BK Virus Encoded MicroRNAs Are Present in Blood of Renal Transplant Recipients With BK Viral Nephropathy

Transcription

1 American Journal of Transplantation 2014; 14: Wiley Periodicals Inc. Brief Communication C Copyright 2014 The American Society of Transplantation and the American Society of Transplant Surgeons doi: /ajt BK Virus Encoded MicroRNAs Are Present in Blood of Renal Transplant Recipients With BK Viral Nephropathy J. Y. Z. Li 1,2,y, K. McNicholas 1,2,y, T. Y. Yong 2,3, N. Rao 4, P. T. H. Coates 4,5, G. D. Higgins 5,6, R. P. Carroll 4,5, R. J. Woodman 2,7, M. Z. Michael 2,8 and J. M. Gleadle 1,2, * 1 Department of Renal Medicine, Flinders Medical Centre, Adelaide, Australia 2 School of Medicine, Flinders University, Adelaide, Australia 3 Department of General Medicine, Flinders Medical Centre, Adelaide, Australia 4 Central Northern Adelaide Renal and Transplantation Service, Royal Adelaide Hospital, Adelaide, Australia 5 Department of Medicine, University of Adelaide, Adelaide, Australia 6 Department of Microbiology, Royal Adelaide Hospital, Adelaide, Australia 7 Flinders Centre for Epidemiology and Biostatistics, Flinders University, Adelaide, Australia 8 Department of Gastroenterology, Flinders Medical Centre, Adelaide, Australia Corresponding author: Jonathan M. Gleadle, jonathan.gleadle@health.sa.gov.au y Both authors have contributed equally to the work. BK viral infection is an important cause of renal transplant dysfunction and failure. Current strategies utilize surveillance for infection with DNA polymerase chain reaction assays and modulation of immunosuppression. Many viruses including polyomaviruses encode micrornas (mirnas). We have detected BK virus (BKV) encoded mirnas in the blood of infected renal transplant recipients, and see a strong correlation between BKV encoded mirna and BKV DNA in blood and a relationship between levels of bkv-mir-b1-5p and the presence of biopsy-proven BK viral nephropathy. Further research is needed to determine whether the detection of this and other virally encoded mirnas may be useful in the diagnosis of active viral replication. Keywords: BK nephropathy, BK virus, microrna, renal transplant Abbreviations: BKV, BK virus; BKVN, BK viral nephropathy; cdna, complementary DNA; Cq, quantification cycle; EBV, Epstein Barr virus; egfr, estimated GFR; JCV, JC virus; mirna, microrna; PCR, polymerase chain reaction; qrt-pcr, quantitative real-time reverse transcriptase polymerase chain reaction; ROC, receiver operating characteristic; RT-PCR, reverse transcriptase polymerase chain reaction; smdrd, simplified Modification of Diet in Renal Disease; TAg, viral protein T antigen Received 31 October 2013, revised 29 January 2014 and accepted for publication 31 January 2014 Introduction Two polyomaviruses, BK virus (BKV) and JC virus (JCV), establish a latent infection in 65 90% of humans (1,2). BKV is the causative agent of BK viral nephropathy (BKVN) in kidney transplant recipients (3). BKVN can lead to deterioration in allograft function and graft failure (4,5). BKV reactivation occurs commonly after renal transplantation and is identified in the urine of 30 50% of patients at 3 months posttransplantation (6). Viral DNA in blood and urine is usually detected by quantitative polymerase chain reaction (PCR) (7). Progression from detection of viral DNA in urine to blood occurs in approximately 15% of kidney transplant recipients and is a harbinger of BKVN and deterioration of renal function (5,7,8). MicroRNAs (mirnas) are noncoding RNAs of nucleotides. mirnas regulate the translation and stability of mrna by binding to complementary target mrna sequences. Lost or enhanced expression of mirnas is associated with fundamental cellular processes (9). Members of several virus families have been reported to encode mirnas (10). The functions of the majority of viral mirnas are unknown, though some viral mirnas regulate host gene expression, viral gene expression (11) and transition from latency to lytic replication and can attenuate immune responses (12 14). Virally encoded mirna expression has been described in viral replication. mirna encoded by the Merkel cell polyomavirus has been detected in specimens of Merkel cell carcinoma (15), and mirnas encoded by Epstein Barr virus (EBV) have been detected in EBVassociated nasopharyngeal carcinoma (16). BKV encoded mirnas have been demonstrated to target the viral protein T antigen (TAg) with perfect complementarity to the 3p coding end of the TAg mrna (11,17). One of 1183

