 |
MB 451 Microbial Diversity |
 Audio recording of this lecture |
Required reading for this lecture :
One of the two following papers (see below):
- Soto C and Castilla J 2004 The controversial protein-only hypothesis of prion propagation. Nature Medicine 10:S63-S67
- Somerville RA 2002 TSE agent strains and PrP: Reconciling structure and function. TiBS 27:606-612
Viruses
 Escherichia coli, being infected with T4 |
What is a virus? Viruses are DNAs or RNAs, usually encapsulated in protein &/or membrane, that infect a cellular host & use the host ribosomes & cellular machinery for replication. Viruses infecting all 3 Domains (Bacteria, Archaea & Eukarya) are common. |
What are viruses in evolutionary terms?
There are 3 general views of what viruses might be:
- genetic offshoots ('satellites') of their hosts
- remnants of precellular life
- degenerate parasites
Since viruses have no ribosomes of their own, they have no rRNA and can't be included in the rRNA-based 'Universal tree'. But it is possible to use other genes present in any particular virus for phylogenetic analysis if they are conserved enough for alignment.
Many genes are very different in viruses because of their specialization and rapid evolutionary rate, & therefore are not useful in phylogenetic analysis. Others are OK - in these cases, the virus is clearly associated as a relative of the host-group (e.g. the tRNA genes of T4), i.e. these viruses are almost certainly off-shoots of the host genetic system. But it is certainly possible that only some of the genes are from the host & that the 'core' of the virus predates the host & acquired those genes from the host recently.
Another reason to believe that viruses are at least predominantly derived from their hosts is that all of them, without known exception, are fundamentally dependent on their hosts for cellular processes for replication. If viruses are remnants of precellular life, they must have had mechanisms for replication independent of cytoplasm. It seems unlikely that only parasitic remnants persist, but it is possible that free-living 'viruses' have yet to be discovered. After all, how would you look for them, or even detect their presence?
The viruses are a collection of various very different kinds of 'organisms', & while some or most of them are probably genetic offshoots of their hosts, some nevertheless may represent remnants of precellular life, or degenerate parasites. There is no reason to believe that viruses are generally related to one another geneologically - different groups of viruses presumably had different origins. Some are apparently genetic offshoots of their hosts. Others may have other origins.
Viruses as genetic offshoots of their hosts
Viruses are similar to transmissible plasmids, but transfer via encapsidation rather than membrane pore proteins. Many plasmids can be transfered to new cells by conjugation, e.g. the F plasmid. These plasmids carry a series of genes that direct replication & conjugative transfer of the plasmid. Since the plasmids often are carrying genes from the host chromosome, these genes are transfered to the new host with the plasmid. This is the mechanism for sexual exchange of genes in Bacteria.
Many viruses do not kill the host; they are essentially transmissible plasmids that are transfered to their new host via an encapsidated form rather than directy via conjugation. An example of this is bacteriophage M13:
M13, then, is a lot like a conjugative plasmid except that the transfer mechanism doesn't involve direct donor-recipient contact.
Some viruses seem to have derived from transposons (mobile genetic elements). Retroviruses are essentially transmissible retroposons (transposons with reverse transcriptase genes, e.g. Ty in yeast, Copia and P-element in flys). Bacteriophage Mu is essentially a transmissible out-of-control transposon.
Some viruses, then, are the ultimate 'selfish genes'.
Viruses as remnants of precellular life
Many believe in an "RNA World" before the evolutionary invention of DNA, protein, or lipid membranes. This is plausable because:
- DNA functions only through RNA intermediates.
- Protein synthesis is fundamentally an RNA-driven process (tRNAs, ribosomes, mRNAs).
- RNA can perform the functions both of DNA (genome) and protein (catalysis).
The complexity of this RNA World is a matter of wildly speculative and largely uninformed debate.
