Research Interests

We are currently pursuing several research directions:

• Isolating novel viruses from environmental samples, including soils and freshwater

• Characterization of soil viral communities (metagenomes)

• Viral persistence, decay, and transport in soils

• Lytic vs. temperate phages and evolutionary success in soil environments

• Assessing the role(s) of viruses within mixed microbial biofilms

Lab Publications 

For a complete listing of publications, please see my CV.  This list focuses on projects that were completed exclusively by my lab group; underline = undergraduate co-author:

2023:

Jacoby, M.L., Hogg, G.D., Assaad, M.R., and Williamson, K.E. (2023) Seasonal trends in lysogeny in an Appalachian oak-hickory forest soil. Applied and Environmental Microbiology.  doi: 10.1128/aem.01408-23  

diPietro, A.G., Bryant, S.A., Zanger, M.A., and Williamson, K.E. (2023) Understanding Viral Impacts in Soil Microbial Ecology Through the Persistence and Decay of Infectious Bacteriophages. Current Microbiology. 80: 276.  https://doi.org/10.1007/s00284-023-03386-x 

2022:

Dar, N., Thompson, C.P., and Williamson, K. (2022). Marker gene analysis reveals novel viral genetic diversity in unsaturated soils. Biology & Fertility of Soils. Published online Dec. 8, 2022. https://doi.org/10.1007/s00374-022-01687-0 

2020:

Payne, A.T., Davidson, A.J., Kan, J., Peipoch, M., Bier, R. and Williamson, K. (2020).  Widespread cryptic viral infections in lotic biofilms.  Biofilm. https://doi.org/10.1016/j.bioflm.2019.100016

2018:

Green, J.C., Rahman, F., Saxton, M.A., Williamson, K.E. (2018). Quantifying aquatic viral community change associated with stormwater runoff in a wet retention pond using metagenomic time series data. Aquatic Microbial Ecology 81(1): 19-35.  Abstract

2017:

Williamson, K.E., Fuhrmann, J.J., Wommack, K.E., and Radosevich, M. (2017).  Viruses in Soil Ecosystems: An Unknown Quantity Within an Unexplored Territory. Annual Review of Virology 4: 201–19.  Abstract

2016:

Saxton, M.A., Naqvi, N.S., Rahman, F., Thompson, C.P., Chambers, R.M., Kaste, J.M., and Williamson, K.E. (2016) Site-specific environmental factors control bacterial and viral diversity in stormwater retention ponds. Aquatic Microbial Ecology 77(1):23-36  Abstract

 2015:  

Green, J.C., Rahman, F., Saxton, M.A., and Williamson, K.E.  Metagenomic assessment of viral diversity in Lake Matoaka, a temperate, eutrophic freshwater lake in southeastern Virginia, USA.  Aquatic Microbial Ecology 75: 112-128.  Abstract

2014:

Williamson, K.E., Harris, J.V., Green, J.C., Rahman, F. and Chambers R.M. (2014) Stormwater runoff drives viral community composition changes in inland freshwaters.  Frontiers in Microbiology.  14 March 2014 | doi: 10.3389/fmicb.2014.00105 Open Access Article

Helsley, K.R., Brown, T.M., Furlong, K. and Williamson, K.E. (2014) Applications and limitations of tea extract as a virucidal agent to assess the role of phage predation in soils. Biology & Fertility of Soils 50 (2): 263-274.  Abstract

2013: 

Williamson, K.E., Corzo, K.A., Drissi, C.L., Buckingham, J.M., Thompson, C.P., Helton, R.R. (2013) Estimates of viral abundance in soils are strongly influenced by extraction and enumeration methods.  Biology & Fertility of Soils 49(7): 857-869.  Abstract

 2012: 

Hardbower, D.M., Dolman, J.L., Glasner, D.R., Kendra, J.A., and Williamson, K.E. (2012) Optimization of viral profiling approaches reveals strong links between viral and bacterial communities in a eutrophic freshwater lake.  Aquatic Microbial Ecology 67: 59-76  Abstract. 

