All animal procedures were approved by the Institutional Animal Care and Use Committee at CDC and in accordance with the guidelines of the Guide for the Care and use of Laboratory Animals. CDC primate protocol number: 1363KRACHIC. The study was done between Sept.23, 2004 and March 22, 2005. The animals were single or pair housed in accordance with the Guide for the Care and Use of Laboratory Animals in an AAALAC accredited facility. Chimpanzees were housed individually for short periods of time due to the restrictions of the scientific protocol involving the infectious nature of hepatitis research. A 12:12-h light:dark cycle was used at a room temperature of 17.8 to 28.9°C and a relative humidity of 30% to 70%. Water was provided ad libitum through an automatic watering system. The diet consisted of a nonhuman primate chow (Lab Diet High Protein Monkey Diet 5045, PMI Nutrition International, LLC, Saint Luis, MO), fruits, and treats (Bio-serv, Frenchtown, NJ). Environmental enrichment plan for chimpanzees consisted of the following: social enrichment, structural enrichment, manipulanda, novel food items/foraging, and sensory enrichment. Potential suffering during liver biopsy and bleeding procedures was alleviate using anesthetic medication using ketamine mixture and tiletamine HCL/Zolazepam HCL (Telazol). None of chimpanzees included in the study was sacrificed. All chimpanzees (CH6413, CH256 and CH1541) were inoculated intravenously with HCV genotype 1a, varying between 10³ to 10^3.5 chimpanzee infectious doses (CID) [22]–[24]. Serum samples were obtained every week at baseline and during 180 days after inoculation. HCV RNA in sera was tested and quantified by Cobas Amplicor HCV v. 2.0 (Roche Diagnostic Systems, Branchburg, NJ). The VITROS Anti-HCV assay (Ortho-Clinical Diagnostics, Rochester, NJ) was used for detection of anti-HCV IgG, and serum ALT levels were quantified with colormetric assay (Drew Scientific, Huston, TX). Cut off for normal ALT activity level values was calculated separately for each chimpanzee from 9 or 10 weekly serum specimens obtained before HCV inoculations by adding the mean value to its 3 standard deviations (99.7% confidence interval).

Liver needle biopsy samples were obtained before inoculation and every week after inoculation. Samples were frozen in liquid nitrogen immediately after the biopsy and subsequently stored at -80°C until further use. For CH256, 6 liver biopsy samples was obtained before inoculation (−56, −44, −37, −31, −24 and 0 days after inoculation [DAI]) and 13 samples after inoculation (DAIs 4, 12, 32, 40, 53, 75, 82, 95, 103, 124, 145, 152, and 180). For CH6413, 5 liver samples obtained before inoculation (DAIs −44, −37, −31, −16, and 0) and 14 samples after inoculation (DAIs 4, 12, 19, 32, 40, 47, 53, 61, 68, 82, 88, 103, 152, and 180) were used. For CH1541, 5 samples were taken before inoculation (DAIs −53, −24, −16, −7, and 0) and 11 samples after inoculation (DAIs 4, 12, 32, 40, 47, 53, 75, 82, 110, 152, and 173). HCV RNA in snap-frozen liver biopsy specimens was quantified by Taqman real-time PCR using 7900 HT fast real-time PCR system (Applied Biosystems, Foster City, CA). Primers and probe sequences were derived from 5′ non-coding region of the HCV genome (forward primer: 5′-GTCTGCGGAACCGGTGAG-3′) reverse primer: 5′- CGACCCAACRCTACTCGGCTAG-3′, probe: 5′-ACACCGGAATTGCCAGGACGACC-3′).

Total RNA from the frozen liver specimens was extracted using the Ribo Pure kit (Ambion, Austin, TX) and 2 pmol of Caenorhabditis elegans miR-39 (Qiagen, Valencia, CA) was spiked to total RNA as quality control RNA for real-time PCR. Quality and quantity of the total RNA was analyzed by the 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA). Total RNA was polyadenylated and reverse transcribed in 10 µl volumes using the miScript reverse transcription (RT) kit (Qiagen, Valencia, CA). All real-time PCR reactions were performed in triplicate on a 384-well 7900 HT fast real-time PCR system (Applied Biosystems, Foster City, CA) including non-template controls. Each real-time PCR reaction was done twice at different times. The reaction mix was incubated for 15 min at 95°C and 10 min at 50°C, then followed by 55 PCR cycles, each cycle comprising 94°C for 15 s, 55°C for 30 s and 70°C for 45 s. Data were analyzed with SDS Software version 2.3 (Applied Biosystems, Foster City, CA) with automatic baseline and threshold settings for cycle threshold (Ct) determination.

