Introduction

The quote "nothing lasts forever" relates to many aspects of our lives. It especially applies to the problem of antibiotic resistance found in bacteria. Concern is mounting about the overuse of antibiotics and the numerous bacterial strains that have acquired the ability to resist the effects of some antibiotics. The use of antibiotics favors the natural selection of strains of bacteria that are resistant to that particular antibiotic.

The progressing problem of antibiotic resistance prompted this research. In order for bacteria to survive a chemotherapeutic agent such as an antibiotic, it must first code for its resistance by accepting R (resistance) plasmids. Plasmids are defined as extrachromosomal circular genetic elements that are capable of self-replication and of transferring genes from one bacterium to another. Plasmid loss is described as an event where plasmids are "lost" from cells dividing during replication. If this process continues, plasmids can eventually become non-existent in a population. The most significant feature of the plasmid is its ability to transfer genetic information to and from different bacterium.

This experiment is concerned mostly with the R (resistance) plasmids that code for antibiotic resistance. R plasmids can be passed not only from one strain to another but also from one bacterial species to another. Antibiotic- resistant strains of bacteria have become a serious health problem because R plasmids can occur in pathogenic bacteria, and the treatment of human bacterial diseases is complicated by the occurrence of those pathogens that are resistant to multiple antibiotics. In this experiment, the presence of R plasmids in bacteria conferred antibiotic resistance. The presence of R plasmids was detectable because it contained the Green Fluorescent Protein (GFP) gene which fluoresced green under black light conditions. Plasmids are lost when a bacterial cell divides to replicate. This raises the question: would antibiotic resistant bacteria lose a significantly greater amount of plasmids in an antibiotic-free environment that in one containing antibiotics? Theoretically, if a strain of bacteria resistant to the antibiotic ampicillin were exposed to ampicillin, it would survive longer and its population would maintain a relatively high percentage of the ampicillin resistance gene, since it is the only strain that can survive to replicate. But, if a strain of bacteria resistant to ampicillin were exposed to no antibiotic, it would still survive although it would retain a lower percentage of the resistance gene over time because of plasmid loss. The results of this experiment were anticipated to be applied to the problem of antibiotic resistance in an attempt to aide the medical community in decreasing the prevalence of antibiotic resistant bacteria. The hypothesis of this experiment is as follows: Compared to ampicillin resistant Escherichia coli (E. coli) bacteria grown in the presence of ampicillin, those not grown in the presence of ampicillin will lose a larger percentage of their resistance genes through plasmid loss.


Experiment

Construction of Strain

The first step to introducing Green Flourescent Protein (GFP) and ampicillin resistance genes into Escherichia coli (E. coli) was to make the bacteria competent to accept the plasmid. First, 10 ml of Luria Broth (LB) media was aseptically added to a 15 ml Falcon tube. Using a micropipette, an aliquot of an overnight culture of E. coli Bacteria was transferred to the 10 ml culture and placed in a 37 degree C incubator for 2 hours. The A600 value for 1 ml aliquot of the culture was determined and the culture was ready for use if it registered an ideal value between 0.2-0.6, indicating that the bacteria were in the log-phase of their growth. The remaining 9 ml culture was left on ice for 5 minutes. The tube was then centrifuged in a cold room at 3000 RPM for 5 minutes. The supernatant was decanted and the cells were resuspended in 4 ml of ice-cold 0.1 M CaCl2 for 20 minutes. The tube was spun again at 3000 RPM in a cold room for 5 minutes. The supernatant was before the cells were resuspended in 100 microliters of ice-cold 0.1 M CaCl2 and keep the cells on ice. The cells where then competent and able to accept the plasmid.

The next step was to transform the E. coli bacteria. Fifty (50) microliters of competent cells were transferred to an ice-cold microcentrifuge tube. Then, 5 microliters of plasmid containing GFP genes and ampicillin resistant genes were transferred to the cells on ice and mixed briefly with a micropipette. The cells were kept on ice for 45 minutes. Next, 800 microliters of liquid LB media was aseptically added to a microcentrifuge tube. The microfuge tube was placed into an open 15 ml Falcon tube and placed into a "shaker" in a 37 degree C incubator for 45 minutes. The tube was then centrifuged for 10-15 seconds and the supernatant was decanted. The cells were resuspended and transferred to the center of an LB agar plate using a sterile rod to evenly spread the bacteria on the plate. The bacteria were then incubated in 37 degree C incubator overnight.

