Howard Temin (1984) - How Some Viruses Cause Cancer

As you can tell by the previous talks, oncogenes seem to be everywhere, both in us and in biological research now. I am going to talk on a similar topic to the last speaker and tell you about the development of the idea of oncogenes and proto-oncogenes. And how this work, originally with animal viruses in non-human hosts, led to our appreciation of the genetic mechanisms in all cancers. There are known viruses involved in human cancer. These are listed on the first slide, if I could get it. These include the DNA viruses, Epstein-Barr virus related to Burkitt’s lymphoma, a nasopharyngeal carcinoma, the hepatitis B virus involved in primary hepatocellular carcinoma, and the recently described human retroviruses HTLV, ATLV involved in adult T-cell lymphoma. However, in the majority of human cancers in the developed countries viruses like these are not important. But other factors like cigarette smoking, radiation, various undefined life cycle factors are responsible for the development of most cancers. However, the study of cancer by viruses called highly oncogenic retro viruses has been avidly pursued for a great number of years, because of the very simple ability of these viruses to cause cancer. In contradistinction to the usual multistage nature of carcinogenesis, these viruses can infect a normal cell and in a single step change the normal cell into a cancer cell. The viruses that can do this differ from normal viruses. Normal viruses either have DNA as their genetic material and replace the genetic material of this cell with the viral DNA – could you focus the bottom of the slide please? – or they have viral RNA as their genetic material; viral RNA which replicates without the use of a DNA intermediate. Most of the viruses which infect us are of this type herpes virus and this type cold viruses, measles virus, so forth. The viruses which are able to cause cancer in this very rapid fashion are known as retroviruses, because they reverse the usual flow of information. Instead of information just flowing from DNA to RNA to proteins, in the case of the retroviruses the information flows in reverse, from RNA to DNA. This DNA then is added to cell DNA and multiplies as a cell gene with the cell DNA. The virus is able to do this because of 3 remarkable genes in the virus. Here we are looking at a diagram of the virus life cycle, with viral RNA indicated as this string. This viral RNA acts as a template. Here we are looking at the genome in DNA. And this gene called polymerase, pol, codes for an enzyme known as reverse transcriptase which, using specific primers, is able to transcribe this RNA into a linear DNA copy. The linear DNA copy contains at both ends, signified by the open boxes, a large terminal repeat with a small inverted repeat. These are the other viral genes. This process takes place in the cytoplasm. In a very specific manner, involving double jumps of the DNA polymerase complex. The linear unintegrated viral DNA is then transported to the nucleus where it is circularised by blunt-end ligation, so that the 2 ends of the linear DNA are juxtaposed, making a circle junction. This circle junction then contains an inverted repeat by putting this inverted repeat next to that one. The pol gene, in addition to coding for reverse transcriptase at its 3-prime end, codes for another activity which we call INT, for integrase, though this is not yet proven. And this activity acts on the circle junction sequence which we have called AT for attachment, to remove 4 base pairs in the centre and then to covalently insert the viral DNA as a co-linear copy minus 2 bases on each ends into the cell DNA. At the same time this enzyme, or another enzyme unknown, brings about a direct duplication of sequences in the cell DNA. So the virus contains an extremely specific and efficient machinery that is able to take external information in the form of RNA and add it to cell DNA as a covalent copy. Now I am not, because of what I was asked to do today, not going to tell you more of the details of this process, though I would be glad to discuss it with anyone who is interested this afternoon. So the first special features of retroviruses that enable them to cause cancer in such an efficient manner are their ability to make a DNA copy of RNA, and then to add this DNA into the cell genome, where it becomes a cell gene, an external gene into a cell gene. Next slide - automatic projectors work better sometimes. Now we’re looking again at a cartoon of the genome of the virus with at the ends the open boxes referring to this large terminal repeat, known as LTR, and in the middle instead of the genes I have just written viral coding sequences. The sequences coding for the proteins of the virus, including those for the coat and those responsible for reverse transcription and integration. And here I have enlarged the end of the viruses. So these large boxes are now the repeat. I have indicated here at the very ends the AT sequences which we have discussed as being responsible for the integration of the virus. In addition there are many other specific sequences in the LTR, and nearby here and here, responsible for control of RNA and DNA synthesis of the virus. And finally there is a sequence here known as E for encapsidation which is responsible for telling the viral proteins to encapsidate the viral RNA into a virus particle. So we can see that in the virus there has been a segregation of its sequences, with the viral coding sequences in the centre and the sequences which control the virus which we speak of as cis-acting sequences, since they act directly on the nucleic acid to which they are attached. As compared to these sequences we speak of as transacting because their product can act on separated molecules. These are segregated with the transacting coding sequences in the centre, the cis-acting sequences at the end. Now some of these viruses can cause