INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

ts1 is a spontaneous temperature-sensitive mutant of Moloney murine leukemia virus (MoMuLV). Temperature sensitivity results from a single Val-25Ile substitution in gPr80^env (39). gPr80^env is a precursor polyprotein for the envelope protein, Env. The mature Env protein consists of gp70, a surface protein, and p15E, a transmembrane protein. At the restrictive temperature, gPr80^env is inefficiently transported from the rough endoplasmic reticulum to the Golgi apparatus (38, 39). As a result, there is an elevated steady-state level of misfolded gPr80^env in the rough endoplasmic reticulum (39), and gp70-deficient virions are released from infected cells (44). Infection is initiated by attachment of the viral Env protein (45) to the host cell receptor protein MCAT-1 (1, 35).

Previously, we showed that the tissues of the central nervous system (CNS) of moribund paralyzed FVB/N mice contained unusually high levels of persistent incomplete physical forms of proviral DNA (36, 37). Due to their migration in agarose gels, these physical forms of DNA were designated high-mobility (HM) DNA. The two major persistent physical forms of retroviral DNA that were found in infected tissues of the CNS consisted of class I HM DNA, minus-sense single-stranded DNA that was shorter than one long terminal repeat (LTR), and class II HM DNA, partial DNA duplexes (Fig. 1A). Except for that of unintegrated linear proviral DNA, form III, the functions of extrachromosomal physical forms of retroviral DNA are not clear (for a review, see reference 9). Form III is the physical form of proviral DNA that integrates into the genome of the host cell (14). In DNA-dependent protein kinase (DNA-PK)-deficient murine scid cells, abortive integration of form III is the trigger for cell death by apoptosis (12). HM DNA is physically similar to damaged DNA. It is possible that high levels of HM DNA are toxic to some types of neural cells in the CNS.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Physical forms of ts1 virus-specific DNA in the spinal cord tissues of paralyzed moribund mice (37). (A) Mobilities of physical forms of unintegrated ts1 virus-specific DNA in agarose gels. The minor free plus-sense single-stranded DNA physical form is not shown. ds, double-stranded DNA. (B) Region of minus-sense single-stranded class I HM DNA that hybridizes to the plus-sense single-stranded DNA probe (+) M13-U3-SS19 (55 nts; nts 8095 to 8149) and regions of partial-duplex class II HM DNA and duplex form III DNA that hybridize to (+) M13-U3-SS19 and the minus-sense single-stranded DNA probe () M13-U3-SS18. HaeIII fragments are H6-H7 (410 bp; nts 8069 to 8264 and 1 to 214) and H6-PBS (341 bp; nts 8069 to 8264 and 1 to 145). PPT-PBS (594 bp; nts 7816 to 8264 and 1 to 145) is the LTR. Nucleotides were numbered by the method of Shinnick et al. (32).

Replication intermediates of the proviral DNA of retroviruses have previously been reported for reactions in detergent-disrupted virions in vitro (17), in membrane vesicles exposed to virions in vitro (34), and for the early stages of cell-free virus infections of cells in cultures (42). In cell cultures and membrane vesicles, the replication intermediates rapidly mature to the full-length proviral DNA form within a few hours (34, 42). The persistent physical forms of DNA that we described for the CNS tissues of moribund paralyzed mice were unique because high levels persisted for weeks after the initial stages of infection. To date, only low levels of persistent replication intermediates of the proviral DNA of MoMuLV have been detected in quiescent cells in vitro by the PCR (18, 28). Some types of quiescent cells contain suboptimal levels of deoxynucleoside triphosphates (dNTPs) for reverse transcription (5, 15, 25). However, high concentrations of exogenous deoxynucleosides in the culture medium of some types of nondividing cells promote the completion of reverse transcription of MoMuLV genomic RNA into full-length proviral DNA (18, 28).

