In vitro AZT Induced Mutations in FIV

In vitro AZT Induced Mutations in Feline Immunodeficiency Virus and the Effects on Growth Kinetics

Randall Forsberg, Shawna Martin

Report by Randall Forsberg

Introduction:

Feline Immunodeficiency Virus (FIV) is a Lentivirus that is the cause of an immunosuppressive disease in cats, homologous to Acquired Immune Deficiency Syndrome (AIDS) in humans (14). Both share a common susceptibility to many antiviral drugs (19,22). FIV, like HIV, has the capacity to develop resistance to drug therapies initiated against it (22). Mutations arising in response to chemotherapy occur in the pol regions of both viral genomes (21,24,25). Therefore, this makes FIV a valuable model for the study of drug and drug resistance related problems encountered in HIV (20,23). Research performed on FlV may have potential applications to HIV research (21).

The high mutation rates of FIV/HIV have presented researchers with a major obstacle when developing effective chemotherapies for combating these viruses (1-5).

The consequence of a high mutation frequency is two-fold. First, deleterious mutations dramatically reduce the number of viable virus produced (1,4). Secondly, and of most importance to researchers, are mutations in regions of the genome frequently targeted by anti-viral chemotherapies (5-12). Traditional drug strategies have frequently targeted a single aspect of the viral life cycle, such as host cell attachment, replication machinery, envelope synthesis, etc. (7). Mutations have allowed the virus to develop resistant strains impervious to these drugs. This has lead to alteration of chemotherapeutic strategies, with multiple agents employed in an effort to counter these drug resistant mutations by the virus (7,22).

The first U.S. government approved anti-m1V drug treatment employed in early clinical trials was 3 '-azido-3 '-deoxythymidine (AZT:) (22). AZT, a synthetic dideonucleoside, is a nucleoside analog of thymidine. AZT functions to prevent

synthesis of pro-viral DNA, which in turn terminates, viral replication (6,11,12). AZT is inserted into the active site of the reverse transcriptase (RT) enzyme (12). Treatment with AZT renders viral reverse transcriptase more susceptible ( ~ 100 times) to inhibition than cellular DNA polymerase (31). Initially, tests utilizing AZT proved successful against HIV infection (5). However, AZT's effectiveness diminishes rapidly and viral levels rapidly reach pre-AZT treatment levels (6,9,11,12). Extensive research has been undertaken, attempting to correlate mutations in regions of FIV/HIV's genome with suspected sites associated with drug activity and linked to viral drug resistance (8,11,12).

Specifically, FIY/HIV resistance to AZT appears to be directly related to mutations occurring in specific regions of the polymerase (pol) gene of both viruses (8,12,18,19,23, 24) .Of particular interest is a study raising the possibility that a single point mutation in FlV RT confers resistance to two separate classes of anti-viral drugs (31). Mutants possessing AZT resistance have been genetically examined and mutations in the RT segment of the polymerase gene documented (9-11,12,29). Specific regions common to both FIV and HIV were evaluated and have been associated with the development of viral drug resistance (19,21). These mutations effectively prevent AZT's insertion into reverse transcriptase. With the exclusion of AZT from R T (20,21), the mutant gains the ability to transcribe and produce R T .The end result shows the mutant developing a higher fitness than wild type in the presence of AZT .In a curious side note, upon termination of AZT treatment, the wild type rapidly regains its position as the predominant strain (9).

The basis for this particular study is to investigate the effects of AZT on the growth kinetics* of FIV-strain 34f10 (a molecular clone of Petaluma). Mutants will begin to develop in response to challenge with AZT. A comparison of the FIV reverse transcriptase region sequences will be performed. Sequences of the wild type will be compared to sequences of mutants after several passes under AZT exposure. This should verify if emergence of the mutants as the predominant strain has occurred.

Viral fitness data will also be collected. An initial decrease in viral fitness is

likely to be observed. But, as mutants develop, exhibiting resistance to AZT, an increase in population should take place. The fitness of the mutants arising as the result of AZT exposure will be directly reflected in the population kinetics. The confirmation of genetic mutations in RT coupled with population kinetics allows for associations between genetic and population dynamics to be established. It is theorized that as the number of AZT resistant mutants increase, a subsequent increase in viral population will also be

observed. Mutations in R T should result in a higher fitness for AZT resistant strains and an increase in population size.

