• Sources of Cancer Inducing Mutations (click pic)

Cancer is a disease, first of all, of cellular mutations. The sources of these mutations are charted in the figure above. Many researchers believe that five to ten critical mutations must accumulate to create a cancer cell line, also known as a clone line. Cancer can be viewed as an accumulation of genetic incidents, that begin with DNA damage, mutation or silencing. These precancerous cells undergo selection for specific growth advantage that ultimately leads to metastasis. These mutations can continue at a higher rate after the clone line has multiplied due to a breakdown in cell cycle quality control checkpoints, much like a runaway train. As in antiobitic resistance scenarios, most methods of treatment kill all the cancer cells except those that have the ability to grow in the more difficult circumstances. The resistant tumor cells are then selected and continue to grow and proliferate in a true Darwinian survival of the fittest fashion. The applicability of Darwin's theories to molecular biology is thus astonishingly accurate. New biologic therapies such as Herceptin and even thalidomide may provide a "Darwinian bypass" to circumvent this process of molecular evolution. Therapies like Tamoxifen circumvent the repeated growth cycling of hormone sensitive tissues, which makes such tissues more vulnerable to the mutations that cause cancer.

There are two categories of genes implicated in cancer. The first category contains tumor suppressor genes, such as p53. Tumor suppressor genes can be likened to the brake pedals in a car, that slow it down when it is moving (growing) too fast. When brakes don't work, the car doesn't stop and the cell keeps growing. The second category contains oncogenes, or growth genes. When these genes are expressed in a cell, the cell grows and divides. Oncogenes can be likened to the accelerator pedal of a car. It is also disaster when the accelerator pedal sticks.

To repeat this useful analogy another way. The brake pedal can fail to actuate the brakes, and the accelerator pedal can stick in the on position. Similarly, a cell can fail to turn off properly, or it can go into an uncontrolled growth mode. Genes that tell a cell when to turn off are called tumor suppressor genes. Genes that tell cells to grow are called oncogenes.

Most cells originate as stem cells, primal undifferentiated cells. On receiving certain chemical signals these stem cells differentiate or morph into the specific tissue type required. It may be possible that mature cells of specific tissue types can revert to more primal forms when signaled to do so, but this reversion is usually limited. In cancer, some fully differentiated cells appear to "back up" to their more primitive states, and then grow out of control. There is some thinking now that some cancers may originate in stem cells for given tissue types, and that it is the duplication of stem cells that make cancer look primitive.

As you likely know genes consist of sequences of DNA bases chosen from the set C,T,A or G. These letters are called nucleotide bases. Groups of three of these letters taken three at a time specify which amino acid will be chosen next for incorporation into a protein. There is a short alphabet of twenty amino acids that make all proteins. Some chemicals that cause cancer, or carcinogens, mimic the shape and charge of these nucleotide bases, and get incorporated into the DNA by mistake. When the DNA is duplicated during cell growth, an error may then occur. Lexically, there are three categories of errors in DNA, deletion, substitution, and insertion. DNA bases can be thought of as letters that must satisify parsing rules.

In the first error category of deletion mutations, one or more of the letters is missing. DNA bases are read in groups of three to produce protein, one amino acid at a time. DNA bases taken in groups of three letters are called codons. Deletion of one or more of these letters throw off the entire protein coding sequence for that gene. This results in nonsense downstream of the mutation. DNA has some built in error correction, due to the double helix. This bears similarity to a Hamming code and reference to the work of Shannon and Norbert Wiener is a useful sidebar. Nucleotide bases pair across the helix with their complementary base. Cytosine, C, always pairs with guanine G, by hydrogen bonding across the rung of the DNA ladder. Similarly adenine, A, pairs with thymine, T. DNA repair enzymes such as DNA polymerase and DNA ligase correct missing or incorrect bases provided that the complementary base on the other strand is correct and provided a carcinogenic subsitution has not occurred. Individuals whose repair enzymes are defective, such as those with XP, Xeroderma pigmentosum, may develop tumors spontaneously on exposure to sunlight. The UV component of sunlight is damaging to DNA. In normal individuals, this is repaired by DNA polymerase. Assymetry of repair is the rule of the day. The leading and lagging strands of DNA are copied and repaired with different enzymes and topological functions resulting in a different likelihood of successful replication for each side of the DNA helix, a 3.4 fold difference in lower organisms.

The second category of mutation is substitution, where one letter is substituted for another. The seriousness of this error depends on whether it is the first, second or third base in the codon. Single letter substitutions frequently result in a codon that specifies the same amino acid. There are 64 codons, because there are 64 ways of ordering a triple of C, T, A, or G. But nature has caused these 64 primitive instructions to map to only 20 amino acids, so the 64 codons are not unique. As in a digital numbering system, the leading letter of the codon is most significant. Two different codons can code for the same amino acid, thus a substitution mutation will frequently make no change in the gene product. Often the protein will function properly if a similar amino acid is substituted. This is not always the case. Substitution mutations are also called point mutations. Sickle cell anemia is caused by a single point mutation in the gene that codes for hemoglobin. Hemoglobin is a globular protein that serves to transport an oxygen diamond in the setting of a porphyrin ring. Sickle cell hemoglobin will polymerize with itself under conditions of low oxygen tension. Sickle cell and cancer are not related, but the mechanisms, history and legacy of mutation are related.