In order to better understand the mechanism of HM DNA synthesis in the CNS in vivo, we developed a method for synthesizing high levels of HM DNA in vitro in cultured cells. HM DNA was synthesized in packed confluent monolayers of SC-1 cells that were chronically infected with ts1. When subconfluent monolayers of chronically infected cells were upshifted from 34 to 37°C, persistent levels of HM DNA were synthesized after the monolayers became confluent. Without temperature upshift, the confluent monolayers of chronically infected cells did not synthesize HM DNA. However, when the packed confluent upshifted cells were downshifted from 37 to 34°C, the incomplete physical forms of HM DNA were extended into full-length physical forms of proviral DNA. These in vitro results suggested that the accumulation of HM DNA both in vitro and in vivo may be the result of superinfection.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells. SC-1 wild mouse embryo fibroblasts (ATCC CRL-1404, passage 73) were maintained in Eagle's minimal essential medium (EMEM) supplemented with 10% fetal bovine serum (Sigma, St. Louis, Mo.) at 34°C (20).

Viruses. A molecular clone of ts1, ts1-92b, was derived from MoMuLV (36). ts1wt-33 is a hybrid virus that is identical to ts1 except for a single base substitution at nucleotide (nt) 5948 which results in an Ile-25Val substitution (39, 40). Both ts1 and ts1wt-33 were propagated in TB cells (36). NB-tropic MoMuLV strain E-286 (ATCC VR-1350) was propagated in NIH 3T3 cells (ATCC CRL 1658) (6). Virus titer was determined by the modified direct focus assay as described previously (36), except that 15F cells were seeded in Falcon Primaria dishes (Becton Dickinson and Co., Oxnard, Calif.) to enhance the physical appearance of foci. Nucleotides were numbered by the method of Shinnick et al. (32).

Purification of total cellular DNA. Total cellular DNA was purified from cultured cells by the method of Sambrook et al. (31). Briefly, the monolayers were washed two times with ice-cold Tris-buffered saline (137 mM NaCl, 2.7 mM KCl, 25 mM Tris [pH 7.4]). After the cells were detached from the flasks by scraping with a rubber policeman, they were washed two times in ice-cold Tris-buffered saline and suspended in TE buffer (10 mM Tris [pH 8.0], 1 mM EDTA) at a concentration of 5 × 10⁷ cells/ml. The cells were lysed by adding 10 ml of lysis buffer (10 mM Tris [pH 8.0], 0.1 M EDTA, 20 µg of RNase A per ml, 0.5% sodium dodecyl sulfate) per ml of cell suspension. After the RNA was digested for 1 h at 37°C, proteinase K was added to a final concentration of 100 µg/ml. Proteins were digested for 3 h at 50°C. Undigested proteins were removed by repeated extraction with phenol-chloroform-isoamyl alcohol (25:24:1, pH 8.0) until proteins were completely removed from the interface. RNA was digested for a second time with 30 µg of RNase A per ml for 1 h at 37°C. RNase A was removed by repeated extraction with phenol-chloroform-isoamyl alcohol (25:24:1, pH 8.0). The DNA was precipitated with 2 volumes of ethanol in 2 M ammonium acetate. After the DNA was pelleted by centrifugation, it was washed once with 70% ethanol and suspended in 5 mM Tris (pH 8.0). Total cellular DNA was purified from the spinal cord tissues of infected mice as described previously (36).

Mung bean digestion of total cellular DNA. Total cellular DNA (360 mg/ml) was digested with 600 U of mung bean nuclease per ml (10, 22) in 30 mM sodium acetate (pH 5.0)-50 mM NaCl-1 mM ZnCl₂-0.001% Triton X-100-5% glycerol for 30 min at 37°C as recommended by the manufacturer (Promega, Madison, Wis.). The digestion was terminated by the addition of EDTA to a final concentration of 5 mM.

