Viral Transformation of Cells

Cancer can be viewed as a genetic disease. It is due to discrete changes in the cellular genome which in some cases are heritable. Virus are probably responsible for about 15% of human cancers and as a risk factor are second only to tobacco. This set of notes is concerned with the phenomenon of cell transformation by viruses. It will discuss the process of transformation, RNA tumour viruses, DNA tumour viruses and viruses which cause tumours in humans e.g. EBV, papilloma, Hepatitis B and HTLV-1.

Transformation:

The cell cycle, (G0)-G1-S-G2-M, is highly regulated in multicellular organisms. The G0/G1 boundary is a particularly important control point because this acts as a commitment to cell division. Tumour formation results from a failure of these regulatory mechanisms; tumour cells continue to divide under circumstances in which their normal cellular counterparts do not. Transformation of cells in culture is the in vitro counterpart of the process by which tumour induction in animals (by viruses) occurs. Many of the properties which can be used to differentiate a transformed cell from a non transformed control are also observed when tumour cells are compared with their normal counterparts. For example:

Growth:

• high/indefinite saturation density
• different, usually reduced serum requirement
• growth in agar-anchorage independent
• tumour formation when injected into animals
• no contact inhibition of movement
• less oriented growth
• growth on normal cell monolayers

Surface:

• increase agglutinability of plant lectins
• changes in glycoproteins and glycolipids
• loss of tight junctions
• foetal antigen expression
• different staining properties
• increased rate of nutrient transfer
• increased secretion of proteases

Intracellular:

• disruption of cytoskeleton
• altered amounts of signalling molecules (cyclic nucleotides, phosphoinositides)

This list highlights regulatory mechanisms acting on normal cells which can be broadly divided into 3 groups:

1. Those involving cell-cell contact i.e. contact inhibition of growth.
   Normal cells exhibit it, transformed cells do not.
2. Those involving dependence on exogenous growth factors.
   Normal cells show such a dependence, transformed cells do not.
3. In some cell types, anchorage dependent growth constitutes a further group of regulatory properties.
   Normal cells exhibit it, transformed cells do not.

These features suggest the basis for selecting cell transformants and assaying transformation. Most cell lines are already transformed and by definition not suitable for transformation assays.

Primary cells have a limited life span in culture. They reach a crisis phase and die after about 50 generations. It is clearly possible to use primary cells in transformation assays but the availability of a characterized cell line behaving in a consistent and identifiable manner is preferable.

Todaro and Green developed a mouse fibroblastic cell line called NIH 3T3 which has been extensively used for studying transformation by cellular and viral oncogenes and other agents. They plated mouse fibroblasts at a low density and replated them every 3 days at this low density ensuring that the cells were never in contact. After more than a year they had cells which were capable of indefinite growth in culture but which still exhibited contact inhibition of growth. They were also aneuploid (chromosomal number is not haploid x N). They undergo spontaneous transformation in vitro in the absence of exogenous agents and exhibit a low tumouroginicity when introduced into mice. It is reasonable to conclude that they have accumulated a mutation or series of mutations required for expression of the transformed phenotype. These features of NIH3T3 cells (which behave "normally" in many ways) make it suitable for use in "transformation assays" because only a limited number of further mutations (perhaps 1) tip the cell over the edge into a fully transformed state. This feature is obviously useful in assays for cellular oncogenes.

Assays for Cell Transformation.

• Focus forming assay. The first assay for viral transformation devised by Rubin and Temin. In mid 1950s Rous sarcoma virus, an avian retrovirus, was found to induce morphological changes and extend the life of chick embryo fibroblast cells in culture. Transformation was rapid and quantitative. Transformation was detected as foci of dense morphologically altered cells in monolayers of chick embryo fibroblasts. This type of assay which depends on the loss of contact inhibition has subsequently been used for a number of DNA viruses e.g. polyoma and SV40 in NIH3T3 cells. It depends on the loss of contact inhibition following transformation so cells must obviously show this feature initially.
• Anchorage independent growth in agar or a methyl cellulose suspension. Primary fibroblasts and many fibroblastic lines must attach to a solid surface before they can divide i.e. anchorage dependence. They fail to grow when suspended in a viscous fluid or gel. Possible to select transformed cells on the basis of growth in the absence of a solid substratum. This is usually considered the most stringent assay. BHK21 cells transformed with polyoma virus show anchorage independent growth.
• Reduced serum requirement The reduced serum requirement of transformed cells can be exploited e.g. NIH 3T3 cells will not grow in media with less than 5% serum. SV40 transformants could be selected for by plating in 0.5% serum. This assay exploits the requirement of cells for growth factors.