2 Li et al the major functions of TAg is driving DNA replication of the viral genome at the origin of replication (18). Furthermore, BKV mirna can target stress-induced protein ULBP3, which is recognized by the natural killer cells receptor NKG2D (13) to potentially avoid NKG2D-mediated elimination. It is not known whether infected renal cells release BKV mirna (bkv-mir-b1) or whether this mirna can be detected in the blood or urine of infected patients. In this study, we report the novel observation that BKV mirna is expressed during replication and can be detected in blood. We also find a correlation between BKV mirna and BKV DNA in blood and increased levels in patients with biopsy-proven BKVN. Materials and Methods Study population The files of the Renal Transplant Patients in South Australia registry were reviewed to identify adult patients (age > 18 years) with BKV DNA in the blood (ascertained by BK DNA PCR) following a kidney transplant between January 2008 and May Thirty-one patients with BKV DNA in blood and stored blood samples were studied. Five patients with previous blood BK DNA positive but now negative were included for comparison. Ten patients who never had detectable BKV in the blood and urine from transplant until May 2013 were selected as a control group. Patients estimated GFR (egfr) was calculated using the simplified Modification of Diet in Renal Disease (smdrd) formula. The diagnosis of BKVN was made by the presence of viral cytopathic effect in renal tubular cells with interstitial mononuclear inflammatory cell infiltrates and/or the presence of homogenous intranuclear inclusion bodies. This was supplemented by immunohistochemistry with Simian virus 40 staining (19). This study was approved by both the Southern and Central Adelaide Local Health Network Human Research Ethics Committees (Approval No: and , respectively). Study method Design of primers and PCR for BKV and JCV DNA: The full method has been described previously (20). Primers and probes for BKV were designed to target the VP2 region. Total RNA was extracted from 0.25 ml of thawed plasma using Trizol LS (Invitrogen, Life Technologies, Carlsbad, CA). The RNA pellet was resuspended in 20 ml of RNase-free water and stored at 808C. Quantitative real-time reverse transcriptase PCR: BKV mirna expression was assessed by quantitative real-time reverse transcriptase PCR (qrt-pcr) using human TaqMan mirna assays (Applied Biosystems, Foster City, CA). The reverse transcriptase reaction using 4 ml of RNA sample was performed with TaqMan mirna Reverse Transcription Kit and TaqMan mirna-specific primers (assay IDs: hsa-mir-16:000391, bkv-mir- B1-5p:007796, bkv-mir-b1-3p:006801; Applied Biosystems). PCR reactions were pipetted in triplicate using a Qiagen QIAgility robot (v4.15.1; Melbourne, Australia). The qrt-pcr reaction contained 1 ml of reverse transcription product, 1 TaqMan Universal PCR mastermix, No AmpErase UNG and 0.5 ml of primer mix. The complementary DNA (cdna) was amplified using a Rotor-Gene Q Thermocycler (Qiagen). Reactions were incubated at 958C for 10 min, followed by 40 cycles at 958C for 15 s and 608C for 60 s. Quantification cycle (Cq) values were calculated using the Rotor- Gene Q software (v2.0.2). A control plasma sample was spiked with a known concentration of synthetic BKV mirna mimic duplex oligonucleotides (bkv-mir-b1; Shanghai GenePharma, Shanghai, China). The sequence of the duplex was AUCUGAGACUUGGGAAGAGCAU and 5 0 -UGCUUGAUCCAUGUCCAGA- GUC. cdna was synthesized from 4 ml of total RNA sample using a specific primer for bkv-mir-b1-5p. Statistical analysis: A one-way analysis of variance was used