Cancer: a case study
Recommended Poem
The Ship Pounding, by Donald Hall
The concepts and examples presented in the previous units should equip you to understand more about news of complex science and scientific findings in the life sciences. In this section, we will consider an example of such complexity: cancer. In considering some of the more reliable findings on the subject, we will find applications for many subjects in the previous units, including proteins, nucleic acids, enzymes, antibodies, metabolic pathways, mutation, and evolution.
Cancer is actually many diseases. Sometimes even a single incidence of cancer results in mixtures of cancer cells that respond differently to treatment. Scientific understanding of cancer is far from complete. If you want to read extensively about cancer, guided by a writer gifted in presenting medical information to a wide range of audiences, read The Emperor of All Maladies: A Biography of Cancer by Siddhartha Mukherjee. Click the title to learn more.
We will consider a hypothetical case of a cancer patient, starting with discovery and diagnosis of cancer. We will follow a common course of actions aimed at curing the cancer: surgery, analysis of cancer progress, treatment, prognosis after treatment, and subsequent surveillance. We will also look at one type of inherited mutation that results in increased cancer risk, including genetic testing to assess risk of further cancer in the patient, and future cancer in the offspring.
Symptoms and discovery
A patient, whom I will call Alice, began to suffer from signs of partial intestinal blockage. X-ray scans were inconclusive due to inadequate clearing of stool from the bowel. During a diagnostic colonoscopy on Alice, the endoscopist’s progress was blocked by a sizable tumor in the bowel. He immediately put Alice in touch with a surgeon for removal of the tumor and diagnosis of the extent of cancer progression. Surgery occurred within a few days of diagnosis.
Surgery, biopsy, prognosis, and recommended treatment
Alice’s tumor lay between the end of the small intestine and the first sharp turn in the bowel, where the ascending colon becomes the transverse colon. Surgery involved removal of the most of the ascending colon, followed by reconnection of the small intestine to the transverse colon. Alice could expect normal bowel function after this kind of surgery.
 |
| Colon resection surgery entails removal of the tumor-containing section of the bowel and reconnection of the new ends. |
 |
| Cancers that are diagnosed as stage III or later require radiation or drug treatment after removal to prevent or stop metastasis. |
All drugs for treating and curing disease, including cancer, are based on a common strategy: find a chemical, signaling, or metabolic difference between disease organisms or diseased cells and the cells of the patient, and find an agent that will destroy cells exhibiting that difference.
For bacterial disease, antibiotics have been successful because they block such functions as the construction of cell walls, and human cells have no comparable cell walls, so they are unaffected by the drug. For cancer, this strategy is trickier, because the cancer cells are human cells, and differences between cancerous cells and normal are likely to be subtle. A vast area of cancer research is focused on discovering key differences between cancerous and normal cells, and finding drugs to destroy only the cancerous ones.
The most widely exploited difference between cancerous and normal human cells is that cancerous cells divide rapidly, often more rapidly than any other cells in the body. Certain drugs that block or disrupt cell division can destroy cancer cells, while doing relatively little damage to normal cells. As you might expect, he normal cells most affected are ones that also divide rapidly. These include cells that line the surface of the digestive tract, and hair follicles, which produce body hair. For this reason, hair loss and digestive upset are among the most common side effects of drugs that employ this strategy.
In Alice’s case, the prescribed chemotherapy involved regular treatments with three drugs, folinic acid, 5-fluorouracil, and oxaliplatin, a treatment nicknamed FOLFOX. Each drug, in a different way, disrupts cell division by interfering with the synthesis of DNA building blocks or by blocking replication. As mentioned earlier, before cells divide, they duplicate all of their DNA (replication), and cells in which replication is blocked will die instead of dividing.
