Tumor cells are a mosaic of different cell types

For the last decade or so, progressive cancer treatments involved taking samples of tumors, testing the cells to determine the genetic makeup, and then prescribing medicines targeted to specific mutations. There are many benefits to this approach, but it doesn’t always work.

It turns out that tumors aren’t uniform; they are mosaics of cells that can be genetically very different. A recent paper in the New England Journal of Medicine showed that a cell in one area may not be the same as a call in another area (a phenomenon called “intratumor heterogeneity”). So a treatment based on a sample from one area may not work for the whole tumor. Some tumor cells may be resistant to the drug so the cancer persists, or even grows.

In this British pilot study, cells from 9 different locations within a primary kidney tumor, and several metastatic tumors, were analyzed using next generation DNA sequencing. Only 34% of the 118 mutations identified were present in all the samples, and several of the major cancer genes were mutated in different ways in different locations. This turned traditional ideas about cancer cells being “clones” of a single, mutated cell on its head.

Previously, it was thought that a tumor develops when a single cell accumulates sufficient mutations over time that eventually lead to it dividing uncontrollably. Therefore if you could find the original mutation, and target treatment to that, then every cell would react to the treatment. But if the tumor is made up of a mosaic of cells, then they could all react differently to the drug. The researchers then created a phylogenetic “tree,” identifying which cells were more persistent, being in the trunk of the tree. They proposed that if those cells were receptive to a targeted medicine, the treatment might be more effective; if not, less so.

Although this study only involved four patients, the results provide a new way of thinking for researchers and clinicians. If we remove the presumption that all tumor cells are identical, we open the way for more creative thinking about how to tackle the problem.

Schizophrenia is a mental disorder characterized by a breakdown of thought processes and by poor emotional responsiveness. According to the WHO the disorder affects around 0.3–0.7% of people at some point in their life, or 24 million people worldwide as of 2011. There is no general cure and the pharmacologic treatment of schizophrenia leaves much to be desired.

Now the National Institute for Health Research in the U.K. is funding a large research trial on Minocycline beginning this April.

Scientists believe schizophrenia and other mental illnesses including depression and Alzheimer’s disease may result from inflammatory processes in the brain. Minocycline, which is a broad-spectrum tetracycline antibiotic, has anti-inflammatory and neuroprotective effects, which could account for the recent positive findings.

The new hope comes after case reports from Japan in which the drug was prescribed to a young male patient with schizophrenia.  The young man had no previous psychiatric history but became agitated and suffered auditory hallucinations, anxiety and insomnia. Blood tests and brain scans showed nothing unusual and he was started on the powerful anti-psychotic drug Halperidol. The treatment had no effect and he was still suffering from psychotic symptoms a week later when he developed severe pneumonia and was prescribed the antibiotic Minocycline to treat the infection. This treatment surprisingly led to dramatic improvements in his psychotic symptoms. When the pneumonia cleared and Minocycline treatment was stopped, the schizophrenia symptoms reappeared.

This serendipitous observation prompted researchers to test Minocycline in patients with schizophrenia around the world. Trials have already been held in Israel, Pakistan and Brazil, with schizophrenic patients showing significant improvement.

Minocycline might be a safe and effective solution to bring symptoms of schizophrenia under control. But more work on how antibiotics alter the inflammation status of the brain remains to be done, as well as trying to find a permanent treatment. Using antibiotics in this way presents a problem. Antibiotics are compounds that are literally ´against life,´ aimed against microbial pathogens. The treatment of infectious diseases by antibiotics is compromised by the development of antibiotic resistance of microbial pathogens. There is a direct correlation over time between antibiotic use and the increase in antibiotic-resistant bacteria. Researchers therefore need to address this issue but the signs are good: Minocycline has been on the market for a while, is broadly described, and resistance seems to have been avoided thus far.

Further reading:

Background information, video links and animations about schizophrenia can be found at the DNA Learning Center’s G2C Online webpage.