* For the purpose of this paper, growth and growth kinetics will be used synonymously with viral infectivity of host cells or enumerated as the titer of infectious colonies observed.

Hypothesis:

Exposure to AZT will result in a mutant with higher fitness levels in AZT than the wild type from which the mutant was derived. Fitness is defined as the ability to replicate and produce viable offspring in the presence of AZT.

Aims:

This goal of this research project is to design and conduct an experiment yielding data to support or disclaim the hypothesis stated above. The biological (fitness) and genetic techniques chosen have come as a result of referencing journal articles pertinent to FIV and selecting procedures best suited for testing the hypothesis.

Experiment Design:

FIV 34F 10 infected CrFK cells provided were passed and reseeded into flasks. These cultures were split into two groups, wild type 34f1 0 and AZT dosed. The wild type culture provided a control that were passed and maintained under the identical conditions as the experimental 34f1 0 infected CrFk cells. Cultures were maintained in DMEM with 10% FBS, 1% Glutamate and 1% Penicillin/Streptomycin. Cells were passed on a four-day rotation and maintained with CrFK media. The AZT dosage schedule established a 0.5 mM concentration for passages one and two, increasing to a 1.0 mM concentration in subsequent passages. With each passage, CrFK cells and supernatent were collected. The supernatent was used for the end point dilution assay for fitness of the mutant vs. wild type, and the CrFK cells provided 34F1O virus for genetic evaluation of mutations. Passages were individually subjected to PCR with selected RT primers and gel electrophoresis to provide confirmation of viral infectivity.

For end point dilution 24 well plates were seeded with 5 x 104 naive CrFK cells and allowed to grow overnight. Supernatent collected from Passages 1,3 and 6 were diluted at concentrations ranging from undilute to 10⁶. AZT was added with the media at a concentration of 1.0 mM. Cells and virus were allowed to grow for 3 days and then observed for syncytia formation. A positive result required that 2 of the 4 wells

contained syncytia. This was done for both the AZT resistant mutant strains and the wild type control for the listed passages.

Genetic analysis involved the preparation of lysates of cell products from selected passages. PCR was performed using 1258F and 12c50R as the first round primers and 1259F and 1261R as the second round primers. The result was a 520 bp segment of the Reverse Transcriptase gene. Clones were produced using TOPO TA Cloning Kits and One Shot TOP 10 Vector Kits. Mini prep was performed using Qiagens QIA prep. DNA sequencing was sent out to WMG Lab., North Carolina. Heteroduplex Mobility Assay was used to compare wild type and mutants to determine extent of genetic changes in the Reverse Transcriptase region.

The purpose of this experiment was to evaluate the fitness of an AZT resistant mutant relative to the wild type 34F10. It was theorized that mutations in the pol region of the FIV genome, specifically the reverse transcriptase segment, helped to provide resistance to AZT and results in increased fitness in the mutant. Specific mutations were documented using DNA sequencing and HMA provided substantiating evidence of mutants. The ability of the 34f10 virus to reinfect CrFK cells from generation to generation and quantification of infection by visualization of syncytia after serial dilutions constituted the basis for the fitness study performed. Consolidation of biological (fitness assay) and genetic (DNA sequencing and HMA) data allowed a viable test of the hypothesis. By correlating the extent of mutation in AZT dosed 34F10 infected cells and the subsequent change in fitness, the goal of the project could be accurately tested.

Results:

Hvpothesis: Exposure to AZT will result in a mutant with higher fitness levels than the wild type from which it was derived. Fitne.~s defined as the ability to replicate and produce viable offspring.

In this study, biology (fitness) was correlated with the incidence of genetic mutations. If the hypothesis is true, than the number of mutants possessing a mutation in the reverse transcriptase region of the FIV pol gene should increase when challenged with AZT, in comparison to non-AZT dosed wild type 34FI0.

End point dilutions (see methods) were performed to determine fitness levels of both the 34FI0 wild type and AZT induced mutant populations. Passages 1,3, and 6 were selected for analysis. These passages would be representative and reflect changes noted in mutations and fitness of the virus. Data collected on fitness is shown in Table I and Figure 1.

