Competitive PCR for Quantification of BM5d Proviral DNA in Mice with AIDS

Abstract

Murine AIDS in C57BL/6 mice is caused by a unique mixture of murine leukemia viruses. We report the use of a competitive PCR to detect and quantitate BM5d proviral DNA. This assay allowed discrimination among endogenous wild-type murine retroviruses and BM5d sequences. Furthermore, the method was subsequently used to evaluate the amount of BM5d in infected mice and in infected AZT (zidovudine)-treated mice, providing an effective way to quantitatively evaluate drug efficacy in the murine AIDS model.

Murine AIDS is a syndrome of lymphoproliferation and immunodeficiency induced by murine leukemia viruses in C57BL/6 mice (13). The etiologic agent in this mixture, known as LP-BM5, has been shown to be a replication-defective virus (BM5d) which requires replication-competent helper viruses, such as ecotropic (BM5e) and mink cell focus-forming viruses (1, 4). This syndrome has many similarities to human AIDS (8, 17). During our studies of BM5d integration in infected C57BL/6 mice (3), and in other reports (4, 10, 11), the presence of multiple p12-related endogenous sequences in the genomes of uninfected mice was demonstrated. Southern hybridization assay proved that these sequences hybridized strongly with a p12^gag gene-specific probe (D30) (1). Furthermore, by using specific primers for the amplification of BM5d gag sequence, an endogenous p12^gag gene homolog (EMBL Data Library accession no. X72930) was found to be present in the DNA of all tissues examined. Since C57BL/6 mice infected with the LP-BM5 virus complex are currently used in the preclinical evaluation of antiretroviral drug combinations (7, 15), we believed that the availability of a quantitative assay of proviral DNA in the genome of infected mice would be advantageous. In this paper, we report a new competitive PCR (cPCR) technique which allows the selective detection of BM5d proviral DNA in treated and untreated infected mice.

For the construction of the competitive template, we chose a fragment of p12^gag gene that showed the highest degree of diversity from those of the other retroviruses and endogenous sequences (1, 4). For this purpose, total RNA (5) was extracted from the lymph nodes of an infected C57BL/6 mouse and cDNA synthesis was performed with the cDNA cycle kit (Invitrogen, San Diego, Calif.), using a 3′-specific primer corresponding to positions 1579 to 1596 of the BM5d genome (4, 6). The following amplification reaction, performed under the same conditions reported elsewhere (10), produced a 141-bp amplified product (positions 1456 to 1596 in the BM5d genome; GenBank accession no. M64096), which was cloned into the pMOSBlue vector (Amersham, Buckinghamshire, United Kingdom). After the transformation reaction, 10 μg of recombinant DNA plasmid (pMOS-141) was cleaved with the unique NcoI and KspI restriction enzymes (Boehringer, Milan, Italy) (recognition sites, positions 1528 and 1565, respectively), generating a 35-bp DNA fragment and a new plasmid (pMOS-106) with a deletion. After ligation the 35-bp deletion was confirmed by sequencing the gag insert by chain termination methods (U.S. Biochemicals, Amersham, United Kingdom). All cloning steps were performed according to the manufacturer’s instructions.