 |
| Oxaliplatin binds to DNA and interferes directly with replication. Folinic acid and 5-flurouracil inhibit an enzyme necessary for making DNA building blocks |
The enzyme affected by these two agents is called thymidylate synthase. It functions to transfer a methyl group (-CH3) from tetrahydrofolate (THF, derived from the vitamin folate, which is one reason you need folate in your diet) to the uracil ring of dUMP, a uracil carrier, to produce dTMP, a thymine carrier, and the precursor of dTTP, which actually incorporates T into DNA during replication. Folinic acid is structurally similar to THF, and interferes with THF binding to the enzyme, thus slowing the enzyme’s action. This type of reversible interference is called competitive inhibition, because the inhibitor competes directly with a substrate to lower enzyme efficiency.
 |
| Folinic acid and 5-flurouracil both block the action of the enzyme thymidylate synthase, which puts a methyl group onto carbon-5 of dUMP to produce dTMP, which is absolutely required for DNA replication |
5-fluorouracil (5-FU) does not just interfere with the enzyme thymidylate synthase, it destroys it. Normally, dUMP is temporarily and covalently bonded to the enzyme, during which time it accepts the methyl group from THF. 5-FU replaces uracil in dUMP, and is able to bond to the enzyme, but not to accept the methyl. It thus remains attached to the enzyme, which can no longer function. This type of inhibition is non-reversible, due to the covalent bonding. Inhibitors like 5-FU are sometimes called suicide substrates, because the enzyme does not survive its interaction with the inhibitor.
The third component of FOLFOX, oxaliplatin, binds to DNA itself, disrupting the double-helical structure wherever two G bases occur in succession. A protein called HMG-1 recognizes and binds to this distorted structure, and this combination of oxaliplatin and HMG-1 constitutes a roadblock that replication cannot pass.
 |
| Model of the complex of DNA (B&S) with HMG-1 (yellow) and a drug similar to oxaliplatin (blue and gray SF). This complex acts like a roadblock to DNA replication enzymes, and thus blocks replication. |
Alice’s chemotherapy involved administering these drugs intravenously, every two weeks for six months. The injection solutions also contained agents to reduce side effects such as nausea and depletion of white blood cells, the latter of which could make Alice more susceptible to infection. To avoid problems with damage to veins punctured repeatedly during the intravenous injections, Alice underwent surgical implantation of a port through which drugs could be administered (and blood samples taken) without the need to puncture veins. The figures show a typical port and the position of implantation.
 |
| A port and catheter (above) allow intravenous injection of chemotherapy drugs, as well as blood sampling, without repeatedly puncturing veins. The injection needle is inserted through the soft yellow center of the port. |
Family history
Alice’s family has a history of cancer, especially bowel cancers. The pattern of cancer incidence suggested to genetic counselors that the family carries a specific mutation that increases cancer incidence. Carriers of this mutation have a 50% chance of contracting bowel cancer before age 70, a much higher than average incidence, and also higher incidences of breast and ovarian cancer. This condition is called Lynch syndrome. It is responsible for about 5% of all bowel cancers, and about 50% of familial ones. Bowel cancers involving the Lynch-syndrome gene are called hereditary non-polyposis colorectal cancer, or more mercifully, HNPCC.
 |
| Alice's family history suggests that she inherited a cancer-predisposing gene from her father. "Now I can no longer say that he left me absolutely nothing," said Alice. |
Genetic testing, using DNA obtained by swabbing some tissue from Alice’s cheek, confirmed the presence of this mutation in one of her two copies of the gene involved in Lynch syndrome (more on this gene later). Because Alice carries this mutation in only one of her two copies of this gene (having two copies of it would have been lethal), there is a 50% chance that Alice’s children have inherited this mutation from her. Alice informed her children of this possibility. With relatively inexpensive and specific genetic tests, Alice’s children can find out whether they inherited this gene. If they have, they are advised to undergo colonoscopies and other cancer tests earlier in life and more frequently than what is normally recommended. Medical insurance often covers this extra surveillance; after all, testing and surveillance are much cheaper than the surgery, chemotherapy, and the extensive surveillance that follows cancer. For the same reason, medical insurance often covers genetic testing for patient and offspring.