to compare the means of the patient characteristics (Table 1). The Spearman s Rho nonparametric correlation coefficient was used to assess correlations between BKV mirna expression, BKV DNA viral load and laboratory measurements. The Kruskal Wallis nonparametric test was used to compare BKV mirna expression across all five clinical groups. A Mann Whitney nonparametric test was used to assess the significance of BKV mirna expression in pairwise comparisons of clinical groups when p < 0.05 across all five groups for the Kruskal Wallis test. For all statistical tests, twosided p < 0.05 was considered statistically significant. We assessed the correlation of plasma BKV mirna, BKV DNA viral load, and the detection of BKVN. Each marker was assessed separately using logistic regression and the C-statistic (area-under receiver operating characteristic [ROC] curve). Tests of equality of the C-statistic between measures were performed using the algorithm suggested by DeLong et al (21). We used plots of the predicted probabilities of the logistic regression to determine the optimal cut-point of each measure in providing both high sensitivity and high specificity and calculated sensitivity and specificity based on these cut-points. Results The detection of BKV mirna BKV encodes one precursor mirna, bkv-mir-b1, which generates two mature mirnas, 3p and 5p. The 3p mirna is conserved between JCV and BKV. However, the sequence of the mature 5p mirna generated by BKV and JCV differs, providing a mean of differentiating between these two viral infections (Figure 1). A nucleotide blast search (Basic Local Alignment Search Tool; Blast.cgi) using a 102-nucleotide query corresponding to the precursor stem-loop sequence bkv-mir-b1 revealed perfect sequence conservation in all (over 100) complete BKV genome sequences in GenBank ( gov/genbank). To examine for bkv-mir-b1, we used qrt- PCR using human TaqMan microrna assays (Applied Biosystems). As a positive control and to enable approximate quantification of oligonucleotide copy number, plasma was spiked with a known concentration of synthetic BKV mirna oligonucleotides (bkv-mir-b1) or with a similar JCV oligonucleotide. Initial assays confirmed the ability of the TaqMan reverse transcriptase PCR (RT-PCR) assays to detect both bkv-mir-b1-5p and bkv-mir-b1-3p. In keeping with the identity of the 3p mirnas, expression of the JCV encoding oligonucleotide could also be detected by the bkv-mir-b1-3p RT-PCR assay but the bkv-mir-b1-5p assay was 500-fold more sensitive (data not shown) in detecting the BK mirna transcript than the JC encoded transcript American Journal of Transplantation 2014; 14:

3 BK Virus MicroRNAs in Renal Transplant Recipients Table 1: Summary of patient clinical details and laboratory results Clinical category Biopsy-proven BKVAN No BKVAN on biopsy BK PCR DNA positive/not biopsied BKVAN No BKVAN BK positive/ no biopsy Previous BK PCR DNA positive now negative Previous BK positive Negative BK DNA PCR BK negative p-value BK PCR DNA detected þ þ þ Renal biopsy undertaken þ þ Renal biopsy showing BKVAN þ Number of patients Age (years) Sex (male:female) 2:5 8:5 9:2 4:1 5:5 Time since transplant (median, months) Serum creatinine 1 (mmol/l) egfr 1 (ml/min/1.73 m 2 ) Serum creatinine 2 (mmol/l) egfr 2 (ml/min/1.73 m 2 ) Hemoglobin 1 (g/l) Platelet count 1 (10 9 /L) BK copy number PCR 1 (/ml) <0.001 BKVN, BK viral nephropathy; egfr, estimated GFR; PCR, polymerase chain reaction. Gender, age, creatinine, egfr (smdrd; simplified Modification of Diet in Renal Disease formula), hemoglobin, platelet count and BK PCR DNA viral load of the 46 patients are shown. Values are means SD. 1 At time of sample acquisition for determination of BK PCR DNA and BK mirna. 