Friederike Roloff and Dr. Ron W. Leavitt, Microbiology
Introduction
Bordetella avium causes a respiratory disease in poultry which especially young turkeys are susceptible to. We are working with the Moroni Feed Company, who supplied us with 12 strains of Bordetella avium. Eleven of the strains are from outbreaks of the disease in the Southern Utah turkey farms. The twelfth strain is the vaccine strain V87. This vaccine was produced by chemically mutating B. avium and selecting a non-virulent temperature sensitive strain. Dr. Jensen, a former professor of Microbiology, BYU, produced this vaccine by chemically mutating B. avium and selecting a non-virulent temperature sensitive strain.
Our goal was to compare the outbreak strains to the vaccine strain by DNA fingerprinting. One of our original hypotheses was that the vaccine strain back mutates and was now causing outbreaks. Further background research indicated though that this is not likely since most outbreaks occurred in non-vaccinated flocks. Only one outbreak in 1999 occurred in a properly vaccinated flock.
Our focus changed to comparing the outbreak strains with each other. A great similarity between them would suggest a common source of B. avium is causing the outbreaks.
Methods
We applied Randomly Amplified Polymorphic DNA (RAPD) analysis to fingerprint all twelve strains. RAPD analysis is a method widely used for quick and inexpensive microbial fingerprinting. It is a polymerase chain reaction (PCR) with arbitrary primers.
In our case we used a set of six different primers, all of them being ten-mers. Gel analysis of the PCR products revealed that eleven of the strains including the vaccine strain are similar. One strain, GLB, showed a different pattern. Due to the fact that we did not get enough reproducible bands for each strain, we decided to apply Amplified Fragment Length Polymorphism (AFLP) analysis.
AFLP is a relatively new method. The DNA is cut by two restriction enzymes and adapters are ligated to the cut sites. Then a PCR is done using primers complementary to the adapters. A dilution of the PCR product will be used in a further PCR, using primers that extend 0 to 2 bases beyond the adapter into the fragment to be amplified. This is done in order to reduce the number of amplified pieces to obtain about 50-100 bands for analysis. A sequencing gel gives the desired resolution.
Challenges
When I first started working on this project we were under the assumption that all outbreaks had occurred in vaccinated flocks. Therefore, our original hypothesis was that the vaccine might be causing them.
Doing more background research by talking to the Moroni Feed Company veterinarian, I found out that all except one strain are from non-vaccinated flocks. This makes our original hypothesis unlikely. We now suspect an outside source such as rodents to be the carrier of B. avium.
One challenge was the irreproducibility of our RAPD data. We were able to compare the strains with each other for every PCR, but when we repeated the PCR with the same strains and primers we obtained different patterns. This might be due to the sensibility of the RAPD to the reaction conditions. Also, B. avium has a GC content of about 67 %, and the PCR conditions we used may not be enough to melt the DNA completely.
For our AFLP analysis we are using the AFLP Microbial Fingerprinting Kit from Perkin- Elmer. Even though we followed their protocol, we have not been able to obtain usable results. Since we did not get any amplification, the problem could be the ligation of the adaptors. Another possibility is that the AFLP kit we have has a production defect. We are setting up different experiments to find out how to make it work.
Future Work
Our goal is to get the AFLP to work and to obtain results that will tell us how related the strains are. Then, we will try to identify the source or the carrier of B. avium in the Southern Utah turkey farms. Eventually we will produce a new vaccine since there was the 1999 outbreak that occurred in a vaccinated flock which indicates that there is at least one strain that is resistant to the vaccine.
We also obtained 30 strains of B. avium from Germany. We will start a comparative study as soon as we get the AFLP to work.
Accomplishments
I presented my RAPD data on a poster at the 49th annual Western Poultry Disease Conference in Sacramento, CA in March 2000.