From trend lines generated from collected data the mutant seemed to maintain a lower fitness than the wild type in the end point dilution assay. This is contrary to the hypothesis generated which stated that the mutant would maintain a higher fitness due to mutations generated while developing resistance to AZT. A slight, steady increase in mutant fitness was noted, but lacked the dramatic increase expected. The wild type, on the other hand, maintained a higher fitness and a more pronounced increase throughout the passages. There is some question as to the validity of this observation though. A significant divergence in the titer level in passage 3 from passages 1 and 3 is highly suspect. The best fit line generated is suspect and ally comparison using it should be used with caution. An accurate conclusion is therefore difficult to ascertain from the data provided. The data seems to lead to the conclusion that the wild type maintains higher fitness than the mutant.

The experiment might have benefited from evaluation of the wild type, not dosed with AZT in the serial dilutions. This could have offered a baseline from which a more accurate reflection of fitness and offered the data more useful for analysis.

Separate experiments and end point dilutions assays were performed by the two researchers and the data was consistent in both cases. This helped to rule out error by the experimenter due to the fact that the data in both experiments pointed toward wild type being the fitter of the two cultures.

The data results were inconclusive and offered little foundation for supporting the emergence of the mutant developing resistance to AZT and maintaining a higher fitness level.

DNA sequencing and the Heteroduplex Mobility Assay were the methods selected for gauging mutations in reverse transcriptase segment of the FIV 34FlO genome. DNA sequencing results follow in Figure 2, amino acid sequence comparison and figure 3, nucleotide sequence comparisons, from passages 4 and 6 respectively. These passages were selected because passage 4 followed the first increase in AZT dosage, following the initial lower dose, and passage 6 was the final passage in the experiment and if our hypothesis was correct, should have shown the greatest number of mutations due to AZT resistance.

Legend for DNA Sequencing

RF1- Passage 6 wild type

RF2- Passage 6 wild type

RF3- Passage 6 wild type

RF4- Passage 6 AZT

RF5- Passage 6 AZT

RF6- Passage 4 AZT

significant number of mutations in the wild type resulted in the passages evaluated. This illustrated the high incidence of mutations in FIV and particularly those in the reverse transcriptase region. Five nucleotide mutations were noted in the sequencing, four were present in the wild type and one in the mutant. Amino acid analysis yielded three amino acid changes, two in the wild type and one in the mutant. This was far from the conclusive evidence sought for correlating fitness and mutations in response to AZT.

The last analysis performed was the Heteroduplex Mobility Assay (Figure 4).

The assay showed the possibility of duplexing in passage 1 and passage 6. The width and intensity of the bands compared with the control (34FI0 wild type lane 1) showed evidence of duplex formation and gave some merit to the possibility that mutations had occurred in response to AZT.

Lane 1-34f10 Wild Type- Homoduplex

Lane 2- Passage 1- Homoduplex

Lane 3- Passage 1- Heteroduplex*

Lane 4- Passage 3- Homoduplex

Lane5- Passage 3- Heteroduplex*

Lane 6- Passage 6- Homoduplex

Lane 7- Passage 6- Heteroduplex*

*Heteroduplexes- Orginal 34F10 Wi1d Type clone duplexed With mutant passage isolate

The genetic results obtained raised some interesting questions but neglected to provide substantial evidence to support the hypothesis. Modifications proposed in the section below would greatly enhance the chances to obtain definitive data and more pronounced results if the experiment were to be restarted. A wealth of knowledge was gained from the experiment along with an increased awareness of experimental methods and techniques. Discrimination and evaluation of journal articles and critical evaluation of other research experiments was a vital benefit gained throughout this experience. Although these factors didn't translate into successful proof of the hypothesis, the

intangibles will contribute to this researcher becoming a better scientist as the result of undertaking this project.

The combination of genetic and fitness data failed to provide evidence to support the claim that higher fitness and AZT resistance due to mutations in reverse transcriptase would be the result of repeated challenge of 34FlO FIV virus with AZT. The hypothesis is valid, we believe, but some modifications in experiment design would yield better definition in the data and the results obtained.

Experimental Changes:

Throughout the course of this research, it has become apparent that several improvements and changes might have enhanced the experiment. Some have come as a result of experiences with the experiment design and others from continuing research in the literature which have provided insight into new methods and techniques. Basically, the hypothesis formed was valid, and the methods and experiment design employed provided an accurate test of the supposition. But, improvements could have been made which may have supplemented the data obtained.