Measuring Plasmid Loss

Two (2) ml of 100 mg/ml ampicillin was added to 2 ml of LB media. The media was then inoculated with a recombinant bacterial colony expressing GFP and the ampicillin resistance gene and grown overnight at 37 degree C. (See figure 1) Next, 2 ml of LB media was added to one tube (LB tube) and 2 ml of LB media was added to another tube with ampicillin (LB/amp tube). Both tubes were inoculated with 20 microliters of recombinant E. coli bacteria that were grown overnight. Twenty-four (24) hours later, cells from the LB and the LB/amp tubes were plated onto LB and LB/amp agar plates. Before the cells can be plated, they were diluted by adding 10 microliters of cells to 0.99 ml of distilled water and mixed. This dilution process was repeated 5 times and then 10 microliters of the diluted cells were transferred to the appropriate LB or LB/amp plates and spread. These plates were grown overnight at 37 degree C. After each plating, two new LB and LB/amp tubes were set up and 20 microliters of the overnight culture from the LB and LB/amp tubes were transferred to this fresh media for growth. The cells were grown in the tubes overnight at 37 degree C. This process was repeated afer 48, 72 and 96 hour intervals. The cells were grown in the tubes overnight at 37 degree C. There were samples available from 24, 48, 72 and 96 hours of growth and the colonies on each plate were counted and the data was recorded. Colonies were examined under a black light and those that fluoresced green and therefore had the ampicillin resistance gene were counted. From this data, the percentage of colonies that kept the plasmid was determined.

Electrophoretic Analysis

To insure cells were properly scored, plasmid DNA was isolated from a 72 hour LB and LB/amp sample and analyzed by electrophoresis. The results of this procedure indicated that the plasmids were definitely lost to the bacteria on non-ampicillin plates.

Plasmid DNA was isolated from the recombinant E. coli bacterial colonies to acquire the samples for electrophoresis. A bacterial colony, scored as loss- of-plasmid, was transferred with a sterile inoculating loop to one 15 ml Falcon tube with 5 ml of LB media and another, scored as with plasmid, transferred to a 15 ml Falcon tube with 5 ml of LB media and 5 microliters of 100 mg/ml ampicillin. The tubes were incubated overnight at 37 degree C. The tubes were then centrifuged at room temperature at 3000 RPM for 5 minutes and the supernatant was decanted and the remaining supernatant was removed with a micropipette. Two-hundred and fifty (250) microliters of cold P1 buffer (QIAGEN) was added to the tubes. The cells were resuspended with a micropipette and 1.5 ml of the cells were transferred to a microcentrifuge tube. Then, 250 microliters of P2 buffer was added and the tube was gently inverted 3-4 times and allowed to sit at room temperature for 15 seconds. Three-hundred and fifty (350) microliters of chilled N3 buffer was added and the tubes were immediately, but gently, inverted 3-4 times. The tubes were then microcentrifuged at room temperature for 10 minutes. While the tubes were spinning, a QIA prep spin-column was placed in a 1.5 ml centrifuge tube. The supernatant was decanted from the tubes and transferred to the 1.5 ml spin-column and centrifuged for 1 minute at room temperature. The spin-column was removed and the flow-through fraction was decanted from the 1.5 ml tube and the spin-column was replaced. Five-hundred (500) microliters of PB buffer was added and the tubes were centrifuged for 1 minute. Seven-hundred and fifty (750) microliters of PE buffer was added and the tubes were then centrifuged for 1 more minute. As before, the spin-column was removed and the flow-through fraction was decanted from the 1.5 ml tube and the spin-column was replaced. The empty spin-column was then centrifuged for an additional minute to remove any residual wash buffer. The spin-column was then placed in a clean 1.5 ml centrifuge tube and 60 microliters of TE buffer (10mM tris/HCI buffer, pH 8.0, 1 mM EDTA) was added to the tube and allowed to sit at room temperature for 1-2 minutes before being centrifuged for 30 seconds. The spin-column was disposed of and the purified plasmid remained in the 1.5 ml tubes. The tubes were then marked properly and placed in a cold room for future use in the electrophoresis analysis.