DNA mismatch repair and cancer induction
The nature of the Lynch syndrome mutation gives interesting insights into how cancer develops, and into such questions as why it can develop so slowly. In Alice’s case, the mutation at the root of her cancer has been in every cell in her body since she was conceived. Yet she was past age 65 when the first cancer appeared.
Replication is often pictured as a deterministic process that produces a perfect match of new strand and template strand. In fact, the process is somewhat error-prone, and many types of “mistakes” occur. Systems of proteins find and correct these mistakes, in processes called editing. Without editing, DNA replication would be so error-prone that replication of our entire, vast genomes would rarely be successful.
One of the more common mistakes is a simple base mismatch (see the illustration of DNA replication in Unit 3). Most such errors are corrected my mismatch-repair systems, which include proteins that recognize and bind to mismatches, proteins that determine which of the two strands has the correct sequence, enzymes that excise the offending strand, and enzymes that replace the excised region by again replicating from the template strand.
 |
| Mismatch repair in mice. The process begins when protein MutS (similar to the MSH complex in humans) detects a base mismatch in newly replicated DNA. Without this detection, the mismatch would not be repaired, and cells receiving the erroneous strand would inherit a new mutation. |
Structure and action of affected protein
 |
| Left: Complex of MSH2 (yellow) and MSH6 (blue) bound to DNA (red and green). In MSH2 endowed by the Lunch-syndrome gene, much of the chain is missing (purple, right), and the complex cannot bind to DNA and detect mismatches. |
The truncated region is shown above in purple. You can see that the remaining yellow portion would be missing important sections that bind it to MSH6 (blue), as well as to DNA (not shown). This truncated protein probably does not even fold properly and is discarded. The result is that half of the MSH6 subunits have no MSH2 partners, and it might appear that the cell has only half the normal capacity to detect mismatches, and that uncorrected mismatch mutations in all cellular genes would be more frequent than in normal cells. But for most genes that help to prevent mutation, which are generally called tumor-supressor genes, it is commonly thought that one good copy of such a gene is enough to maintain a fairly normal error rate for replication. It is not clear whether this is true for mismatch repair genes. But it is clear that a Lynch-syndrome cell that suffers an additional mutation that disables the normal copy of the mismatch-repair gene would lack mismatch repair altogether, and would produce mutated offspring at a significantly higher rate. This higher mutation rate would hasten progress toward cancer (next section).
A model of cancer induction
As mentioned earlier, the mutation that disables one copy of the MSH2 gene is present in every cell of Alice’s body since conception, yet she lived 65 years before cancer appeared. (Not quite true: she suffered numerous precancerous skin lesions, but she was vigilant in monitoring the condition of moles and other skin spots, and abnormal ones, sometimes including abnormal surrounding tissue, were removed upon detection.) In other types of cancers, the same picture emerges. Exposure to carcinogens in cigarette smoke; exposure to radiation, such as radiation treatment of tonsils during youth; exposure to asbestos—all can take decades, even most of a lifetime, to produce cancer.
Here is a simple model of what might have been going on in Alice’s slow progress to cancer.
Imagine that cell division, which goes out of control in cancer, is normally restrained by several gene products that allow cell division only when the genes receive appropriate signals. Cell division is also restrained by DNA-repair systems, such as mismatch-repair genes, because they prevent mutational damage to all kinds of genes, including genes that restrain cell division. I will oversimplify a bit, and lump all of these genes together, calling them division-braking genes (DBGs). As I said, these include DNA-repair genes, but also others that exert control by activating or suppressing specific events in or leading to cell division.
Think of all these genes and their protein products as several feet pressing on the brake pedal of cell division. Each one contributes some of the brake-pedal pressure that keeps cell division under proper control. Each time a random mutation disrupts or destroys the function of one such gene product, one foot is lifted from the brake, and the resulting cell becomes a faster divider, or one that is poorer at responding to restraint signals. That cell becomes the founder of a new subpopulation of cells (a cell line) that outgrow the ones surrounding them. Mutations to DNA-repair genes turn the resulting cell line into sloppy copiers, which accumulate more DNA errors each time they divide. In a cell, faster division and sloppy copying make a bad combination. In Alice's case, with one disabled MSH2 gene already present in every cell, mutation in the second, working, MSH2 gene, disables mismatch repair completely, and copying becomes very sloppy.