In retrospect, the initial 34FlO cells could have been exposed to a wide range of AZT dosages creating a more severe bottleneck. Selection of a mutant with a much lower survivability would likely have yielded many more mutations in reverse transcriptase and the fitness more defined than was evidenced in the current research project. The greater stress placed on the virus could have greatly enhanced the results. Higher dosages in the end point dilutions used to quantify fitness would also yield more pronounced results with increased AZT dosage.

Initially a plaque assay could also have been utilized to select for AZT resistant mutants. By isolation of a single clone the data obtained would more accurately reflect the mutations occurring in the reverse transcriptase region. This is due to the fact that mutation rates are so high in FIV and analysis of the' samples obtained in the current project reflected the products of many clones. Originally, the 34FlO cultures provided for this experiment were derived from a single clone but much diversity was already present in the 34FlO cells due to the fact that transfected cells were utilized to start the experiment. Isolation of a single clone and establishment into uninfected CrFK cells would greatly benefit the experiment and aid in the credibility of results.

PCR might have been performed on every passage to document infection. Genetic analysis, specifically sequencing or HMA, on the PCR products would have provided the benefit of constantly monitoring the experiment to accurately gauge the changes made by the virus in response to AZT challenge. This would have enabled the researcher to provide a substantial challenge to the virus while still assuring survival. Time was a consideration in this project and the extension from one term to two terms would have given more latitude to the project in incorporating other methods and techniques omitted as the result of a limited time schedule.

The passage schedule of the virus would also be altered. In the current study the virus was passed and AZT dosed every four days. The constant stress on both the CrFK host cells and the 34FlO may have adversely affected the experiment. An interspersing of3-4 feedings, with AZT in mutants, between passages might have provided a stronger infection of the host and allowed a greater opportunity for mutant proliferation. This would be especially desirable if the initial virus was subjected to the more severe bottleneck described above as a probable change in protocol if a new experiment was designed. In the current study six passages were performed, in a new experiment a

doubling or tripling of the number of passages would be undertaken to give more time for mutations to be manifest in the FIV genome.

As a method for quantification of fitness, (infectivity) the Focal Infectivity Assay would be used (19,31). Tagging with an antibody specific for FIV would provide a more accurate means of identifying infected CrFK cells than the visualization of syncytia, which is very prone to experimenter error. Any methods that reduce the possibility of errors greatly enhances the credibility of the study and provides more accuracy in the data collected.

In determining fitness, an additional category would have been added. In the current experiment, wild type virus and AZT dosed 34F10 virus were subjected to AZT with serial dilutions and a titer established. The addition of a third group, representing wild type 34F10 not exposed to AZT in the end point dilution, would provide a base line infection rate and an accurate reflection of the impact of AZT on the wild type 34F10 virus.

Methods and Materials-

Virus and Cells-

FIV 34f10, a molecular clone of FIV Petaluma strain, was used in this project. Crandall Feline Kidney Cells (CrFK) were utilized for the growth and maintenance of virus cells. CrFK nutritional media was made in 250 mi batches and contained:

90% DMEM

8% Fetal Bovine Serum

1% Glutamine

1% Pen/Strep

Passage of confluent cells involved trypsinizcition. Dilutions of 1 : 5 were reseeded into 5 mi CrFK media in 25 cm2 flasks and incubated at 37° C in 6% CO2 atmosphere. Cells were passaged weekly and fed in the interim.

Two separate cultures were maintained. The first was fed Mondays and passed Thursdays. The second was fed on Tuesday and passed on Friday. AZT was administered in appropriate dosages with the feeding and passage of cells.

AZT Dosage and Administration-

AZT provided for the experiment was diluted by addition of 10:1 AZT (@ 10 rnM) to 1000 :1 dH20, a 1: 100 dilution. For the first and second passage, 0.5:M concentrations were administered. This assured adequate survival of infectious virus in order to continue experiment. AZT dosages for subsequent passages consisted of 1.0 :MAZT.

Virus Ouantification-

End Point dilution:

End point dilution was used to detect and quantify infectious viral particles. Serial dilutions were performed and an Elisa performed to determine infectious viral particle population size.

