Introduction

The non-enveloped virion of parvoviruses has a diameter of 20-26 nm and contains a linear, single strand DNA genome of approximately 5 kilobases encapsidated with a simple icosahedral protein coat. Parvoviruses are divided into 2 groups: (1) Adeno-associated viruses, which need a coinfection with a helper virus, such as adenovirus or herpesvirus for a productive infection of host cells; (2) Autonomous parvoviruses1, which require no coinfection and rely on proliferation and differential states of host cells for a productive life cycle2. Autonomus parvoviral DNA does not integrate into its host genome3. H-1, an osteolytic virus of hamsters, belongs to the autonomous parvoviruses4.

Many papers have ldentified an oncosuppressive effect of parvoviruses. The viruses lowered the incidences of some spontaneous and induced tumors5, 6. Parvoviruses were able to kill many kinds of transformed cells. Minute virus of mice (MVM) infection decreased markedly the viability of simian virus 40 (SV40) transformed MVM-resistant BALB/3T3 variant as measured by cell colony-efficiency on plastic plates7. H-1 virus suppressed the growth of human fibroblast transformed by SV40 8 and human keratinocytes transformed by Ha-ras oncogene9. Our previous studies have documented that H-1 virus displayed a selective antineoplastic activity on human gastric cancer or hepatoma tissues cultured directly after surgical operation, while the killing activity on their juxta-cancerous tissues could not be observed10, 11. There was a report about the suppression of H-1 to the formation and growth of tumors in nude mice by transformed mammary epithelial cells12.

The mechanism of the preferentially oncosuppressive activity of parvovirus has not yet been completely elucidated. However, at earlier stages of viral life cycle there were no detectable differences whether in normal cells or in their transformed counterparts, while at later stages, significant differences did appear. In most transformed cells, amplification of replicating form DNA (RF DNA) maintained at a much higher level than that in their normal parent cells. Thus, RF DNA amplification level seemed to be a good mark to ascertain whether host cells have transformed or not8. It has been documented that the nonstructural protein type 1 (NS-1) of MVM displayed a cytotoxic activity13, and that NS-1 and NS-2 might act synergistically in the cytopathogenicity of parvovirus MVM14. The amount of NS-1 accumulated in transformed cells was several times higher than that in the normal. This might be a useful index to imply that the tumors or the transformed cells were inhibited15, 16.

Our previous studies have reported the preferentially cytolytic effect by parvovirus H-1 in 4 kinds of cultured hepatoma cell lines10. To elucidate the mechanisms, we assessed quantitatively the killing effect, H-1 DNA amplification and host cell DNA synthesis, and investigated this effect on the heterotransplanted human hepatoma in nude mice.

Materials and Methods

1. Cell line and virus

Human hepatoma cell line QGY -770317 was cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), in 5% CO2 and saturated vapor at 37 °C. Tumorigenic characteristics of the cells was determined with high colony-forming efficiency (46%) in 0.3% soft agar and 100% tumorigenicity in nude mice after s c injection or transplantation.

H-1 virus was propagated in NB-K cells (a SV40-transformed new-born human kidney cell line) and purified according to the published method18. Virus titration by plaque assay was mearsured as described by Ledinko19. Virus infection was performed as the published method20, and determined with multiplicity of infection (MOI), the ratio of plaque-forming unit (PFU) to the amount of infected cells.

2. Plasmid

Plasmid pULB 3514 (a gift of Prof J Rommelaere, Universite Libre de Bruxelles) was constructed with BamHI fragment of pUC12 and Bg1II fragment of H-1 DNA. The construct of the plasmid was outlined in Fig 1. Amplification, purification and identification were performed as described in “Molecular Cloning”21.

Fig 1
figure 1

Construct of plasmid pULB3514

Full size image

3. Colony-forming efficiency in soft agar

For measurement of anchorage-independent growth, 500 QGY -7703 cells in 4 ml of RPMI 1640 medium containing 10% FCS and 0.3% agar were seeded onto each 6 cm diameter plastic dish (Nunc) overlayed previously with 4 ml of RPMI 1640 medium supplemented with 10% FCS and 0.6% agar. Colonies were counted 6-7 d later by microscopic examination.

4. Survival of the cells after infection

The 500 QGY -7703 cells infected with H-1 virus of mock infected in 4 ml of RPMI 1640 medium were plated onto each 6 am-diameter Nunc dish. After 6-7 days culture, cells were fixed with Carnoy solution and stained with Giemsa and colonies were counted under light microscope. Colony-forming efficiency of the infected cells in RPMI 1640 medium containing 0.3% agar was also examined according to Method 3.