The third category of mutation is an insertion mutation. In this mutation, one or more letters are inserted into a coding sequence. If three letters or a multiple of three letters are inserted, the protein may still function properly. If a non multiple of three letters is inserted, nonsense will again result for all codons downstream of the insertion mutation. Viruses can cause insertion mutations, since their DNA is incorporated into the host. An interactive gene sorting illustration of the genetic wreckage carried by the BRCA1 breast cancer gene is shown below.

There are two kinds of DNA sequence, those stretches that code for protein product, called exons, and those that do not code for protein, called introns. The latter are sometimes called, "junk DNA", because no function is known for this genetic material. About 96.4% of the human genome is "junk DNA". This is quite remarkable.

Examples of these three kinds of mutations can be seen in the BRACA1 gene rollover graphic below. Placing your mouse over the graphic, causes the components of BRACA1 to be sorted. A common repeated intron is called an ALU repeat. The ALU repeats in BRACA1 can be seen to be profoundly frame shifted and mutated. Fortunately these repeats do not code for gene product, but they portend the mutations that do occur in the exonic regions, that do contribute to breast cancer.

Deletion and insertion mutations are also called frame shift mutations, because they alter the codon reading frame and produce nonsense. Mutations rarely result in a gain of function. Mathematically, loss of function mutations are much more likely.

Mutations are not the only way that important genes can be damaged or silenced. Genes have switches called promoter and repressor regions. When certain proteins bind to the promoters or repressors the amount of protein produced from the gene can be turned up, turned down, or turned off. Another mechanism for silencing genes is methylation. In gene methylation, a cystosine base in the gene's promoter region is endowed with a methyl group. Like protein binding to promoters, methylation is an important natural mechanism for gene regulation. But when a tumor supressor gene is silenced by methylation, serious trouble may result. See link.

Most cancer can ultimately be traced to DNA damage, mutation or gene silencing by methylation. The source of unwanted methylation is unknown, but one agent that causes methylation is the enzyme methyl transferase. Thus tracking down sources of methyl transferase, for example in viruses that are known to cause cancer, would be a possible productive line of reasoning. Looking at mutations that affect the expression of methyl transferase would be another.

Mutations can be inherited from a previous generation, if the mutation appeared in the germ-line, or they can be acquired along the way in somatic cells. The same is true of methylation.

Cancer is also a disease of age. When DNA is replicated, there must be extra length at the ends of the strands to allow the duplicating enzymes to stay on the tracks, so to speak. These extra lengths are called telomeres and are maintained by an enzyme called telomerase. Telomerase is a reverse transcriptase (much like the HIV virus). When telomerase is absent, the ends of the chromosomes fray. Important gene products coded by these fraying ends then fail to be made and the deterioration of aging continues.

The Human Genome & Cancer
Click On Any Chromosome
-- SKY image kindly provided by Dr. Jeff Sawyer, Arkansas Children's Hospital

The human body consists of 100 trillion cells divided into several organ systems and tissue types. These specialized tissues consists of groups of similar cells, e.g. liver cells, skin cells, nerve cells etc. Each cell contains its own copy of the human genome, the chromosomes that carry the instructions for how to make a complete human being. Chromosomes are made of DNA and 96.4% of the DNA does not carry instructions for anything useful. Only 3.6% of our DNA contains useful genetic information. At the end of the cell cycle the genome consists of 46 chromosomes, 23 provided by each parent. Each chromosome contains a few thousand genes separated by long deserts of non-coding DNA. In the human genome there are 32,997 confirmed genes. In differentiated, mature tissues, most chromosomes are turned off, except for a few genes that are necessary for housekeeping and essential cell function. Genes are the instructions to make a given protein. Proteins can be enzymes, which enable chemical reactions. Proteins can be structural materials such as tubulin which give cells internal form. Proteins affect shape as in collagen which connects cells to each other.

In reproductive cancers, hormone sensitive cells derived from epithelial cells grow rapidly in an uncontrolled manner forming a mass of unwanted tissue called a tumor. Cancer cells express proteins that dissolve collagen and enable them to break away and travel to other parts of the body. These proteins are called serine proteases, or MMP's. They are similar to digestive enzymes.

When certain growth genes and cell regulation genes such as p53 are broken from mutation, the cell will divide before its DNA has passed certain "quality control" checkpoints. When this occurs, frayed ends of the chromosomes in the cell can rearrange into new configurations such as the one shown above. Pieces of one chromosome can fuse to another in a process known as translocation.

A technique called Spectral Karyotyping is informative about translocations that occur in a variety of cancers.

Rapidly cycling hormone sensitive tissues such as breast, ovary and prostate, are much more likely to contain mutated DNA. Hormones themselves do not cause cancer, but hormone induced DNA transcription and cell cycling increases the probability of mutation.

 

Characteristics of hereditary breast cancer:

  • Multiple cases of early onset (< age 45) cancer
  • Associated ovarian and/or colon malignancies

  • Vertical transmission (maternal or paternal lines)
  • Bilateral breast cancer
• Five to Ten Mutations = A "Full House"

Each of the 100 trillion cells in the body carries its own unique configuration of mutations.

Most of the time, these mutations are harmless because they occur in noncritical sections of DNA, or are corrected by the DNA repair enzymes before replication occurs.

Sometimes therapy can help and hurt at the same time. For example, chemotherapy kills cells that are highly mutated, at the cost of increasing the mutation level of normal cells.


Mutations enable:

Oncogenes found in biopsy specimens of breast carcinoma cells include:

Immortalization is considered to be one important transformation in a cancer cell line. Without immortalization the cell can only multiply 50 times before the chromosomes ends. How many cells can be made in 50 divisions? How much would a tumor made of such cells weigh?
 

200 Breast Cancer Genes:
Grouped By Function
Tumor Supressors And Cell Growth (Oncogenes)

• The BRCA1 Breast Cancer Gene

In 1990, through genetic linkage studies, a gene(BRCA1) was localized on chromosome 17 at band q21 that predisposes to a significant proportion of early-onset breast cancer. Later studies found BRCA1 increased susceptibility for ovarian cancer as well as breast cancer. A study of 214 families linked to 17q21 the cumulative risk for breast cancer associated with BRAC1 to be about 59% by age 50 and 87 % by age 70.

Another gene (BRCA2) has been implicated in hereditary breast cancer. This gene has been localized to a region on chromosome 13q12-q13.

Several families linked to BRCA2 show breast cancer in women as well as men, whereas no breast cancers in men have yet been observed in families linked to BRCA1.

The p53 gene also has been linked to breast cancer. This gene is a tumor suppresser gene which normally acts as a transcription factor that regulates the expression of other genes and has been mapped to chromosome 17p13.

A variety of different types of mutations in this gene have been found.

Investigators are now showing cautious optimism for using p53 as a prognostic tool; the over expression of p53 correlates with a poor prognosis in node-negative breast cancer.

The OMIM morbidity map shows this and other cancer genes. Genes implicated in the cancer process are called oncogenes.

 

A = aqua C = black G = gold T = Teal

The gene called BRCA1 actually contains instructions for four different proteins, IFP35, BRCA1, RHO7 and IFB35. These instructions are called exons. If you place your mouse on the image you can see these exons at the very top in the sorted version. The next region down with a regular pattern consists of noncoding DNA called ALU repeats. These make up 35% of the BRCA1 gene. These are the regions in the middle. The smeared appearance is due to deletion and point mutations that have occurred over the entire history of this gene.

• Gene Chips
Characterizing Mutations In Breast Cancer

Breast cancers that were formerly reported as being the same, are being screened using gene chip technology to create a personal portrait of an individual's specific cancer. These arrays may used in the future to customize therapeutic regimens basd on the information on the specific cancer clone. In this way, patients can be treated according to the specific pathways that are active in their specific tumors. Data on how they respond to these treatments can be incorporated into treatment databases to further improve therapy.

Groups of genes implicated in carcinogenesis can be clustered into representative "eigengenes" that characterize the activity of entire pathways that are up or down regulated in the disease state.

To explore a map of breast cancer gene expression click here.

This will be one of the most powerful new research tools in the fight against breast cancer. It is now possible to compare gene expression in a patient's tumor with those in databases and in cell lines. By using combinatorial chemotherapy techniques on cell lines, we can know in advance which chemotherapy drugs have the best chance to attack the tumor effectively.

If gene chips follow the same growth law that computer chips do then these advances should come relatively quickly. At this writing it costs $2500 to test an array that profiles 6000 genes. If Moore's Law applies it will cost $40 to profile 100,000 genes by 2009. That is twice the number of genes expected to be found in the human genome.

In the time that this section was being developed, gene chips were applied to breast cancer to create the personal cancer portrait.

The links below provide more information on genes involved in cancer.

Knowledge Mapping

We now live in the era of a sequenced genome, gene chips, monoclonal antibodies, fast computers and the internet. With this armada of powerful weapons a new approach is emerging to solving the cancer riddle. Applying computers to the cancer problem is yielding unexpected avenues of progress. One such avenue is called knowledge mapping. Two knowledge maps that would yield great insight are the oncogene knowledge map (200+ maps) and the gene expression knowledge map (1763 maps). The completion of these first maps will greatly increase our understanding of cancer and will enable simulation of cellular processes. In addition to knowledge mapping there are interspecies comparative genomics projects from the field of gene ontology.
Knowledge Mapping

New and computationally aggressive approaches are being explored in the new field of bioinformatics. These approaches are using data-mining techniques developed in disciplines that are far removed from traditional oncology. A step by step demonstration of these techniques is available.

by L. Van Warren (c) 2000 Warren Design Vision * All Rights Reserved
Featuring contributions by Dr. Jeff Sawyer, Marilyn Fulper, Lynn Warren & Nick Warren.
Special Thanks to the NIH Online Mendelian Inheritance In Man Project