Probes specific for ecotropic and endogenous viral DNA sequences. The U3-SS probe is a 55-bp (nt 8095 to 8149) SinI fragment that hybridizes to the U3 region of the LTR (37). The endo probe is a 455-bp BglII-EcoRI restriction fragment that hybridizes to endogenous viral DNA sequences (36). Endo was radiolabeled with [-³²P]dCTP by use of a Prime-a-Gene kit (Promega) (36). The plus-sense single-stranded DNA probe, (+) M13-U3-SS19, and the minus-sense single-stranded DNA probe, () M13-U3-SS18, were radiolabeled with [-³²P]dCTP by primer extension as described previously (36, 37). For in situ hybridization, plasmid pU3-SS was linearized with EcoRI. An antisense riboprobe, () U3-SS, was radiolabeled with [-³³P]UTP by in vitro transcription with T7 RNA polymerase. The U3-SS probe did not hybridize to sequences in HaeIII-digested total cellular DNA of uninfected SC-1 cells.

Southern blot hybridization. The method used for Southern blot hybridization was described previously (36).

In situ hybridization. Cells were seeded in plastic Lab-Tek tissue culture chamber slides (Nalge Nunc International Corp., Naperville, Ill.) containing EMEM at 34°C. After being incubated for 3 days, the cells were washed twice with phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.4 mM KH₂PO₄ [pH 7.3]) at room temperature and fixed in 4% paraformaldehyde-phosphate-buffered saline for 10 min (16). The samples were prehybridized in 50% formamide-2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) for 30 min at 37°C in a humid chamber (43). The prehybridization solution was removed by aspiration. A sample of 10 µl of hybridization solution [50% formamide, 5% dextran sulfate, 2× SSC, 1× Denhardt's solution, 200 mM Tris (pH 7.5), 25 µg of poly(A)/ml, 25 µg of poly(C)/ml, 250 µg of yeast tRNA/ml, 10 mM dithiothreitol, 0.5% sodium pyrophosphate] containing 3 × 10⁵ cpm of ³³P-labeled () U3-SS riboprobe was placed on the cells and covered with a siliconized 15-mm coverslip.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Establishing lines of SC-1 cells that were chronically infected with ts1, ts1wt-33, or MoMuLV E-286. SC-1 cells were used for the following experiments because this cell line is highly susceptible to infection by MoMuLV (20, 29). SC-1 cells (passage 78) were infected with ts1, ts1wt-33, and MoMuLV E-286 at a multiplicity of four focus-forming units per cell (36). Lines of chronically infected cells were established by incubating the infected cells for 3 weeks at 34°C in EMEM supplemented with 10% fetal bovine serum. The chronically infected cells were then frozen at 80°C. SC-1 cells that were chronically infected with ts1, ts1wt-33, and MoMuLV E-286 were designated ts1-SC-1, 33-SC-1, and Mo-SC-1, respectively.

View larger version (157K): [in this window] [in a new window]   FIG. 2.   In situ hybridization of chronically infected cells. Cells were seeded in tissue culture chamber slides at 34°C. Three days later, ecotropic virus-specific RNA sequences were detected by in situ hybridization with the () U3-SS riboprobe. (A) 33-SC-1 cells. (B) ts1-SC-1 cells. (C) Uninfected SC-1 cells. (D) Mo-SC-1 cells.

Time course of HM DNA synthesis after temperature upshift. Samples that contained 3 × 10⁶ chronically infected ts1-SC-1 cells were seeded in tissue culture flasks with a growth surface area of 150 cm² (T-150), and the flasks were incubated at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium. The subconfluent monolayers of cells were shifted to an elevated temperature of 37°C. Total cellular DNA was prepared at 0, 1, 2, or 3 days after temperature upshift and analyzed for the presence of HM DNA by Southern blot hybridization with the single-stranded (+) M13-U3-SS19 probe. On the day of temperature upshift, the monolayers were subconfluent, and a trace level of form III DNA was detectable (Fig. 3, lane 1). One day after temperature upshift, the cells were confluent, and there were 6.1 × 10⁶ cells per flask. Unintegrated forms of virus-specific DNA were not detectable (Fig. 3, lane 2). Two days after temperature upshift, there were 6.8 × 10⁶ cells per flask. Low levels of form III DNA and HM DNA were detectable (Fig. 3, lane 3). The cells continued to pack. Three days after temperature upshift, there were 8 × 10⁶ cells per flask. There was a significant increase in both form III DNA and HM DNA levels (Fig. 3, lane 4). These results showed that high levels of HM DNA could be synthesized in vitro by upshifting chronically infected ts1-SC-1 cells from 34 to 37°C. However, the HM DNA was not detected immediately after temperature upshift. After a delay of 2 days, a low level of HM DNA was detectable. The level of HM DNA was highest 3 days after temperature upshift. Detectable levels of HM DNA were correlated with both time and increased cell density after temperature upshift.

View larger version (27K): [in this window] [in a new window]   FIG. 3.   Time course of HM DNA synthesis after temperature upshift. Samples that contained 3 × 106 chronically infected ts1-SC-1 cells were seeded in T-150 tissue culture flasks at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium, and the incubation temperature was shifted to 37°C. Total cellular DNA was prepared on the day of or 1, 2, or 3 days after temperature upshift. Ecotropic virus-specific sequences in 18-µg samples of DNA were detected by Southern blot hybridization with the [-32P]dCTP-labeled (+) M13-U3-SS19 probe.

Assay for the synthesis of HM DNA by three different viruses after temperature upshift to two higher temperatures. Samples that contained 3 × 10⁶ uninfected SC-1, ts1-SC-1, Mo-SC-1, and 33-SC-1 cells were seeded in T-150 tissue culture flasks at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium. The cells were incubated at 34, 37, or 39°C. Total cellular DNA was prepared 3 days after temperature upshift and analyzed for virus-specific unintegrated DNA by Southern blot hybridization with the single-stranded (+) M13-U3-SS19 probe.

View larger version (86K): [in this window] [in a new window]   FIG. 4.   Synthesis of HM DNA after temperature upshift for three different viruses. Samples that contained 3 × 106 cells were seeded in T-150 tissue culture flasks at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium, and the incubation temperature was shifted to 34, 37, or 39°C. Total cellular DNA was prepared 3 days after temperature upshift. Ecotropic virus-specific sequences in 18-µg samples of DNA were detected by Southern blot hybridization with the [-32P]dCTP-labeled (+) M13-U3-SS19 probe. SC-1 cells were uninfected.

Comparison of physical forms of HM DNA synthesized in vitro after temperature upshift to physical forms of HM DNA synthesized in spinal cord tissues. Previously, we characterized the HM DNA in the spinal cord tissues of paralyzed moribund FVB/N mice by Southern blot analysis of detailed nuclease and restriction endonuclease digests of total cellular DNA (36, 37). HM DNA that was synthesized in vivo consisted mostly of short minus-sense single-stranded DNA (class I molecules) and partial DNA duplexes (class II molecules) (Fig. 1A). To determine if the HM DNA that was synthesized in vitro after temperature upshift had a composition similar to that of the HM DNA that was synthesized in vivo, nuclease digestion products from both sources of HM DNA were compared to one another by Southern blot hybridization with the (+) M13-U3-SS19 and () M13-U3-SS18 probes (Fig. 5). In Fig. 5, total cellular DNA that was prepared from the spinal cord tissues of paralyzed moribund mice was designated "T" whereas total cellular DNA that was prepared from ts1-SC-1 cells after temperature upshift was designated "C."

View larger version (63K): [in this window] [in a new window]   FIG. 5.   Comparison of physical forms of HM DNA synthesized in vitro after temperature upshift to physical forms of HM DNA synthesized in spinal cord tissues in vivo. Total cellular DNA was prepared from the spinal cord tissues of paralyzed moribund mice (T) and from ts1-SC-1 cells 3 days after temperature upshift to 37°C (C). Total cellular DNA was either undigested or digested with nucleases. Ecotropic virus-specific sequences in 18-µg samples of DNA were detected by Southern blot hybridization with the (+) M13-U3-SS19 and the () M13-U3-SS18 probes. (A) Assay for class II HM DNA molecules, +, hybridization to the (+) M13-U3-SS19 probe. Lanes: 1, undigested spinal cord tissue DNA; 2, undigested ts1-SC-1 DNA; 3, mung bean nuclease (mb)-digested spinal cord tissue DNA; 4, mung bean nuclease-digested ts1-SC-1 DNA. , hybridization to the () M13-U3-SS18 probe. Lanes: 5, undigested spinal cord tissue DNA; 6, undigested ts1-SC-1 DNA; 7, mung bean nuclease-digested spinal cord tissue DNA; 8, mung bean nuclease-digested ts1-SC-1 DNA. (B) Assay for class I HM DNA molecules. +, hybridization to the (+) M13-U3-SS19 probe. Lanes: 1, HaeIII (H)-digested spinal cord tissue DNA; 2, HaeIII-digested ts1-SC-1 DNA; 3, spinal cord tissue DNA digested with HaeIII followed by digestion with mung bean nuclease; 4, ts1-SC-1 DNA digested with HaeIII followed by digestion with mung bean nuclease. , hybridization to the () M13-U3-SS18 probe. Lanes: 5, HaeIII-digested spinal cord tissue DNA; 6, HaeIII-digested ts1-SC-1 DNA; 7, spinal cord tissue DNA digested with HaeIII followed by digestion with mung bean nuclease; 8, ts1-SC-1 DNA digested with HaeIII followed by digestion with mung bean nuclease. The arrowhead indicates a band of single-stranded DNA.

Unintegrated sequences of endogenous retroviral DNA were not detectable in ts1-SC-1 cells by Southern blot analysis. The chromosomal DNA of SC-1 cells contains sequences of endogenous ecotropic, xenotropic, polytropic, and modified polytropic murine retroviruses. In cell cultures, uninfected SC-1 cells constitutively express xenotropic, polytropic, and modified polytropic RNAs (21). Endogenous retroviral RNA transcripts are not packaged into replication-competent virus particles (2). It is possible that the HM DNA molecules that were synthesized in the ts1-SC-1 cells after temperature upshift were incomplete products of reverse transcription of endogenous viral RNAs.

View larger version (23K): [in this window] [in a new window]   FIG. 6.   Unintegrated endogenous retroviral DNA was not detectable in SC-1 or ts1-SC-1 cells. Samples that contained 3 × 106 cells were seeded in T-150 tissue culture flasks at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium, and the incubation temperature was shifted to 37°C. Total cellular DNA was prepared 3 days after temperature upshift. Endogenous nonecotropic virus-specific sequences in 18-µg samples of DNA were detected by Southern blot hybridization with the [-32P]dCTP-labeled endo probe. Lanes: 1, undigested SC-1 DNA; 2, undigested ts1-SC-1 DNA; 3, HaeIII (H)-digested SC-1 DNA; 4, HaeIII-digested ts1-SC-1 DNA.

Temperature downshift of ts1-SC-1 cells after temperature upshift. High levels of HM DNA were synthesized in vitro after temperature upshift of ts1-SC-1 cells (Fig. 3, lane 4). The ts1-SC-1 cells could not be maintained in cell cultures for longer than 3 days after temperature upshift, since the sheet of packed cells would peel off the growth surface of the tissue culture flask (20). Three days after temperature upshift, the packed ts1-SC-1 cells were most likely not dividing. To determine the fate of the HM DNA in the upshifted cells, samples containing 3 × 10⁶ packed ts1-SC-1 cells were seeded in T-150 flasks to yield a lower cell density. The cells were then downshifted to 34°C. Total cellular DNA was prepared at 0, 1, 2, 3, or 4 days after temperature downshift and analyzed for the presence of unintegrated virus-specific DNA by Southern blot hybridization with the single-stranded (+) M13-U3-SS19 probe.

View larger version (34K): [in this window] [in a new window]   FIG. 7.   Downshift of ts1-SC-1 cells after temperature upshift. Samples that contained 3 × 106 chronically infected ts1-SC-1 cells were seeded in T-150 tissue culture flasks at 34°C. Twenty-four hours later, the medium was decanted and replaced with fresh medium, and the incubation temperature was shifted to 37°C. For one set of flasks, total cellular DNA was prepared 3 days after upshifting of the temperature to 37°C (lane 1). The cells in the remaining flasks were subcultured at a cell density of 3 × 106 cells per flask 3 days after upshifting of the temperature to 37°C. These flasks were downshifted to 34°C. Total cellular DNA was prepared 1, 2, 3, or 4 days after temperature downshift (lanes 2 to 5, respectively). Ecotropic virus-specific sequences in 18-µg samples of DNA were detected by Southern blot hybridization with the [-32P]dCTP-labeled (+) M13-U3-SS19 probe.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Physical forms of incomplete reverse transcription products are not usually detectable in murine leukemia virus (MuLV)-infected cells (34, 42). When a naive host cell is infected by an extracellular MuLV, the viral RNA genome is reverse transcribed into form III DNA. Within 24 h, form III DNA integrates into the genome of the dividing host cell (30). The productively infected cell develops immunity to superinfection. Reverse transcription is rapid, and incomplete products of reverse transcription are not usually detectable during all stages of the process of infection. In this report, we have produced persistent high levels of incomplete products of reverse transcription in vitro after temperature upshift of chronically infected ts1-SC-1 cells. The production of HM DNA after temperature upshift may be due to unique virus-cell interactions in the chronically infected cells.

The synthesis of HM DNA after temperature upshift indicated that reverse transcription was occurring in the chronically infected ts1-SC-1 cells (Fig. 3, lane 4). Since endogenous viral RNA is constitutively expressed in SC-1 cells (21), the source of the RNA template for reverse transcription could have been endogenous viral RNA or ts1 viral RNA. In some cell systems, endogenous viral RNA transcripts can be packaged by MuLVs (4), but endogenous viral RNA transcripts are not packaged in uninfected SC-1 cells (2). The endo probe did not detect HM DNA sequences in either upshifted uninfected SC-1 cells (Fig. 6, lane 1) or upshifted ts1-SC-1 cells (Fig. 6, lane 2). In addition, ecotropic HM DNA sequences were not detected in upshifted uninfected SC-1 cells (Fig. 4, lane 2). Therefore, the HM DNA sequences that were detected after temperature upshift of ts1-SC-1 cells (Fig. 3, lane 4) resulted from reverse transcription of RNA templates of ts1.

In MoMuLV-infected cells, the proviral DNA is transcribed by host cell DNA-dependent RNA polymerase II into full-length viral RNA molecules that either are packaged into immature virus particles or serve as mRNA in protein synthesis. After budding from the infected cell, the virus particle matures. After gag-pol is cleaved, the encapsidated viral RNA may be reverse transcribed into proviral DNA (46). Reverse transcription of viral mRNA is very rare. Our data indicated that full-length viral mRNA molecules of ts1 did not serve as templates for reverse transcription into HM DNA. HM DNA was not detectable in ts1-SC-1 cells that were incubated for 3 days at 34°C (Fig. 4, lane 10). Furthermore, HM DNA was not synthesized on the day of or 1 day after temperature upshift (Fig. 3, lanes 1 and 2, respectively). In addition, if ecotropic full-length viral mRNA was being constitutively reverse transcribed, Mo-SC-1, 33-SC-1, and ts1-SC-1 cells should have contained similar levels of HM DNA at all three temperatures (Fig. 4). Only ts1-SC-1 cells contained high levels of HM DNA (Fig. 4, lanes 11 and 12). Therefore, it was unlikely that full-length viral mRNA molecules of ts1 were constitutively reverse transcribed into HM DNA.