2 One-year after sample acquisition. p-value generated from one-way analysis of variance; significant p-values (<0.05) shown in bold. The detection of BKV mirnas in blood from patients with BK replication The presence of BK mirna (bkv-mir-b1-3p and bkv-mir- B1-5p) was examined by TaqMan qrt-pcr in RNA purified from blood and urine of renal transplant recipients with BK DNA detected by PCR and in controls (renal transplant recipients with no detectable BK DNA). We were able to detect bkv-mir-b1-3p and bkv-mir-b1-5p in both blood and urine of some, but not all, patients with detectable BK DNA (Figure 2) with Cq values ranging from 24 to 34. A significant negative correlation was found between the qrt-pcr Cq value and the detection of BKV DNA in blood and urine suggesting a higher level of bkv-mir-b1-5p and bkv-mir- B1-3p in patients with higher BKV DNA levels (Figure 2). Cq values are inversely proportional to the amount of target nucleic acid in the sample, that is, the lower the Cq, the greater the amount of BKV mirna in the sample. The bkvmir-b1-3p mirna was detected in the urine of two control patients with positive urine BKV DNA but negative blood BKV DNA. We were unable to detect BK mirna in control patients without BKV infection other than amplification of bkv-mir-b1-3p with a Cq of 30.3 in the urine from a single control patient with negative BKV DNA (Figure 2). Optimization of bkv-mir-b1-5p qrt-pcr assay Given the ability of the bkv-mir-b1-5p qrt-pcr assay to detect BKV mirna but not the JCV encoded 5p mirna, we focused on this assay. Stem-loop primed TaqMan assays for mirna detection are specific for mature forms of bkvmir-b1-5p, and in keeping with design of the assay to detect RNA, positive samples assayed without the reverse transcriptase did not show bkv-mir-b1-5p PCR Figure 1: BKV and JCV mirna precursor sequences. Precursor mirna sequences generated by BKV and JCV with mature mirna sequences underlined and illustrating the perfect conservation of the 3p sequences but the differences in the 5p sequences providing a potential route to distinguishing between the viruses. BKV, BK virus; JCV, JC virus; mirna, microrna. American Journal of Transplantation 2014; 14:

4 Li et al Figure 2: Pilot examination of bkv-mir-b1-5p mirna and bkv-mir-b1-3p in blood and urine. A significant negative correlation was found between the qrt-pcr quantification cycle (Cq) value and the BKV DNA viral load (number of copies per milliliter) in blood and urine indicating a higher level of bkv-mir-b1-5p mirna in patients with higher viral loads. Patients with positive BKV DNA viral detection in blood are shown as solid dots: bkv-mir-b1-5p in blood, p < , r ¼ and urine p < , r ¼ 0.834; bkv-mir-b1-3p in blood, p < , r ¼ 0.85 and urine p ¼ 0.02, r ¼ PCR reactions with Cq values >35 or with poor amplification efficiency were treated as a negative result and assigned a Cq value of 35 in the analysis. Negative results are shown slightly offset from the origin of the x-axis (gray-dotted line) to avoid over plotting. Blood (n ¼ 5 BKV DNA-positive patients; n ¼ 7 control patients); urine (n ¼ 4 BKV DNA-positive in blood and urine of patients; n ¼ 6 control patients). BKV, BK virus; mirna, microrna; qrt-pcr, quantitative real-time reverse transcriptase polymerase chain reaction. amplification, in accord with reports of inability to amplify genomic DNA sequences using this method (22). To provide an approximate quantification of the number of molecules of bkv-mir-b1-5p, we undertook a titration series of synthetic oligonucleotide duplex BKV mirna mimic. A control plasma sample was spiked with a known concentration of the BKV mirna mimic and the resultant cdna was serially diluted. A standard curve of mean Cq values, versus cdna concentration, was used to extrapolate the concentration of BKV mirna in patient samples (Figure 3). The limit of detection under these conditions corresponded to a Cq of <31 and 1000 molecules of bkv-mir-b1-5p. Validation of bkv-mir-b1-5p detection in plasma from renal transplant recipients To determine if detection of bkv-mir-b1-5p was associated with BKVN we examined archived samples from renal transplant patients, which had previously been examined for BKV DNA by PCR because of deterioration of graft function. Stored blood samples were obtained from 46 patients in five groups: Group 1: patients with a positive BKV DNA PCR and biopsy-proven BKVN (BKVN) (n ¼ 