Skip to main content
Premium Trial:

Request an Annual Quote

UCLA s William Pardridge on the Blood-Brain Barrier and RNAi In Vivo


At A Glance

Name: William Pardridge

Position: Professor, medicine/ endocrinology, University of California, Los Angeles

Background: Assistant professor, UCLA — 1978-1985; Fellow in endocrinology, UCLA — 1976-1978; Medical resident, Boston University — 1974-1976; MD, Pennsylvania State University — 1974; BS, chemistry, UCLA — 1969

Like the work of so many other researchers, William Pardridge’s foray into RNA interference came about as a result of his realization that the gene-silencing technology — being developed by others — could be highly complementary to his own work.

Pardridge, who is developing approaches to get drugs through the blood-brain barrier, recently spoke with RNAi News about his work.

Could you give a little background on [the] focus of your work?

Our work started in the 1980s when we sought to develop new approaches to the delivery of drugs across the blood-brain barrier. We were the first to show that there were receptor systems on the blood-brain barrier that acted as transport systems.

For example, the blood-brain barrier expresses an insulin receptor or a transferrin receptor. These receptors cause receptor-mediated transcytosis of the circulating transferrin or insulin from the blood into the brain. We realized that based on that finding in basic biology, one could innovate new approaches to drug delivery using what we call molecular Trojan horses. These would be either endogenous peptides or a peptidomimetic monoclonal antibody that binds the blood-brain barrier receptor systems and piggybacks across on this transport system carrying the drug that was attached to it.

In the 1990s we reduced this to practice for over half a dozen recombinant proteins or antisense agents. When I say reduce to practice, I mean in vivo pharmacologic practice.

The blood-brain barrier is the paradigm of translational research, and to me translational research [involves] those areas of basic science that allow you to move from Petri dish to people. There are a lot of areas of biomedical research that are reduced to practice strictly in Petri dishes and have no chance of ever being reduced to practice in people because of delivery issues — RNA interference is almost at the head of the line at that.

In 2000, we sought to use our molecular Trojan horses to carry non-viral plasmid DNA into the brain and we brought in liposome technology in addition to the molecular Trojan horse technology. What we created were pegylated immunoliposomes, or PILs. A PIL is an 85- to 100-nanometer anionic liposome that has the plasmid DNA fully encapsulated in the interior of the liposome. This is to be contrasted with conventional cationic liposomes — our PILs are completely different.

The surface of the liposome is coated with several thousand strands of 2,000 Dalton polyethylene glycol, or PEG. This creates a pegylated liposome and essentially coverts the surface of the fatty nano-container from that of a ping-pong ball to that of a tennis ball. Changing the consistency of the surface eliminates absorption by serum proteins, eliminates rapid uptake by cells lining the reticulo-endothelial system, and allows the pegylated lipsome to have a prolonged blood residence time with optimized pharmacokinetics. Pharmacokinetics, by the way, is also another aspect of translational science that is completely absent in the Petri dish but plays a prominent role in drug development in animals and people.

Now, this pegylated liposome itself would be biologically inert — it has no targeting properties, and one should not expect that significant gene expression could be achieved in vivo by just simply injecting plasmid DNA encapsulated inside a pegylated lipsome.

The pegylated liposome has to be targeted, and here we conjugate to the tips of 1 to 2 percent of the PEG strands a receptor-specific targeting monoclonal antibody, and we use monoclonal antibodies to either the transferrin receptor or the insulin receptor. This antibody binds that receptor [onto] the blood-brain barrier, or on any cell, and triggers receptor-mediated uptake of the entire liposome construct, and the DNA goes along for the ride not being recognized as such by the endothelial cell.

Once it gets inside the brain and crosses what we call barrier one, which is the capillary endothelial barrier or blood-brain barrier, the antibody acts a second time to trigger transport across the brain cell or tumor cell membrane, owing to expression of the insulin or transferrin receptor on the plasma membrane of the brain cell or brain tumor cell.

In the case of [the] transferrin system, the liposome, which is comprised of fusogenic lipids, fuses with the endosomal membrane to release the plasmid DNA into the cytosol, where it then defuses into the nucleus.

In the case of the insulin receptor, the insulin receptor normally serves to take its endogenous ligand, insulin, all the way to the nuclear compartment. It also mediates direct nuclear transfer of the PIL carrying the DNA to the nuclear compartment. For this reason, we always get about a 10- to 50-fold higher level of gene expression using the insulin receptor rather than the transferrin receptor, although the transferrin receptor works very well.

I might point out as a side [note] that these molecular Trojan horses that we use are species-specific. So, we use a specific transferrin receptor antibody for gene delivery in mice [and] a different transferrin receptor antibody for gene delivery in rats. For gene delivery in Old World primates such as Rhesus monkeys, we use the insulin receptor antibody, and we’ve genetically engineered the insulin receptor antibody to allow for use in humans.

