Benjamin Nelson Kartchner and Dr. Richard A Robison, Microbiology
Burkholderia mallei the etiological agent of the disease glanders, has the capability to cause disease in both animals and humans. Infection is usually caused through inhalation of organisms present in aerosols which results in serious pulmonary and systemic conditions that results in 40% mortality even with vigorous antimicrobial intervention and rises above 90% without such therapy(2). In addition, because of their dangerous nature both organisms have emerged as possible agents which may be used in the production of biological weapons. Recent research conducted in Dr. Richard Robison’s lab has analyzed genomic relatedness in various strains of B. mallei and the non-virulent species B. thailandensis using amplified fragment length polymorphism. Not only did these data create a phylogenetic tree for the organisms, but also revealed segments of DNA that are present in some species, but absent in others. My project centered on isolating these bands, sequencing them and then comparing their sequences to known protein sequences using a BLAST search (1,2,3).
After localizing the bands that were specific only to the virulent species of Burkholderia, it was necessary to isolate and purify these bands for sequencing. Initially, I theorized that it would be necessary to clone the band into a plasmid vector, multiply it and then extract it again from the vector and plasmid. However, I realized that because the band had the correct end bases needed to mark it in sequencing–generated during AFLP work, I could cut the band from a gel and use primers specific for these correct end bases, coupled with PCR to multiply the bands. Most purification protocols suggest the use of a plasmid for purification; I therefore include the protocol used for my study, as an equal substitute in this process, for those who may wish to pursue this in the future. I apologize for its technical nature.
Original AFLP+0 DNA was amplified using 15μL PE Core mix, 4 μL sample, .25μL of the +0 primers Apa and Taq and then brought to a total volume of 20.0 μL with sterile water. The samples were run in the thermocycler cycle used for AFLP work (4), and run on a 3% agarose gel. Bands in question were then cut from the gel and separated from the agarose using a chloroform extraction. DNA was then amplified a second time using the following volumes: 22.5μL supermix, 2μL primer (Apa and Taq) and .5μL sample. Total volume was brought to at least 50 μL using sterile water and run in a thermocycler using the AFLP cycle (4). Bands were then visualized on a 3% agarose gel. If they appeared pure enough, they were sent directly to the sequencing center for analysis using Apa and Taq primers already attached. To further purify the bands, if necessary the process was repeated.
Sequence data revealed from the two bands revealed a 250 base pair sequence and a 450 base pair sequence. A BLAST search was performed using two databases that would match the sequences from the bands with known sequences from other organisms.
Blast search results using one database revealed over 1000 possible matches ranging from human chromosomes to types of fungus; however no significant results were obtained. Results from a second search using the TIGRE database which contains sequences from the unfinished genome of B. mallei gave 100% matches for both of my segments. The 252 base pair sequence matched with a known sequence that lies on chromosome 1. Likewise the 450 base pair band matched with a known sequence on chromosome 2.
Although the results of my work were not directly useful in finding virulence proteins, methods established and data obtained would be useful in further studies. Initially it was thought that band purification using a plasmid vector would be necessary. However, I found that the second and much shorter method using PCR was just as efficient and solved many of the mini prep difficulties I encountered. I would recommend that this become the standard isolating procedure for AFLP bands, given that primers are already attached. It provides a fast, easy way to analyze many unique bands encountered in AFLP work.
Most usable proteins are coded for by genes that are at minimum 2-3 kbps in length. It is therefore quite understandable that we would not get a BLAST match for a sequence that is merely 250-450 bps in length. However, the data obtained here could be every useful in future experiments. Because we have verified the location of these segments in the genome, extended sequences could be obtained through TIGRE and by looking for natural start and stop codons, genes could be located. In addition, because these sequences have primers attached, chromosome walking methods could be used to search and sequence right and left from this point, again looking for natural start and stop codons. Further studies using this data would be needed to fully realize the usefulness of this genome analysis and I recommend that further steps be taken (as discussed above) to locate the complete genes that these sequences are a part of.
References
- Brett, Paul J., DeShazer, David, Woods, Donald E. (1998). Burkholderia thailandensis sp. Nov., a Burkholderia pseudomallei-like species. International Journal of Systematic Bacteriology. 48: 317-320
- David A.B. Dance. (2000). Melioidosis as an Emerging Global Problem. Acta Tropica. 74: 115-119
- Woo PC, Leung PK, Wong SS. 2001. groEL Encodes a Highly Antigenic Protein In Burkholderia pseudomallei. Clinical Diagnostic Laboratory Immunology. 8(4): 832-6
- Huber, Doug, Robison R. A., Pitt T.L. Genetic Typing of Burkholderia Isolates Using Amplified Fragment Length Polymorphism. Presented at the 37th Annual Meeting of the American Society of Microbiology