blood microbubbles—gas-filled contrast agents used in medical sonography—which interact with the focused ultrasound (into the left hippocampus, in this case) and serve as a contrast agent to enhance imaging. Once the BBB is open (it usually remains so for a day), they intravenously injected fluorescent-tagged dextrans of 3, 70, and 2,000 kDa. The 3- and 70-kDa dextrans made it into the brain—specifically into the left hippocampus and not the right. But there was a cutoff; the 2,000-kDa compound did not go through. Concluding his talk, Choi said, “We’ve delivered into the brain pharmacologically sized agents that were localized to the target region using a noninvasive technique.” The scientists are now optimizing the ultrasound parameters for uniform BBB opening; specifically, they are trying to identify exactly how and when it opens, and to target specific brain regions associated with AD and PD (such as the hippocampus and substantia nigra), Konofagou wrote in an e-mail to ARF.
Jessica Jordao is a Ph.D. student working with Isabelle Aubert and Kullervo Hynynen at Sunnybrook Health Sciences Center of the University of Toronto, Ontario, Canada. She used a similar focused ultrasound method for passive immunotherapy in an AD mouse model (TgCRND8). The researchers injected microbubbles, along with a magnetic resonance imaging (MRI) contrast agent and anti-Aβ antibodies into the tail veins of four-month-old mice. The TgCRND8 strain develops plaques by three months. They then applied focused ultrasound to four cortical spots in the right hemisphere of the brain while the left side served as the negative control, and sacrificed the mice at four hours, two days, and four days after treatment. Using immunoprecipitation and Western blot analysis, they confirmed that the Aβ antibodies only found their way into the right side of the brain. By four days, the treatment had brought a 12 percent reduction in number of plaques and mean plaque size, as well as a 23 percent decrease in Aβ surface area, Jordao said.
One difference between the two ultrasound protocols is that the Columbia method does not require MRI for targeting, and is hence less expensive, Konofagou said.—Esther Landhuis.