So when did RNAi come into the picture?

Well, we started a logical series of events where we first worked with reporter genes, beta-galactosidase and luciferase, and we demonstrated that you could get global expression of a transgene throughout the entire primate brain with a simple intravenous injection of the non-viral formulation.

We then went from reporter genes to two different types of models: One, experimental Parkinson’s disease, and second, experimental human brain cancer in SCID mice. We first developed an expression plasmid that produced a 700-nucleotide antisense RNA against the human epidermal growth factor receptor, which plays a prominent role in solid cancer including brain cancer. We showed that we could deliver this EGFR-antisense RNA expression plasmid to intracranial brain cancer with weekly intravenous injections, and we could double the survival time of the mice with a very large human brain cancer.

We published that paper in 2002, [meanwhile] RNAi was just coming into its own — Science magazine named it [breakthrough] of the year at the end of 2002 — and adapting RNAi methodology to our delivery system was a logical extension.

We first set up a luciferase brain tumor model where we permanently transfected rat glioma cells with [the] luciferase gene. We then implanted those tumor cells into the brain of Fisher rats, and these animals developed luciferase-producing intracranial brain tumors. We then developed an expression plasmid driven by the U6 promoter that would produce a short-hairpin RNA directed against a defined sequence of the luciferase mRNA. We packaged that in our PILs, injected it into the tumor-bearing rats, and we found that we could cause a 90 percent knockdown of the luciferase gene with a single intravenous injection that persisted for at least five days. That was really the first demonstration of RNA interference in [the] brain following an intravenous injection.

I might point out that since we’re a delivery lab, we realize that … one could [take] two different approaches with RNAi — that being an RNA pathway versus a DNA pathway. Although you could get RNAi in Petri dishes with RNA-based therapeutics, that would never work in vivo because the RNA is so degraded. So, if you really are serious about solving delivery problems in vivo, then you are almost obligated to go to a DNA-based form of RNAi, which is expression plasmids producing short-hairpin RNA. [This] is analogous to microRNAs, and replicates probably what goes on in the nucleus with humans having maybe hundreds of microRNA genes.

Anyway, after we reduced to practice the luciferase brain tumor study, we wanted to get a more definitive endpoint and show that one could use intravenous non-viral RNA interference-based gene therapy to prolong survival time in brain cancer.

Our paper will be published in the June issue of Clinical Cancer Research, and it shows that you can double the life span of mice with weekly intravenous RNAi-based gene therapy directed against the EGFR providing that you use an expression plasmid that produces a short-haripin RNA against an appropriate or biologically active sequence of EGFR mRNA. We found that only about one out of five potential sequences gave the maximal knockdown.

After seeing those kind of data come back, what’s next? Where do you go with this program?

It should go into human clinical trials in conjunction with what we call polygenic gene therapy. In other words, you should use RNAi to knock down tumor-causing oncogenes and then use gene delivery to knock in mutated tumor-suppressor genes. We think that RNAi should be coupled with other forms of gene therapy that knock in genes, such as p53 or p10, that are mutated in cancer.

Are human trials something that you’re looking to get involved in?

I founded a biotechnology company called ArmaGen Technologies … and one of our products is called AGT-2000, which is a form of gene therapy for brain cancer that is antisense based.

In terms of any RNAi-based therapy, is that something ArmaGen is going to be working on?

We’d be happy to collaborate with partners that are interested in delivery of RNAi.

There are a few companies that are looking into this approach of expression of shRNA …

Yeah, but they’re using delivery systems that will never work in vivo. Viral vectors are extremely problematic.

There are some that are using DNA plasmids …

The only way you can deliver DNA plasmid other than with the technology we’ve developed is with cationic liposomes, and those aggregate in serum, precipitate, and embolize in the lung.

Nobody is ever going to achieve DNA-based RNAi with cationic liposomes following intravenous injection unless your target is the pulmonary endothelium.

Have you actually talked to anybody at this point about a possible collaboration?

No. It’d be up to them to contact us.

The Scan

Pig Organ Transplants Considered

The Wall Street Journal reports that the US Food and Drug Administration may soon allow clinical trials that involve transplanting pig organs into humans.

'Poo-Bank' Proposal

Harvard Medical School researchers suggest people should bank stool samples when they are young to transplant when they later develop age-related diseases.

Spurred to Develop Again

New Scientist reports that researchers may have uncovered why about 60 percent of in vitro fertilization embryos stop developing.

Science Papers Examine Breast Milk Cell Populations, Cerebral Cortex Cellular Diversity, Micronesia Population History

In Science this week: unique cell populations found within breast milk, 100 transcriptionally distinct cell populations uncovered in the cerebral cortex, and more.