Bob Williamson AO, FRS
Professor of Medical Genetics
University of Melbourne
Melbourne, Australia

When the cystic fibrosis gene (CFTR) was finally isolated and described by Lap Chee Tsui and his colleagues in 1989, everyone thought that it would be easy to use a normal copy of the gene to treat the disease. We should have realised that the human body resists the incorporation of genes from outside the body very strongly, to protect itself from gene invasion by other animals or plants. In spite of the efforts of many groups using viruses or fats to carry the normal CFTR gene into cells of CF patients, we still have not succeeded in developing this new form of treatment. Having said that, we also know that therapy is now much better than 20 years ago, not only in terms of life expectancy, but in better quality of life for most of those with CF.

“I believe that we will see gene and cell therapy in the coming few years, but not immediately.”

Anyone who studies drug development knows that it takes about twenty years from the idea of any new medicine to the marketing of a product, so perhaps we should not be surprised at this slow rate of progress. Scientists and doctors, just like patients, get excited at new ideas and forget the time it takes to prove the drug works, find the best dose, define how stable it is at high and low temperatures and in wet and dry conditions, and most important, prove it is safe. The time scale is just as relevant for new proposals to use genes or stem cells for therapy. I believe that we will see gene and cell therapy in the coming few years, but not immediately.

The first attempts at gene therapy used “onco-retroviruses” (based on mouse cancer viruses) or “adenoviruses” (based on human common cold viruses). When used to treat children with immunodeficiency (“bubble tent children”), retroviruses were shown to cause cancer, and while there are reasons to think this might not happen with CF, the lungs would be very difficult to infect with these viruses in any case. Two groups in the United States used adenoviruses to carry the CFTR gene into the cells lining the lungs. The trials were not successful, and the adenoviruses caused problems for some patients with CF when doses were increased. Complications from adenovirus led to a tragic death in a trial for another disease.

In any case, the normal copy of the CF gene that can be taken into the cells by the virus is very small and only contains the “coding sequence”, or instructions for the protein, with none of the regulatory regions. It is a bit like putting an engine into a car that could only go forward at one speed, with no brakes or accelerator. At the same time, groups in London (St Mary’s and the Brompton Hospital) attempted CFTR gene therapy using non-viral fatty coatings to get the gene into the cells lining the nose and upper airways. While this was shown to be safe, the improvement in the cells' ability to correct the CF defect by transporting salt and water to the surface of the airways was small, inefficient and short-lived.

After these trials, carried out by many groups in different countries, much of the hard work moved away from patients and into the laboratory with animal and cell models, to develop new ways to move forward with more chance of success. Now, ten years later, there is movement again in the gene and cell therapy field.

“Now, ten years later, there is movement again in the gene and cell therapy field.”

NEW ADENOVIRUS AND AAV APPROACHES

The first adenovirus experiments caused problems because transfer of CFTR to lung cells is inefficient, and continuing expression of genes from the virus, which still made virus proteins, caused inflammation. To solve these problems new “gutted” adenoviruses were designed, which had no viral genes at all (they were made in the lab using cells with “helper viruses”, which supply the missing functions but are not there in the final product). When these new tools are used to carry genes into cells, along with human start and stop signals on the CFTR gene instead of virus signals, much better results are obtained. There is little if any inflammation of the nose or lung cells and long term expression of CFTR has resulted in mouse models. These hopeful results still have to be repeated in human trials, where getting the large amounts of virus into all parts of the lung is a problem. It is also possible that there will be an immune reaction to the proteins that coat the virus, in which case only one dose can be delivered.


“…the high safety record of AAV continues to make it a tempting vector for use in CF.”

One of the few viruses that do not appear to cause disease is the adeno-associated virus, or AAV. It is a very small virus that usually requires adenovirus to multiply, and it has the interesting property of integrating its DNA into the human genome at a specific site on chromosome 19. This is of value because if we know where a CFTR gene, for instance, goes in, we can be reasonably sure it is remote from genes that can cause cancer if disturbed. Unfortunately, it loses this ability to integrate into this site when it is carrying a human gene in its DNA!