Many RNA viruses (especially plant viroids) seem similar to what these pre-biotic RNAs might have been like, in that they direct their own replication process without DNA intermediates and preform at least some of their own replication functions (self-cleavage & ligation of replication forms by ribozymes). About all these viruses need from their host to replicate is a primase, RNA polymerase, NTPs, and a decent physical/chemical environment. But, it's hard to imagine that these RNAs, which are functional only in cytoplasm (that of it's host) could be directly descended from independent RNAs that lived in the precellular RNA world that may or may not have existed.
Viruses as degenerate parasites
Viruses may be the simplified remnant of microbial intracellular parasites, similar to Bdellovibrio, Chlamydia, or Rickettsia. It is certainly true that parasites often become extremely simplified, shedding anything unneeded. Starting with an obligate intracellular energy parasite such as Chlamydia, all that need happen to get to a virus is the fusion of the cytoplasms of the host and parasite so that the virus can use the translational apparatus, and perhaps even the hosts transcription and/or DNA replication machinery. The genome of the parasite then become even more drastically simplified, perhaps even dispensing even with DNA itself, leaving only RNA.
The clearlest potential example of this might be Mimivirus, a virus of amoeba, which has a capsid 0.4um in diameter (as big as the smallest of prokaryotes, e.g. Nanoarchaeum or Ultramicrobium) and a genome of 1.2Mbp, as big as many prokaryotic genomes and bigger than many. The genome encoded a slew of translational proteins, such as EF-Tu, release factor 1, tRNA-synthetase (tRNA charging emzymes), IF-1, topoisomerases, etc. However, this organism lacks genes for ribosomal RNAs or proteins, as well as genes for central metabolism.
Prions
Transmissible spongiform encephalopathies (TSEs) are infectious neurodegenerative diseases, leading ultimately to disabling dementia and death. Scrapie is a well-known and historical problem in sheep herding, coming in a variety of versions with symptoms ranging from disabling itching (and resulting in poor wool yield and quality) to dementia (how would you know, in sheep?). Bovine spongiform encephalopathy (mad cow disease; BSE) is a recent problem in cattle, in which cows have become infected by scrapie by feeding them sheep carcasses in their feed. The problem here is that it can be transmitted to humans who eat infected cows. Creutzfeld-Jacob disease (CJD) is a human disease, found in cannabalistic populations, where the human brains are ceremoniously eaten. TSE-like diseases can also occur spontaneously in individual (very rarely), and if in the germ line and not lethal too young, can result in familial/inherited forms of the disease.
TSE brains have a degenerate spongy appearance, and there is also an accumulation of "plaque", an insoluable, protease-resistant form of a normal brain protein of unclear function. Other than being transmissible, this looks a lot like Alzheimers.
At first, of course, it was assumed that these diseases were caused by a slow-acting virus. But no viral DNA or RNA has ever been found to be associated directly with TSEs, and the fact that you could heat or irrradiate infective samples without destroying their infectivity suggested that they probably weren't viral in nature.
In 1982, Stanley Prusiner suggested that TSEs were caused by infectious particles made entirely of protein - no nucleic acid - and dubbed these "prions". This idea met heavy resistance, at least some of which was dogmatic. But very quickly it was discovered that plaque was composed of a protein (Protease-resistant Protein; PrP) normally found in the CNS, but in a different conformation. The normal conformation of this protein (PrP^c) is soluble, primarily alpha-helical, and is a fragment of a larger pre-protein released by a specific proteolysis. The pathological conformation of this protein (PrP^Sc) is insoluable, mostly beta-sheet, and protease-resistant. Amazingly, the introduction of small amounts of PrP^Sc into a solution of PrP^c in vitro causes the PrP^c to change into PrP^Sc. In other words, PrP^Sc catalyses the transformation of PrP^c into PrP^Sc. It looked like the "Prion" had been found!
If the idea of a protein with a "normal" and "abnormal" conformation, in which the abnormal conformation has the ability to convert normal protein to it's abnormal form, seems odd, then think of it this way. If you have crystals of salt and a container of supersaturated salt solution, these represent two forms of the same chemical, sodium chloride. If you add a single salt grain to the supersaturated solution, it will convert much of the dissolved salt into new crystals. One of these new crystals could then be transfered to another container of saturated brine, causing it to crystalize out. Thought of this way, TSEs aren't so much an infection as they are a communicable metabolic disorder.
In the 20 years since then, the study of TSEs has focused on PrP and prions. Those who don't entirely buy the prion theory might argue that the prion theory has ironically become the new dogma! The argument over the cause of TSEs has been loud and heated over the years, and of course the reason is simple; because nobody can definitively prove what the infectious agent is. The leading candidate is PrP/prions, but the required experiment has never succeeded. Although PrP^Sc extracted from diseased brains (that could presumably still contain virus) has been shown to be infectious, nobody has ever caused an infection from PrP^Sc created in vitro. In other words, Koch's Postulates have not been fulfilled. This leaves the door open for a variety of doubts, especially since the original reasons for doubting the presence of a virus are weak. And although PrP is clearly a component of the disease, it still isn't clear whether it is the central player, or a side-show.
The crux of the argument about the cause of TSEs has focused on one aspect of these diseases that is very difficult for the prion theory to explain; the fact that these diseases come in clearly distinguishable "strains" that transmit true from host to host. For example, scrapie comes in two major types; the itchy form and the dementia form, and many sub-types can be distinguished pathologically. Given that PrP is a single host protein, how could you have more than one form of the disease? Shouldn't the form of the disease depend on the host (which alleles of the PrP gene it has) rather than the source of the infection? But it's not the host alleles of PrP that determine the type of infection; any breed of sheep can get either form of the disease. These "strains" transmit faithfully even from species to species. Neither is it the route of "infection" that determines the type of disease. Those who believe in the prion theory argue that there are multiple types (conformations) of PrP^Sc, each of which can catalyse normal PrP^c to adopt it's conformation. Given that there are something like 2 or 3 dozen strains of scrapie, and that no other protein in the body can do this even one way, this is a lot to ask for, to say the least. Many believe this demands some kind of genetic information independent of the host, i.e. a nucleic acid.
However, going back to the saturated salt analogy, imagine it's not salt but something a bit more complicated - silica. Although NaCl has only one common crystal form, silica can form many crystal types. Supersaturated silica solutions seeded with one type of crystal will grow that type of crystal. If something as simple as silica can do it, why not a protein? But if it were this simple, why don't other proteins do it at all?
And so, for some of us, the jury will remain out until TSEs either are shown to be transmitted by PrP^Sc grown in vitro (or Koch's postulated are fulfilled in some other way), or a virus (or viroid, etc) is identified and it fulfills Koch's postulates. My favorite quote from Charles Darwin is "Ignorance more frequently begets confidence than does knowledge,...", referring to scientific dogmas.
And so, in this light, the required reading for this lecture will require you to have a look at the other side, whichever side that may be. If you don't buy the prion theory, just can't imagine a non-nucleic acid transmissible disease with strains, then read:
to get a good explanation of the consensus perspective. On the other hand, if you do like the prion theory, and feel like it does a good job explaining TSE's and TSE strains, then read:
to get the dissenting view. These are both good descriptions of the issue. There are many others, but I tried to pick these out to be clear and non-fanatical (and recent). There are lunatic fringes on both sides of this fence!
Questions for thought:
- Are viruses alive? (There's a great Scientific American article with this title, in the December 2004 issue.) How would you argue your point to a critical audience?
- How would you go about attempting to figure out the origin of a particular virus (choose one)?
- What do you think the limit might be for how small and simple a virus could be? Why not smaller?
- What do you think the limit might be for how big and complex a virus might be? Why not bigger?
- Do you buy the Prion Theory? Why, or why not?