The feasibility of cPCR assay was first investigated by comparing the amplification kinetics of the wild-type (pMOS-141) and competitor (pMOS-106) DNA plasmids. cPCR was carried out in a 50-μl final volume containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1 mM MgCl₂, 200 μM (each) deoxynucleoside triphosphate, the primers (400 nM each) reported previously (6), and 2.5 U of Taq DNA polymerase (Perkin-Elmer, Branchburg, New Jersey). The reaction mixture was subjected to 1 cycle of denaturation at 95°C for 3 min and 50 cycles of denaturation at 95°C for 30 s, annealing at 58°C for 30 s, and extension at 72°C for 30 s, followed by a final extension at 72°C for 10 min. Twenty-five-microliter aliquots of the amplification reaction mixtures were analyzed on 3% NuSieve GTG (FMC Bio Products, Rockland, Maine)–1% pure agarose gel (Bio-Rad, Hercules, Calif.), transferred onto a nylon membrane, and hybridized in a standard solution with ³²P-labeled D30 excess probe (at least 2.5 × 10⁶ cpm/ml) (6). The filters were then washed at high stringency as described in the procedure for DNA hybridization supplied by Amersham. The intensities of the 141- and 106-bp radiolabeled bands were analyzed and quantified in a GS-250 molecular imager (Bio-Rad). The values obtained were then plotted as a function of the log₁₀ of the known competitor DNA copy number. The point of equivalence was that at which the amounts of radioactivity incorporated into the competitor and target were equal and represented the number of copies of BM5d proviral DNA in the starting sample. In our cPCR method, the 141- and 106-bp products were quantitated by radioactivity associated with the respective bands. This procedure provides an additional level of specificity and is more sensitive (9, 19) than ethidium bromide staining, which is frequently used in DNA quantitation (16). The D30 probe hybridized with the 141- and 106-bp sequences, and thus, no correction factor was needed because the amount of radioactivity was proportional to the amount of DNA. In repeated experiments very similar amplification efficiencies, in a linear range of 10¹ to 10⁶ DNA copies, were found when both plasmids were separately amplified (data not shown). In other experiments we examined whether there was competition between the pMOS-141 and pMOS-106 DNA plasmids when they were coamplified in the same reaction tube. Figure 1 shows a representative result of one of these experiments. The autoradiographic film shows a progressive competition between a fixed amount of wild-type DNA (5,000 molecules per reaction) and variable quantities of competitor DNA (from 100 to 125,000 molecules). In addition, we observed an unexplained low-molecular-weight band clearly resolved from both the 141- and 106-bp bands that may consist of a heteroduplex of the two expected products. The presence of such a heteroduplex band should not interfere with quantitation of the target and competitive templates (18). In the graph, the equivalence point at the intersection of the two curves corresponds to 5,054 molecules of pMOS-141, a number very similar to the input. Three experiments were performed in order to confirm the reproducibility of the assay (5,418 ± 373 [mean ± standard deviation {SD}]; coefficient of variation, 6.9%). Subsequently, we applied this cPCR method to a murine model of AIDS to evaluate the changes in BM5d proviral load during the course of the disease, in response to drug therapy. For this purpose 15 C57BL/6 mice were infected with the retroviral complex LP-BM5. The sources of the animals and the virus strain have been reported previously (2, 6). Three of the mice received 0.25 mg of AZT (zidovudine)/ml in drinking water for the duration of the experiment, beginning 24 h after infection, as described previously (6). Three of the infected mice were killed at a time, at 3, 7, and 10 weeks postinfection (p.i.). The three treated mice were also killed at 10 weeks p.i., and the remaining three infected mice were killed at 15 weeks p.i. Three uninfected mice served as controls. BM5d proviral DNA in infected mice and infected treated mice was detected by semiquantitative PCR (sPCR), described in detail elsewhere (6), and the new cPCR method. In this case a constant amount of target DNA (usually 0.3 μg) and a variable known copy number of pMOS-106 plasmid (from 10¹ to 10⁶ molecules) were used. Total DNA was isolated from a 1-mm³ section of each tissue sample by the sodium iodide-isopropanol method described previously (12). At first we evaluated the relationship between the increase in the levels of BM5d proviral DNA, the weeks of infection, and the lymph node weights. The competitive assay was performed at 3, 7, 10, and 15 weeks p.i. As shown in Table 1, there was a good correlation up to the 10th week between the BM5d virus DNA copy number, the lymph node weight, and weeks of disease progression. Thus, since enlargement of lymph nodes was correlated with viral infection, we were able to use this cPCR as a marker for disease progression. At 15 weeks p.i. a further increase of lymph node weight, but not of DNA proviral copies, was observed. These results were in agreement with the hypothesis that the expansion of the mass of the lymph nodes could be due to noninfected reacting cells that had migrated into the nodes (8). In the last series of experiments, the cPCR method was used to evaluate the effect of treatment with AZT on proviral DNA content. In previous studies we had evaluated the therapeutic benefits of this antiretroviral treatment in murine AIDS (6). Figure 2 illustrates a representative competitive experiment in which DNAs from the lymph nodes and spleens of one infected mouse and one infected AZT-treated mouse were coamplified with different sets of competitor dilutions. As documented, after 10 weeks p.i. AZT administration was effective in reducing the number of proviral DNA copies in all of the tissues investigated. The same DNA samples were also analyzed by a semiquantitative assay and detected as described previously (6). The percentages of inhibition of BM5d proviral DNA in infected AZT-treated mice were calculated by the two methods. The cPCR results correlate with the results obtained with the semiquantitative amplification assay on lymph node (inhibitions of 83% ± 3% and 77% ± 10% were obtained with cPCR and sPCR, respectively [mean ± SD of three mice]; P > 0.1) and spleen (inhibitions of 76.5% ± 11% and 70% ± 4% were obtained with cPCR and sPCR, respectively [mean ± SD of three mice]; P > 0.1) tissues. A significant difference between the results obtained with the two methods was found for bone marrow (inhibitions of 74% ± 8% and 48% ± 5% were obtained with cPCR and sPCR, respectively [mean ± SD of three mice]; P = 0.006). This difference was likely due to the small amount of BM5d in bone marrow. In fact, the BM5d DNA copy number in the bone marrow of infected mice is 13,031 ± 606, compared to 248,250 ± 23,200 in the lymph nodes (mean ± SD of three mice). Furthermore, after treatment all tissues showed a further decline in proviral DNA levels of at least 1 log unit. Thus, one possible explanation for the lack of correlation could be that the cPCR assay is more sensitive than the semiquantitative technique and therefore provides more thoroughness in the detection of low levels of proviral DNA.