When AAV was used in clinical trials with about 200 persons with CF, it did transfer the gene safely, but unfortunately expression did not last long, and it seems that repeated doses would not work because the patients developed an immune reaction to the proteins that coat the virus. Although gene therapy has been slowed by the disappointing results, the high safety record of AAV continues to make it a tempting vector for use in CF, provided that its efficiency can be increased or it can be made to integrate at a safe site. Some of the newer AAV vectors which have different coats (“serotypes”) seem to give much more efficient gene transfer to lung cells.

“Unlike viruses, liposomes are fairly inert and have a large safety margin,”

LENTIVIRUS APPROACHES

The best known lentivirus is the one that causes AIDS. (Lentiviruses such as HIV are also “retroviruses”, but quite different from the onco-retroviruses discussed above.) Obviously, the use of any virus that can cause a devastating disease has to be approached with great caution, to be sure that the vector is safe. However, lentiviral vectors carrying the CFTR gene have one great advantage, which is that they infect lung cells far more efficiently than adenoviruses or onco-retroviruses. (Many viruses only infect dividing cells, and lung cells divide very slowly). Although lentiviruses are not big enough to take a full size CFTR gene, they can accept more of the virus with its control regions than many of the other viruses that have been tested. It has been shown that after the CFTR gene is transferred into CF cells in culture, the ion transport defect is corrected, but the need for guarantees of safety has meant that clinical trials are not yet under way. Instead of using human HIV, laboratories are now experimenting with sequences from cat and woodchuck lentiviruses, although it is not clear whether these will pose lower or higher risks to humans (they may be more deadly because human cells have never been infected by them before and have no resistance mechanisms).

NON-VIRAL GENE TRANSFER

Liposomes are packets of biodegradable natural fats (“lipids”) that can wrap around genes and carry them into cells. They were first used in a clinical trial for CF in London in 1995, and proved to be safe but very inefficient. The CF Trust in the UK decided to put a great deal of effort into developing new liposomes that could take the CF gene into cells quickly and safely. One challenge is that traditional liposomes could only wrap around small bits of DNA, but the newer lipids can take the entire gene, control sequences and all, into the human cell in an unscrambled state. Unlike viruses, liposomes are fairly inert and have a large safety margin, and (unlike viruses) they don’t seem to lose their potency when used over and over again. Clinical trials in Britain are likely to be under way in the coming year.

“…Someone born today with CF has the prospect of living into their 40s, or even 50s, so an experimental treatment early in childhood must be completely risk-free”

FETAL GENE THERAPY

One of the major problems for gene therapy is timing. Ideally, any form of therapy that tries to treat the lung should take place at birth, or even before birth, when the lungs have not yet become infected or damaged. However, someone born today with CF has the prospect of living into their 40s, or even 50s, so an experimental treatment early in childhood must be completely risk-free. An Imperial College London group has shown, using animal models, that gene therapy may work after birth when the treatment is given late in fetal life (at the equivalent of eight months), but fetal therapy must be shown to be safe before it will be tried in humans.

STEM CELLS

Everyone, everywhere, has heard of stem cells by now. The derivation and use of embryonic stem cells causes major ethical dilemmas for some scientists and patients, while many others argue that every natural approach to curing CF should be followed. I won’t go into the ethical debate about embryonic stem cells, but I will explain why one approach to stem cell therapy (“somatic cell nuclear transfer”, or therapeutic cloning) is particularly important for cystic fibrosis.

Every one of us started as a single stem cell, composed of a sperm and an egg. After fertilisation, this stem cell gives rise to every tissue in the body, and in the developing fetus and a newborn baby, there are many stem cells in almost every tissue in the body, including the lung. However, the stem cells we know how to use in medicine best, at this time, are the ones in bone marrow that make white cells and red cells in the blood. In bone marrow transplantation (for a child with leukaemia or sickle cell anaemia) these stem cells from a relative, or a matched donor, can replace the original stem cells of the child and give a whole new set of blood cells. There are also therapies being developed to use stem cells from patients to treat burns (as an alternative to skin grafts), and to use stem cells to treat eye disease (because the eye doesn’t reject “foreign” cells as actively as bone marrow or lung does).

