Irwin McLean's filaggrin readthrough drug could be revolutionary

Recently I predicted that nothing resembling a cure for eczema would appear for at least 25 years. I followed that up with a prediction that if a surprise cure were to emerge, it would be an anti-itch therapy.

At least one reader disagreed, and pointed me to a strategy now being developed at the University of Dundee in Scotland: drugs to stimulate or enable filaggrin expression in patients with one or two defective copies of the corresponding gene, FLG. At least one such drug is in the very earliest stages of drug development, toxicology studies in animals. If the drug succeeds in human clinical trials, we might see it in clinics in about 15 years. Potentially, such a drug could help some of the patients most severely affected by eczema.

The best review on filaggrin I’ve seen was co-written by three authors, two of whose names I am familiar with as among the biggest in the field of eczema research: Irwin McLean, who led the team that linked mutations in FLG to increased risk of developing ichthyosis vulgaris and eczema, and Donald Leung, principal investigator of the Atopic Dermatitis Research Network. (The first author is Alan Irvine, a colleague of McLean’s who works in Ireland.)

FLG is a giant, and unusual, gene, one of the last to be sequenced by the Human Genome Project. It encodes an enormous protein, profilaggrin, which contains from 10 to 12 repeats that are cleaved off into individual filaggrin units. Filaggrin itself has several important roles in the upper layers of the skin: it flattens skin cells into their characteristic final shape; it helps bind these cells together into a barrier; and it breaks down into the acidic “natural moisturizing factor.”

Many mutations have been found in FLG. Interestingly, if you take any particular ethnic group (say Japanese), there will be a characteristic profile of mutations for this group that is likely to be different than the profile for another group (say Scottish).

FLG (read from left to right; red hexagons are filaggrin units.) Common mutations in Asian and European populations. Look how early the nonsense mutation R501X appears. Figure 3A from Irvine, McLean, and Leung's NEJM review

All filaggrin variants are either “nonsense” or “frameshift” mutations in DNA that encodes protein (as opposed to so-called “junk DNA”). In a nonsense mutation, the correct DNA base has been replaced with a wrong one, and the upshot is that the protein-making machinery, known as the ribosome, runs into a code that it doesn’t recognize; it can’t add an amino acid to the growing protein, and it stops decoding profilaggrin at that point.

If the nonsense mutation occurs early enough, before the first filaggrin unit in FLG, no filaggrin gets made at all.

I believe that with a frameshift mutation (in which one or more DNA bases are missing or added) the protein is very likely to be terminated soon afterward. The end result is the same: little or no filaggrin.

In his 2006 Nature Genetics paper, McLean and his group identified two mutations, R501X (nonsense) and 2282del4 (frameshift), which have turned out to be the most common in Caucasian populations. Both R501X and 2282del4 occur early in the first filaggrin repeat. That means that if you have one such mutation, you will have one good copy of FLG and one bad copy; and if you have mutations on both your copies of FLG, you won’t have any filaggrin at all, and there is a high chance that you have eczema.

Soon after their 2006 discovery, McLean and first author Frances Smith applied for a US patent,  “Prevention/treatment of ichthyosis vulgaris, atopy and other disorders.” The patent, number 8,338,386, was granted only recently, on December 25, 2012. It makes many claims, all relating to the ability of five antibiotic drugs, or potentially tRNA molecules, to force the ribosome to read through nonsense mutations.

McLean and Smith’s patent is aimed at the nonsense mutation R501X (they say so in the patent), because it occurs so early in the gene. Such a drug would also work for nonsense mutations later in the sequence.

The drug or agent would not work on frameshift mutations such as 2282del4, as I understand it.

One cool thing about a readthrough drug would be that it would have its strongest effect on people who had nonsense mutations on both copies of FLG. A readthrough drug would theoretically make both copies functional. People with only one mutated copy of FLG would still benefit from having that number increased to two.

In the patent,“atopy” is mentioned in the title and the background information, but neither “atopy” nor “atopic dermatitis” appear in any of the 14 explicit claims. I don’t know whether this matters. I'm not an IP lawyer. It seems curious that the inventors left it out though.