The cell survival rate was calculated as follows:

5. H-1 virus DNA amplification and host cellular DNA synthesis

Total viral DNA within host cells was measured by dispersed cell assay22. Plasmid pULB 3514 was labled with α-32 P -d CTP by nick translation. Labeled plasmid was separated from unincorporated precursors by chromatography with Sephadex G-50 column. An amount of 1 × 106 cells infected with H-1 virus or the mock infected cells was dispersed onto nitrocellulose membrane (Schleider and Schuell BA85, 0.45 μm, ф 25 mm) and hybridized with labeled plasmid. Radioactivity was measured with a liquid scintillator. In the determination of viral DNA synthesis, the background radioactivity inherent in the hybridization method and the input radioactivity by uptake of viral DNA were measured with samples uninfected and harvested at 2 h postinfection, respectively; while the peak of viral DNA synthesis was shown at 30 h postinfection 8, when viral DNA had completed the conversion from double-strand monomers to multimers23. Viral DNA amplification level was obtained from the ratio of intracellular H-1 DNA contents at 30 h vs 2 h postinfection.

For measurement of the effect of H-1 virus on cell cycle, especially on the percentage of S-phase cells, QGY -7703 cells infected with H-1 virus at MOI 5 were harvested at various time intervals and fixed immediately with 70% ethanol solution. Then, the fixed cells were treated with 1% Triton X -100 solution , digested by RNase, stained with propidium iodide solution, finally filtered with 200-net nylon membrane. After the above procedures, the percentages of difierent cell phase were determined with an EPICS -751 fluorescence analyzer24.

6. Progeny virus in cells

Progeny virus was detected by infectious cell center assay. After 5 h incubation p.i, cells were washed with PBS 3 times, then plated onto the indicating NB-K cells, the most susceptible cells to H-1 infection. After 1 wk culturing, plaques which indicated cells releasing infectious progeny virions were formed and counted.

7. Inhibitory effect of H-1 on the formation and growth of tumors by QGY-7703 cells in nude mice

Cells were infected with H-1 virus at various MOIs, 10, 25, 50 for 1 h, then incubated with RPMI 1640 medium plus 10% FCS for one more hour. The infected cells were trypsinized and suspended in RPMI 1640 medium, then 1×107 cells were injected s c at the flanked fat pad of BALB / C nude mice aged 4-6 wk old. Cell viability was determined with trypan blue under the same condition as in in vitro assay. Three dimesions of tumors (L, W, H) formed in nude mice were measured twice weekly after injection. The volume of tumors was caculated as follow:

V= 4/3 π × (L/2 × W/2 × H/2) 25

Results

A. Effects on QGY -7703 cells by H-1 virus

1. The survival rate of QGY -7703 cells decreased due to H-1 infection

The survival rate became lower at a higher MOI given, either in RPMI 1640 medium (Fig 2A) or in RPMI 1640 medium containing 0.3% agar (Fig 2B), both showing the relation between the amount of virus used and growth inhibition of QGY -7703 cells.

Fig 2
figure 2

H-1 effect on survival of QGY -7703 cells. Survival of QGY -7703 cells infected (• - •) with different MOI 1,5,10 of H-1 and mock infected (○) by colony-formed ability assay (A), and anchorage-independence assay (B). Average valued from 3 experiments, bars represented SEM.

Full size image

2. H-1 DNA amplified in QGY -7703 cells and cellular DNA synthesis decreased

Three kinds of cell samples were collected, including 2 postinfection samples, i e, at 2 h to determine the uptake of viral DNA, and 30 h to indicate the peak of viral DNA amplification and the uninfected control sample. In infected cells, H-1 DNA amplification at 30 h postinfection was 7.4 times higher than that at 2 h (Tab 1). These results showed that QGY -7703 cells supported the viral DNA amplification. Meanwhile, H-1 infection greatly reduced the percentage of S-phses cells in the cell cycle (Tab 2) and cells were hindered in G1 or Go phase, indicating that cellular DNA synthesis was inhibited. The reduction of S-phase cells maintained from 2 h p.i. to 30 h p.i. Apparently, parvoviral DNA replication exhausted DNA replication enzyme system of host cells on a large scale of amplification, and cellular DNA synthesis itself was thus disordered and inhibited.

Tab 1 H-1 DNA amplification in QGY -7703 cells
Full size table
Tab 2 Effect of H-1 virus on QGY -7703 cell cycle
Full size table

3. Percentage of cells releasing infectious progeny virions increased at higher MOI

Infection progeny virions released from QGY -7703 cell formed a plaque on the NB-K monolayer, that stood for an infectious cell center. At a higher MOI given, the percentage of cells releasing progeny virions increased (Tab 3). These results showed that higher MOI infection improved the lytic efficacy to QGY -7703 cells.