“This is attractive because these cells, since they are from the affected person, would not be rejected”

Of course, anyone with cystic fibrosis has the same genetic mistake in the CFTR gene in every cell in his or her body. There is no point in taking a stem cell from someone with CF and trying to use it in one of their lungs, since it will still “have CF”. However, is there a possibility that the mutation could be corrected in the test tube in stem cells and then the corrected cells put back into the lung? This is attractive because these cells, since they are from the affected person, would not be rejected. Groups in San Francisco and Rome are trying to see if they can correct the CF mutation in stem cells, but leave the cells healthy and fit to be put back into the patient. This approach seems to me to be more likely to work than taking stem cells from a cell bank or an embryo, because cells from an unrelated person would be rejected unless powerful immuno-suppressive drugs are used.

If it were possible to correct the CF defect in a stem cell from a patient, a new and exciting possibility arises. The easiest way to obtain lung stem cells from a patient is probably by "somatic cell nuclear transfer," or therapeutic cloning, where the DNA from a skin cell of the patient is placed in an egg from which all the genes have been removed. The stem cells, which will be "CF cells" at first, can be grown in the laboratory and then corrected! We already know from research in France and North Carolina that this type of cell can be pushed down the path to give lung epithelial cells. This is not ethically acceptable to everyone, although it is now legal in the United States, United Kingdom and most European countries, and Australia. I would argue that a stem cell from a patient cannot be an embryo because it has not been created as a result of the union of an egg and a sperm, but from the body cells of the patient, but I know that not everyone agrees with me on this.

“...amniotic fluid and cord blood. These seem to have very early stem cells that can multiply easily and give epithelium”

ADULT STEM CELLS

A fetus has more stem cells than a baby, and a baby, in turn, has more than an adult. Confusingly, stem cells from any of these are called “adult stem cells,” because they do not come from an embryo. Not only are there more cells as you go back in time, but also the cells can do more things; they are more “pluripotent” (can make lots of different types of cell). One source of cells that is only just beginning to be studied is the placenta, and its associated amniotic fluid and cord blood. These seem to have very early stem cells that can multiply easily and give epithelium, but of course we would have to know they would not be rejected before they could be used with CF patients.

Lung stem cells have never been isolated in large numbers, are poorly understood, and because lung is a “structured tissue” (unlike bone marrow, which is more like a bag of loose cells) stem cells cannot be taken outside the body for treatment. The lung is also well protected against taking up living cells from the airways, and it is still not clear if cells can get into the damaged parts of a CF lung from the blood supply. There is hope that cells from bone marrow might change to give lung cells if treated correctly in the lab, and it is known that embryonic stem cells definitely can give lung cells. However, one conclusion that everyone should agree with is that a lot more work will have to be done to prove reliability and safety before stem cell therapy for CF becomes a reality.

COMBINATIONS, COMBINATIONS

One of the most dramatic advances in medicine over recent years has been the demonstration that several approaches together often work better than any one treatment in isolation (or, as one of my friends put it, “Miracles don’t happen all that often!”) One mistake that many of us (including me) made twenty years ago was to assume that gene therapy would be a cure that would replace all the other therapeutic advances: antibiotics, physiotherapy, diet and so on. I no longer believe that. I think that gene therapy and cell therapy will take their place alongside other approaches. Even if gene or cell therapy only allow yet another small advance, we may soon be talking about average life expectancies of 50 or 60, with good quality of life, for persons born with CF in the coming decade. Wouldn’t that be wonderful!

REFERENCES
For those who want a more detailed description of current developments, there have been four very comprehensive reviews published recently. They are written for scientists and doctors, but from experience I know that many persons with CF, and their families and friends, will be able to follow the complex arguments they outline.