cancer very rapidly and efficiently. This is an example of a chicken injected 2 weeks earlier with one of these viruses and the chicken is dead of a fulminating leukaemia. This is an example in cell culture of a cell infected 10 days previously by a virus and transformed into a cancer cell seen here on the dish. Now when we look at the structure of viruses able to bring about this very rapid transformation – and an example of one of these is shown at the bottom - we find that the structure of the virus is very different from the structure of the viruses I have been talking about. In these top viruses, which we call replication-competent viruses, we see the familiar open boxes of the large terminal repeat. Here various cis-acting sequences responsible for DNA synthesis and encapsidation and the 3 genes coding for viral proteins. In the highly oncogenic retrovirus we see that the terminal sequences at each end are the same. In other words the highly oncogenic retrovirus maintains the cis-acting sequences responsible for RNA and DNA synthesis and encapsidation of viral RNA. But it has sustained a deletion of coding sequences and a substitution of new sequences. To establish the role of these new sequences, we carry out genetic experiments using techniques of recombinant DNA, which allows us to make specific deletions in viral genomes and then using transfection of animal cells related to the transformation process described by Doctor Smith, we can recover the viruses and ask if they’re still capable of causing cancer. Here we have the genome of one of these highly oncogenic retroviruses containing the new sequences, the deletion and other viral sequences. If we delete more of the viral sequences - and it’s best to look at this construct -, so that all that is left are the cis-acting sequences at both ends and the new substituted sequences, we see the virus still transforms. However, if we maintain the viral sequences and make any kind of an alteration in the substituted sequences, the viruses which are recovered are no longer tumorigenic. This is the strict definition of an oncogene, an oncogene or oncogenes or genes in highly oncogenic retroviruses, the code for protein product required for neoplastic transformation. Similar types of experiments have been carried out with about 20 different viruses. And about 20 different oncogenes have been recognised. All from their appearance in retroviruses and their ability then to transform normal cells to neoplastic cells after addition to the cell genome by retrovirus infection. Of course you can’t read them and the names src, yes, fps, fes, abl, ros, fgr, erb, fms, mas, K-ras, H-ras, myc, myb, fos, raf, skirail and sis are merely 3 letter mnemonics. Listed here are the species of origin of the viruses. They have been found in chickens, cats, mice, rats, turkey and monkey. And in some case the same oncogene has been isolated from a virus from 2 different species. Furthermore the oncogenes are clustered into groups on the basis of similarities in sequence and function. So that these 6 oncogenes all have a tyrosine protein kinase activity. These 3 have no such activity but have a DNA sequence related to src. These 2 are related by sequence and by protein making a GTP-binding activity. These have a nuclear location and these have unknown functions and locations. So we see that viral oncogenes are not an infinite set, but are actually a small set with the same gene being isolated repeatedly from different species. And furthermore we see that even the different viral oncogenes fall into related families. So this tells us that at least from this definition of oncogene that only a small number, somewhere we would guess less than 40 genes, are able to be turned into these powerful cancer genes. And not any kind of a gene can become a cancer gene. Now as Professor Ochoa indicated, the viral oncogene comes from a normal cell gene known as proto-oncogene. In a number of steps a retrovirus picks up the cell DNA and makes it into an oncogene. This is again a specific cartoon comparing a replication-competent virus, a highly oncogenic retrovirus, and a proto-oncogene whose sequence is homologous to the sequence of the viral oncogene. And you can immediately see that the proto-oncogene does not resemble the viral oncogene. The proto-oncogene is split into exons and introns. The proto-oncogene has its own promoter and 3 prime sequences, while the viral oncogene uses the viral promotor and termination sequences. The proto-oncogene is expressed in normal cells and, as I guess Doctor Holley told you, at times can have a very important function controlling normal cell growth. So the proto-oncogenes are cell genes whose DNA sequences are homologous to viral oncogenes. And so as there are 20 to 40 oncogenes, we imagine there are only 20 to 40 proto-oncogenes, normal cell genes whose mutation in the way of transduction by a retrovirus can cause cancer. Now what is the difference between the proto-oncogene which does not cause cancer and the viral oncogene which does cause cancer? I have already indicated to you that there is a very big difference in regulation because in the one case the proto-oncogene is controlled by cell transcriptional control signals. In the case of the oncogene the expression is controlled by viral transcriptional controls. Furthermore, when the sequences are compared of the protein produced by the proto-oncogene, which is illustrated by the heavy black line as if the amino acid sequence was written out. And since this is the normal homologue, the precursor is written out as a continuous bore. When this is compared with the protein sequence of the viral oncogene, indicated here with circles, Xs, triangles and thin lines, we see there are numerous differences. First we’ll look at the similarities. Wherever the thin line is continuous the proteins are the same. Wherever there are triangles there are amino acid differences in the middle of the