For determination of miR-122 levels in serum, total RNA was extracted from 200 µl of each serum sample by TRI-reagent (Life Technologies, Grand Island, NY) according to the manufacturer's recommendations. Prior to RNA extraction, 2 pmol SV-40 miRNA (Qiagen, Valencia, CA) was added to each serum sample for normalization. Then, cDNA synthesized from 1/10 of reverse transcription reaction from serum specimens was used for real-time PCR using miScript SybrGreen PCR kit (Qiagen, Valencia, CA) according to the manufacturer's recommendations.

The NormFinder program [25] was used to select unbiased endogenous control genes for normalization in both non-infected and HCV-infected liver tissues separately for each chimpanzee. Candidate genes used included six small RNAs [RNU6 (MS00033740), SNORD61 (MS00033705), SNORD68 (MS00033712), SNORD96A (MS00033733), SNORD95 (MS00033726), SNORD72 (MS00033719] and six miRNAs [miR-191 (MS00003682), miR-103a (MS00031241), miR-17 (MS00029274), let-7b (MS00003122), miR-15a (MS00003178), C. elegans miR-39 (MS000197890)]. All of these candidate control RNAs have been reported to be expressed in the liver and six miRNAs were used as endogenous control genes in previous studies [21], [26]. Primers for the small RNAs and miRNAs including miR-122 (MS00003416) were obtained from Qiagen. Relative quantification of the miRNA and mRNA expression was calculated using the 2^−ΔΔCt method [27]. Greater than 2-fold increases of genomic expressions were considered to be significant. miR-122 values in serum samples were normalized by the levels of added SV40 tracer [28]. miR-122 values in liver tissue and serum samples obtained before HCV inoculations were used as calibrators. Two groups of liver samples were included, those taken before HCV inoculation (n = 6 in CH256, n = 5 in CH6413 and CH1541) and those after HCV inoculation (n = 13 in CH256, n = 14 in CH6413, and n = 11 in CH1541). For each small RNA and miRNA control, the Ct values were calculated separately for each chimpanzee and then the relative quantities (RQ) for each control gene across all samples were calculated from Ct values scaled to a calibrator sample (lowest Ct) by using the equation: RQ = 1/(2^{(Ct sample−Ct min)}) [29]. Only the small RNA or miRNAs controls with the lowest intergroup variation and intragroup variation were selected for normalization. Geometric mean Ct values from these selected control RNAs were calculated and then subtracted from the miR-122 Ct (ΔCt miR-122). Liver samples obtained before HCV inoculation were used as calibrator samples, and the arithmetic mean Ct of each calibrator sample (n = 6 in CH256, n = 5 in CH6413 and n = 5 CH1541) was calculated. Fold change of miR-122 expression relative to the endogenous control genes was determined by the 2^−ΔΔCT method. The ΔΔCt was obtained by mean Ct of the calibrator samples (Ct miR-122 – Ct endogenous control genes) subtracted from ΔCt miR-122 (Ct miR-122 – Ct endogenous control genes).

cDNA was synthesized with 1 µg of total RNA using a high capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA) and used as template for real-time PCR performed by the TaqMan method. The reaction mix was incubated for 5 min at 95°C and 10 min at 50°C, followed by 40 PCR cycles, each cycle comprising 95°C for 15 s and 60°C for 1 min. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 4326317E) was used as internal control and a preinoculation sample used as calibrator. Primers and internal probes for IFN-α (Hs00256882_s1) and IFN-β (Hs02621180_s1) were obtained from Applied Biosystems. All of real-time PCR reactions were performed in triplicate on the 7900 HT fast real-time PCR system (Applied Biosystems, Foster City, CA). Transcripts of each gene were calculated based on the 2^−ΔΔCT method [27].

Statistical analyses were carried out using SPSS software 20 (Chicago, IL). Differences were detected using analysis of variance (ANOVA). A positive or negative correlation was indicated by the Spearman correlation coefficient and verified by two-tailed significance test. For all statistical tests, with a p value <0.05 considered significant. The correlations were calculated separately for each chimpanzee and also a single analysis was performed with data from all three animals.