The most likely source of the viral RNA template for the synthesis of the HM DNA was extracellular virions of ts1 that were superinfecting the chronically infected ts1-SC-1 cells. Although the preinfected ts1-SC-1 cells should have been immune to superinfection, superinfection immunity is not absolute. Odawara et al. (27) have shown that NIH 3T3 cells that are preinfected with MuLV are susceptible to low levels of superinfection until the cells attain a threshold level of integrated proviral DNA copies. These investigators hypothesized that interference was determined by the intracellular proportions of the Env protein and the host cell receptor protein MCAT-1. ts1 is a leaky mutant because elevated temperatures do not totally inactivate the virus. Elevated temperatures only cripple the virus, since virions with decreased amounts of gp70 are released from infected cells at the restrictive temperature (44). After temperature upshift, the defective Env protein of ts1 appears to be ineffective in establishing superinfection immunity. Defective Env proteins probably do not participate in establishing superinfection immunity. Therefore, ts1-SC-1 cells were probably more susceptible to superinfection after temperature upshift from 34 to 37°C than ts1-SC-1 cells that were maintained at 34°C. These results suggested that superinfection in the packed chronically infected ts1-SC-1 cells might have been responsible for the synthesis of HM DNA after temperature upshift. Superinfection might be one requirement for the synthesis of HM DNA.

HM DNA synthesis did not begin immediately after temperature upshift (Fig. 3). There was a delay of 2 days before HM DNA was synthesized (Fig. 3, lane 3). This 2-day delay could be the time it took for the ts1-SC-1 cells to adapt to the elevated temperature of 37°C. Since the gPr80^env protein of ts1 is temperature sensitive, new intracellular proportions of gPr80^env, gp70, and p15E were probably established at 37°C. Another possible reason for the delay in HM DNA synthesis after temperature upshift is that the cell density had to increase to a certain threshold level in order for HM DNA synthesis to occur. The cell density was highest 4 days after temperature upshift, and the cells at this time contained the highest levels of HM DNA (Fig. 3, lane 4). The packing of the ts1-SC-1 cells might have enhanced the amount of superinfection by cell-to-cell contact. Cell-to-cell transmission of retrovirus infection is much more efficient than cell-free virus-to-cell transmission because the short half-lives of retroviruses limit the distance that they can effectively travel in solution by Brownian motion (11). Our results do not distinguish between superinfection by cell-free virions of ts1 or superinfection by cell-to-cell contact of ts1-SC-1 cells, but the results do suggest that the accumulation of HM DNA in ts1-SC-1 cells requires superinfection. In vivo, it is unclear if retroviruses ever fully establish receptor interference to superinfection (35). The HM DNA that we previously described for the spinal cord tissues of moribund paralyzed mice might have resulted from superinfection of neural cells by virions of ts1.

Although increased cell density might have increased the rate of superinfection by virions of ts1, the packing of the ts1-SC-1 cells might have caused the genomes of the superinfecting virions to be incompletely reverse transcribed into HM DNA. Reverse transcription requires specific levels of dNTPs to synthesize full-length proviral DNA. If retrovirus-infected cells contain suboptimal levels of dNTPs, the products of reverse transcription are incomplete (5, 15, 25). Cultured cells may contain low levels of dNTPs if they are not dividing (18, 28) or if they are density inhibited (41) because the intracellular pools of dNTPs fluctuate during the cell cycle (7, 26). Three days after temperature upshift of the ts1-SC-1 cells, the packed cells were probably not dividing, and low levels of dNTPs in the packed cells might have contributed to the production of high levels of HM DNA (Fig. 3, lane 4).