protein. And the Xs at the end and the circles and Xs at the beginning indicate that the viral oncogene uses sequences from the virus to code the amino and carboxy-termini of its protein. In other words the proto-oncogene has been severely modified on becoming a viral oncogene. It has had its regulation changed by losing its own control sequences, and having viral control sequences it has changed the ends of its protein molecule and it has changed some amino acids in the middle. Now these are correlational changes. I am just going to indicate to you, this one kind of experiment which tells you how we determine which of these changes are significant in the transforming process. Again there is an enormous amount more I could tell you about this but this is just to indicate how these are done. Again we’re looking at viral genomes with the boxes and the lines representing the cis-acting sequences which are all the same at the end. These are now what we call retrovirus vectors where we’ve removed all the virus-coding sequences and in this case replaced them with the thymidine kinase gene, or viral oncogenes, or cellular proto oncogene. So these are highly oncogenic-like viruses that have been constructed, as we say, by cut and paste in the laboratory. Where we make viruses at will, viruses of different kinds for different purposes. Now the purpose here is to look at 2 of these, this one and this one. This one contains a viral oncogene, the Harvey RAS gene. The virus is produced in good yield. It is able to transform normal chicken cells immediately upon infection. But it is not able to transform rat cells. Even though the virus is present in the rat cells and the gene in the virus is expressed at a low level. This difference is related to the regulation of the gene activity. The viral promotors are very active in chicken cells. But these particular virus promotors are not active in rat cells. In that case there is not sufficient product of the oncogene to transform the cells. Similarly, I will tell you, but not illustrate, with other viral oncogenes if the oncogene is expressed at a high level the cells are killed. And it is only when the oncogene is expressed at a low level that the cells are transformed. Thus we can conclude experimentally that the change in regulation has been important in the transforming process. We speak of this as a quantitative change. Now at the bottom we’re looking at a similar construct. For various technical reasons there had to be a deletion involving the proto-oncogene which, as Doctor Ochoa told you, just differs in a single amino acid from the viral oncogene. Again the virus is recovered at full yield. But it is unable to transform chicken cells even though the viral oncogene can transform them. Thus this experiment indicates that even when there is misregulation of the normal cell gene, it is unable to transform the cells. Thus we can answer the question of our title, how some viruses cause cancer. That highly oncogenic retroviruses cause cancer by introducing, by adding to the genomes of sensitive target cells at new locations a quantitatively and qualitatively altered cellular gene. The cellular gene is the proto-oncogene. It is qualitatively altered by base-pair mutations and fusion with viral coding sequences. It is quantitatively altered by being under viral regulation rather than cellular regulation. It is at new locations because the viral integration machinery is completely specific for the attachment sequence of the virus, but is not at all specific for the recipient sequences, the target sequences in the cell. There is no homology involved in integration. And this is another topic we can discuss later. The cell must be an appropriate cell. This has been a special kind of carcinogenesis carried out by special viruses. Such viruses exist in domestic cats, where they cause solid tumours in this very rapid fashion. Fortunately such viruses do not exist in man. And most cancer in man, as Doctor Ochoa indicates in this cartoon just confirms, arises in a multi-step fashion. Here we are considering a row of normal cells and watching this cohort of cells through time. Where first a single mutation appears and then replicates, giving an altered clone of cells. Then a second mutation appears giving a further altered clone of cells - now 2 altered clones. Then a third mutation appears. And somewhere between 3 and 7, we’re not sure of the number, mutations occur and the cell becomes cancerous. So we have 2 apparently very different kind of processes leading to cancer. But as Doctor Ochoa again told you, the same proto-oncogenes are apparently involved. How do we reconcile these differences? Over here we’re looking at proto-oncogenes and here at active oncogenes. In the case of carcinogenesis by a highly oncogenic retrovirus, the proto-oncogene has been altered through evolution in multiple steps to give a multiply altered product mutated, fused, differently regulated. So the highly oncogenic retrovirus is able to cause cancer so efficiently by adding the already multiply altered gene into the cell DNA. Now in the case of the non-viral cancers - could you focus at the bottom please - in the case of the non-viral cancers, again in some cases proto-oncogenes have been found to be altered: a base-pair mutation, an amplification, a translocation. However, in the case of the non-viral cancers there are several proto-oncogenes each which has been altered one time. And the multi-step process has been the accumulation of one change, a second change, a third change. The multi-step process here took place in time. But because of the nature of the viral vectors, all of the changes involved the single gene, the viral oncogene. So we see that the same process is involved in carcinogenesis by highly oncogenic retroviruses and by other kinds of agents. There are several genetic changes to normal cell genes which change them into active cancer genes. Thank you. Applause.

Howard Temin (1984)

How Some Viruses Cause Cancer

Howard Temin (1984)

How Some Viruses Cause Cancer

Comment

Howard Temin participated in two Lindau Meetings and gave lectures at both of them. The first lecture, in 1981, is rather technical and the present lecture is more easily understood. It is an unusually clear presentation of the way that certain so-called retroviruses act and an explanation of how they can cause cancer. Central to the understanding is the process of reverse transcription, which was discovered independently by Temin and David Baltimore around 1970. With transcription is meant the process when the double stranded DNA makes a single stranded RNA copy, which then can be used as a messenger to the protein factory of the cell, ordering it what kind of proteins to produce. Some viruses only contain RNA, but can anyway modify the DNA of a cell into which the virus has entered to transform it into a cancer cell. How this can be done was a mystery until Temin and Baltimore, using methods developed by Temin’s teacher Renato Dulbecco, made their discovery of reverse transcription. In his lecture, Temin goes one step further in the understanding of how retroviruses act and mainly talks about oncogenes, proto-oncogenes and viral oncogens. An oncogene is a gene which causes a cell to become a cancer cell and which, e.g., may have been introduced into the cell by a virus. As shown by two other Nobel Laureates, Michael Bishop and Harold Varmus, the virus first picks up the gene that eventually will become an oncogene in a normal cell. The gene that is picked up is called a proto-oncogene. Temin, in his lecture, explains how this proto-oncogene first is modified by the virus to become a viral oncogene. When this viral oncogene has been introduced into the DNA of a cell, it becomes an oncogene and the cell becomes a cancer cell. It turns out that this has to do with the process of regulation. A proto-oncogene in the DNA of its original normal cell is regulated by other genes of the cell. When the virus picks up the proto-oncogene, it doesn’t pick up the regulator genes. This means that the regulation changes, so that the proto-oncogene becomes a viral oncogene. According to Temin, the number of proto-oncogenes is rather small and certainly below 40. Different viruses pick up different proto-oncogenes, causing different kinds of cancer. All this is of great interest for basic cancer research, but as Temin points out, the majority of cancers in human beings are not caused by virus infections. Instead other factors, such as environmental pollution or smoking, are important and a large proportion of human cancers are lung cancers caused by smoking. Howard Temin was not a smoker, but he acquired lung cancer anyway and passed away at age 59.Anders Bárány

Cite


Specify width: px

Share

COPYRIGHT

Cite


Specify width: px

Share

COPYRIGHT


Related Content