7); Group 2: patients with a positive BKV DNA PCR but no BKVN on biopsy (No BKVN) (n ¼13); Group 3: patients with detectable BKV DNA but in whom a renal biopsy was not undertaken (BK positive/no biopsy) (n ¼ 11); Group 4: patients who were BKV DNA PCR positive but had become negative (previous BK positive, mean time to negative 12 months) at the time of this study (n ¼ 5); and Group 5: a control group consisting of patients who were consistently negative for BKV DNA posttransplant until May 2013 (BK negative; n ¼ 10). For patients who underwent renal biopsy, samples were selected that had been obtained closest in timing to the renal biopsy. Patients characteristics in each group are summarized in Table 1. In keeping with the usual approach to management of patients with 1186 American Journal of Transplantation 2014; 14:

5 BK Virus MicroRNAs in Renal Transplant Recipients Figure 3: Quantification of bkv-mir-b1-5p microrna using synthetic oligonucleotide titration curve. A standard curve with 95% CI was constructed from serial dilutions of synthetic bkv-mir-b1-5p duplex in triplicate quantitative real-time reverse transcriptase polymerase chain reaction reactions (R 2 ¼ ; Y ¼ 4.173x þ 44.77). potential BKVN, there was a greater likelihood of undertaking renal biopsy in patients with higher levels of plasma BKV DNA and more impaired renal function. As in the pilot study, we were able to detect bkv-mir-b1-5p in the plasma of patients who were positive for circulating BKV DNA and there was a significant correlation between bkv-mir-b1-5p and BKV DNA assays (Figure 4). By contrast, we were unable to detect bkv-mir-b1-5p in samples in which BKV DNA had not been detected. The highest levels of bkv-mir-b1-5p were apparent in patients with biopsy-proven BKVN. The Cq value and derived copy numbers (Figure 5A) of bkv-mir-b1-5p in the biopsy-proven BKVN (Group 1) were significantly higher than those with high plasma viral load but biopsy negative for BKVN (Group 2) and those with high plasma viral load who did not undergo biopsy (Group 3). Furthermore, there was no bkv-mir-b1-5p detected in serum from patients with very low circulating BKV DNA, control patients without BKV replication or those who had previously had BKV DNA detected in blood but were now negative. In keeping with previous studies, plasma BKV DNA viral load in patients with biopsy-proven BKVN was significantly higher than in patients without BKVN on renal biopsy (Figure 5B). Our data demonstrated a 10-fold higher mean expression of BKV mirna in the blood of patients with biopsy-proven BKVN compared with patients positive for BKV DNA without evidence of BKVN on biopsy (p ¼ ). American Journal of Transplantation 2014; 14: Figure 4: Correlation between bkv-mir-b1-5p and plasma BKV DNA viral load. A significant positive correlation was found between the number of copies of microrna bkv-mir-b1-5p in plasma and the plasma DNA viral load (number of copies per milliliter); p < ; r ¼ The copy number was extrapolated from the bkv-mir-b1-5p standard curve (Figure 3); negative results are shown slightly offset from the origin of the x-axis (gray-dotted line) to avoid over plotting. The regression line was calculated after excluding negative quantitative polymerase chain reaction results; p ¼ 0.006; n ¼ 31 (BKV DNA-positive patients); n ¼ 15 (control patients). BKV, BK virus. Given the unlikely possibility that other clinical features might influence the detection of BKV mirna, we examined for correlations between the level of bkv-mir-b1-5p and other measurements (Table 2). However, while the Cq values of bkv-mir-b1-5p were strikingly inversely correlated with the BKV DNA levels (r ¼ 0.766; p < ), variations in renal function, hemoglobin, platelets or the circulating levels of a human mirna (mir-16) were not significantly associated with bkv-mir-b1-5p detection. Correlation between plasma bkv-mir-b1-5p, BKV DNA viral load and the presence of BKVN To examine the ability of bkv-mir-b1-5p and BKV DNA levels to determine the presence of BKVN, we undertook ROC analyses (Figure 6) on the 46 