Optimization of Virus Extraction from Soils

With much of the enumeration work temporarily halted by a lack of filters, I have become much more involved in the molecular biology of the Lake Matoaka ecology project. If I cannot measure virus abundance (while simultaneously fulfilling the legwork of this blog’s titular project), then I can still measure diversity.

As evidenced by the schematic above, viruses are a diverse group indeed, in terms of form alone, leave alone methods of infectivity. In an ideal world, there would be a single gene common to them all, like the 16s rRNA gene in bacteria, which could be targeted for molecular work. Not so! “Virus” as a broad term is not limited by any genetic boundaries; they are as diverse as all life on earth, given that they have likely been co-evolving with cellular life for as long as cellular life has existed. There are certainly a few more prominent genes, but none that are all inclusive. For example, genes which code for specific capsid proteins may cover a broad range of bacteriophages, but even considering that gene specifically would not be inclusive enough.

For those of you not steeped in PCR, whether for class or research, allow me to clarify. A common gene can be instrumental in obtaining a quick census of what is present in a community. All bacteria have the gene necessary for making a ribosome; it is so necessary, it has been conserved in the face of all other variation. Via the polymerase chain reaction (PCR), the gene of interest can be amplified (copied) across all species present in a sample. However, the 16s rRNA gene, while highly conserved, is not exactly identical amongst all bacteria. By detecting subtle differences in the amplified genes, one can associate those differences with known bacterial species. Even if the variation is only a few nucleotides of DNA, that can be enough to determine whether an amplified 16sRNA gene fragment belongs to “Bacteria Species A” or “Bacteria Species B.” Again, since this gene is so ubiquitous, this can be extrapolated across all bacteria! Below is a highly hypnotic model of the small subunit 16s Ribosome, courtesy of wikipedia:
To avoid glossing completely: the 16s rRNA gene I have reference only codes for the small subunit of bacterial ribosomes, which acts like something of a scaffold for ribosome structure (along with the “large subunit”). This gene is common given that bacteria all possess ribosomes, which translate DNA into protein.Again, viruses have no such gene, which would allow such a complete census in a single shot. Taking a parallel but not identical approach, we have RAPD (Random Amplification of Polymorphic DNA) PCR. The only difference between RAPD PCR and ordinary PCR is that the amplification is random, rather than deliberately chosen. In PCR, one typcially selects DNA primers which induce the amplification of a specific gene, such as the 16s rRNA gene which I’ve referenced continuously. In RAPD PCR, a single primer (which can act as both forward and reverse) is mixed with the template DNA in the reaction tubes; the primer will attach and amplify, wherever possible.

This method might not reveal specific virus identity, but one can compare RAPD PCR results from samples taken in different places or times. Our lab has been sampling from various soil and water sites around Lake Mataoka for several months. We can compare PCR products amplified by the same primer between months or locations; the degree to which the bands resemble one another is a measure of how similar the viruses are between the two samples. The gels above are not ours, but from a study of enterobacteria common to humans and pigs Leal et al, 1999), but it seemed to demonstrate the point. The truth is, I have been working for a few weeks to make PCR work with soil samples at all; it seems that various factors present in soil (such as accumulated humic acids) make molecular work inherently difficult. In case you don’t believe me:

As you can imagine, all sorts of things build up in soil, making it a chemical mixture that is hellishly difficult to analyze. Even now, the chemical structure of humic acids has not been clearly defined. The structure above is such an approximation. Furthermore, all of our filtration methods do not exclude either these acids or tiny colloid particles in soil; I will report on my findings! So far, I have gotten a single successful amplification of soil samples, along with many successful amplifications of water samples.

Virus enumeration has been halted for some weeks now due to what was for some time an ambiguous source of contamination. After any given extraction method from soil or water, we prepare slides of the extract by filtering the extract through filters attached to a vacuum manifold. The point is to draw through the liquid part of the sample but retain the viruses; the pore size of these filters is on the order of 0.02 micrometers across. Below is an extreme close up of the Whatman “Anodiscs” we had been using for quite some time. They advertise extremely high pore density, which is made our work very efficient. Until complications arose!

Unfortunately, across all samples and extracts, we consistently saw large, ambiguously shaped noise on our slides. This was with a brand new box of filters, which had always worked fine, so we automatically assumed the source of error was somewhere on our end, whether our materials or human error. After trying again with completely fresh sterile water, DNA stain, anti-fade solution, extracts, pipets, and tubes, along with sterilizing the vacuum manifold (among other things), we eventually ruled out every possibility of contamination other than the filter discs themselves, straight from the company. This was verified by comparing them to other filters, which did not exhibit this strange noise. After contacting their tech support, it became clear we were not the only lab to experience this difficulty! They have offered free replacements, but unfortunately these will be pending for some time as Whatman needs to determine the source of contamination in their production line. I do not envy the engineers who need to find whatever tiny problem is leading to the contamination we have seen.Undaunted, we moved forward with virus extractions, deciding to preserve our samples until enumeration could be achieved. The most proven way to preserve viruses without destroying them (or preventing future molecular biology work) is freezing them in liquid nitrogen, or “snap freezing.” The point is to preserve the sample in amorphous ice; whereas ordinary ice, with its crystal lattice, is rather jagged and destructive on the molecular level (freeze/thaw is a method for cutting cells open), amorphous ice forms too quickly to arrange into a lattice, and is comparitively  gentle with biological samples.
Ordinary ice:

Amorphous ice:

The moral of the story: if you’re going to get cryogenically frozen, make sure they snap freeze you; ordinary ice will cause untold damage to your bodily tissues. Unless, in the future where you are going, they also have armies of nanomachines which could execute those repairs.For now, this freezing method is definitely our prescribed solution, given that it is the most effective method of preservation we know of. One might wonder why we haven’t tried other filter varieties; we have! We ordered some 0.01 micrometer filters from Sterlitech. When we attempted slide making, the sample didn’t seem to draw through, or if it was, at an agonizingly slow rate, perhaps requiring many hours of constant vacuum pressure. Whatman’s high pore density seemed to make quite a difference after all! For now, this lack of analysis somewhat inhibits my ability to draw conclusions about our various extraction methods. There is still the possibility of creating wet mounts directly from the extracts, which is not the best solution for reasons I will discuss in the future.

Virus ecology isn’t exactly a walk in the woods; my lab mates and I cannot go out and count all of the foxes or trees by eye, at least not without substantial assistive technology! Very few scientists have considered the population dynamics, say, of soil viruses, which in my view is key to understanding viruses on the whole. We should not neglect the part of the viral “life” cycle when viruses exist outside of hosts! Some very simple questions need to be answered; how much is out there, and what are we looking at?For assessing how many viruses exist in an environment, we count things, as in any ecology study, with some modification. Our principle method of measuring virus abundance is epifluoresence microscopy. Which, by the way, is just a method which allows us to literally count the viruses present in a sample. Just as one could hardly be expected to count every tree in a forest as part of a study, we cannot count every virus in a sample. Instead, we quantify a number of small subsets of a sample, and calculate relative abundance from there. I’ve posted here some pictures of fields taken with the epifluourescence microscope. The samples in question are dilutions of treated soil samples. The field includes numerous points of light; the small and symmetrical among these are most likely viruses, whereas the larger blotches may be any number of things which the DNA stain has attached to. For more about epifluorescence:  http://en.wikipedia.org/wiki/Epifluorescence_microscopy(In fact, their picture of an Olympus scope on the top right looks almost exactly like ours. )It was quite an experience seeing these fieldsfor the first time. I have to say, I was reminded of outer space.

There are ten times more viruses on this planet than officially living things, continuously co-evolving with cellular life. There is no such thing as an organism which cannot be infected by at least one virus, to our knowledge; even thermophilic archaea, which inhabit otherwise inhospitable hot springs. Even microbes under arctic ice or what life can exist adjacent to hydrothermal vents. Indeed, you and I, despite our fancy shmancy (and generally quite effective) immune systems, are no exception. In fact, multicellular life has a particularly interesting time of it: not only are there viruses which infect our human cells with a high degree of specificity, but there are also numerous viruses which can infect the bacteria that live inside of us. Almost any gastrointestinal problem you’ve ever had or ever will is because the bacterial communities in your digestive tract are thrown off balance. Invasive bacteria can do this, but so can bacteriophages, the viruses which infect bacteria. The study of infectious disease is to a large degree a highly focused ecology study; the human body is an ecosystem  in its own right, and one that must be regulated with care. (For the record, there are about five times as many bacteria in you as there are human cells. We would be foolish to ignore them!)Viruses are not alive, per say, but there’s quite a lot of debate about it. They are generally some form of genetic information, encased in protiens. In the image above, you can see the distinctive bacteriophage “heads” (protien capsids which contain the genetic information, usually in the form of double stranded DNA) and “tails,” which actually act like syringes for inserting DNA into host cells. Viruses have fantastic potential for genetic engineering! There already exists in nature the means to insert new genes into anything alive. Such therapies are generally in the devopemental stages, but viruses could be a great way to write over harmful genes, e.g., cancer inducing oncogenes, inhereted genetic disorders, and so on.One reason I am so interested in virus ecology is because I think it is always possible to find superior vectors for all of our purposes. For example, take any bacteria we don’t like. Tuberculosis! Which kills millions of people each year globally. Viruses exist that can infect any type of cell; the potential here is obvious. If we could isolate bacteriophages which infect tuberculosis, then we would have in our hands an excellent method of both detecting and potentially eliminating this crippling pathogen. A fine idea, but we need to find an appropriate phage.Hence, we go virus hunting! Which is a challenging practise, especially with regard to soil. Above is a sonicator probe, which has proven to be a useful tool in this process. So far, ultra high intensity sound has proven effective in extracting viruses from soil because it breaks apart soil and “desorbs” virus particles from that mixture. In a typical gram of soil from one’s backyard, you could easily expect there to be millions of viruses. The goal here is to make those viruses accesible for study. From there, specific, useful strains can be isolated; it would be exceedingly easy to discover entirely new viruses, simply because this field has recieved such little attention. I will explain more in due time; wish us luck.