In analyzing the data a titer, or TCID50 ( 50% tissue culture infectious dose ) protocol, was applied. The highest dilution in the four separate samples yielding two dilutions positive for syncytia were used to determine the titer, which is simply the calculated as the inverse of the dilution (example- dilution=l: 1000, titer=10 ).

Genetic analvsis-

Preparation of samples for PCR analysis:

Preparation of cells for PCR involved treatment with PBS, lysis buffer and proteinase K. It is necessary to lyse cells prior to analysis by PCR to liberate the DNA.

Polvmerase Chain Reaction (PCRj:

PCR allows amplification of defined DNA sequences between two different primers. Polymerase chain reaction (PCR) was utilized throughout the experiment. Initially it provided proof of viable 34f10 virally infected CrFK cell lines. Two separate rounds of PCR were performed for the initial confirmation of virus infectivity. Primers used for the first round of PCR were forward 1258 and reverse 1260. Round two primers used were forward 1259 and reverse 1261.

PCR was also utilized for FIV mutant detection. A Primer specifically designed for FIV polymerase reverse transcriptase allowed PCR detection of the mutant strain emergence in response to AZT application. The predominance of the mutant strain post- AZT exposure in comparison to wild type is the purpose of this analysis.

Heteroduolex MobilitY Assay lHMA):

Heat denatured DNA from separate samples of DNA is combined and cooled. The DNA will either form homo duplexes (similar DNA sequences) or heteroduplexes (dissimilar sequences). Migration through a gel will differ between the two complexes. Kinking and bending in the heteroduplex will impede movement through the gel and it will migrate more slowly than homoduplexes. HMA is used for sequence comparison. This was the first exposure to HMA encountered in college and the plan is to perform an HMA mid-point through the procedure.

The major reason for performing HMA will be to give the researcher experience in this new technique. If the results prove beneficial, it may be included in the actual experiment data. Since one purpose of this lab exercise is to gain experience in genetic testing methods, this is an opportunity to explore new horizons.

References

1) Preston BD, Poiesz BJ, Loeb LA. Fidelity of HIV-1 reverse transcriptase. Science 1988;242:1168-1171.

2) Roberts JD, Benebek K, Kunkel TA. The accuracy of reverse transcriptase from HIV-I. Science 1981;242:1171-1173.

3) Nowak MA, Anderson RM, McLean AR, Wolds TFW, Goudsmit J, May RM. Antigenic diversity thresholds and the development of AIDS. Science 1991 ;254:963-969.

3) Wainberg MA, Gu 7, Gao a, et at. Clinical correlates and molecular basis of HIV-1 drug resistance. J AcquiT Immune Defic SyndT 1993;6(suppl.1 ):836-846.

4) Richmann 00. Resistance of clinical isolates of human immunodeficiency virus to antiviral agents. Antimicrob Agents Chemother 1993: 37: 1207 -1213.

5) De La Torre, Juan Carlos and Holland, John J.. Quasispecies populations can suppress vastly superior mutant projeny. J Viral 1990;6278-6281.

6) Coffin JM. HIV population dynamics in vivo: implications for genetic variation, pathogenesis, and therapy. Science 1995:267:483-489.

7) Remington KM, Chesebro B, Wehrly K, Pederson NS, North TW, Mutants of feline immunodeficiency virus resistant to 3'-azido-3'- deoxythymidine. J Viro/1991;65:308-312.

8) Remington KM, Zhu y -a, Phillips TR, North TW. Rapid phenotypic reversion of AZT -resistant mutants of feline immunodeficiency virus without loss of drug-resistant reverse transcriptase. J Viro/1994;68:632- 637.

9) Gobert JD, Remington KM, Zhu Y -Q, North TW. Multiple drug-resistant mutants of feline immunodeficiency virus selected with 2',3'- dideoxyinosine alone and in combination with 3'-azido-3'-deoxythymidine. Antimicrob Agents Chemother 1994;38:861-864.

10) Larder BA, Oarby G, Richmann 00. HIV with reduced sensitivity to zidovudine (AZT) isolated during prolonged therapy. Science 1989;243: 1731-1734.