 Research Techniques

My lab group uses a number of nifty tools to address questions about viral ecology.  Here are a few:

Culture-based approaches:

     Plaque Assays. The plaque assay is one of the oldest techniques available for detecting, isolating, and enumerating infectious virus particles, developed by Felix d'Herelle in 1917.  While originally developed for bacteriophages (viruses that infect bacteria), many virus-host systems have been adapted for plaque assays (e.g., insect cells).  To conduct a plaque assay, we need a sample that contains (or putatively contains) viruses, a line of host cells that the virus is able to infect, solid media plates, and a molten, low-percentage agar medium known as "top agar."  Generally, we dilute the putative virus sample to generate a continuum of high to low concentration.  Each virus dilution is mixed with a set volume of host cell culture, the top agar is mixed in, and then the mixture is poured over an agar plate and allowed to harden.  The plates are incubated under environmental conditions that are conducive to host cell growth.  After the host cells grow (generally 18 - 72 h), intact cells are usually visible growing on the surface of the plates in what is called a "lawn."  If infectious virus particles were present in any of the virus dilutions, discrete zones of clearing or "plaques" will be evident in this lawn of host cells.  

For interested parties, here are additional resources on how fluorescence works and how the fluorescence microscope works.

     Transmission Electron Microscopy.  The first commercial transmission electron microscope (TEM) was developed by Siemens in 1939. Also in 1939, the tobacco mosaic virus became the first virus particle to be visualized using TEM.  Since then, TEM has been an essential tool in virology for enumeration of virus particles, and perhaps most importantly, revealing virion structure.  TEM is the standard technique for morphological classification of virus isolates and has been used to assess morphological diversity of environmental virus assemblages.  Viruses are generally invisible under light microscopy (unless certain tricks are applied, such as EFM, described above) because most virus particles are smaller than the wavelengths of visible light that we can see.  So, virus particles do not deflect enough light into our retinas to form a clear picture.  But the wavelength of an electron is much smaller than the smallest wavelength of visible light - so electrons will enable us to form a picture of an individual virus particle ... if only we could see electrons.  We can't see electrons but we can see photons.

     To solve this problem, the electron microscope features a special plate that is coated with phosphorescent material.  When electrons strike this plate, the electrons excite the phosphorescent coating and it emits photons.  To our eyes, the plate appears to glow green.  Now, if you put your sample in between the electron source and this phosphorescent plate, the materials in your sample will block some of the electrons from being transmitted to the plate and dark shadows will appear on the plate, where no photons are being generated.  This is the basic operating principle of TEM: you are looking at the electron shadow cast by your sample.   For more details on TEM, click here.

     To obtain TEM data, we apply a drop of our virus sample (which could be an isolate or an environmental sample) to a standard TEM grid.  Samples may also be centrifuged onto grids placed at the bottoms of ultracentrifuge tubes.  Once virus particles are on the grid, they are stained with a heavy metal salt, such as uranyl acetate.  The stain is then wicked away, leaving behind trace salt residue around the virus particles.  This gives sharp contrast under the electron microscope, since the beam of electrons will not be transmitted through the electron-dense salt residue, showing up as those dark shadows mentioned above.  Once we have electron micrographs of the virus particles in our samples, we can measure dimensions of various structures (capsids, tails, filaments) and compare particle morphology.  

Molecular approaches:

     RAPD-PCR.  Randomly Amplified Polymorphic DNA - Polymerase Chain Reaction (RAPD-PCR) is a type of amplified fragment length polymorphism (AFLP) analysis.  We use this approach is used to compare viral community composition over time and space.  The primer we use in RAPD-PCR does not target any specific viral taxonomic group or marker gene, but allows us to reliably and reproducibly compare viral assemblages without prior sequence information.  Whenever this primer can serve as both forward and reverse primer on the same template, we will obtain an amplified fragment.  While the amplified targets are arbitrary, they are not actually "random" as the name of this technique implies.  Any template sequence that produces an amplified fragment of defined size will consistently produce that size fragment.  Thus, specific fragment sizes are diagnostic for comparing viral operational taxonomic units (or OTU) across samples.  This phase, "operational taxonomic unit" (OTU), is a human convention that acknowledges the shortcomings of the technique.  For example, we assume that identically sized fragments found in different samples are indicative of the same viral taxon - but in reality, this may not always be the case.

     At any rate, the resulting fragments amplified by RAPD-PCR are then separated under gel electrophoresis using a high-resolution agarose (below, left).  The visualized bands are then sorted into specific size classes and band sets are compared across samples.  Similarity scores are generated based on the number of shared band sizes between sample pairs, enabling statistical comparisons, as well as graphical analyses like cluster dendrograms (below, right).  These analyses allow us to draw conclusions about how viral communities change over time and space, and in response to specific environmental drivers.  

RAPD-PCR of viral concentrates generated from Lake Matoaka water samples in 2012.  Left: gel electrophoresis of RAPD_PCR products.  Bands were sized and pairwise comparisons of band presence/absence were conducted.  Right: cluster dendrogram showing results of pairwise comparisons.  Replicate samples are highly similar (> 80% shared OTU).  Monthly changes in viral community composition are evident, and can be correlated to environmental factors using statistical analyses.

     Metagenomics.  By sampling the collective genomes of (theoretically) all viruses within a given sample, viral metagenomics has allowed researchers to survey the genetic diversity within entire environmental viral assemblages. The first viral metagenome was published by Mya Breitbart et al. in 2002 and described previously unknown diversity in marine viral communities.  In the subsequent ~15 years, more than 100 viral metagenomes have been described, the vast majority of which have been from marine environments.  Freshwaters, and in particular, temperate soils remain underrepresented in the literature.  

     With metagenomics, we start with an environmental sample (for example, pond water, soil, or sediments) and extract and purify the virus particles away from everything else in the sample.  Then we crack open the virus capsids and purify the viral nucleic acids.  We take all of those released virus genomes and break them up into small fragments, generally somewhere between 250 and 600 base pairs long, depending on which sequencing platform is being used.  The collection of genome fragments, or metagenome, is then sequenced, typically resulting in hundreds of thousands of individual nucleotide sequences or reads.      

     Reads are then analyzed in a number of ways.  We can use BLAST-based algorithms to compare our reads against databases containing sequences of known taxonomy and function, allowing us to determine what viruses might be present in our samples and what they might be doing.  For example, MG-RAST allows users to compare reads against multiple databases to extract the closest taxonomic and functional homologues, while MetaVir enables more targeted taxonomic searches of known virus sequences. 

     A number of database-independent analyses have been developed to try to get around this limitation.  Phage Communities from Contig Spectrum (PHACCS) is one approach that predicts viral community diversity statistics (e.g., richness, evenness, dominance) without relying upon database matches.  Instead, PHACCS tries to put together, or assemble, individual reads into larger contiguous stretches, or contigs.  Based on the contig spectrum, or array of different length contigs that result from assembly, PHACCS can determine rank-abundance models that best fit a particular viral metagenome.

     MaxiPhi and crAss are other database-independent tools that make use of cross-contigs to estimate similarity between two different viral metagenomes.  In this approach, we attempt to assemble reads from two different metagenomes.  Based on the number and length of the resulting contigs that are formed only from reads coming from both metagenomes (cross-contigs), we can extract similarity/dissimilarity scores.  Essentially, we can determine how similar two viral communities are to each other, even if we don't know 100% of which specific viruses are in each community.

     BLASTing entire viral metagenomes against each other is also a powerful way of comparing multiple viral communities to each other without depending upon database matching to known sequences.  Unlike cross-contig analysis, which attempts to assemble reads across different metagenomes, BLAST-based cross-comparisons uses the BLAST algorithm to determine if reads from one metagenome can be matched to homologues in another metagenome.  A higher number of closer matches will result in closer clustering of metagenomes, while lower numbers of matches, or more distant matches will result in more distant clustering of samples.  

Because of the intrinsic strengths and weaknesses of these different analytical approaches, most viral metagenomics projects incorporate both database-dependent and database-independent approaches to develop the clearest and most accurate picture of the viral communities in their samples.