To make the agarose gel, 1.0 g of agarose was dissolved in 50 ml of Tris- Acetate buffer (40 mM tris, 20 mM Na acetate, 2mM EDTA pH 8.3) by bringing it to a boil in an autoclave. Next, when the agarose was completely dissolved, 49 ml of room temperature Tris-Acetate buffer was added and mixed thoroughly. One (1) microliter of 50 mg/l stock solution of ethidium was carefully added to the agarose solution. To form the mini-gel, approximately 20-30 ml of the agarose solution was poured into the gel mold and was left to harden by cooling it to room temperature. A comb was placed at the end of the gel and removed when hardened to form the wells into which the isolated plasmids were placed. The isolated plasmid DNA were then transferred to separate wells. The lid was placed on the electrophoresis chamber so that the negative electrode was on the opposite end of the wells and the power source was turned on and the voltage set at 80 volts for 1 hour. The gel was removed from the apparatus and the image captured in a photograph.


Discussion

This experiment successfully demonstrated that compared to E. coli bacteria that are grown in the presence of ampicillin, those not grown in the presence of ampicillin lost a larger percentage of its ampicillin resistance genes through plasmid loss.

In all four trials, virtually 100% of the recombinant E. coli bacteria that were grown in the presence of ampicillin retained the plasmid over a 96 hour period. (See figure 2) These results were expected because the ampicillin acted as a selective agent. The only strain of the recombined E. coli bacteria that could survive after replication was the strain that retained the R plasmid. Therefore, the preliminary count of colonies present virtually equaled the number of scored colonies retaining the plasmids. This resulted in a near 100% retention of the plasmid in each 24 hour sample over a 96 hour period.

In all four trials, a large percentage of the recombinant E. coli bacteria that were not grown in the presence of ampicillin lost its plasmids over a 96 hour period. (See figure 2) The average percentage of colonies in the non- ampicillin environment that retained the plasmid decreased from 70.50% after 24 hours to 12.53% after 96 hours. (See figure 2) These data are evidence of substantial plasmid loss in the non-ampicillin populations.

Another trend in the data that was evident is the overall number of colonies that grew on the ampicillin plates and non-ampicillin plates. (See figure 3) Compared to the plates with ampicillin, more colonies grew on the plates with only LB media because any strain could grow on those plates without being terminated by ampicillin. Consequently, less bacteria grew on the plates with ampicillin because only the bacteria that retained the plasmid could survive. If a colony on an ampicillin plate lost plasmids during replication, those offspring lacking the plasmid would be terminated.

To prove the cells were properly scored, plasmid DNA was isolated from a 72 hour LB and LB/amp sample and analyzed by electrophoresis (See figure 4). The electrophoretic results demonstrated that the colony grown on a non-ampicillin plate after 72 hours retained barely any plasmids compared to the colony grown on an ampicillin plate after which, after 72 hours, displayed a high retention of the plasmid.

Plasmid loss was detected in the bacterial colonies by determining the number of colonies that fluoresced green, indicating that the colonies still had the plasmid that contained the Green Flourescent Protein (GFP) and ampicillin resistance genes. If the colonies did not fluoresce green, plasmid loss had occurred. GFP is a protein that, when fused to other proteins, is an effective labeling and tracking device used frequently in microscopy and pathology. In nature, GFP is made by the jellyfish Aequorea victoria, which is widely found in the northwest region of the Pacific Ocean. In response to shaking or attack, the jellyfish emits a bright green flash resulting from change in calcium-ion levels that causes the GFP to fluoresce, presumably blinding the attacker. GFP is a very valuable labeling device because it only needs to be bound to certain proteins to fluoresce; therefore, no outside aid is necessary. Because of its abilities, GFP was used to determine the plasmid loss in E. coli bacteria and thus obtain the data and results in this experiment.

There were some minor variations evident in the data but the overall trend in each trial was similar. Nearly 100% of the colonies grown in an ampicillin environment retained the plasmid. The minor variations occurred in the non- ampicillin environment in which there was always a decreasing percentage of colonies that retained the plasmid. But, each 24 hour sample had a different percentage of colonies that kept the plasmid. This variation was expected because of the possible different dilutions of the cells and/or slight contamination.