Van Os J, Kapur S. Schizophrenia. Lancet. 2009;374(9690):635–45. doi:10.1016/S0140-6736(09)60995-8. PMID 19700006

Some students I was teaching thought that this might be possible: to engineer the insulin-producing cells with a correctly functioning gene, a type of gene therapy.  While this has been a goal for researchers, and they have successfully made insulin-producing cells in the lab from embryonic stem cells, they are not appropriate for transplant because they do not release the insulin in response to glucose levels.  Plus, the immune system might still recognize these cells as foreign and destroy them.

So a new study is looking at transforming cells of the gut that don’t have a specific job yet.  These cells receive signals throughout the life of an individual to become many different types of cells that are used for normal gut function.  So could they engineer these cells to receive the signals to become insulin-producing cells?  Also, would the cells only release the insulin in response to blood glucose levels?

Two Columbia University researchers have started finding possible answers to these questions.  Once they turned off a gene that normally plays a key role in the fate of a cell, insulin-producing cells were generated.  Having cells in the gut that make insulin can be dangerous if they did not release insulin in response to blood glucose levels, but these “new” gut cells have glucose-sensing receptors to allow them to do just that. Another remarkable feature was that the gene could be turned off either early on in development, or later on in adulthood, so it wouldn’t matter how old the patient was.

The next step is to take the research that has been done on mice so far, and see if they can mimic this in humans with the use of a drug or chemical.  This method will also need to prove to be safe and more effective than current methods of treatment, not just to avoid the burden of daily injections.

More importantly, Bexarotene is already in use in human patients, making it easier to determine if the drug will have similar benefits for Alzheimer’s patients.

According to Alzheimer’s Foundation statistics, 5.4 million Americans suffer from this debilitating disease. Alzheimer’s devastates patient’s cognitive abilities, with the most notable symptom being profound and worsening memory loss. While some amount of memory loss is to be expected with old age, Alzheimer’s patients lose the ability to recognize family members, their surroundings, and ultimately require 24-hour care and hospitalization.

Alzheimer’s is thought to be the result of the building up of beta-amyloid proteins, which coalesce into plaques within the brain. In addition to effects on the mouse immune system which could help clear plaques, treatment with Bexarotene seems to modulate the production of ApoE (Apolipoprotein E), an important cholesterol transporter that can interact with and clear these beta-amyloid proteins.

If this treatment works as well in humans as in mice (where effects were noted within hours of treatment) we should know soon if there is a new hope for those affected by this devastating ailment.

This has recently been reported for patients with Hemophilia B.  Hemophilia B, also known as Christmas disease, is due to a deficiency of the clotting factor IX (FIX).  The first reported case of Hemophilia B due to a decrease in FIX was in 1952, and was called “Christmas Disease” after the first patient diagnosed was named Stephen Christmas.  Without this clotting factor, the blood does not form clots and results in severe bleeding episodes, especially in the joints and muscles.

Bettert reatment for this disorder began back in the 1960’s where they would inject FIX concentrates into the blood of patients with hemophilia B.  This increased the average age of death of 24 to a median lifespan of 63 years of age.  So with the success of the protein therapy, why try to fix the genes?  With each treatment costing $150,000 to $300,000, a patient needing clotting factors for hemophilia could incur a lifetime cost of $20 million.

So there needs to be a way that a patient can have a more effective treatment option that will cost less.  This new treatment option offers some hope.  Using a new virus for the administration of the gene, patients have seen an increased production of FIX protein for longer periods of time, and were able to stop or decrease the amount of concentrate injections they would need.  With one injection of the virus only costing about $30,000, dramatic cost savings have already been seen.  While this does offer new hope for the treatment of clotting disorders, follow-up with a larger number of patients and for longer periods of time will be needed to fully weigh the benefits and risks of this technique.  Once this has been done, hopefully we will see gene therapy used more in practice and maybe even for more than just clotting disorders.