11) Larder BA, Kemp SO. Multiple mutations in HIV-1 reverse transcriptase confer high-Ievel resistance to zidovudine (AZT 1989;246:1155-1158.). Science

12) Novotney C, English RV, Housman J, et al. Lymphocyte population changes in cats naturally infected with feline immunodeficiency virus. AIDS 1990;4:1213-1218.

13) Pedersen NC, Barlough JE. Clinical overview of feline immunodeficiency virus. J Am Vet MedAssoc 1991;199:1298-1305.

14) Ishida I, Tomoda I. Clinical stages of feline immunodeficiency virus infection. Jpn J Vet Sci 1990;52:645-648.

15) Yamamoto JK, Sparger E, Ho EW, et al. Pathogenesis of experimentally induced feline immunodeficiency virus infection in cats. Am J Vet Res 1988;49:1246-1258.

16) Torten M, Franchini M, Barlough JE, George JW, Mozes E, Lutz H, Pedersen NC. Progressive immune dysfunction in cats experimentally

infected with feline immunodeficiency virus. J Viro/1991 ;65:2225-2230.

17) Cronn RC, Remington KM, Preston BD, North 1 W. Inhibition of reverse transcriptase from feline immunodeficiency virus by analogs of 2'- deoxyadenosine-5'-triphosphate. Biochem Pharmaco/1992;44: 1375-1381.

18) North TW, Cronn RC, Remington KM, Tandberg RT. Direct comparisons of inhibitor sensitivities of reverse transcriptase from feline and human immunodeficiency viruses. Antimicrob Agents Chemother 1990;34: 1505-1507.

19) North TW, Cronn RC, Remington KM, Tandberg RI, Judd RC. Characterization of reverse transcriptase from feline immunodeficiency virus. JBiolChem 1990;265:5121-5128.

20) North TW, North GL T, Pedersen NC. Feline immunodeficiency virus, a model for reverse transcriptase-targeted chemotherapy for acquired immune deficiency syndrome. Antimicrob Agents Chemother 1989;33:915-919.

21) North TW, LaCasse IRA. Testing anti-HIV drugs in the FIV model Nature Med 1995;1 :410-411.

22) Talbott RL, Sparger FE, Lovelace KM, et al. Nucleotide sequence and genomic organization of feline immunodeficiency virus. Proc Natl Acad Sd USA 1989:86:5743-5747

23) Coffin JM. Genetic diversity and evolution of retroviruses. Curr Topics Microbiollmmuno/1992; 176: 143-164.

24) Coffin JM. Genetic variation in retroviruses. In: Kurstak F, Marusyk RG, Murphy FA, Van Regenmortel MHV, eds. Applied Virology Research. New York: Plenum Press, 1990;2: 11-33.

25) Collier AC, Bozzette S, Coomb RW, et al. Low dose zidovudine in HIV infection. N Engl J Med 1990;321 :1626-1631.

26) Hamilton JD, Hartigan PM, Simberkoff MS, et al. A controlled trial of early versus late treatment with zidovudine in symptomatic human immunodeficiency virus infection: results of the Veteran Affairs Cooperative Study. N Engl J Med 1992;326:437-443.

27) Frick LW, Nelson DJ, St. Clair MH, Furman PA, Krenitsky TA. Effects of 3'-zido-3'-deoxythymidine on the deoxynucleotide triphosphate pools of cultured human cells. Biochem Biophys Res Commun 1988;154:124-129.

28) Furman PA, Fyfe JA, St Clair MH, et al. Phosphorylation of 3'-azido-3'. deoxythymidine and selective interaction of the 5'-triphosphate with human immunodeficiency virus reverse transcriptase. Proc Natl Acad Sd USA 1986;83:8333-8337.

29) Meuth M. The molecular basis of mutations induced by deoxyribonucleoside triphosphate pool imbalances in mammalian cells. Exp Cell Res 1989; 181 :305-316.

30) Brooks GF, Butel JS, Morse SA. Medical Microbiology. Appleton and Lange, 1998.

31) North TW, Smith RA, Remington KM, et al. A novel point mutation at position 156 of reverse transcriptase from feline immunodeficiency virus confers resistance to the combination of -)-2-2',3'-dideooxy-3'-thiacytidine and 3'-azido-3'-deoxythymidine. J Viro/1998;72:2335-2340.