Interactive Cancer Map (click here)
Barely twenty years ago, most researchers thought that cancer would be impossible to understand. There were hundreds of types of cancer arising in countless tissues and organs of the body--perhaps each with a specific cause. We know now that there are really only a small number of genetic defects which lead to cancer. Most of these defects are in master genes which control cell division and cell growth. Defective genes leading to cancer are called oncogenes. The normal, non-defective forms of these genes are called proto-oncogenes. Oncogenes are dominant; that is, only one copy of the defective gene is sufficient to cause cancer. There are other kinds of genes, called tumor suppressor genes, that act to kill renegade cells with oncogenes (a process called apoptosis). Tumor suppressor genes are recessive; that is, both copies of the genes must be mutated or lost before cancer can occur.
"Normal" cells grown in culture attach to the surface of the culture dish and divide until they touch each other. The cells form a single layer of cells, called a monolayer, covering the surface of the dish. This behavior is called contact inhibition. However, when cancer cells are grown in culture, they continue to divide after touching and pile up on each other to form multiple layers of cells. Loss of contact inhibition can be used as a genetic screen for oncogenes. Human tumor DNA encodes an oncogene as shown by a gene transfer experiment (Figure CA1). Mouse fibroblasts that have lost contact inhibition by acquiring a human oncogene form a focus of piled-up cells (the cells in the focus are called "transformed" cells, and the process is called transformation). "Transformed" cells can also form tumors when injected into mice. On the other hand, normal human DNA put into mouse fibroblasts does not lead to transformation.
Several experimental approaches have been used to clone oncogenes. The method shown here (Figure CA2) exploits the fact that human DNA contains Alu sequences, but mouse DNA does not. There are roughly a million nearly identical copies of Alu sequences spread randomly throughout the human genome. Human tumor DNA is used to transform mouse fibroblasts. The DNA from transformed mouse cells is used in another cycle of transformation. DNA extracted from the second cycle of transformed cells now contains very little human DNA (only the small segment of human DNA harboring an oncogene). All of the other segments of human DNA have been lost during serial gene transfer since entry of such segments into mouse cells does not lead to focus formation. Clones containing human DNA can be identified by screening a library of transformed mouse DNA with an Alu probe. The clones containing human DNA have potent biological activity. Whereas two micrograms of the original human tumor DNA induced an average of one focus (colony of transformed mouse cells), a comparable amount of the cloned oncogene induced as many as 50,000 foci (this is known as the specific activity = number of foci per amount of DNA; the higher the specific activity, the higher the purity of the oncogene). The normal version of the oncogene (the proto-oncogene) could also now be isolated by using the cloned oncogene to probe a library of normal human DNA.
The alteration of the proto-oncogene leading to its conversion into an oncogene could now be determined by comparing the DNA sequences. Unfortunately, the clones containing the proto-oncogene and the oncogene were both still very large. In order to focus on the pertinent sequences, recombinant DNA techniques were used to construct hybrid molecules containing pieces of the two clones (Figure CA3). Each hybrid molecule was tested for the ability to transform mouse fibroblasts. When the region still causing the recombinant molecule to act as an oncogene was small enough, the DNA sequence was determined. There was a single nucleotide alteration (point mutation) changing a glycine codon into a valine codon at position 12 of the encoded protein. We know today that altering glycine at the 12th position to any other amino acid results in an oncogenic form of the protein. This protein is called Ras (for rat sarcoma). Mutated Ras is estimated to cause 20-30% of human cancers including half of all colon cancers and one quarter of lung cancers. Many breast cancers involve mutated Ras. So do almost all pancreatic cancers. Even this high incidence of Ras involvement is probably an underestimate. As we discuss below, Ras is a central participant in signaling pathways controling cell division and cell growth. Mutated proteins involved above and below the level of Ras in these pathways can lead to chronic up-regulation of the pathways even in the absence of mutations in Ras itself.
Ras mediates its effects on cellular division and growth in part by initiating a cascade of protein kinases (enzymes that donate phosphate from ATP to particular proteins) that ultimately activates specfic transcription factors to turn on transcription of target genes (Figure CA4). Ras is a member of the superfamily of G-proteins, proteins regulated by a GDP/GTP cycle (note: remember that some of the initiation, elongation, and termination factors involved in protein synthesis are also members of this superfamily). Ras is active when bound to GTP, but inactive when bound to GDP. The oncogenic form of Ras always has GTP bound, and is therefore always active. The active state of Ras is terminated by converting GTP to GDP mediated by GTPase activating protein (GAP). Reactivation of Ras requires the removal of GDP by a guanine nucleotide exchange factor called SOS (son of sevenless). Since the concentration of GTP is 10-times higher than the concentration of GDP in cells, GTP binds to unoccupied Ras. Ras genes have been found in organisms such as yeast, fruit flies, round worms, and man.