Cidofovir Inhibits Polyomavirus BK Replication ... - Wiley Online Library

American Journal of Transplantation 2008; 8: 1413–1422 Blackwell Munksgaard

C 2008 The Authors C 2008 The American Society of Journal compilation Transplantation and the American Society of Transplant Surgeons

doi: 10.1111/j.1600-6143.2008.02269.x

Cidofovir Inhibits Polyomavirus BK Replication in Human Renal Tubular Cells Downstream of Viral Early Gene Expression E. Bernhoffa,b , T. J. Gutteberga,b , K. Sandvika , H. H. Hirschc,d and C. H. Rinaldoa, ∗ a Department of Microbiology and Infection Control, University Hospital of North Norway, Tromsø, Norway b Department of Microbiology and Virology, Institute for Medical Microbiology, University of Tromsø, Tromsø, Norway c Transplantation Virology, Institute for Medical Microbiology, Department of Biomedicine, University of Basel, Basel, Switzerland d Infectious Diseases & Hospital Epidemiology, University Hospital Basel, Basel, Switzerland ∗ Corresponding author: Christine H. Rinaldo, [email protected]

The human polyomavirus BK (BKV) causes nephropathy and hemorrhagic cystitis in kidney and bone marrow transplant patients, respectively. The anti-viral cidofovir (CDV) has been used in small case series but the effects on BKV replication are unclear, since polyomaviruses do not encode viral DNA polymerases. We investigated the effects of CDV on BKV(Dunlop) replication in primary human renal proximal tubule epithelial cells (RPTECs). CDV inhibited the generation of viral progeny in a dose-dependent manner yielding a 90% reduction at 40 lg/mL. Early steps such as receptor binding and entry seemed unaffected. Initial large T-antigen transcription and expression were also unaffected, but subsequent intra-cellular BKV DNA replication was reduced by >90%. Late viral mRNA and corresponding protein levels were also 90% reduced. In uninfected RPTECs, CDV 40 lg/mL reduced cellular DNA replication and metabolic activity by 7% and 11% in BrdU and WST-1 assays, respectively. BKV infection increased DNA replication to 142% and metabolic activity to 116%, respectively, which were reduced by CDV 40 lg/mL to levels of uninfected untreated RPTECs. Our results show that CDV inhibits BKV DNA replication downstream of large T-antigen expression and involves significant host cell toxicity. This should be considered in current treatment and drug development. Key words: BK nephropathy, BK polyoma virus, BK virus, BK virus allograft nephropathy, BK virus nephritis, bone marrow transplantation, cidofovir, kidney allograft, kidney transplantation, proximal tubule, real time polymerase chain reaction (PCR), real time RTPCR, viral infection, viral load, viral therapy

Received 08 February 2008, revised 13 March 2008 and accepted for publication 17 March 2008

Introduction Polyomavirus BK (BKV) infects up to 90% of the general population and establishes latent infection in the renourinary tract (1). Spontaneous reactivation and shedding of BKV have been observed in 5–10% of immunocompetent individuals. In immunosuppressed individuals, the rate of urinary shedding increases to 20–60% with viral loads > 107 copies/mL (1,2). Two major diseases have been associated with BKV, namely polyomavirus-associated nephropathy after kidney transplantation and hemorrhagic cystitis following bone marrow transplantation (1). Polyomavirus-associated nephropathy results from high-level replication of BKV in renal tubular epithelial cells and leads to reduced kidney function and graft loss (3–6). The primary treatment is reduction of immunosuppression although this might increase the risk of graft rejection (7). Hemorrhagic cystitis is associated with high-level replication of BKV in urothelial cells, and current treatment is symptomatic with pain relief and bladder irrigation. There are no drugs with proven efficacy for BKV replication (8). Cidofovir (CDV), a nucleoside analogue of deoxy cytidine mono phosphate (dCMP), has been sporadically used for the treatment of BKV replication in patients with nephropathy or hemorrhagic cystitis (8–12) However, it is not clear if the usually favorable outcome of these reports can be attributed to the anti-polyomavirus activity of the drug or rather involved recovery of BKV-specific immunity. CDV is licensed to treat ganciclovir resistant cytomegalovirus (CMV) retinitis in patients with HIV/AIDS (13), but an important drawback of systemic CDV use is its nephrotoxicity (1,14–19). So far, only two in vitro studies have examined the effect of CDV on BKV replication. The first study of human embryonic lung fibroblasts (WI-38 cells) found that CDV concentrations of 36.3 ± 11.7 lg/mL reduced intra-cellular BKV DNA levels by 50% after 7 days, while higher CDV concentrations of 63.9 ± 17.2 lg/mL also reduced cellular DNA by 50% (20). In the second study, seven BKV isolates from urine were inoculated onto human embryonic lung fibroblasts (HEL) and scoring the microscopic appearance of 1413

Bernhoff et al.

cytopathic effects until 30 days. At lower BKV inocula, termed 10 × TCID 50 , CDV concentrations of 31.25 lg/mL inhibited the cytopathic effect of five isolates, whereas two isolates required higher CDV concentrations (21). At higher BKV inocula, only two strains seemed inhibited by CDV. In herpes and poxvirus the CDV inhibition results from the diphosphorylated CDV, which has higher affinity for the viral DNA polymerase than for the cellular enzyme (22–24) and blocks excision repair of the growing viral genome. The effects of CDV on BKV are less clear since polyomaviruses do not encode a viral DNA polymerase. Here we report on a detailed study of the effects on CDV on the BKV lifecycle in primary human renal proximal tubule epithelial cells (RPTECs) as well as the effects on host cell DNA replication and metabolism.

Materials and Methods

by a robot (GenoM-48, Qiagen, www.qiagen.com). Intra-cellular DNA was extracted from cells using the Qiagen DNA mini kit. Cells were first washed, trypsinized and pelleted at 220 g for 10 min, then resuspended in phosphatebuffered saline (PBS) and frozen at −80◦ C until extraction.

Western blotting Cells were lysed in Cell Disruption buffer (mirVana PARIS kit, Ambion) and stored at −80◦ C until separation on SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were transferred onto Immobilon-FL PVDF membrane (Millipore, www.millipore.com), blocked with blocking buffer (Odyssey, www.licor.com) and incubated with polyclonal rabbit sera antilarge T-antigen (1:2000) (26), anti-VP1 (1: 10000) (25) and anti-agnoprotein (1:10000) (26) followed by anti-rabbit infrared dye-labeled secondary antibodies (IR Dye 800, Rockland, www.rockland-inc.com) (1:5000) before detection with Licor Odyssey Infrared detection system. Membranes were stripped in NewBlot Stripping Buffer PVDF (Odyssey) before reprobing with monoclonal anti-GAPDH (1:5000) (ab8245, Abcam, www.abcam.com) and anti-mouse (Alexa Fluor 680, Invitrogen, www.invitrogen.com).

Cell culture and viruses

Immunofluorescence staining

Primary human renal proximal tubule epithelial cells (RPTECs) (Lonza, www.lonzabioscience.com) were propagated as described by the manufacturer. All experiments were performed with RPTECs passage 4. BKV(Dunlop) supernatants from infected Vero cells were used with a multiplicity of infection (MOI) determined on HUV-EC-C (ATCC CRL1730) (25).

Three days post-infection, cells were washed in PBS, fixed in methanol and blocked with 3% serum in goat-serum in PBS for 30 min and then treated as earlier described (27). In addition, we used polyclonal rabbit serum anti-VP1 (1:800).

Infection and CDV-treatment About ∼60% confluent RPTECs were infected with BKV(Dunlop) at MOI of ∼1. After 2-h incubation at 37◦ C the virus was replaced with fresh medium with or without CDV (VISTIDE 75 mg/mL, anhydrous CDV, Pfizer Enterprises SARL, Luxembourg). CDV was freshly diluted in medium to 5, 10, 20, 40 and 80 lg/mL and added to BKV-infected and uninfected RPTECs.

BrdU and WST-1 assays DNA synthesis was quantified by the colorimetric measurement of BrdU incorporation into DNA in proliferating cells using the ‘Cell proliferation ELISA, BrdU’ kit (Roche, www.roche-applied-science.com). The metabolic activity was monitored by the colorimetric WST-1 assay (Roche) of the mitochondrial dehydrogenases in viable cells. For both assays, RPTECs were seeded in 96 well plates and infected with BKV (Dunlop) or mock infected as described above. BrdU incorporation and WST-1 cleavage was measured at 24, 48, 72 and 96 h p.i. according to the manufacturer’s protocol. Absorbance at 370 nm (sample) and at 450 nm (background) was determined 15 min after addition of the substrate (Spectramax plus, Molecular Devices). The WST-1 cleavage product was measured at 450 nm (sample) and at 650 nm (background). WST-1 plus medium alone served as blank.

RNA extraction and cDNA synthesis Total cellular RNA was harvested 24, 48, 72 and 96 h p.i. using the mirVana PARIS kit (Ambion, Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. RNA samples were treated with DNase turbo (Ambion) to remove residual DNA before RNA concentration and assayed by Nanodrop (Nanodrop Technologies, Thermo Fisher Scientific, Waltham, MA), and by microcapillary electrophoresis on a Bioanalyser 2100 (Agilent Technologies, Santa Clara, CA). cDNA was generated from 280 ng RNA per sample using the High Capacity cDNA Kit (Applied Biosystems).

DNA extraction To assay extracellular DNA, cell culture supernatants were harvested at 24, 48, 72 and 96 h p.i. and frozen in −80◦ C until automatic extraction

1414

Quantitative PCR for BKV DNA and cellular gene detection To quantify intra-cellular or extracellular BKV DNA load, a quantitative PCR (qPCR) with primers and probe targeting the large T-antigen gene was used (Table 1). Reactions were performed in a total volume of 25 lL using 2× Taqman Fast Universal PCRmaster mix (Applied Biosystems), 5 lL DNA template, 300 nm of each primer and 125 nm probe. For normalization of intra-cellular BKV DNA each sample was analyzed in parallel by the qPCR for the gene for aspartoacylase (ACY) (28) to correct for cellular DNA (Table 1). For ACY qPCR, 300 nm of each primer, and 250 nm probe were used. Amplification was performed in 96 well PCR plates using the 7500 HT cycler (Applied Biosystems) with the following PCR program: denaturation at 95◦ C, 20 and 45 cycles of 95◦ C, 3 ; 60◦ C, 30 .

Quantitative reverse transcriptase PCR for BKV and host genes To quantify BKV early gene (T-antigen) and late (VP1 and agnoprotein) gene transcription expression, reverse transcription-quantitative PCR (RT-qPCR) was performed (Table 1). RT-qPCR was performed in 25 lL using 2× Taqman Fast Universal PCR master mix, 5 lL cDNA template and 20× Expression Assay Mix (Applied Biosystems). Results are presented as the changes (n-fold) in T-antigen- or late-transcript levels, compared to untreated sample 24 h p.i. The 2−C(T) method (29) were used for normalization to the levels of endogenous control transcript human Hypoxanthine PhosphoRibosylTranserase (huHPRT).

Endogenous control To identify host cell control transcripts, we used the TaqMan Endogenous Control Plate (Applied Biosystems) with 11 human housekeeping genes. cDNA were made from: (a) BKV-infected CDV (80 lg/mL) treated RPTECs 24 h p.i., (b) BKV-infected CDV (80 lg/mL) treated RPTECs 48 h p.i., (c) BKVinfected untreated RPTECs 48 h p.i. and (d) Uninfected untreated RPTECs 48 h p.i. used as calibrator sample. Amplification was performed in duplicate according to the manufacturer’s protocol. The results were presented as Ct and 5 genes had Ct < than 0.5 with huHPTR showing the least variation.

American Journal of Transplantation 2008; 8: 1413–1422

Cidofovir Inhibition of BKV Replication in Kidney Cells Table 1: Primers and probes used Gene/ Transcript T-antigen

Asparto-acylase (ACY)

T-antigen cDNA

VP1/Agno cDNA

huHPRT cDNA

Primer and probe sequences (5 to 3 )

Dye

F: TAG GTG CCA ACC TAT GGA ACA GA R: RMR GGA AAG TCY TTR GGG TCT TCT ACC P: TGT TGA GTG TTG AGA ATC TGC TGT TGC TTC TTC F: CCC TGC TAC GTT TAT CTG ATT GAG R: CCC ACA GGA TAC TTG GCT ATG G P: CCT TCC CTC AAA TAT GCG ACC ACT CG F: ACT CCC ACT CTT CTG TTC CAT AGG R: TCA TCA GCC TGA TTT TGG AAC CT P: TTG GCA CCT CTG AGC TAC F: GGC CTC TTT GTA AAG CTG ATA GC R: CTG TTG TGT TCC AGA GCT GTT AGT A P: CAG CTG CTG ATA TTT G P: TGG TCA AGG TCG CAA GCT TGC TGG T

FAM

Fragment length 185 bp

TAMRA

77 bp

FAM

68 bp

FAM

83 bp

FAM

Reference

(28)

www.appliedbiosystems.com

F = forward primer; R = reverse primer; P = probe.

Results Determination of CDV inhibitory concentration IC90 To examine the replication cycle of BKV(Dunlop) strain in RPTECs, cells were infected at MOI ∼1.0. Supernatants and cells were harvested 2–96 h p.i., and the corresponding extracellular and intra-cellular BKV load was measured. Increases of intra-cellular BKV load were detected from 24 h p.i onward while extracellular BKV loads increased from 48 h p.i. (Figure 1A). Thus, completion of the first lifecycle of BKV(Dunlop) in RPTECs takes approximately 48–72 h. To investigate the effect of CDV on BKV progeny, increasing concentrations of CDV were added 2 h p.i. and supernatants harvested at 72 h p.i. We observed that CDV reduced the extracellular BKV load in a concentrationdependent manner (Figure 1B). CDV 5 lg/mL reduced the BKV load by only 66% or 0.47 log. CDV 40 lg/mL reduced the BKV load by an average of 90% therefore defining the inhibitory concentration IC90. Immunofluorescence staining at 72 h p.i. demonstrated less agnoprotein expressing cells at 40 and 80 lg/mL, whereas 5 lg/mL had no discernable effect compared to untreated cells (Figure 1C). To investigate whether or not the inhibitory effect of CDV was dependent on levels of input virus, RPTECs were infected with decreasing MOI of BKV and BKV loads in supernatants 72 h p.i. were examined. Diluting the input MOI over three orders of magnitude, a reduction of 94–98% was observed (Figure 1D). Effects of CDV on BKV-infected and uninfected RPTECs Inspection of RPTECs by phase contrast microscopy did not reveal any signs of impaired viability during the 4-day exposure to CDV 40 lg/mL. To use more sensitive assays we investigated host cell DNA replication and metabolic activity using BrdU incorporation and WST-1 assays in BKVAmerican Journal of Transplantation 2008; 8: 1413–1422

infected and uninfected RPTECs. Compared to uninfected untreated RPTECs, BKV infection by itself increased cellular DNA replication and metabolic activity to 142 and 116%, respectively (Figure 2A). Addition of CDV reduced both DNA replication and metabolic activity of BKV-infected RPTECs in a concentration-dependent manner (Figure 2A). Compared to BKV-infected untreated RPTECs, CDV at 5 to 80 lg/mL decreased DNA replication by 13% to 33%, respectively. CDV concentrations up to 20 lg/mL did not affect the metabolic activity, but a 10–40% decrease was seen with 20–80 lg/mL. Of note, CDV 40 lg/mL reduced DNA replication in BKV-infected cells by 36% and metabolic activity by 23%, to activities similar to untreated uninfected cells. In uninfected cells, CDV 40 lg/mL also reduced BrdU incorporation and metabolic activity with 7 and 11%, respectively. When we measured BrdU incorporation and metabolic activity from 24–96 h p.i., DNA replication was decreased at all time points, while effects on metabolic activity became prominent after 48 h p.i. (Figure 2B). Taken together, CDV significantly reduced host cell replicative and metabolic activity. Effect of CDV on early steps of the BKV lifecycle To investigate whether CDV affected early steps of the BKV lifecycle, CDV 40 lg/mL was added 2 h before, together with and 2 h after adding BKV to RPTECs. No significant difference in the extracellular BKV loads were observed at 72 h p.i. (Figure 3) suggesting that CDV did not significantly interfere with early steps such as receptor binding and viral entry. To study the effects of CDV on BKV early gene expression, we compared T-antigen mRNA levels in CDV-treated and untreated RPTECs at 24–96 h p.i. by RT-qPCR. Large T-antigen mRNA levels were normalized to the housekeeping gene huHPRT and presented as changes (n-fold) of the levels found in untreated cells at 24 h p.i. No differences in T-antigen transcripts were seen in CDV-treated 1415

Bernhoff et al.

Figure 1: BKV (Dunlop) and CDV titration. (A) Characterization of the life cycle of BKV(Dunlop) in RPTEC. Supernatants and cells were harvested 2–96 h p.i. and extra- and intra-cellular BKV load was measured by qPCR. The 2-h time point show input virus. Data are presented as BKV load in Geq/mL (Geq=genome equivalents). (B) Effect of increasing concentrations of CDV on BKV load. RPTEC supernatants were harvested 72 h p.i., that is 70 h post start of treatment with indicated CDV concentrations. BKV load was measured by qPCR and DNA load in untreated cells (1.67E + 09 Geq/mL) was set as 100%. Log Geq/mL is shown above the bars. (C) Effect of increasing concentrations of CDV on BKV-agnoprotein expression. Indirect immunofluorescence of BKV-infected RPTECs methanol fixed 72 h p.i. and stained with rabbit anti-agnoprotein serum (green) for visualization of the late agnoprotein (a) untreated, (b) CDV 5 lg/mL, (c) CDV 40 lg/mL and (d) CDV 80 lg/mL. (D) Influence of CDV 40 lg/mL on BKV-infection using decreasing amounts of input virus. Supernatants from CDV-treated and untreated BKV-infected RPTECs were harvested at 72 h p.i and extracellular BKV load measured by qPCR. Data are presented as BKV load in Geq/mL.

and untreated cells at 24 h p.i. At 48 and 72 h p.i., a moderate reduction of 2.3- and 1.5-fold of T-antigen transcripts was noted, respectively (Figure 4A). When protein levels of large T-antigen were compared, we found no differences of T-antigen levels in CDV-treated and untreated cells 24 h p.i. At 48 and 72 h p.i., however, a five- and three-fold reduction was seen (Figure 4B). These results indicated that initial large T-antigen expression up to 24 h p.i. was not affected by CDV. From 48 h to 72 h p.i., however, a moderate reduction in viral early gene expression became discernible. We concluded that CDV must decrease BKV progeny in supernatants at a step downstream of the initial BKV early gene expression.

sured intra-cellular BKV load at 24–96 h p.i. by qPCR. The intra-cellular BKV load was normalized to the cell number using the aspartoacylase (ACY) gene as described (30). Compared to untreated RPTECs, CDV 40 lg/mL reduced the intra-cellular BKV load by more than 99% at 48 h, 72 h and 96 h p.i., whereas only a small effect was seen at 24 h p.i. (Figure 5). Thus, intra-cellular BKV genome replication is the first step of the BKV lifecycle where we could identify a significant inhibitory effect of CDV. This step is known to require large T-antigen expression, which also increases viral late gene expression by increasing the DNA templates for late gene transcription and by activating transcription from the late promoter (31).

Effect of CDV on BKV genome replication To investigate whether the next step of the BKV lifecycle BKV genome replication was affected by CDV, we mea-

CDVs effects on BKV late transcription and expression To study the effect of CDV on BKV late gene transcription, VP1 and agnoprotein mRNA expression was measured by

1416


Cidofovir Inhibition of BKV Replication in Kidney Cells

Figure 2: Influence of CDV on DNA replication and metabolic activity of BKVinfected and uninfected RPTECs. Cellular DNA replication was examined with BrdU incorporation and metabolic activity as WST1 cleavage. Absorbance was measured as described in Material and Methods. (A) Medium with indicated CDV concentrations was added 2 h p.i. and absorbance measured 72 h p.i. Absorbance for untreated uninfected cells was set as 100%. (B) Medium with or without CDV 40 lg/mL was added 2 h p.i. and absorbance measured at indicated time points. Absorbance for untreated BKVinfected cells at each time point was set as 100%.

RT-qPCR 48–96 h p.i. and normalized to huHPRT transcripts as described above. The primers and probe for late mRNA expression are located in the VP1 gene but detect mRNA for VP1 and agnoprotein because the transcripts are bici-

ctronic. In untreated cells, late transcripts increased from 24 h p.i. more than 1000-, 5000- and 8000-fold at 48 h, 72 h and 96 h p.i., respectively (Figure 6A). With CDV 40 lg/mL treatment the corresponding increase was only 100-,

Figure 3: Effect of CDV 40 lg/mL addition before, simultaneously and after BKV adsorbtion. Supernatants were harvested 72 h p.i. and BKV load measured by qPCR. Data are presented as BKV load in Geq/mL.


1417

Bernhoff et al.

Figure 4: Influence of CDV 40 lg/mL on BKV early transcripts and early expression. (A) RNA was extracted from CDV-treated and untreated BKV-infected RPTECs at indicated time points. T-antigen mRNA expression was measured by RTqPCR and normalized to huHPRT transcripts. Results are presented as the changes (n-fold) in T-antigen transcript levels, with the level in untreated sample 24 h arbitrarily set to 1. (B) Western blot on cell extracts from untreated and CDV-treated BKV-infected RPTECs at indicated time points was performed with polyclonal rabbit antilarge T-antigen serum. The lowest band is caused by cross-reaction to a cellular protein.

700- and 800-fold. Thus the expression was 85–90% reduced in CDV-treated cells. When we examined viral late protein expression, a significant reduction of VP1 and agnoprotein was observed in CDV-treated RPTECs. At 72 h p.i., VP1 and agnoprotein were both found to be 90% reduced in CDV-treated cells compared to untreated controls (Figure 6B). We conclude that CDV significantly reduces late protein expression. Morphologic examination of BKV early and late gene expression in CDV-treated RPTEC Given the significant difference between early and late gene expression, we examined large T-antigen, agnopro-

tein and VP1 expression at the single-cell level by immunofluorescence staining at 72 h p.i. We found a slight reduction in large T-antigen positive cells of 8% compared to a large reduction of 90% in agnoprotein-positive cells (Figure 6C). Similar results were found for VP1 staining (results not shown). Thus, the overall inhibition of CDV on BKV late gene expression was detectable at the single-cell level. Interestingly, immunofluorescence revealed some cells in the CDV treated culture expressing agnoprotein at levels comparable to untreated cells. Even when increasing the CDV concentration up to 300 lg/mL, some cells seemed refractory to exposure of CDV.

Figure 5: Influence of CDV 40 lg/mL on BKV genome replication. CDV-treated and untreated BKVinfected RPTECs were harvested at indicated time points and intra-cellular BKV DNA load per cell was measured by qPCR as described in Material and Methods.

1418



Figure 6: Influence of CDV 40 lg/mL on BKV late transcription and late expression. (A) RNA was extracted from CDV-treated and untreated BKV-infected RPTECs at indicated time points and late mRNA expression (VP1 and agno) measured by RT-qPCR and normalized to huHPRT transcripts. Results are presented as the changes (n-fold) in T-antigen transcript levels, with the level in untreated sample 24 h arbitrarily set to 1. (B) Cell extracts from CDV-treated and untreated BKV-infected RPTECs at indicated time points were harvested and western blot performed with rabbit anti-BKV VP1 and anti-agnoprotein serum. The membrane was stripped and reprobed with a monoclonal antibody directed against the housekeeping protein GAPDH. (C) Indirect immunofluorescence of BKV-infected RPTECs methanol fixed 72 h p.i. and stained with rabbit anti-agnoprotein serum (green) for visualization of the late agnoprotein. The early protein T-antigen (red) is shown by staining with the SV40 large T-antigen monoclonal Pab416. Left: untreated; right: treated with CDV 40 lg/mL.

Time course of CDV on extracellular BKV load To examine the effect of CDV on BKV progeny over time, supernatants of treated and untreated cells were harvested at the indicated time points. The reduction of BKV loads in supernatants reached 95% (1.74 × 109 Geq/mL) at 96 h p.i. (Figure 7A). We next examined whether or not CDV had a persistent or transient effect on BKV replication in RPTECs. Removal of CDV containing medium after 24 h reduced BKV loads in the 96-h supernatants by only 0.89 log Geq/mL. CDV removal after 48 h caused a reduction of 1.62 log Geq/mL (Figure 7B) while 96-h treatment gave a 1.98 log Geq/mL reduction. We concluded that CDV exposure the first 24 h had only a limited inhibitory effect that further increased when the drug was present for 48 h or 96 h.


Discussion High-level BKV replication is the common hallmark of hemorrhagic cystitis and nephropathy jeopardizing the outcome of hemopoietic stem cell transplantation and kidney transplantation in 1–10% of affected patients. The urgent need of anti-viral drugs, has led to the tentative use of CDV in small uncontrolled case series. However, given the potential for serious, partly irreversible side effects, particularly for kidneys, there is a clear need to better characterize the effects of CDV on BKV replication in clinically relevant experimental systems. In this study, we report that CDV inhibits BKV replication in primary human renal tubular epithelial cells at a step downstream of large T-antigen expression and is associated with significant host cell toxicity.

1419

Bernhoff et al.

about 90%. Importantly, immunofluorescent staining allowed the analysis of viral protein expression at the singlecell level indicating that the nuclear large T-antigen expression was largely similar to untreated cells, whereas the late agnoprotein expression was virtually undetectable in at least 90% of treated cells. Interestingly, about one-tenth of CDV-treated cells seemed refractory to inhibition and this could not be overcome by higher concentrations. The underlying mechanism of this seemingly stochastic resistance is presently unclear and requires further study. Most likely, these refractory cells are at least partly responsible for generating residual BKV progeny, at least in tissue culture.

Figure 7: Influence of CDV on extracellular BKV load. (A) Effect of CDV 40 lg/mL over time. Supernatants from CDV-treated and untreated BKV-infected RPTECs were harvested at indicated time points after infection and BKV load measured by qPCR. Data are presented as BKV load in Geq/mL. (B) Effect of CDV wash-out. BKV-infected RPTECs were either untreated, continuously treated with CDV 40 ug/mL from 2 h p.i. or medium was replaced with medium without CDV 24 or 48 h p.i. Supernatants were harvested 96 h p.i. and extracellular BKV load measured by qPCR. Data are presented as BKV load decrease in log Geq/mL.

With regard to BKV replication, we found that CDV at 40 lg/mL caused a 90% reduction of extracellular BKV loads. Presence of CDV 2 h before and during incubation of RPTECs with BKV did not affect rate of inhibition suggesting that early steps of the viral replication cycle as receptor binding and entry were not affected. In fact, our detailed study of the BKV replication cycle revealed that the initial early gene expression up to 24 h p.i. was not significantly reduced by CDV as measured by large T-antigen mRNA and protein levels. However, CDV significantly inhibited intracellular BKV genome replication. This step occurs after 24 h p.i. and is dependent on prior large T-antigen expression providing DNA helicase functions as well as recruiting host cell replication factors. This notion was also supported by wash-out experiments indicating that the presence of CDV was required for more than 24 h to reach near-maximal inhibition of the first replication cycle. In line with the known role of amplification of DNA templates for late viral gene transcription by viral DNA replication (31), CDV reduced overall late mRNA and proteins levels by 1420

With regard to the host cells, we observed that CDV significantly inhibited the overall cellular DNA replication and metabolic activity. This effect was more pronounced in BKV-infected than in uninfected RPTECs. Interestingly, both, DNA replication and metabolic activity was increased following BKV infection which is a consequence of large T-antigen inactivating the tumor-suppressor proteins Rb and p53 and stimulating cell progression into S-phase (32). Thus, CDV at 40 lg/mL decreased BrdU and WST-1 activities to levels seen in uninfected untreated cells. These data document significant toxicity of CDV on human RPTECs, which was not visible by microscopic evaluation. Part of the toxic effects by CDV are likely to come from its incorporation into host cell genomic DNA as reported in a study of human papillomaviruses that, akin to BKV, lack a viral DNA polymerase (33). The IC90 of 40 lg/mL in RPTECs is similar to the IC50 reported for the inhibition of BKV replication in lung fibroblast lines (20,21). We suspect that this is partly the result of the active transport system for organic compounds present in RPTEC (14–18) yielding higher intra-cellular CDV concentration than in lung fibroblasts. It is not clear if CDV has a similar effect on DNA replication and metabolic activity in lung fibroblasts, but our study emphasizes the importance of choosing pathologically relevant in vitro models. In view of the IC90 of 40 lg/mL, do our findings support clinical use of CDV? In a clinical study by Kuypers and colleagues (34), CDV peak concentrations in serum of eight nephropathy patients treated with reduced immunosuppression and low-dose CDV (0,5 mg/kg or 1,0 mg/kg), were found to be 0,77–3,07 lg/mL. This is 13- to 52-times less than the IC90 determined here in vitro. However, active uptake of CDV into tubular epithelial cells may allow for some increased intra-cellular concentration over what is found in serum. Assuming that uptake is similarly active in RPTECs in vitro, we note that we observed no relevant inhibition of BKV replication or late gene expression at CDV concentration 5 lg/mL. We wonder, therefore, how much of the favorable results in that study can in fact be attributed to CDV. When CDV was administered to HIV patients at 7.5 mg/kg with concomitant high-dose probenicid, maximal CDV concentrations of 43 lg/mL could be American Journal of Transplantation 2008; 8: 1413–1422


reached, but probenicid may prevent uptake into tubular epithelial cells (35). Can a 90% inhibition be sufficient to clear the virus in view of the high-level BKV replication of 107 –1010 copies/mL in patients with nephropathy or with hemorrhagic cystitis? Results from mathematical modeling of BKV replication in urothelial and renal tubular epithelial cells suggest that a more than 80% reduction of renal BKV replication must be maintained for up to 10 weeks to observe clearing of plasma and urine viral loads (36). Based on these data, long-term maintenance of inhibitory CDV concentrations may be necessary. Thus, we see at least three caveats. First, long-term exposure and intra-cellular accumulation increases the risk of nephrotoxicity. Even though uninfected RPTECs seem less affected than BKV-infected cells, considerable toxicity must be expected from CDV acting on proliferating tubular epithelial cells replacing the approximately 106 –107 cells lost each day due to fast BKV replication dynamics (37). Second, the immunofluorescence data shown here revealed that CDV inhibition was only partial, with some cells being completely refractory which may allow for continuous replication and emerging resistance. Third, in cells with inhibited BKV replication, the large T-antigen was expressed and these cells may continue the BKV lifecycle once CDV concentrations become insufficient as observed in our wash-out experiments. Transient inhibition by CDV has been reported for papillomavirus (33), but not for herpes and vacciniavirus, which encode viral DNA polymerases (23,24,38). These caveats do not preclude the clinical use of CDV, but should be factored into the assessment of the drug. In conclusion, our study of CDV effects on BKV infection of primary human RPTECs demonstrate inhibition of BKV DNA replication downstream of initial large T-antigen expression with significant host cell toxicity. This needs to be considered in current treatment strategies and the development of future drugs.

Acknowledgments This research was supported by ‘Fondet for forskning om nyresykdommer og organtransplantasjon, Landsforeningen for nyrepasienter og transplanterte’, Norway.

References 1. Hirsch HH, Steiger J. Polyomavirus BK. Lancet Infect Dis 2003; 3: 611–623. 2. Polo C, Perez JL, Mielnichuck A, Fedele CG, Niubo J, Tenorio A. Prevalence and patterns of polyomavirus urinary excretion in immunocompetent adults and children. Clin Microbiol Infect 2004; 10: 640–644. 3. Randhawa PS, Finkelstein S, Scantlebury V et al. Human polyoma virus-associated interstitial nephritis in the allograft kidney. Transplantation 1999; 67: 103–109.


4. Hirsch HH, Knowles W, Dickenmann M et al. Prospective study of polyomavirus type BK replication and nephropathy in renaltransplant recipients. N Engl J Med 2002; 347: 488–496. 5. Ramos E, Drachenberg CB, Portocarrero M et al. BK virus nephropathy diagnosis and treatment: Experience at the University of Maryland Renal Transplant Program. Clin Transpl 2002;143–153. 6. Fishman JA. BK virus nephropathy–polyomavirus adding insult to injury. N Engl J Med 2002; 347: 527–530. 7. Hirsch HH, Brennan DC, Drachenberg CB et al. Polyomavirusassociated nephropathy in renal transplantation: Interdisciplinary analyses and recommendations. Transplantation 2005; 79: 1277– 1286. 8. Rinaldo CH, Hirsch HH. Antivirals for the treatment of polyomavirus BK replication. Expert Rev Anti Infect Ther 2007; 5: 105–115. 9. Benavides CA, Pollard VB, Mauiyyedi S, Podder H, Knight R, Kahan BD. BK virus-associated nephropathy in sirolimus-treated renal transplant patients: Incidence, course, and clinical outcomes. Transplantation 2007; 84: 83–88. 10. Rajpoot DK, Gomez A, Tsang W, Shanberg A. Ureteric and urethral stenosis: A complication of BK virus infection in a pediatric renal transplant patient. Pediatr Transplant 2007; 11: 433–435. 11. Cheerva AC, Raj A, Bertolone SJ, Bertolone K, Silverman CL. BK virus-associated hemorrhagic cystitis in pediatric cancer patients receiving high-dose cyclophosphamide. J Pediatr Hematol Oncol 2007; 29: 617–621. 12. Andrei G, Fiten P, Goubau P et al. Dual infection with polyomavirus BK and acyclovir-resistant herpes simplex virus successfully treated with cidofovir in a bone marrow transplant recipient. Transpl Infect Dis 2007; 9: 126–131. 13. De Clercq E. Acyclic nucleoside phosphonates: Past, present and future. Bridging chemistry to HIV, HBV, HCV, HPV, adeno-, herpes-, and poxvirus infections: The phosphonate bridge. Biochem Pharmacol 2007; 73: 911–922. 14. Ho ES, Lin DC, Mendel DB, Cihlar T. Cytotoxicity of antiviral nucleotides adefovir and cidofovir is induced by the expression of human renal organic anion transporter 1. J Am Soc Nephrol 2000; 11: 383–393. 15. Izzedine H, Launay-Vacher V, Deray G. Renal tubular transporters and antiviral drugs: An update. AIDS 2005; 19: 455–462. 16. Ortiz A, Justo P, Sanz A et al. Tubular cell apoptosis and cidofovirinduced acute renal failure. Antivir Ther 2005; 10: 185–190. 17. Cihlar T, Ho ES, Lin DC, Mulato AS. Human renal organic anion transporter 1 (hOAT1) and its role in the nephrotoxicity of antiviral nucleotide analogs. Nucleosides Nucleotides Nucleic Acids 2001; 20: 641–648. 18. Miller DS. Nucleoside phosphonate interactions with multiple organic anion transporters in renal proximal tubule. J Pharmacol Exp Ther 2001; 299: 567–574. 19. Wolf DL, Rodriguez CA, Mucci M, Ingrosso A, Duncan BA, Nickens DJ. Pharmacokinetics and renal effects of cidofovir with a reduced dose of probenecid in HIV-infected patients with cytomegalovirus retinitis. J Clin Pharmacol 2003; 43: 43–51. 20. Farasati NA, Shapiro R, Vats A, Randhawa P. Effect of leflunomide and cidofovir on replication of BK virus in an in vitro culture system. Transplantation 2005; 79: 116–118. 21. Leung AY, Chan MT, Yuen KY et al. Ciprofloxacin decreased polyoma BK virus load in patients who underwent allogeneic hematopoietic stem cell transplantation. Clin Infect Dis 2005; 40: 528–537. 22. Magee WC, Aldern KA, Hostetler KY, Evans DH. Cidofovir and (S)HPMPA are highly effective inhibitors of vaccinia virus DNA polymerase when incorporated into the template strand. Antimicrob Agents Chemother 2008; 52: 586–597.

1421

Bernhoff et al. 23. Ho HT, Woods KL, Bronson JJ, De BH, Martin JC, Hitchcock MJ. Intracellular metabolism of the antiherpes agent (S)-1-[3-hydroxy2-(phosphonylmethoxy)propyl]cytosine. Mol Pharmacol 1992; 41: 197–202. 24. Neyts J, Snoeck R, Schols D, Balzarini J, De Clercq E. Selective inhibition of human cytomegalovirus DNA synthesis by (S)-1-(3hydroxy-2-phosphonylmethoxypropyl)cytosine [(S)-HPMPC] and 9(1,3-dihydroxy-2-propoxymethyl)guanine (DHPG). Virology 1990; 179: 41–50. 25. Grinde B, Gayorfar M, Rinaldo CH. Impact of a polyomavirus (BKV) infection on mRNA expression in human endothelial cells. Virus Res 2007; 123: 86–94. 26. Hey AW, Johnsen JI, Johansen B, Traavik T. A two fusion partner system for raising antibodies against small immunogens expressed in bacteria. J Immunol Methods 1994; 173: 149–156. 27. Leuenberger D, Andresen PA, Gosert R et al. Human Polyomavirus type 1 (BK virus) Agnoprotein is abundantly expressed, but immunologically ignored. Clin Vaccine Immunol 2007; 14: 959–968. 28. Randhawa PS, Vats A, Zygmunt D et al. Quantitation of viral DNA in renal allograft tissue from patients with BK virus nephropathy. Transplantation 2002; 74: 485–488. 29. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001; 25: 402–408. 30. Randhawa P, Shapiro R, Vats A. Quantitation of DNA of polyomaviruses BK and JC in human kidneys. J Infect Dis 2005; 192: 504–509.

1422

31. Cole CN. Polyomavirinae: The viruses and their replication. In: Fields BN, Knipe DM, Howley PH, eds. Fields Virology. New York: Lippincott-Raven, 1996: 1997–2043. 32. Harris KF, Christensen JB, Imperiale MJ. BK virus large T antigen: Interactions with the retinoblastoma family of tumor suppressor proteins and effects on cellular growth control. J Virol 1996; 70: 2378–2386. 33. Spanos WC, El-Deiry M, Lee JH. Cidofovir incorporation into human keratinocytes with episomal HPV 16 results in nonselective cytotoxicity. Ann Otol Rhinol Laryngol 2005; 114: 840–846. 34. Kuypers DR, Vandooren AK, Lerut E et al. Adjuvant low-dose cidofovir therapy for BK polyomavirus interstitial nephritis in renal transplant recipients. Am J Transplant 2005; 5: 1997–2004. 35. Cundy KC, Petty BG, Flaherty J et al. Clinical pharmacokinetics of cidofovir in human immunodeficiency virus-infected patients. Antimicrob Agents Chemother 1995; 39: 1247–1252. 36. Hirsch HH, Gosert R, Funk GA. The 3rd International Transplant Infectious Disease Conference September 29–30, 2007, Prague, Czech Republic, Available from: http://www.intmedpress. com/tid/. 37. Funk GA, Steiger J, Hirsch HH. Rapid dynamics of polyomavirus type BK in renal transplant recipients. J Infect Dis 2006; 193: 80– 87. 38. Neyts J, De Clercq E. Efficacy of (S)-1-(3-hydroxy-2phosphonylmethoxypropyl)cytosine for the treatment of lethal vaccinia virus infections in severe combined immune deficiency (SCID) mice. J Med Virol 1993; 41: 242–246.