Transformation by individual genes:
The development of high efficiency transfection protocols in the 1970s was an important technical advance. This led to the observation that transfected DNA could confer on the cell biochemically detectable characteristics. This prompted studies of the transfection of cellular DNA from transformed cells into normal cells which demonstrated that such DNA was capable of eliciting transformation. Using these techniques a considerable number of cellular and viral oncogenes were identified. NIH3T3 cells were often used in these studies because they take up DNA efficiently. NIH3T3 cells do not detect all classes of transforming genes - only 50% of tumour derived DNAs gave rise to transformants. This is due either to their recessive nature or to a lack of activity in the NIH3T3 system particularly. It seems evident that the way in which these cells were established would influence the types of oncogenes they could be used to detect. Thus although the cell line exhibits growth control it is capable of indefinite growth which could be considered an essential feature of any transformation process.

Mechanism of Transformation:

In vivo and epidemiological studies suggest that transformation is a multistep process involving initiation, promotion and progression events. Chemically induced carcinogenesis (e.g. a polycyclic hydrocarbon) on mouse skin results in an initiation event which is the activation of a cellular ras gene. The initiation event is irreversible and phenotypically undetectable. Application of a phorbol ester such as TPA promotes clonal expansion of cells with the ras mutation. This can occur 12 months after the initial ras activation. There is an initial period of reversibility in this clonally expanded cell population but finally irreversible genetic changes occur leading to the formation of a malignant tumour. The promotion and progression stages involve further cellular mutations.

The molecular basis of transformation in NIH3T3 cells is similar. The (viral) v-myc gene from the avian retrovirus MC29 and activated (cell) c-N-ras derived from a human bladder tumour cooperate to transform NIH3T3 cells. Neither are sufficient on their own. It is possible to transfect the oncogenes together or sequentially in either order transform the cells.

In contrast to this, some DNA viruses are apparently capable of a single step transformation. The multifunctionality of their viral oncogenes goes some way to explaining. There are cases where a single step transformation does a take place. An example presented below are the transducing retroviruses.

There is also a class of cellular genes known as tumour suppressors e.g. p53 and the retinoblastoma gene product Rb. These encode proteins that exert a negative control on cell replication. Interference with the function of these genes seems to be a major factor in some types of DNA viral oncogenesis. In many cell types progression from a low grade non malignant to a high grade malignant tumour involves mutations in the p53 gene. The function of these genes is discussed here.

How do viruses transform cells?

The frequency of infection by agents such as EBV, HBV or HTLV-1 are far greater than the incidence of the cancers with which they are linked. Viruses may act in concert with other factors in the evolution of cancer. The other factors might be genetic, immunological or environmental. Cell transformation by viruses is accompanied by the persistence of all or a part of the viral genome. It is also accompanied by the continual expression of a limited number of viral genes. Viral oncogenes disrupt the normal cellular gene expression and signal transduction pathways. The signal transduction pathway is responsible for altering cellular gene expression in response to a wide range of external and internal signals. It is a complex net of regulatory proteins which means that a given gene product can be (de)activated by many different stimuli and that a single stimuli can (de)activate many different genes. Viral oncogenes disrupt the normal functioning of this net. They do this by virtue of altered structures or overproduction. The end result is an altered transcriptional profile which favours cell division. This can be divided this into 4 broad groups, with many known examples of oncogenes for each grouping:

1. Cell surface where receptors interact with growth factors and components of the extracellular matrix.
2. Transmembrane signalling apparatus including the cytoplasmic domain of receptors and submembranous components that are functionally linked to surface receptors. conveying signals from the outside of the cell to the interior.
3. Cytosolic elements i.e. soluble proteins and second messengers.
4. Nuclear proteins including DNA binding proteins and factors which interact directly and indirectly in gene regulation and replication.

At the biochemical level, communication between compartments often involves protein phosphorylation. Thus the initial step in transmembrane signalling involves the interaction of an extracellular ligand (peptide hormone or growth factor with its receptor to trigger protein phosphorylation within the cell. This phosphorylation is catalysed by the cytoplasmic domain of the receptor. Other components that might be involved include submembranous kinases and G proteins, these latter are not integral but associate with the membrane. In this way events happening outside of the cell are signalled to the cytoplasm. Cytosolic protein factors are often serine/threonine protein kinases some of which may be activated directly by events happening at the membrane, e.g. protein kinase C family. They also include small soluble second messengers which again may be generated at the membrane e.g. cyclic nucleotides and inositol phosphates.

Nuclear proteins comprise a large and functionally diverse group includes proteins which bind directly to cis-acting regulatory elements in DNA. Proteins which form large multifunctional complexes in transcriptional regulation and DNA replication e.g. T3 receptor and steroid hormone receptors. Many of these factors are phospho-proteins whose activity again may be regulated in part by phosphorylation. If any of these become permanently switched on, either because of a dosage effect or a mutation then the normal pathway is bypassed and transformation may result.