There are also some possible sources for errors in this experiment. Some samples could have been diluted more than others and thus have a higher or lower number of colonies that could have contained the plasmid. The plates could have been infected and bacteria other than E. coli could have grown on the plates, distorting the percentage of colonies that retained the plasmids. Also, a colony could retain anywhere from 5-20 copies of a plasmid. Colonies with less plasmids did not fluoresce as much as a colonies with many plasmids and therefore might not have been scored as retaining the plasmid. An uncontrolled event that might have affected the results was the different degree of crowding seen on the plates. It was difficult to accurately count the number of bacteria on a plate if there were thousands of colonies present. Several colonies could have combined to look as one and the total number of colonies could have been distorted. A large population on a plate was most likely due to insufficient dilution.

A few changes would be suggested to this experiment if it were to be replicated in the future. First, the number of colonies grown on each plate would be kept to a minimum to aide in the counting process. Second, the time intervals for the samples and the total time for the experiment would increase to allow a more precise analysis of the rate that plasmids are lost in a population. An increase in the overall time of the experiment would also analyze the bacteria when they are out of their log-phase and their growth is slower. This bacterium was only analyzed over a 96 hour period when the bacteria was in its most rapid growth phase, resulting in a higher rate of plasmid loss because of the more frequent cell divisions. But if the bacteria is analyzed for a longer amount of time after it has passed its log-phase, a more accurate determination of the overall rate of plasmid loss could result. Finally, additional trials would be run to gather more comprehensive data. Aside from replicating this experiment, there is another experiment that can be conducted to further analyze the rate of plasmid loss in a living organism. Two groups of mice would be infected with non-virulent bacteria containing a resistant gene to penicillin. Then, one group would be injected with penicillin and the other group with tetracycline. After a period of time, bacteria would be extracted from both groups of mice and an analysis could be performed to determine whether there was a higher percentage of the antibiotic resistant strain of bacteria in the mice injected with tetracycline than the mice subjected to penicillin, thus suggesting a higher rate of plasmid loss without a selective agent. This future experiment could be conducted in a manner similar to this experiment and it might further the theory that its predecessor established. This future experiment would surpass a limitation to this study in that the bacteria would be analyzed in vivo and not just in vitro.

On its own, this experiment could lead to a possible solution to the problem of increasing antibiotic resistance found in bacteria. Because the results of this project indicate that plasmid loss is very frequent when there is no selective agent, it is possible that resistant bacteria could lose their resistance genes if they are not given a specific antibiotic for a certain time period. It could be suggested that the medical community make a concerted effort to lower bacterial resistance to an antibiotic by restricting the use of that antibiotic in the medical field for a certain time period. This is a possibility for bacteria with two or more antibiotic cures because the antibiotic still in use could kill the bacteria while the remaining antibiotics are held in abeyance for the specified period of time. If a specific time for 100% plasmid loss is discovered for specific antibiotics, then a rotation of permissible antibiotics could be arranged and adhered to by the medical community. This would allow for the same effective antibiotic cures while decreasing the prevalence of antibiotic resistance in bacteria.


Conclusion

This experiment proves the hypothesis that compared to E. coli bacteria grown in the presence of ampicillin, those not grown in the presence of ampicillin will lose a larger percentage of its ampicillin resistance genes through plasmid loss. The recombinant E. coli bacteria that grew on the LB/amp agar plates completely retained the plasmid over a 96 hour period. But the recombinant E. coli bacteria that grew on the LB agar plates retained less and less plasmids as time progressed. More colonies grew in a non-ampicillin environment than an ampicillin environment because there was no selective agent. These results, tested for accuracy with the electrophoretic process, can be analyzed to devise methods to decrease the incidence of antibiotic resistant bacteria in our environment.


Acknowledgments

I would like to extend my gratitude to Dr. Michael Bumbulis of Baldwin-Wallace Collage for aiding me in this experiment. I would also like to thank my advisor Mr. Glen Novotony


References

Atlas, Ronald M.. Microorganisms in Our World. St. Louis: Mosby-Year Book, Inc., 1995. Boxer, Steven G.. "Another Green Revolution." Nature. October 10 1996, Vol. 383, pp. 484-5. Brock, Thomas and Madigan, Michael. Biology of Microorganisms. Englewood Cliffs, New Jersey: Prentice Hall, 1991. Levy, Stuart B.. "The Challenge of Antibiotic Resistance." Scientific American. March 1998. Online. Available FTP: http://www.sciam.com. Namecek, Sasha. "Beating Bacteria." Scientific American. February 1997. Online. Available FTP: http://www.sciam.com


About the author

Awards won by thie project