Ponder, Katherine P.  Merry Christmas for Patients with Hemophilia B. The New England Journal of Medicine 10.1056; December 10, 2011.  Nathwani A.C., Tuddenham E.G.D., Rangarajan S., et al.

So researchers in the UK brought in the big guns – bioinformatics.

Cancer Research UK has set up seven British centers to start collecting 9,000 tumor samples from a wide range of cancer patients to create a DNA database. Researchers will extract DNA from these tumors and scan them for a series of key genes involved in tumor development. The results will then be cross-checked against a range of cancer treatments, to create a map of which treatments work best for cancers associated with which particular genes.

This is based on the concept of pharmacogenomics: certain genes predispose people to respond to certain drugs in certain ways. We can already test a cancer patient for a single gene, knowing how tumors with that gene respond to a particular drug. However currently we don’t have a way of testing a broad panel of genes. And to compound the problem, we don’t have a way of quickly and accurately sharing information between labs in the same city, across the country or internationally.

Again, enter the power of bioinformatics.

With the proposed cancer DNA database, a doctor might analyze a patient’s tumor sample and prescribe a tailored treatment plan within a very short period of time, perhaps as little as two weeks.

As Professor Matthew Seymour, director of the National Cancer Research Network (NCRN) in the UK, recently stated, “We have to get clever about how to target drugs. Medications for cancer have to be personalized because no two cancers are identical.”

Bioinformatics research is increasing at an exponential rate. DNA sequences are available to anyone with an Internet connection – along with free bioinformatics tools to explore sequence data, predict the presence of genes, and compare features shared between organisms.

The DNALC has been working in DNA sequencing and bioinformatics for over a decade, developing intuitive, visually appealing computer tools for teachers and students to quickly learn the rudiments of gene analysis and integrate bioinformatics with biochemistry labs.

If you want to find out more, check out:

• G2C Online Bioinformatics section
• DNAi: Applications > Genes and medicine > Genetic profiling
• Gene Boy, a fun, intuitive Flash interface to analyze DNA sequences.
• Sequence Server, a database and personal workspace for students to conduct phylogenetic analyses using their own DNA sequences.
• DNA Subway, a platform that uses the metaphor of a subway network to provide students access to various bioinformatics workflows.

Researchers at the University of California Los Angeles and Ohio State University Comprehensive Cancer Center are investigating a new treatment for glioblastoma, the deadliest form of brain cancer. Their paper, out this week in Cancer Discovery, shows how blocking a mechanism involved in cell metabolism and triggered by a cancer gene can reduce brain tumors.

Glioblastoma affects about 18,500 Americans each year, with less than a third surviving. The brain tumors are very difficult to remove as the cancer cells invade surrounding brain tissue. To make matters worse, some people are genetically predisposed to resisting chemotherapy or radiotherapy.

The researchers looked at the cellular mechanism that involves an over-active PI3K signaling pathway. This pathway is stimulated by a gene variant called EGFRvIII, which is present in nearly half of all glioblastomas. The gene variant also switches on a transcription regulator, increasing the activity of the low-density lipoprotein (LDL) receptor. This increases the uptake of LDL, providing more cholesterol for the brain tumor cells to feed on, grow and survive.

The number of LDL receptors was reduced in these experiments by activating an alternative receptor, the nuclear Liver X Receptor. This then caused the cholesterol to be transported back out of the tumor cells using an ABCA1 protein pump. Without the extra cholesterol, the greedy brain tumor cells eventually starve and die.

The good news is that this signaling pathway is not just confined to glioblastomas so this therapy may eventually be used to treat other forms of cancer.

So it’s yet another reason to cut out the eggs, cream and butter and have oatmeal for breakfast!

Guo D, Reinitz F, Youssef M, et al. An LXR agonist promotes glioblastoma cell death through inhibition of an EGFR/AKT/SREBP-1/LDLR-dependent pathway. Cancer Discovery 2011; early online.

(For more on signaling pathways in cancer cells, check out “Pathways to Cancer” @ the Inside Cancer website.)

To find these driver mutations, scientists look for the ones that occur frequently. Until recently, this was very difficult to do. However, new sequencing technologies now let scientists look for mutations in genes at an incredible rate. The cost of sequencing is dropping dramatically; to the point where in the near future sequencing the DNA from a cancer may be sequenced as a diagnostic. Soon, it may be the cost of computing that limits our sequencing efforts.

Improvements in technology allow scientists to look at the genomes of many tumors, and there is an international effort to look at 25000 cancer genomes. This will provide the data that will let them find the mutations that lead to cancer, even if they occur in a small proportion of tumors of a particular kind. Already, hundreds of tumors have been studied in detail, which is giving scientists a good feel for the patterns of mutations that happen in cancer cells. So far, over 400 genes directly linked to cancer have been identified in this and other studies. Figuring out how these many genes contribute to cancer is likely to lead to huge advances in diagnosis and treatment, although the task remains gargantuan.

Another reader wanted to know why there are cures for other diseases, but not for cancer. There are many reasons for this, of which I’ll mention a few. First, cancers are the result of our own cells acting abnormally. This means that many of the treatments we might want to use to kill cancer cells would also kill our normal cells. The challenge is to identify the differences between our normal cells and their cancerous relatives and then to identify weaknesses in the cancer cells. This in itself is very difficult. However, it is much more difficult because cancers are not all the same. As I mentioned above, cancers in different tissues arise by distinct changes, so a drug for one cancer may have no effect on another. Even worse, different cancers within a particular tissue are different. In fact, within one tumor, the different cells can have different mutations, and these can affect how the cells respond to therapy or allow the cancer to develop drug resistance. So, cancer isn’t just one disease- it is many related diseases. In fact, calling for “a cure” for cancer isn’t really fair; what will be needed are many cures for this family of diseases.

I know I’ve probably produced more questions than answers, but I hope that this helps some of you.

In a bone marrow transplant, a patient’s own marrow is destroyed and replaced with tissue from a donor. The donor marrow contains healthy hematopoietic stem cells (HSCs, adult stem cells in the blood) which repopulate the patient’s body with healthy red and white blood cells for oxygen transport and immune defense. Just as with other varieties of organ donation, tissue-type matches are critical. In the case of the AIDS patient, another screen was also applied: his doctors searched for donors whose cells were also resistant to HIV infection.

In order to infect a white blood cell, HIV must latch onto 2 receptors on the cell’s surface: the cd4 receptor and the CCR5 receptor. Some people have no CCR5 receptors due to mutations in the genes encoding the protein—those people are highly resistant to HIV infection. The AIDS patient received bone marrow from a donor whose HSCs (and subsequent white blood cells) could not produce the CCR5 receptor: his new cells cannot be infected by the HIV that decimated his old cells. The AIDS patient was, rather ironically, cured by receiving “defective” cells. The absence of a CCR5 receptor does not appear to affect normal physiology.

Bone marrow transplants are, unfortunately, not a feasible treatment option for the 33.4 million people infected with HIV worldwide. Patients receiving the transplant are at major risk for infection as they wait for their immune systems to regenerate: between 10 and 30% of patients die from the procedure. In addition, there are very few HIV resistant donors relative to the number of infected individuals. On average, 1 of every 1,000 Europeans carries 2 copies of the defective gene while the mutation (a 32 base pair deletion) is very rare in people of Asian and African descent.

The tidings are not all together grim, however. This case shows that HIV can be treated by inhibiting CCR5 expression. If a patient’s HSC could be removed, rehabilitated via gene therapy, and returned to the patient, future AIDS treatment might not require life-long drug courses or dangerous transplants.