Tab 3 Proportion of QGY -7703 cells releasing infectious progeny virus particles after H-1 infection
Full size table

B. Effect of H-1 on the formation and growth of tumors by QGY -7703 cells in nude mice

1. Cell viability in treated and control groups remained at a high level within 2 h postinfection

The QGY -7703 cells were injected into nude mice just 2 h p.i.. The viability of the cells at this time obtained from trypan blue staining did not show any obvious differences among the control and 3 treated groups (Tab 4). Hence, any difference of tumor formation and growth in nude mice among 4 groups could not be due to the culturing and viral infecting steps in vitro.

Tab 4 QGY-7703 cell viability 2 h after H-1 infection
Full size table

2. H-1 virus inhibited the formation and growth of hepatoma by QGY-7703 cells in nude mice

Tumors grew much more rapidly in control group than those in the 3 treated groups (Fig 3). In the control group, tumors expanded at such an exponential rate that the overlying skin became denuded with tumors burgeoning on d 31 after trasplantation. Thus, mice in the control group were killed on that day, with an average tumor weight of 2.3 g. In the MOI 10 group, tumors remained non-palpable throughout the first 20 d, then regrew noticeably at a much lower exponential rate than those in the control group (Fig 3). Tumors did not appear at all in other 2 treated groups, MOI 25 and MOI 50. These results suggested that H-1 virus harbored an obvious oncosuppression in vivo as well as in vitro.

Fig 3
figure 3

Tumor growth curve of QGY-7703 cells in nude mice.

1 × 107 of QGY-7703 cells, mock-infected (• -•) and infected with H-1 at MOI 10 (○ - ○) , were injected s c at flanked fat pad of nude mice. The size of tumors was measured twice weekly after transplantation. Each point averaged from 4 values, bars represented SEM.

Full size image

Higher MOIs of H-1 virus exhibited a more significant inhibitory effect on tumorigenicity as well as on growth. Within the 3 treated groups, MOI 25 or 50 tumor fromation was completely suppressed and no growth occurred in the 56 d period of observation after transplantation. It might be concluded that oncosuppressive effect was greatly affected by virus uptake strategy, i e, the MOI chosen, or the total amount of virions for infection.

Discussion

As far as parvoviral inhibition of tumor growth was concerned, there are 3 possibilities. First, suppression occurred during cell transformation. For instance, adenoassociated virus inhibited the transformation of adenovirus or herpes virus26. Second, an indirect effect by triggering the host immune surveillance system against tumors, including production of immunogenic viral proteins or cytotoxic factors, and inducement of interference factors such as the interferon and tumor necrosis factor27, and /or stimulation of new antigens on cell surface so that they could easily be recognized by the host immune system 28. Finally, a direct cytolytic effect. Our previous studies10, 11 showed that H-1 acts cytolytically on various human cancer cells of different tissues (or origins), but has no obvious inhibitory effect on normal cells of the same origin. Moreover, this viral killing effect on malignant cells is only selective. Recently, Lin et al discovered that the human fetal liver cell line, HuL-1, which had spontaneously transformed to (pre)-neoplastic cells through several scores of passages in our lab, became much more sensitive to H-l infection, thus offered an evidence for selective killing of H-1 (data to be published). In this study, a suppressive effect of H-1 virus on tumorigenicity and growth of hepatoma by QGY -7703 cells in nude mice, as well as inhibition to colony formation in vitro have been demonstrated. Parvovirus H-1 exhibited direct killing effect on hapatoma cells whether in vivo or in vitro. This conclusion was supported by another report, which suggested that the recognition of parvovirus-specific antigens on the surface of infected tumor cells as a target for an increased immune rejection almost certainly does not constitute a major component of parvoviral oncosuppression29.

The mechanism of oncosuppression of H-1 virus indeed needs further study. However, it would become clearer at the DNA replication level. On the one hand, viral DNA amplification was supported by host cells, as was shown above. On the other hand, cellular DNA synthesis deregulated to a much lower level several hours after infection. H-1 virus seemed to interfere with the cellular DNA replication by exhausting the replication enzyme system for its own DNA amplification. This might be one aspect of the so-called direct killing effect. With regard to the viral gene expression in hepatoma cells, recent studies showed that NS-1 protein was accumulated to a higher level in hepatoma cells than that in other non-permissive cells (Lin et al., unpublished data). It seemed that NS-1 protein was cytotoxic, though the metablic mechanisms of this protein on tumor cells remained to be explored.

Analyzing the results from nude mouse experiment, it was interesting to note that the exponential growth rate in the MOI 10 group was much lower than that in the control group. What is its mechanism? We might suppose that cells uninfected or infected beyond S phase may regrow after a latent period; that the progeny virions may attack the hepatoma cells successively; or through a combined effect of both. Further studies on the distribution of virus in tumors and main tissues might be more fruitful.

H-1 virus did not show any potent ability to act as a carcinogen5, 30. Studies on the oncosuppression of H-1 virus might lead to some new approaches in human hepatoma prevention and therapy.