FIG. 1.

Quantitation of the known copy number of wild-type pMOS-141. cPCR was performed with increasing copy numbers of competitor pMOS-106 (lanes: 1, 0.1 × 10³; 2, 1 × 10³; 3, 5 × 10³; 4, 25 × 10³; 5, 125 × 10³) to compete against a fixed amount (5,000 molecules) of wild-type pMOS-141 per reaction. After the hybridization procedure, the radioactivity present in each band was determined and the values obtained were plotted as functions of the log₁₀ of the competitor copy number. The intersection of the two curves shows the point of equivalence. The determined value of 5,054 molecules for the input wild-type sequence is in excellent agreement with the actual input of 5,000 copies.

TABLE 1.

Relationship between BM5d proviral DNA copy number, lymph node weight, and weeks p.i.

Wk p.i.	BM5d proviral DNA copy no.a,b	Lymph node wt (g)b
3	<400	0.192 ± 0.015
7	123,560 ± 10,910	0.28 ± 0.04
10	248,250 ± 23,200	1.84 ± 0.6
15	197,390 ± 18,785	4.0 ± 0.72

^aProviral DNA molecules per 0.3 μg of total cellular DNA, as assessed by cPCR. 

^bAll values are the mean ± SD for three infected mice. 

FIG. 2.

Application of cPCR for the determination of BM5d proviral DNA copy number in tissues from one infected mouse and one infected and AZT-treated mouse killed 10 weeks p.i. DNAs (0.3 μg) from the lymph nodes and spleens were coamplified with a selected range of copy numbers of competitor pMOS-106. The autoradiographic profile of coamplification is shown above each graph. The following ranges of pMOS-106 plasmid molecules were used. (Upper left) Infected lymph node. Lanes: 1, 1 × 10⁶; 2, 4 × 10⁵; 3, 8 × 10⁴; 4, 2 × 10³. The equivalence point corresponds to 299,340 copies of DNA. (Upper right) Infected AZT-treated lymph node. Lanes: 1, 4 × 10⁵; 2, 8 × 10⁴; 3, 2.5 × 10⁴; 4, 1 × 10³. The equivalence point corresponds to 41,306 copies of DNA. (Lower left) Infected spleen. Lanes: 1, 1 × 10⁶; 2, 5 × 10⁵; 3, 1.5 × 10⁵; 4, 2 × 10³. The equivalence point corresponds to 112,000 copies of DNA. (Lower right) Infected AZT-treated spleen. Lanes: 1, 5 × 10⁵; 2, 1.5 × 10⁵; 3, 3 × 10⁴; 4, 2.5 × 10³. The equivalence point corresponds to 12,125 copies of DNA.

In conclusion, we have demonstrated the optimization and the reliability of cPCR and the possibility of monitoring BM5d proviral DNA levels during drug therapy, in this case with AZT, in a murine model of AIDS. In previous studies we found greater efficacy with combination therapy (6, 14), and in the future we will investigate the effect of new combined treatments with drugs acting on different steps of the virus life cycle. The extraordinary sensitivity of cPCR will be very useful in evaluating BM5d content as a marker of the response to these new antiretroviral treatments, particularly in tissues characterized by relatively low levels of infection, such as the brain.