Other types of genes in which mutation leads towards cancer (DBGs in a slightly different sense) include genes that promote self-destruction (called apoptosis) of abnormal cells, preventing them from proliferating; and genes that normally promote cell division in response to appropriate signals (called proto-oncogenes), and which, by mutation, become stronger and less selective promotors (called oncogenes).
Each time a mutation occurs in a DBG, another foot is lifted off the brake, or even moved to the accelerator, and the offspring cells are all even faster dividers and perhaps sloppy copiers, so the next damaging mutation is even more likely.
The result of each mutation in a DBG starts a new cell line, or subpopulation of cells, that are even faster dividers, or even sloppier copiers, or both. Each time a mutation produces a cell that divides even faster, that cell becomes the founder of a new line that outgrows normal cells, and takes the lead in the random, halting, lurching race toward cancer. This figure illustrates the process:
 |
Progress towards cancer, with 5 mutations needed for cancerous growth. Normal cells would maintain a constant number with time (gray), through cell division and cell death. One cell division produces a mutant call containing one of five mutations necessary for cancerous growth. This cell reproduces, initiating a growing population of cells containing that mutation (widening band of blue). Among these cells, a second cancer-causing mutation produces one cell of type 2, which contain an additional mutation necessary for cancerous growth, and initiating a growing population of cells containing two mutations (widening band of green). Each successive mutation produces a new cell type that contains more of the mutations necessary for progress towards cancer. In this example, the fifth mutation produces a cancerous cell, which gives rise to a rapidly growing population of tumor or metastasizing cells (red). Illustration kindly provided by Professor Ken Weber, University of Southern Maine, and used with his permission.
|
For people without Lynch syndrome, it would take two mutations in the same cell line, one in each of the two normal MSH2 genes, to disable mismatch repair, and hasten progress toward cancer. In Lynch syndrome, however, it takes only one. For such rare events, one is much more likely than two. Here is why. If mutation in a particular gene occurs once in, say, one million cell divisions, then the chance of two mutations in the same cell line, disabling both copies of the same gene in the cell that receives the second mutation, is (1/1,000,000) times (1/1,000,000), or only one in a trillion cell divisions. For the Lynch patient, disabling mismatch repair is much more likely, a one-in-a-million possibility. One is a million might sound unlikely, but given the number of rapidly dividing cells in the intestines, and given a lifetime of 70 or 80 years, it becomes quite likely. Normal patients, waiting for a one-in-a-trillion event, will probably die from something else first.
In a deep sense, evolution by natural selection is occurring in a tumor. The faster dividers are producing offspring faster than are surrounding cells, becoming more common, and any offspring that can out-divide them will soon out-populate them and dominate the tumor. Also, realize that new and different mutants might appear in several places in the same tumor, making the tumor-cell population diverse—one tumor becoming several different diseases.
This model is somewhat speculative, and it is specific to Alice’s cancer, which might or might not be aided by the sloppy copying in all of her dividing cells, and is certainly aided by the greater chance that mismatch repair might be completely disabled by a single mutation. Without this inborn mutation, the first insult that moves a more typical patient towards cancer might also be an un-repaired mismatch, or it might be chemical or radiation damage to one of those DBGs, leading to a cell that divides somewhat more often than signals from surrounding cells might warrant. But mutations that increase mutation rates, like Alice’s inborn mutation, are probably an important element of any progress towards cancer. Higher mutation rates simply increase the likelihood of mutations in genes that 1) release the brakes on cell division and 2) further increase mutation rates.
Summary
This unit makes use of many ideas from previous units, including proteins, nucleic acids, enzymes and inhibitors, antibodies, metabolic pathways, mutation, and evolution. I hope that the ideas in these units will equip you to understand many other complex, news-making topics in life, medical, and environmental science.
