VIRUSES AS CAUSES OF CANCER

26 May

Viruses are ubiquitous obligate intracellular parasites. Because viruses replicate in and are de- pendent upon their host cells, they use the rules, signals, and regulatory pathways of the host cell. Viruses subvert and perturb normal cellular mechanisms and pathways as a means of repli- cating. These perturbations can have dire consequences for the host cell. It is not an uncommon consequence of a viral infection for the host cell to die. Though less common, viral infection can change or transform a normal cell into a neoplastic one, ultimately leading to a cancer. In fact, there is compelling evidence that several different human cancers are caused by viral infection (Chapter 12). Clearly, appreciation of this relationship can be critical in the epidemiological con- trol of cancer. Prevention or cure of a viral infection may lower the incidence of the cancer in- duced by a given viral agent. Knowledge  of cancer-causing  viruses has served a second very important function. These viruses cause cancer by perturbing normal cellular processes or path- ways. Understanding  specifically  how different viruses do this has led to an appreciation  and understanding of various molecular pathways in the host cell that can contribute to the develop- ment of cancer. Moreover, these studies have led to an appreciation  and understanding  of the normal functions of these same processes and pathways. Many insights into cancer cell biology and normal cell biology have resulted from the study of viruses that either cause cancer in ex- perimental animals or transform cells in culture. Viruses with oncogenic potential in humans, animals,  or cell culture are known collectively  as tumor viruses (for a review, see Vogt and Nevins, 1996).

Animal viruses can be divided into two broad groups: those with DNA genomes and those with RNA genomes.  The DNA viruses with oncogenic  potential  are from six distinct virus groups: hepadnaviruses,  papillomaviruses,  polyomaviruses,  herpesviruses,  adenoviruses,  and poxviruses. Two different families of RNA viruses have been found to have oncogenic potential: retroviruses and a flavivirus, hepatitis C virus (Figure 4.1, Table 4.2). Some viruses can act as carcinogens when infecting their natural host, either human or animal. Others, such as adeno- viruses or SV40 (a polyomavirus), show their oncogenic potential only in experimental settings, such as infection of cell cultures (Chapter 14). The time it takes different tumor viruses to cause neoplasms  can vary widely (Flint et al., 2000). Some induce tumors rapidly, within days or weeks (e.g., the transducing retroviruses), while others take months if not years for cancer devel- opment (e.g., human hepatitis B virus). Some tumor viruses, such as adenovirus or SV40 (poly- omavirus), do not lead to neoplasms in cells in which they replicate, but only in cells that do not support their replication.

Retroviruses

Reports in the early 1900s described the induction of neoplasms in chickens inoculated with a “cell-free  filtrate” (Ellerman  and Bang, 1908; Rous, 1911). In retrospect,  these early reports were describing the induction of cancer via infection with avian retroviruses. Later in the twenti- eth century, with the advent of cell culture methods, it became possible to induce oncogenic transformation  of individual cells in culture via retroviral infection. The retrovirus, Rous sar- coma virus (RSV), was the first reported to transform cells in culture (Temin and Rubin, 1958). The study of retroviruses has a rich and long history that has yielded many seminal discoveries.

Figure 4.1 Artist’s  conception  of shapes  and sizes of viruses  from different  families  that cause neo- plasms in vertebrates. On top are representations  of five different DNA virus families that can cause neo- plasia. Below are the two RNA viruses families associated with cancer. Polyomaviruses and papillomavirus belong to the same family, the Papovaviridae.  (Adapted from White and Fenner, 1994, with permission of authors and publisher.)

