Aug 2 2006
Scientists at St. Jude Children's Research Hospital have developed a mouse model of a severe disease of the immune system that helps explain why gene therapy used to treat children with this disease at an institution in Europe caused some of them to develop leukemia.
The disease, called X-linked severe combined immunodeficiency (XSCID), is caused by a mutation in a gene called gamma C that prevents the immune system from forming B and T lymphocytes. B lymphocytes produce antibodies, and T lymphocytes perform a variety of tasks such as helping B cells and killing cells that are infected with viruses.
XSCID was made famous by the story of the so-called "Bubble Boy" who lived inside a plastic "bubble" to shield him from infections.
Based on these recent studies, the St. Jude researchers concluded that XSCID itself makes mice--and by extension, children with this disease--particularly susceptible to cancer caused by gene therapy.
One important implication of this finding is that gene therapy for other forms of genetic blood diseases will pose significantly less risk for causing cancer than was previously thought, according to Brian Sorrentino, M.D., director of St. Jude Experimental Hematology Division and co-director of Transplantation and Gene Therapy. Sorrentino is senior author of a report on these findings that appears in the August 1 issue of Proceedings of the National Academy of Sciences.
"Our current findings with this new mouse model offer real hope that we can make gene therapy for X-linked SCID safe as well as effective," said Yan Shou, Ph.D., the paper's first author and the major contributor to this work. XSCID was the first disease ever successfully treated with gene therapy. In 2002, French researchers inserted normal gamma C genes into bone marrow stem cells removed from young children with the disease. The clinicians then infused the genetically modified stem cells back into the patients, restoring the ability of nine children to develop full immune systems. However, three of the patients eventually developed leukemia, an event that caused the field of gene therapy to reconsider the safety of this approach. Researchers later determined that the gamma C gene added to the stem cells had inserted itself into oncogenes (cancer-causing genes). The genetic on-switch that was part of the gamma C gene had turned on the oncogenes in the blood cells and caused them to multiply uncontrollably, causing leukemia. To study this complication of gene therapy, Sorrentino's group developed a mouse model that lacked the gene for gamma C (gamma C-/-) as well as Arf (Arf -/-). Arf is an important "tumor-suppressor" gene that forces cells that multiply abnormally fast to commit suicide. By using a mouse lacking Arf, the team ensured that leukemia caused by gene therapy in this model would not be suppressed by the gene's activity.
The researchers removed bone marrow stem cells from these new mouse models (Arf -/- gamma C -/- ) and inserted gamma C into them using specially modified viruses called vectors. The researchers then infused the treated cells back into the XSCID mice. The gene therapy cured the genetic disease, but 13 out of 15 of the treated XSCID mice went on to develop leukemia, the researchers reported. The team removed leukemic cells from mice that developed cancer and found that gamma C genes had inserted themselves into a variety of oncogenes.
In another experiment, the St. Jude investigators used mouse models that lacked Arf but had normal gamma C genes and therefore did not have XSCID. The team used this model to determine whether gene therapy would cause cancer even if a mouse model did not have XSCID. The researchers found that the rate of cancer development in non-XSCID mice getting gene therapy was significantly lower than among mice that had XSCID. "This finding suggests the presence of XSCID disease is a strong risk factor for developing leukemia from gene therapy," Shou said. "Although several XSCID patients treated elsewhere with gene therapy have developed leukemia, patients with other types of genetic blood disorders have not developed cancer following gene therapy. So our discovery suggests that patients with XSCID are especially vulnerable to this serious complication but that people with other genetic blood diseases may not be at as high a risk."
Sorrentino's team discovered a clue to what makes XSCID mouse models--and children with XSCID--particularly vulnerable to leukemia caused by gene therapy. Specifically, they analyzed the bone marrow cells of XSCID mice to determine the size of the population of primitive stem cells that had not yet produced B and T cells. This population of cells is the target for gene therapy, Sorrentino noted. The researchers found that this population of primitive stem cells was significantly expanded.
"The many primitive cells stalled in their development greatly increases the odds that gamma C genes will insert themselves into oncogenes in at least a few of them," Sorrentino said. "That is probably is a major reason that XSCID patients are particularly vulnerable to developing leukemia from gene therapy."
In addition to this large population of primitive cells in XSCID patients, the ability of gamma C to insert into an oncogene and act as an on-signal contributes to the risk of gene therapy for this immune deficiency disease, Sorrentino said. Another risk factor is the lack of the Arf gene to put the brakes on the overgrowth of cells that turn cancerous because of gene therapy.
"Without the Arf gene to force cancerous cells to commit suicide, these cells continue to divide and cause leukemia," he said. "This suggests that the patients with XSCID in France who developed leukemia following gene therapy may also have abnormalities in either Arf or another tumor suppressor gene."
The apparent lower risk of cancer due to gene therapy of other blood disorders is an important piece of good news that comes out of this study. "If we could find a way to eliminate the excess cells, and the ability of gamma C to insert into oncogenes and turn them on, we could substantially reduce the risk of gene therapy for XSCID as well," Sorrentino said. He and his colleagues at St. Jude are currently studying ways to "insulate" genes used for gene therapy in order to prevent them from inserting themselves into oncogenes and turning them on. The St. Jude XSCID mouse models will then be used to test these new vectors.
The (Arf -/- gamma C -/- ) mouse models were developed using Arf -/- mice provided by Charles Sherr, M.D., Ph.D., co-chair of the St. Jude Department of Genetics and Tumor Cell Biology and a Howard Hughes Medical Institute investigator. Other authors of the paper include Zhijun Ma and Taihe Lu.
Genetic diagnosis solves medical mystery for 30 undiagnosed patients