To identify the most stable endogenous RNA as normalizers in hepatic miR-122 expression, twelve small RNAs and miRNAs were selected as candidate endogenous control RNAs. Ct values of C. elegance miR-39, used as a spike-in RNA control to assess the quality control and confidence of the real-time PCR performance, were similar in all samples from all three chimpanzees included in the study (Table S1). The twelve endogenous control RNAs were expressed in all samples, with Ct values ranging from 22.36 to 36.74 in CH256, 18.08 to 44.88 in CH6413, and 22.24 to 35.42 in CH1541. miR-103 was not used for further calculation in CH6413 due to its low expression (from 32.62 to 44.88, average Ct 41.43) (Table S1). miR-122 Ct values were observed in range from 20.68 to 29.01 for CH256, 23.74 to 29.03 for CH6413, and 22.79 to 27.78 for CH1541, i.e., in range of Ct values of the endogenous controls. The ANOVA-based model to estimate intra- and inter-group variation was used to select endogenous control genes for unbiased normalization across all samples individually in each chimpanzee [25]. It determined a stability value for each endogenous control gene and indicated the single most stable endogenous control for each animal and endogenous control gene pair for which the stability of the latter was greater than that of the single endogenous control. Two Ct values of each endogenous control obtained from two separate reactions were combined and the mean Ct value was used for calculation. The combination of SNORD95 and SNORD72 was selected for two chimpanzees, CH256 and CH6413, and the combination of SNORD95 and miR-191 was selected for CH1541 (Table 1).

Normalized levels of miR-122 in the liver increased from 2.2 to 3.9 times those of pre-inoculation samples: from DAIs 4 to 32 for CH256, DAIs 4 to 19 for CH6413, and on DAI 4 for CH1541 (Fig. 1A). During the peak of HCV replication, miR-122 expression had either decreased or remained unchanged (DAIs 40 to 95 in CH256; DAI 32 to 88 in CH6413; and DAIs 12 to 53 in CH1541). In all three animals, hepatic miR-122 levels were on the rise coinciding with the period when HCV RNA titers were declining (Fig. 1A and B).

All three were HCV RNA-positive in serum from DAIs 4 (CH1541) or 7 (CH256 and CH6413) and viral titers were maintained until DAI 71 (CH1541) or DAI 95 (CH256 and CH6413) (Fig. 1B). A similar profile was observed for HCV RNA in the liver; about 2-log decline in liver viral RNA levels was detected at DAI 75 (CH1541), 131 (CH256) or 88 (CH6413) (Fig. 1B). ALT activity was significantly elevated at DAI 71 (CH1541), 95 (CH256) or 75 (CH6413) (Fig. 1C, Table S2). A statistically significant inverse correlation was found between hepatic miR-122 expression and liver HCV RNA titers in CH256 and with both liver and serum HCV RNA titers in CH1541 (Fig. 2A, B, and D, Table S3A). A statistically significant negative correlation between hepatic miR-122 expression and the levels of serum miR-122 was found in CH1541 (Fig. 2F, Table S3A). Serum miR-122 levels were positively correlated with serum HCV RNA titer in CH256 and CH1541 (Fig. 2G and H, Table S3B). When hepatic and serum miR-122 data and all other parameters from all three chimpanzees were included in one analysis, a statistically significant inverse correlation between hepatic miR-122 and HCV RNA level in the liver and serum was observed (Fig. 2C and E, Table S3C). Levels of serum miR-122 were positively correlated with HCV RNA titer in the liver and serum (Fig. 2I and J, Table S3C). Expression profiles of serum miR-122 detected in three chimpanzees were similar to those of ALT activity (Fig. 1C); the correlation between those two measures was statistically significant in CH256 (Fig. 2K, Table S3B).

High levels of hepatic IFN-α and IFN-β mRNAs were detected during the period when levels of hepatic miR-122 expression were in decline (Fig. 1D, Table S2). Hepatic IFN-α and IFN-β mRNA expressions were correlated with hepatic miR-122 levels in CH256 (Fig. 3A and B, Table S3A). In CH6413, the levels of IFN-α and IFN-β mRNAs were correlated with serum miR-122 levels (Fig. 3E, F, and Table S3B). When all parameters from all three chimpanzees were included in one analysis, the moderate, but statistically significant positive correlation between hepatic IFN-α and IFN-β mRNA and hepatic miR-122 levels was observed (Fig. 3C and D, Table S3C).