Table 4.2 Oncogenic Viruses and Cancer

Families                                                      Associated Cancers

RNA viruses

Flaviviridae

Hepatitis C virus         Hepatocellular carcinoma

Retroviridae                    Hematopoietic cancers, sarcomas, and carcinomas

DNA viruses

Hepadnaviridae              Hepatocellular carcinoma

Papovaviridae

Papillomaviruses        Papillomas and carcinomas

Polyomaviruses          Various solid tumors

Adenoviridae                  Various solid tumors

Herpesviridae                 Lymphomas, carcinomas, and sarcomas

Poxviridae                      Myxomas and fibromas

For example,  the discovery  of oncogenes  (see below) came from the study of retroviruses. Retroviruses  are a large and diverse family of viruses, and not all family members are tumor viruses. The virion contains two identical copies of a single-stranded  RNA. Depending on the family member, the RNA is 7 to 12 kb (Coffin et al., 1997). Retroviruses are enveloped (Figure 4.2). Their exterior is comprised of a lipid bilayer derived from the host cell. Embedded within this lipid bilayer are virally encoded proteins called envelope  or env proteins.  The envelope protein plays a major role in entry of the virus into a susceptible cell. Removal of the lipid bi- layer reveals the nucleocapsid,  a proteinaceous  structure that contains the genomic RNA. The viral gag gene encodes the proteins comprising  the nucleocapsid.  Inside the nucleocapsid  is found not only the genomic RNA, but two virally encoded proteins, reverse transcriptase  and integrase, which are encoded by the pol gene. Entry into a susceptible cell is dependent not only on the virion envelope protein, but also on a cell surface receptor (Figure 4.3). Upon entry into the cell, the nucleocapsid is uncoated and released into the cytoplasm. At this time, reverse tran- scription takes place, the process in which the single-stranded  RNA is converted into double- strand DNA by the action of reverse transcriptase.  Next, the linear double-stranded  DNA is transported into the nucleus of the cell and is integrated into the host DNA via the enzymatic action of the viral integrase protein. At this point the viral DNA, referred to as a provirus, has

Figure 4.2 Generalized  structure  and organization  of retroviral  particle.  The exterior of the virus is a cell-derived  lipid bilayer or membrane.  Embedded  within the membrane  are two viral proteins,  SU and TM, which are encoded by the env gene. Env precursor polypeptide is proteolytically processed to give rise to SU and TM. Underneath  the lipid envelope are viral structural proteins which are encoded by the gag gene. They are the matrix (MA), capsid (CA), and nucleocapsid  (NC) proteins. CA is the subunit of the capsid,  which is represented  by a hexagon.  Inside the capsid are two identical  copies of RNA genome (which are coated with NC protein), the reverse transcriptase, RT, and the integrase, IN. The viral protease, PR, is responsible  for processing  the gag and gag-pol  precursor  polypeptides  to yield mature  proteins. (Adapted from Coffin et al., 1997, with permission of authors and publisher.)

Figure 4.3 Retroviral replication cycle. The virus, via the SU portion of the viral Env protein, binds to a specific receptor on the surface of the susceptible cell. This interaction initiates a series of steps resulting in entry of the virus into the cell. Fusion of viral and cellular membranes is an example of one mechanism by which the viral capsid is uncoated  and deposited  into the cytoplasm.  The process of reverse transcription generates a double-stranded  DNA copy, called proviral DNA, of the RNA genome. Long terminal repeats, called LTRs, are at the 5′ and 3′ end of the proviral DNA. Proviral DNA is transported  to the nucleus and integrated into chromosomal DNA. The integrated provirus is transcribed by cellular RNA polymerase. RNA transcription begins within the 5′ LTR and ends, via polyadenylation,  within the 3′ LTR. Viral RNAs can be either full-length or spliced, and are mRNAs for translation of viral proteins. The full-length RNA is also the genomic RNA and becomes encapsidated into progeny virions. Viral proteins and two copies of the genomic RNA assemble and progeny virus bud from the plasma membrane of the cell, followed by proteolytic matu- ration of the virus. (Adapted from Coffin et. al., 1997, with permission of authors and publisher.)

become a stable component of the host chromosome and will be inherited by daughter cells. The proviral DNA contains long terminal repeats (LTRs), which are repeated sequences (up to sev- eral hundred nucleotides), found at the 5′ and 3′ ends of the proviral DNA, and generated during reverse transcription (Goff and Skalka, 1993; Varmus and Brown, 1989). The genomic organiza- tion of all retroviruses has general, common features (Figure 4.4). Between the 5′ and 3′ LTRs are the viral genes. All retroviruses capable of replicating have three genes in common: gag, pol, and env. Each of these genes gives rise to multiple polypeptides as a consequence of posttransla- tional processing. gag encodes the structural proteins that compose the caspid; pol encodes the reverse transcriptase and integrase, enzymes that function in the synthesis and integration of the provirus; env encodes the envelope glycoproteins that interact with cell surface receptors to me- diate virus entry. The viral genome contains a number of cis-acting sequences that are important for viral replication. The LTRs contain regulatory sequences for the synthesis of the viral mR- NAs. The 5′ LTR contains enhancer and promoter sequences that are used by the host RNA tran- scription  machinery.  The 3′ LTR contains  sequences  necessary  for polyadenylation  of the mRNA transcript.

Other cis-acting sequences include signals for packaging RNA into virus particles, sites of initiation of DNA synthesis, and specific sequences for correct integration of the proviral DNA. Once the proviral state is established, the cell can produce progeny virus. First, RNA transcrip-

Figure 4.4 Generalized  genetic organization  of retrovirus. The proviral DNA has integrated into chro- mosomal DNA. LTRs are comprised of three domains, U3, R, and U5. The viral genes, gag, pol, and env, are between the LTRs. All replication-competent retroviruses have the above-mentioned  genes. A subset of retroviruses (e.g., HIV-1) have additional genes, termed accessory genes, which are found in the 3′ half of the genome. The 5′ LTR contain enhancers and promoter elements, while the 3′ LTR contains the polyade- nylation signal. Translation of genomic RNA gives rise to gag and gag-pol polyproteins,  while the spliced RNA is translated to yield the env precursor polypeptide. PBS and PPT are the sites of initiation of the first and second strands of DNA, respectively.  SA and SD are splice donor and acceptor sequences.  (Adapted from Coffin et al., 1997, with permission of authors and publisher.)

tion, mediated by the host machinery, begins in the 5′ LTR and terminates in the 3′ LTR to pro- duce a primary transcript that is essentially the genomic viral RNA. This will be the mRNA for the gag- and pol-encoded proteins as well as genomic RNA for progeny virus. A spliced version of the genomic RNA is synthesized and will be translated for the synthesis of env glycoprotein. Upon synthesis and accumulation of all of the necessary viral proteins, a new virus particle will assemble at the plasma membrane and bud from the cell, yielding a new progeny virus.

Oncogenic  retroviruses  can be isolated from neoplasms  and, upon introduction  into the appropriate  host, cause malignant  and benign tumors. They can be divided into two general groups on the basis of the time after infection at which the cancer phenotype appears: rapid (1 to a few weeks) or slow (6 months to 1 year) (Coffin, 1996). Members of the rapid group will effi- ciently transform cells in culture. This group of oncogenic retroviruses are also called the trans- ducing retroviruses. The significance of this name is discussed below. A well-studied member of the transducing retroviruses is RSV. Within 1 to 2 weeks of infection with RSV, a chicken will develop several large sarcomas. In addition, infection of chicken embryo fibroblast cells in culture with RSV will result in nearly all cells becoming transformed. The second group of oncogenic retroviruses take a longer time to induce cancer in animals and are not transforming in cell cul- ture. Members of this group are called the nontransducing retroviruses. The prototype member of this class is the avian leukosis virus (ALV). As can be inferred from their different cancer pheno- types, the two groups of oncogenic retroviruses induce cancer via different mechanisms.

Molecular analysis of the RSV genome reveals, in addition to the gag, pol, and env genes, the presence of a fourth gene (Figure 4.5) (Schwartz et al., 1983). This gene is located between the env gene and the 3′ LTR. Mutants of RSV that have this gene deleted are not transforming but can still replicate, indicating that the new gene is not required for viral replication. But the fourth gene, called v-src, is required and is sufficient for the development of cancer. v-src was the first example of what is now called a viral oncogene or v-onc. Introduction and expression of the v-src gene in cells in culture can lead to their transformation.  It was found that normal chicken cells contained a gene similar to v-src (Stehelin et al., 1976). In fact, sequences related to v-src are present in other vertebrate species, including humans. The cellular gene that is a homologue  of v-src is called c-src (short for cellular-src)  or proto-src, a proto-oncogene,  and contains introns, unlike its viral counterpart (Figure 4.6). The protein encoded by c-src can be found in cells, and its expression does not lead to cell transformation  in culture (Collet et al.,1978; Iba et al., 1984). The c-src gene has biological and biochemical properties different from v-src. Both c-src and v-src encode plasma membrane-associated  proteins with tyrosine kinase activity.  Comparison  of the amino acid sequence  of v-src and c-src proteins  reveals several changes or mutations within v-src that are responsible for its transforming activities (Takeya and Hanafusa, 1983). We now know that the v-src gene in RSV was captured or derived from its cellular counterpart. The capture of an oncogene such as v-src is believed to happen during re- verse transcription at a relatively low frequency. But because the selection and amplification of these infrequent events is very powerful (i.e., clonal expansion of the neoplasm), capture of a proto-oncogene by a retrovirus has been detected many times.

In addition to RSV, more than 60 other examples of transducing retroviruses that encode over 30 different oncogenes have been isolated from a variety of birds and mammals (but not from humans) (for a review, see Coffin et al., 1997) (Table 4.3). Each of these viruses has at least one oncogene that is not required for viral replication but that is responsible for the virus’s can- cer phenotype.  For each example of a v-onc, a corresponding  c-onc has been identified.  The existence of multiple v-onc genes and their corresponding c-onc genes suggested that there are multiple ways to cause a cell to become cancerous. Proto-oncogenes  have roles in the normal functioning of a variety of cellular processes such as proliferation, differentiation, and develop- ment. The v-onc gene almost always has changes or mutations that are responsible for the vi-

Figure 4.5 Rous sarcoma virus (RSV) contains an additional gene, src, not present in other retroviruses. The src gene is located between the env gene and the 3′ LTR. In comparison, avian leukosis virus (ALV), a closely related retrovirus, contains only the normal complement of genes, gag, pol, and env. RSV expresses Src protein from a spliced mRNA. (Adapted from Cooper, 1995, with permission of author and publisher.)

rus’s ability to cause cancer. The role of oncogenes in cancer is further underscored by the fact that some nonviral cancers are due to mutation of proto-oncogenes  and formation  of cellular oncogenes (Table 6.2; Parada et al., 1982).

Other transducing retroviruses differ from RSV in one key respect (Figure 4.7). Whereas RSV can replicate on its own, all other examples of transducing retroviruses are replication de- fective (for a review, see Coffin et al., 1997). They are incapable of replicating on their own.

Figure 4.6 Comparison  of the chicken src proto-oncogene  and the RSV src oncogene.  The cellular src proto-oncogene is comprised of 10 exons (filled boxes) that are separated by introns and spans 8 kb of chicken chromosomal DNA. The viral src oncogene does not contain introns, as a consequence of splicing and reverse transcription of a src mRNA. (Adapted from Cooper, 1995, with permission of author and publisher.)

They require coinfection  with a replication-competent  or helper retrovirus  to be propagated. These viruses are replication defective because one or more of the viral replication genes (gag, pol, or env) are missing. In their place is the viral oncogene. As a consequence of replacing part of the viral genome, it is not uncommon for the v-onc to be expressed as a fusion with one of the viral replication proteins such as gag. In some cases, the fusion partner of the v-onc gene con- tributes to its oncogenic potential. There are two general mechanisms by which a v-onc of the transducing retroviruses contributes to the cancer phenotype. As mentioned above, v-onc genes contain mutations that confer an altered biochemical activity that in turn leads to an altered bio- logical activity.  In addition,  in general, v-onc genes are expressed  at levels higher than their c-onc counterparts as a consequence of being under the transcriptional control of the viral LTR (for a review, see Cooper, 1995). Higher levels of the v-onc mRNA can lead to higher levels of the v-onc protein, which can contribute to the cancer phenotype.

The group of retroviruses that take longer (6 months to 1 year) to induce cancer do not possess an oncogene.  This group of oncogenic  retroviruses  induces neoplasms  by integrating near a proto-oncogene.  This phenomenon  is termed proviral  insertional  mutagenesis  or cis- activation (Figure 4.8). The presence of the provirus next to the proto-oncogene leads to its dys- regulated expression, which plays a central role in induction of neoplasia (for a review, see Kung and Vogt, 1991). The prototypic  member of cis-activating  retroviruses  is ALV, which causes lymphomas in chickens. Within a given ALV-induced lymphoma, all the cells contain an ALV provirus at the same chromosomal site, indicating that the neoplasm was clonal. When different independent lymphomas were examined, the ALV proviruses were found integrated into the same region of cellular DNA, adjacent to the c-myc locus (Hayward et al., 1981). c-myc is the cellular counterpart of a viral oncogene initially isolated from the transducing retrovirus, MC29. Insertion of the ALV provirus results in overexpression  of the c-myc gene. RNA transcription  from the ALV LTR, which is now next to the c-myc gene, leads to upregulation of its expression, which in turn leads to neoplastic transformation. ALV can upregulate expression of other proto-oncogenes, such as erbB, fos, and H-ras, by insertional activation to contribute to the induction of cancer (Cooper, 1995). A second example of a non-transducing oncogenic retrovirus is mouse mammary tumor virus (MMTV). MMTV has several preferred integration sites close to cellular genes, Int-1, Int-2, and Int-3 (for a review, see Nusse, 1991; see Chapter 8). These three genes are located on different chromosomes and are not related to one another. In mammary tumors the MMTV provi- rus increases expression of the adjacent int locus, which likely contributes to tumor development. Int-1, now called Wnt-1, belongs to a family of genes that play a role in pattern formation during embryo development in organisms as diverse as Drosophila and mammals. Int-2 gene encodes a member of the fibroblast growth factor family. The Int-3 gene codes for a protein believed to have developmental function and is a member of the notch family of developmental regulators.

In summary, both transducing and nontransducing retroviruses contribute to the cancer pheno- type via oncogenes. Proto-oncogenes are cellular genes that are important for a variety of cellular processes such as proliferation,  differentiation,  and development.  There are two general mecha- nisms by which these oncogenic retroviruses contribute to the cancer phenotype: (1) by capture, delivery, and expression of a variant version (v-onc) of a proto-oncogene and (2) by integrating next to a proto-oncogene and causing its overexpression via the viral LTR. The study of oncogenic retro- viruses has been revolutionary in that it led to the discovery of the oncogene. To date, no example of a transducing or a cis-activating oncogenic retrovirus has been identified in humans.

Random Posts

Comments are closed.