Another curious fact is that only five specific drugs are mentioned: gentamicin, paromomycin, neomycin, tobramycin and negamycin. There is no discussion of the drug discovery process—any tweaking of molecules for greater effect or less toxicity—that I can see. I don’t know whether this matters either.

From what I can tell, McLean’s group has at least one of these compounds, which has most likely been altered from its original structure, in “toxicology” (which means testing in mice, rats, etc.) (see page 16). If you look at this handy graphic provided by the FDA, you will see that toxicology studies are step 2 of 12 in the drug discovery-to-market process. So it is extremely early days and you have to keep in mind that, as I keep saying, almost all drug candidates fail at some stage of clinical testing.

A relevant factor is also that currently there is no FDA-approved drug that acts by this mechanism—by causing the ribosome to ignore nonsense mutations. Ataluren, a drug to cure Duchenne muscular dystrophy and cystic fibrosis, is a readthrough drug in phase III clinical trials. The FDA is apparently especially careful when approving drugs that are “first-in-class,” or the first of their general type. So it is worth following Ataluren’s progress closely; its success could mean that we might see an eczema readthrough drug sooner.

Hat tip to Anonymous (you know who you are)

Where to look for a surprise eczema cure to emerge

As I wrote in the previous post, the outlook is bleak for new eczema therapies that might qualify as a “cure.” On the fronts of barrier protection and repair and anti-inflammatories, nothing revolutionary is in the works apart from, perhaps, dupilumab, Regeneron’s antibody to IL-4. I can’t see anything emerging from research and entering and successfully exiting clinical trials for at least 25 years.

What might I have left out of this discussion? Where could a surprise come from?

Itch. Itch was the area that occurred to me. Imagine being able to break the itch-scratch cycle in eczema. You know what it’s like: your skin flares up and the itch becomes unbearable. You scratch to get relief. Sometimes you scratch in your sleep. Then your skin is torn up, which for a start can be embarrassing, but also often leads to infection. If there were no itch to begin with, eczema might never become anything more than a minor rash. Its impact on quality of life would be greatly minimized.

I believe we might see a convergence of two major trends that would result in a new anti-itch drug that patients could take in pill or cream form.

The first trend: In the past few years I have seen a number of papers describing newly identified neurons that transmit the sensation of itch, distinct from pain. The experiments were done on animals such as mice and cats; I don’t think these neurons have been found in people yet. But you can bet there are many scientists beavering away to be the first in the field.

Turning on or blocking neural receptors is what drugs do best. Think anesthetics. These itch neurons, if found in humans, are likely going to have receptors similar to those in other animals, and the search will be on to find drugs that block the receptors.

(You could also imagine a therapy using RNA interference to prevent neurons in the skin from making itch receptors in the first place.)

The second trend: scientists are developing powerful new techniques to speed the drug discovery process. While it does take around 15 years to take a new drug all the way through clinical trials to FDA approval, the path is shorter for “repurposed” drugs (such as Viagra, originally planned as a heart medication). The barrier is lower because the drug has already been proven nontoxic. Repurposed drugs have been approved as treatments for one condition but have side effects that, depending on your perspective, qualify as primary effects. There could well be an FDA-approved anti-itch drug out there already. It’s just being used to treat toenail fungus.

A company I am familiar with (I know the founders), SeaChange Pharmaceuticals, developed a rigorous way to search through databases of drugs and identify potential side effects or secondary uses, based on the chemistry of the protein targets for the drugs. (Wired magazine named SeaChange’s technology one of the top 10 breakthroughs of 2009.)

The idea would be that scientists would identify itch neurons in humans, and pin down the itch receptor; then somebody at Pfizer or Novartis or whatever would use a SeaChange-like technique to find FDA-approved drugs that block the receptor. Presto: no more itch. Conceivably this might happen within a decade.

Now, evidently these new drug discovery techniques could be applied in the areas of anti-inflammatories, or barrier repair. I think, though, that itch is a prime candidate for a surprise eczema “cure” because it’s likely that the itch sensation comes down to a single receptor. Blocking that receptor by a conventional drug will be a relatively simple task, compared to controlling inflammation without leaving the patient vulnerable to infection, or taking on the dubious task of compensating for a defective skin barrier in infants.

That’s my opinion.