patients. The optimum cut-point for bkv-mir-b1-5p was at a predicted probability for biopsy-proven BKVN of 17.3%, which corresponded to a bkv-mir-b1-5p Cq of This provided a sensitivity of 100.0% and a specificity of 94.9%. The area under the ROC was The optimum cut-point for plasma DNA viral load was at a predicted probability for biopsy-proven BKVN of 7.7% corresponding to a plasma DNA viral load of This provided a sensitivity of 100.0% and a specificity of 92.3%. The area under the ROC was There was no significant difference between the C-statistic for bkv-mir- B1-5p and plasma BKV DNA viral load (p ¼ 1.00). Utilizing the bkv-mir-b1-5p test with a Cq cutoff of 31.9 gave a positive-predictive value of 77.8% and negative-predictive value of 100% for BKVN, while utilizing plasma viral DNA 1187

6 Li et al Figure 5: (A) Examination of plasma bkv-mir-b1-5p mirna expression by clinical groupings. Highest levels of mirna bkv-mir-b1-5p were observed in plasma samples from patients with biopsy-proven BKVN. p < generated from Kruskal Wallis nonparametric test for all groups; p-values between groups generated using Mann Whitney nonparametric t-test; qpcr reactions with Cq values >35 or with poor amplification efficiency were treated as a negative result and assigned a Cq value of 35 in the analysis. (B) Examination of plasma BKV DNA viral levels by clinical groupings. The viral load, as determined by qpcr assay of viral DNA, is shown for patient plasma samples grouped by clinical classification. BKV, BK virus; BKVN, BK viral nephropathy; Cq, quantification cycle; mirna, microrna; qpcr, quantitative polymerase chain reaction. load with a cut-point of gave a positive-predictive value of 70.0% and negative-predictive value of 100% for BKVN. No patients with BK positivity who did not undergo renal biopsy had a Cq less than Discussion We demonstrate for the first time that mirnas encoded by BKV can be detected in blood and urine of patients with BKV infection. However, we were unable to detect these molecules in individuals without infection or in those no longer positive for circulating BKV DNA. The detection of high levels of bkv-mir-b1-5p in blood was strongly associated with high levels of BKV DNA and with biopsyproven BKVN. These samples were examined retrospectively, and in keeping with the stability of these molecules, we were able to detect significant expression despite longterm freezer storage. This study shows a very substantial difference between the levels of bkv-mir-b1-5p in patients with biopsy-proven Table 2: Summary of correlation (r) and significance (p) of associations between BKV microrna, BK PCR DNA viral loads and laboratory measurements BK PCR viral load bkv-mir-b1-5p hsa-mir-16 r p r p r p egfr (ml/min/1.73 m 2 ) Creatinine (mmol/l) Hemoglobin (g/l) Platelet count (10 9 /L) BK PCR viral load (per ml) < bkv-mir-b1-5p < hsa-mir BKV, BK virus; Cq, quantification cycle; egfr, estimated GFR; mirna, microrna; qpcr, quantitative polymerase chain reaction; smdrd, simplified Modification of Diet in Renal Disease. Correlations between the Cq values of mirnas hsa-mir-16 and bkv-mir-b1-5p with measurements of BKV DNA viral load, egfr (smdrd), creatinine, hemoglobin and platelet for n ¼ 46 samples; qrt-pcr reactions with Cq values >35 or with poor amplification efficiency were treated as a negative result and assigned a Cq value of 35 in the analysis. Spearman s nonparametric correlation two-tailed p-value; significant p-values (<0.05) shown in bold American Journal of Transplantation 2014; 14:

7 BK Virus MicroRNAs in Renal Transplant Recipients mirnas can have diagnostic utility and distinguish latent from active replication. Evidence for this would require a large-scale study examining the detection of bkv-mir-b1-5p in blood and urine of renal transplant recipients. It could also address whether these molecules are detectable at differing stages of infection compared with existing DNA assays, whether they persist differently following remission or whether urine testing is useful. Furthermore, the half-life of BKV DNA and mirna molecules may be very different. Figure 6: Receiver operating characteristic (ROC) curves showing discrimination of patients with biopsy-proven BK nephropathy (BKVN) using bkv-mir-b1-5p mirna or plasma BKV DNA viral levels. ROC curves demonstrating diagnostic accuracy of detecting patients with biopsy-proven BKVN for BKV encoded mirna (long-dash line) and plasma DNA viral load (shortdash line). The C-statistic is represented by the area under the curve (AUC). There was no difference in the AUC between measures (p ¼ 1.00). BKV, BK virus; BKVN, BK viral nephropathy; mirna, microrna. BKVN and those with positive BKV DNA in the blood without biopsy evidence of nephropathy or those whose clinical features did not lead to biopsy. Furthermore, BKV mirna is not detectable in patients without BKV replication. There was a strong correlation between the plasma levels of bkv-mir-b1-5p and BKV DNA determined by PCR. It remains to be established if BKV mirna levels are solely indicative of high viral loads or have specificity for nephropathy. In patients with severe chronic kidney disease, we recently found that circulating mirnas were reduced in comparison to patients with normal renal function (23) suggesting that kidneys do not function in the clearance of circulating mirnas. Given the potential for other clinical features to influence the detection of BKV mirna, we examined for correlations between bkv-mir-b1-5p and other measurements. However, variations in renal function, hemoglobin or mir-16 levels were not significantly associated with bkvmir-b1-5p detection. This is most in keeping with the levels of bkv-mir-b1-5p being associated with the severity of BKV replication. However, this study has its limitations as it was a retrospective study of available archived blood specimens in a small number of patients. It is not capable of providing evidence that bkv-mir-b1-5p detection can perform as a potential diagnostic test for BKVN, nor that such mirnas have a definite pathophysiological role. It does suggest the need for further research to determine whether in this and other viral infections the detection of virally encoded American Journal of Transplantation 2014; 14: The definitive diagnosis of BKVN can be made by renal allograft biopsy and histological findings on biopsy have been characterized and correlated with graft outcome (24). However, changes of BKVN can be focal or isolated to the medulla and missed on one-third of biopsies if only a single core is evaluated (24) and the histological finding of acute rejection can be similar to BKVN. The progression from BKV in the urine to BKV in the blood to nephropathy is often regarded as a stepwise transition and provides an opportunity for early detection and early intervention with reduction of immunosuppression (7). The recent American Society of Transplantation Infectious Diseases Guidelines recommended that in patients with sustained plasma BKV DNA loads of >10 4 copies/ml, a diagnosis of presumptive BKVN should be made in absence of demonstrable BKV replication in biopsies (25). However, a wide variety of primers, probes and standards for quantitative PCR assays are in use, which limits the usefulness of generalized quantitative viral load cutoffs in screening and diagnostic protocols. Furthermore, there is a broad range of sequence variation in different clinical isolates of BKV so that PCR assays vary in their ability to detect viral DNA (26) and can lead to false negative tests. In this study, high levels of both BKV DNA and BKV mirna are strongly predictive of BKVN. However, this study is not designed or powered to directly address whether BK mirna detection has a better (or worse) predictive value in the noninvasive diagnosis of BKVN. Indeed, while the detection of BK mirna had a good specificity for the presence of biopsy-proven BKVN, it appeared less sensitive for the detection of BKV DNA in the blood and assayed by this method would not usefully replace BK viral DNA screening protocols. Furthermore, the lower sensitivity of RT-PCR for RNA compared with DNA PCR may account for the correlation between higher plasma viral loads and BKV mirna levels, and for the apparently high specificity of BKV mirna levels in predicting BKVN. This is the first demonstration that circulating BKV encoded mirna bkv-mir-b1 is detectable in the plasma of patients with biopsy-proven BKVN and those with significant BKV DNA in the blood. It has also shown that bkv-mir-b1-5p levels are highest in patients with biopsy-proven BKVN. It suggests that circulating virally encoded mirnas may act as indicators of severity of virus replication. 1189