HM DNA molecules are incomplete products of reverse transcription. Incomplete reverse transcription could result from inhibition of the reverse transcriptase enzyme by low levels of dNTPs or from reverse transcription of viral RNA templates that are shorter than the 8-kb full-length viral RNA molecules. By downshifting ts1-SC-1 cells that contained HM DNA (Fig. 7, lane 1) from 37 to 34°C, we showed that the incomplete forms of HM DNA could be extended to the full-length form III molecules (Fig. 7, lanes 2 and 3). This result showed that the HM DNA molecules in the ts1-SC-1 cells were genuine incomplete products of reverse transcription of full-length viral RNA templates. After temperature downshift, most of the form III molecules appeared to be converted to other physical forms, such as one- and two-LTR form I molecules and the low-mobility physical forms in band Y (Fig. 7, lane 3). Since form III is the physical form that integrates into the genome of the host cell, most form III molecules that were generated from extension of HM DNA molecules after temperature downshift did not appear to integrate into the genome of the host cell.

In this report, the HM DNA that we previously described for the spinal cord tissues of paralyzed moribund mice was synthesized in vitro. The HM DNA contained both single-stranded class I molecules and partial-duplex class II molecules. Temperature upshift was the stimulus that was responsible for the synthesis of the HM DNA. Previously (36), we identified extremely high levels of HM DNA in the spinal cord tissues of paralyzed moribund mice that were infected with ts1. The severity of the neurodegenerative disease was correlated with the level of HM DNA in the spinal cord tissues of the infected mice. Presently, the role that the HM DNA plays in the process of the neurodegenerative disease is not known. We have now produced large amounts of HM DNA in cultured ts1-SC-1 cells in vitro. This model system can now be used to test the effect of HM DNA on the physiology of the host cell.

HM DNA is physically similar to damaged DNA. Damaged DNA is toxic inside some types of cells. In DNA-PK-deficient murine scid cells, abortive integration of form III unintegrated DNA was the trigger for cell death by apoptosis (12). The HM DNA that we described for the spinal cord tissues of paralyzed moribund mice was physically similar to damaged DNA. It is possible that high levels of HM DNA are toxic to some types of neural cells in the CNS by activation of apoptotic pathways in DNA-PK-deficient cells. We can now determine if persistent levels of HM DNA in ts1-SC-1 cells activate enzymes in apoptotic pathways. Alternatively, it is also possible that HM DNA is toxic to neural cells by activation of DNA-PK. Intracellularly, DNA-PK is inactive until it is activated by single-stranded DNA ends (19). Once activated, DNA-PK phosphorylates many cellular proteins (3). The phosphorylated protein substrates include nuclear DNA-binding proteins and other proteins involved in transcription. It is possible that the single-stranded DNA ends of class II HM molecules activate DNA-PK in some types of neural cells. As a result, alteration of transcription in neural cells that are infected with ts1 may induce neurodegeneration in cells of the surrounding CNS tissue. We now have an in vitro system that enables us to determine if HM DNA activates DNA-PK.

Moreover, producing the HM DNA in vitro has identified a type of cell and a physiological perturbation that may allow study of the early steps of replication of MoMuLV, with the accumulation of large amounts of preintegration complexes. The accumulation of high levels of early replication intermediates after an abrupt cessation of reverse transcription in packed infected cells following temperature upshift may be useful for studying reverse transcription complexes (13), preintegration complexes (23), and the effect of the sizes of dNTP pools on the kinetics of reverse transcription in quiescent cells (28). In gene therapies that use MuLV vectors, the vectors are not effective in quiescent cells. Through study of the high levels of early intermediates of viral replication in SC-1 cells, new methods to increase the effectiveness of MuLV vectors in quiescent cells may be forthcoming.