Slowacting Transforming Retroviruses

Slow-acting transforming retroviruses are as a rule replication-competent and do not transduce oncogenes. Instead, they cause transformation by integrating within or in the vicinity of cellular genes and altering their expression. In this fashion they convert cellular genes from proto-oncogenes into oncogenes in a cis-acting manner.

Integration of a slow-acting retrovirus disrupts negative regulatory elements of the targeted gene and/or activates its transcription by the transcriptional regulatory sequences contained in the retroviral LTR. Several mechanisms are conceivable by which the viral regulatory sequences could cause gene overexpression. Overall, the predominant mechanism seems activation of the cellular gene promoter by the enhancer in the retroviral LTR. This mechanism is most effective, if the retrovirus integrates in inverse orientation upstream from the promoter (Figure 4.2).

Figure 4.2 Oncogene activation by retroviral insertion The most frequent mode of oncogene activation is by enhancer activity of the 'idle' 3' provirus LTR on the proto-oncogene promoter shown in the figure. Often, provirus insertion additionally disrupts or separates negative regulatory elements of the host.

Figure 4.2 Oncogene activation by retroviral insertion The most frequent mode of oncogene activation is by enhancer activity of the 'idle' 3' provirus LTR on the proto-oncogene promoter shown in the figure. Often, provirus insertion additionally disrupts or separates negative regulatory elements of the host.

In the figure, this mode of activation is shown for the cellular myc gene (initially called c-myc). This gene consists of three exons and contains several negative regulatory elements upstream of its two transcriptional start sites and in the first intron. Transcription from a physiological start site proceeds into the first intron and pauses there until further signals arrive, similar as during transcriptional regulation by attenuation in bacteria or by the tat protein of HIV. A typical retroviral insertion of the myc gene disrupts this negative control mechanism (as well as others) and at the same time elicits strong transcription from an otherwise inactive ('cryptic') promoter near the end of intron 1. In this fashion, myc transcription becomes independent of extracellular signals. Specifically, the expression of the gene is not down-regulated in response to differentiation signals.

The c-myc gene activated by slowly transforming retroviruses is very similar to the oncogene carried by the myelocytomatosis virus (Figure 4.3). Apparently, the cellular gene is the precursor of the viral gene and has been picked up by an evolutionary precursor of the myelocytomatosis virus. This may have occured by recombination of the c-myc mRNA with the retroviral genomic mRNA followed by transduction. Even more likely, the recombination may have involved a retroviral genomic transcript and a transcript from a c-myc locus into which a retrovirus had inserted. In this fashion, slow-acting transforming retroviruses may give rise to the rarer acutely transforming types.

Figure 4.3 The v-myc protein In avian and murine transforming retroviruses, the v-myc protein retains all functional domains of its cellular ortholog. It is always overexpressed, usually as a fusion protein with viral gag sequences at the N-terminus plus linker amino acids. Thr61 is consistently mutated. Several other amino acids in the N-terminal part are mutated in individual viral strains.

Figure 4.3 The v-myc protein In avian and murine transforming retroviruses, the v-myc protein retains all functional domains of its cellular ortholog. It is always overexpressed, usually as a fusion protein with viral gag sequences at the N-terminus plus linker amino acids. Thr61 is consistently mutated. Several other amino acids in the N-terminal part are mutated in individual viral strains.

When a slowly transforming retrovirus integrates into the c-myc gene, the target gene becomes deregulated and over-expressed (Figure 4.2). The v-myc gene contained in the myelocytomatosis retrovirus is likewise strongly expressed. In addition, there are several changes in the amino acid sequence of v-myc compared to the avian c-myc gene. These are not random. For instance, a change in most viral strains removes a threonine which is required for inactivation of the myc protein, further enhancing its oncogenicity (cf. 10.3). So, in summary, the acutely transforming retrovirus overexpresses an altered cellular protein, whereas a slowly transforming retrovirus deregulates the endogenous protein, which may remain unchanged, at least initially.

As in the example of c-myc/v-myc, retroviral oncogenes are almost always altered compared to their cellular orthologs, from which they were derived. These alterations are often more severe than in the case of myc. They comprise truncation, mutation or fusion to viral proteins which increase the activity, affect the regulation and alter the localization of the oncoproteins within the cell. For instance, the v-erbB product is derived from the cellular erb-B1 gene which encodes a growth factor receptor, EGFR (Figure 4.4). This receptor is basically composed of three domains: an extracellular domain binding the growth factor ligands, a transmembrane domain, and a cytoplasmic tyrosine kinase domain, which is controlled by an autoinhibitory loop. The viral product lacks most of the extracellular domain, but contains a small gag segment which causes aggregration of the protein, as would normally be induced by the ligand. Furthermore, a point mutation and a c-terminal truncation in the cytoplasmic domain relieve auto- and feedback inhibition. So, in summary, the virus encodes and overexpresses a constitutively active protein.

Figure 4.4 Activation of erbBl to the v-erbB oncogene In the oncogenic receptor, truncation of the extracellular domain abolishes ligand binding, while fusion to a gag peptide leads to constitutive oligomerization. A truncation at the C-terminus and a point mutation in the autoinhibitory loop further enhance constitutive activity.

Figure 4.4 Activation of erbBl to the v-erbB oncogene In the oncogenic receptor, truncation of the extracellular domain abolishes ligand binding, while fusion to a gag peptide leads to constitutive oligomerization. A truncation at the C-terminus and a point mutation in the autoinhibitory loop further enhance constitutive activity.

The different time-courses of transformation by acutely and slow-transforming viruses may be accounted for by these additional alterations in the transduced oncogene. Further differences may also contribute. Acutely transforming retroviruses transduce the activated oncogene into each cell they infect and thereby create a large pool of potentially transformed cells. In contrast, slowly transforming retroviruses integrate into many different sites in different infected cells and only rarely 'hit' a cellular proto-oncogene, yielding only a few potentially transformed cells. One reason for the time lag in tumor development is thus the time required for expansion of a tumor cell clone.

There are likely two more reasons. The first is that transduction of an activated oncogene into a large number of cells is bound to have a substantial effect on their interaction with each other and host cells. For instance, it could significantly alter the levels of autocrine or paracrine cytokines secreted by these cells or overwhelm an antitumor immune response. The second reason is that a larger pool of cells containing an oncogene increases the probability of a second mutation that causes complete transformation and subsequent tumor progression. There is indeed very good evidence that transformation by slowly transforming retroviruses often requires a second hit. Occasionally, this is provided by insertion of a second retrovirus elsewhere in the genome. Some acutely transforming retroviruses also carry two oncogenes, thereby achieving the 'two hits' at one stroke.

A selection of genes that have been found activated by retroviral insertion is shown in Table 4.2. A comparison with Table 4.1 reveals that several of these genes - like c-myc - possess viral homologs. This is expected, if one assumes that acutely transforming viruses arose from slowly-acting precursors. However, many genes activated by viral insertion have never been observed to be transduced. In some cases, this may be due to the size than can be accomodated in a retrovirus, but functional limitations are also conceivable.

Importantly, there is a cellular homologue for each and every retroviral oncogene found so far. So, all retroviral oncogenes are thought to have evolved from cellular precursors. For instance, the homologue of the v-src gene of RSV is c-src, a protein kinase located at focal adhesion points, at which the actin cytoskeleton is attached to the cell membrane. The c-src kinase relays signals from cell adhesion to the cytoskeleton and to other kinases that control cell proliferation. Such signals may be mimicked by the viral oncoprotein, which is altered towards the cellular protein by several point mutations and the replacement of the C-terminal 17 amino acids by an unrelated peptide.

0 0

Post a comment