8 Li et al Acknowledgments This study was supported by a seeding grant from the Flinders University Faculty of Health Sciences. We thank the patients and their clinicians for supporting this study. Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation. References 1. Stolt A, Sasnauskas K, Koskela P, et al. Seroepidemiology of the human polyomaviruses. J Gen Virol 2003; 84: Eash S, Manley K, Gasparovic M, et al. The human polyomaviruses. Cell Mol Life Sci 2006; 63: Ahsan N, Shah KV. Polyomaviruses and human diseases. Adv Exp Med Biol 2006; 577: Nickeleit V, Hirsch HH, Binet IF, et al. Polyomavirus infection of renal allograft recipients: From latent infection to manifest disease. J Am Soc Nephrol 1999; 10: Hirsch HH. BK virus: Opportunity makes a pathogen. Clin Infect Dis 2005; 41: Bressollette-Bodin C, Coste-Burel M, Hourmant M, et al. A prospective longitudinal study of BK virus infection in 104 renal transplant recipients. Am J Transplant 2005; 5: Nickeleit V, Klimkait T, Binet IF, et al. Testing for polyomavirus type BK DNA in plasma to identify renal-allograft recipients with viral nephropathy. N Engl J Med 2000; 342: Ramos E, Drachenberg CB, Wali R, et al. The decade of polyomavirus BK-associated nephropathy: State of affairs. Transplantation 2009; 87: Bartel DP. MicroRNAs: Target recognition and regulatory functions. Cell 2009; 136: Grundhoff A, Sullivan CS. Virus-encoded micrornas. Virology 2011; 411: Seo GJ, Fink LH, O Hara B, et al. Evolutionarily conserved function of a viral microrna. J Virol 2008; 82: Cullen BR. Viral and cellular messenger RNA targets of viral micrornas. Nature 2009; 457: Bauman Y, Mandelboim O. MicroRNA based immunoevasion mechanism of human polyomaviruses. RNA Biol 2011; 8: Broekema NM, Imperiale MJ. mirna regulation of BK polyomavirus replication during early infection. Proc Natl Acad Sci USA 2013; 110: Lee S, Paulson KG, Murchison EP, et al. Identification and validation of a novel mature microrna encoded by the Merkel cell polyomavirus in human Merkel cell carcinomas. J Clin Virol 2011; 52: Wong AM, Kong KL, Tsang JW, et al. Profiling of Epstein Barr virus-encoded micrornas in nasopharyngeal carcinoma reveals potential biomarkers and oncomirs. Cancer 2012; 118: Sullivan CS, Ganem D. MicroRNAs and viral infection. Mol Cell 2005; 20: Fanning E, Zhao K. SV40 DNA replication: From the A gene to a nanomachine. Virology 2009; 384: Drachenberg C, Hirsch HH, Papadimitriou JC, et al. Cost efficiency in the prospective diagnosis and follow-up of polyomavirus allograft nephropathy. Transplant Proc 2004; 36: Rao N, Schepetiuk M, Choudhry M, et al. JC viremia in kidney transplant recipients: To act or not to act? Clin Kidney J 2012; 5: DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 1988; 44: Chen C, Ridzon DA, Broomer AJ, et al. Real-time quantification of micrornas by stem-loop RT-PCR. Nucleic Acids Res 2005; 33: e Neal CS, Michael MZ, Pimlott LK, et al. Circulating microrna expression is reduced in chronic kidney disease. Nephrol Dial Transplant 2011; 26: Drachenberg CB, Papadimitriou JC, Ramos E. Histologic versus molecular diagnosis of BK polyomavirus-associated nephropathy: A shifting paradigm? Clin J Am Soc Nephrol 2006; 1: Hirsch HH, Randhawa P. BK polyomavirus in solid organ transplantation. Am J Transplant 2013; 13(Suppl 4): Hoffman NG, Cook L, Atienza EE, et al. Marked variability of BK virus load measurement using quantitative real-time PCR among commonly used assays. J Clin Microbiol 2008; 46: American Journal of Transplantation 2014; 14: