Journal of General Virology (1997), 78, 2747–2750.

Printed in Great Britain

............................................................................................................................................................................................................... SHORT COMMUNICATION

Hepatitis C virus negative strand RNA is not detected in

peripheral blood mononuclear cells and viral sequences
are identical to those in serum: a case against
extrahepatic replication
Tomasz Laskus,1 Marek Radkowski,2 Lian-Fu Wang,1 Janusz Cianciara,2,4 Hugo Vargas3
and Jorge Rakela1,3
1,3
Division of Transplantation Medicine, Thomas E. Starzl Transplantation Institute, University of Pittsburgh, Lhormer Bldg 301 1, and
Division of Gastroenterology and Hepatology3, University of Pittsburgh Medical Center, 200 Lothrop Street, Pittsburgh, PA 15213,
USA
2,4
Institute of Infectious Diseases2 and Municipal Hospital for Infectious Diseases4, Wolska 37, 01-201 Warsaw, Poland

et al., 1994, 1995). In the current study the latter

Peripheral blood mononuclear cells (PBMCs) technique was employed for an extensive search for HCV
from 27 hepatitis C virus (HCV)-infected RNA negative strand in PBMCs from patients with
patients were analysed for the presence of HCV chronic hepatitis C. Synthetic RNA was used to optimize
negative strand RNA with strand-specific Tth- the sensitivity and strand-specificity of the assays.
based RT–PCR. No negative strand RNA was Furthermore, HCV sequences amplified from PBMCs were
detected in any sample, and positive strand HCV compared with those amplified from serum, assuming that
sequences amplified from PBMCs were identical in the presence of passive virus
to those found in serum. These findings suggest adsorption and/or contamination of these cells by
that HCV does not replicate in PBMCs, and the circulating viral RNA the amplified sequences should be
presence of HCV sequences at this site is identical, while in
compatible with passive virus adsorption and/or the presence of independent replication they might be
contamination by circulating virus. different. The sequences were compared by single strand
conformation polymorphism (SSCP), as well as by direct
sequencing.
The existence of extrahepatic reservoirs of hepatitis C Twenty-seven patients with chronic hepatitis C were
virus (HCV) replication remains controversial. Several subjects of the study. None of the patients had received
groups have described the detection of HCV negative any antiviral therapy prior to the study, and none had
strand, a viral replicative intermediate, in peripheral blood serological evidence of hepatitis B virus or HIV infection.
mononuclear cells (PBMCs) (Wang et al., 1992 ; Muller et PBMCs were isolated by Ficoll–Hypaque (Pharmacia)
al., 1993 ; Gabrielli et al., 1994 ; Saleh et al., 1994), and it density-gradient cen- trifugation, washed three times with
has been reported that a human T-cell line is capable of PBS pH 7·4, and stored frozen at —80 ºC until use. RNA
supporting a productive infection (Shimizu et al., 1993). was extracted from 5×10ª cells or 100 µl serum by means
However, the strand specificity of RT–PCR, which is of a modified guanidinium thiocyanate–phenol–chloroform
currently the only suitable method for the detection of technique using a commer- cially available kit (RNAzol B;
HCV RNA, has recently been questioned: this assay has Biotecx Laboratories) and finally dissolved in 20 µl water; 10
been demonstrated to be prone to false priming of the µl of this RNA solution was reverse transcribed as further
incorrect strand or to self-priming related to RNA described.
secondary structures (Lanford et al., 1994). To generate synthetic positive and negative HCV RNA
Several methods of avoiding the mispriming events strands, PCR products encompassing the 5’ untranslated
have region were cloned into a plasmid vector (pGEM-3Z;
been proposed: tagged PCR (Lanford et al., 1994); using Promega) and, after plasmid linearization, transcribed
primers specific for a region devoid of strong secondary with T7 polymerase (Riboprobe Transcription System,
structures (Lerat et al., 1996); or conducting cDNA synthesis Promega). The orientation of the insert was checked by
at high temperature with the thermostable enzyme Tth direct sequencing of the plasmid.
(Lanford The template was removed by digestion with DNase I (1
U/µg DNA for 60 min at 37 ºC), and the absence of
significant
Author for correspondence: Tomasz Laskus. amounts of residual DNA was ascertained by routine
Fax ‡1 412 647 9672. inclusion of control PCR without the RT step.
Two different RT–PCR assays, one employing MMLV-
RT and the other employing Tth, were used. For the MMLV-
based
0001-4871© 1997 SGM CHEH
 Laskus and
T. Laskus andothers
others

Fig. 1. Sensitivity and specificity of RT–PCR using MMLV-RT and Tth. Synthetic positive and negative strands were
generated by in vitro run-off transcription with T7 RNA polymerase from a vector (pGEM-3Z) containing the 5’ untranslated
sequence of HCV and serially diluted in water. The number of target template copies was calculated from absorbance
readings and gel electrophoresis. A positive-sense primer was present during cDNA synthesis, after which the enzyme
was inactivated either by
heating for 10 min at 99 ºC (MMLV) or chelating with Mn2+ (Tth) and then negative-sense primer was added. Samples were
amplified for 50 cycles using Taq polymerase (MMLV-based assay) or Tth (Tth-based assay) as described in the text. A 20 µl
volume (20 %) of the reaction was fractionated on a 2·5 % agarose gel, transferred to nylon membrane by Southern
blotting and subsequently hybridized to a 32P-labelled probe internal to the amplification primers. Significant strand
specificity could be demonstrated for Tth-based RT–PCR, but not for MMLV-based assay. When 1 µg of total cellular RNA
extracted from normal human livers was added, the sensitivity of the reactions was lowered by no more than one log.

detection of negative strand, extracted RNA was incubated
for 20 min at 42 ºC in a 30 µl reaction containing 50 pM
of the positive-sense primer (5’
A/GAC/TCACTCCCCTGTGAG-
GAAC 3’ ; nt 35–55), 1× RT buffer (Gibco), 5 mM DTT,
5 mM MgCl , 1 mM dNTP, 20 U RNase inhibitor
(RNasin,
¹ ACCCGCTGCTTTGA 3’ (nt 8245–8275 ; sense) and the
Promega) and 20 U MMLV RT (Gibco). After heating to
99 ºC
for 10 min, 50 pM of the negative-sense primer (5’ TGA/
GTGCACGGTCTACGAGACCTC 3’ ; nt 342–320), 7 µl
of
10× PCR buffer II (Perkin Elmer) and 2·5 U Taq DNA
polymerase (Perkin Elmer) were added and the volume
was adjusted to 100 µl. For the detection of positive
strand, the primers were added in reverse order.
Amplification was run in DNA Thermal Cycler 480
(Perkin Elmer) as follows: initial denaturing at 94 ºC for 4
min followed by 50 cycles of 94 ºC for 1 min, 58 ºC for 1
min and a final extension at 72 ºC for 7 min. The final
product (20 µl) was analysed by agarose gel
electrophoresis and Southern hybridization with a ◆¹P-
labelled internal oligoprobe (5’
ACTGTCTTCACGCAGAAAGCGTC 3’ ; nt 57–79).
For Tth-based RT–PCR, the cDNA was generated in a
20 µl reaction mixture containing 50 pM of sense primer,
1× RT
buffer (Perkin Elmer), 1 mM MnCl , and 200 µM (each)
dNTP ¹
and 5 U Tth (Perkin Elmer). After 20 min at ºC, Mn¹+
65 was
chelated with 8 µl of 10× EGTA chelating buffer (Perkin
Elmer). 50 pM of antisense primer was added and the
volume
adjusted to 100 µl, and MgCl concentration
¹ was adjusted
to 2·2 mM. The amplification was performed in a Perkin
Elmer
at 72 ºC for 7 min. PCR products were analysed as
described above.
For amplification of the NS5 region, the primers were
5’ GGCGGAATTCCTGGTCATAGCCTCCGTGAA 3’ (nt
8645–8616 ; antisense) and 5’ TGGGGATCCCGTATGAT-
probe was 5’ CTCAACCGTCACTGAGAGAGACAT 3’ (nt
8276–8299). To increase the specificity and sensitivity of
our assays, wax beads (Ampliwax, Perkin Elmer) were
routinely employed for hot start of all PCR after the RT
step.
The results of the analysis of the serial dilution of
synthetic RNA are presented in Fig. 1. The MMLV RT–
PCR assay for negative strand was capable of detecting 10
equivalent genomic molecules (Eq) of the respective
template. However, when using the same protocol, a
positive signal was observed in the presence of “ 10◆
Eq, and occasionally even in the presence of “ 10 Eq,
of the positive strand. Use of Tth at the 70 ºC RT step
increased the specificity of RT–PCR by seven logs since
the incorrect strand was now detected only at 10¹O Eq;
however it also lowered sensitivity 100-fold. Since this loss
of sensitivity could have been partly related to the high
temperature of the RT step, the experiments were repeated
with the RT step conducted at 65 ºC. This increased the
GenAmp PCR System 9600 thermocycler as follows: initial
denaturing for 1 min at 94 ºC, 50 cycles of 94 ºC for 15 s, 58
ºC for 30 s and 72 ºC for 30 s followed by a final extension
sensitivity of the assay 10-fold, but simultaneously lowered contamination by transcriptional vector DNA, but rather
the specificity 100-fold. These temperature-dependent reflects persistent incorrect priming. To mimic the
changes suggest that the loss of strand specificity at high conditions encountered in amplification of biological
concentrations of incorrect template is not due to trace samples, 1 µg RNA extracted from

CHEI
 Extrahepatic HCV replication
replication

Fig. 2. Lack of detection of HCV RNA negative strand in PBMCs from an HCV infected patient. Tenfold serial
dilutions of extracted RNA were tested for the presence of positive and negative HCV RNA strands using MMLV-RT-based
and Tth-based RT–PCR, as described in the text. Both positive and negative HCV RNA strands were detected by the
MMLV assay; however, only positive strands were detected by the Tth-based strand-specific assay. The amount of RNA
loaded into reaction at dilution
10−0 corresponds to 2·5×106 cells. Negative controls (lane N) consisted of RNA extracted from PBMCs from uninfected
subjects, and positive/sensitivity controls (lane P) consisted of end-point dilutions of the correct synthetic strand (10 Eq for
MMLV assay and 100 Eq for the Tth assay).

uninfected human liver was added into each reaction. This 5’ untranslated
lowered the sensitivity of the assays by no more than one
log (not shown). The sensitivity of our assays for the
detection of the positive strand was identical to that for
the detection of the negative strand.
The sensitivity and specificity of the Tth protocol was
even higher than that described by Lanford et al. (1994),
probably reflecting such factors as the simplified hot start
procedure with wax beads, and the short ramp times
achievable in the Perkin Elmer GenAmp PCR System
9600 which are likely to lower nonspecific amplification
and increase sensitivity.
During optimization of the MMLV-RT-based assay,
we found that hot start of the PCR step increased
sensitivity 10◆–10ª times (not shown). This effect was
probably related to the strong secondary structure of the
template in the 5’ un- translated region, as it was not
observed using RNA templates devoid of hairpin
structures (unpublished observations).
Tth assays with primers specific for the NS5 region
were not analysed as rigorously for sensitivity and strand
specificity with synthetic RNA. However, when tested on
serial dilutions of two positive liver samples, they were no
more than one log less sensitive in the detection of
positive and negative strands than assays for the 5’
untranslated region (not shown).
Taking into account the results of the initial
experiments with synthetic template, only Tth-based assay
(RT at 65 ºC) was used for HCV negative strand search.
This assay seemed to be a good compromise between
sensitivity and strand specificity. All RT–PCR runs
included positive controls consisting of end-point
dilutions of respective RNA strands; negative controls
included normal PBMCs and normal sera.
PBMCs and sera from all 27 patients were negative for
the presence of the minus strand when tested with Tth-
based assays in both 5’ noncoding and NS5 regions. The
positive strand was detected by MMLV assay in 17 and
12 PBMC samples and in 27 and 20 serum samples in the
and NS5 regions, respectively. Positive strand titres,
which were calculated by assuming that the end-point
10-fold serial dilution contains 10 Eq, ranged from 10¹–
10& Eq per reaction for serum (mean 10ª·¹) and 10¹ to
10ª Eq per reaction for PBMCs (mean 10&·¹). Thus, if
negative strand were present in PBMCs it would have to
be present at 10& lower levels than positive strand, at
least in those PBMCs with the highest virus titres. A
representative presence of high titre HCV RNA
positive strand and lack of evidence for the presence of
the negative strand are shown in Fig. 2.
To determine the proportion between the positive
and negative strands in liver cells, two explant livers
from HCV infected transplant recipients were studied.
By testing of liver RNA dilutions with the Tth-based
assay, the titre of the
positive strand was determined to be 10& and 10ª/1 µg
RNA while the titre of the negative strand was 10ª and
10&/1 µg RNA, respectively. The titres were calculated
assuming that
the end-point 10-fold serial dilution contains 10¹ Eq.
Considering the strand specificity of our assays as de-
termined on synthetic RNA, it is not surprising that false
positive detection of the negative strand was not
encountered. Nonspecific detection of the incorrect strand
might be expected when the latter is present at high
numbers, at least 108 Eq per reaction, which was not
encountered in the samples studied.
Twelve patients were positive for positive strand HCV
RNA in both serum and PBMCs with primers specific for
the NS5 region. Since this part of the viral genome is
considerably more variable than the 5’ untranslated
region, we considered it appropriate for comparison of
amplified sequences. It was assumed that in the presence
of passive virus adsorption
and/or contamination of the PBMCs by circulating viral
RNA, the amplified sequences would be identical to those
in serum,
while in the presence of independent replication they
might be different. The sequences were compared by
SSCP, which is appropriate for the detection of minor
sequence differences, as
CHEJ
 Laskus and
T. Laskus andothers
others
Fig. 3. Analysis by SSCP of HCV sequences, amplified by RT–PCR,
from PBMCs and serum from 12 patients with chronic hepatitis C. Lanes
A and B represent viral sequences from PBMCs and serum,
respectively. All patients show identical band patterns for PBMCs and
serum; the presence of identical viral sequences was verified by
direct sequencing.
well as by direct sequencing. We were encouraged by
results of our earlier study on HBV infection, in which
differences were detected between viral sequences from
serum and PBMCs (Laskus et al., 1997), and by a recently
published work documenting discrepancies between the
replicating and circu- lating HCV quasispecies (Cabot et
al., 1997).
For SSCP analysis, 0·1 µg of purified PCR product from
the NS5 region was subjected to non-denaturing PAGE in
1× Tris–borate–EDTA buffer and the bands were visualized
with silver staining as described (Laskus et al., 1996).
As illustrated in Fig. 3, SSCP analysis of PCR
products revealed the presence of indistinguishable band
patterns from PBMCs and serum, compatible with the
presence of identical viral sequences. The presence of
identical viral sequences was subsequently verified by
direct sequencing as described elsewhere (Laskus et al.,
1996). Based on the sequence of the NS5 region, all
genotypes were assigned to subtype 1b (Simmonds et al.,
1993).
The existence of extrahepatic sites of HCV replication
would have broad implications for antiviral treatment and
liver transplantation. The presence of independent
replication in PBMCs has recently been questioned by
Lanford et al. (1995), who did not detect negative strand
viral RNA when employing strand-specific RT–PCR. Our
study, done on a large scale and employing very sensitive
and highly specific assays, corro- borates and extends
these findings by showing that HCV sequences amplified
from PBMCs are identical to those found
in serum, which is compatible with passive virus adsorption on
the cell and/or contamination of cell samples with
circulating
virus. Taking into consideration the efficiency of the
techniques employed, any existing virus replication in
PBMCs would have to be extremely low to elude
detection.
References
Cabot, B., Esteban, J. I., Martell, M., Genesca, J., Vargas, V.,
Esteban, R., Guardia, J. & Gomez, J. (1997). Structure of
replicating hepatitis C virus (HCV) quasispecies in the liver may
not be reflected by analysis of circulating HCV virions. Journal of
Virology 71, 1732–1734.
Gabrielli, A., Manzin, A., Candela, M., Caniglia, M. L., Paolucci,
S., Danieli, M. G. & Clementi, M. (1994). Active hepatitis C virus
infection in bone marrow and peripheral blood mononuclear cells
from patients with mixed cryoglobulinaemia. Clinical and Experimental
Immunology 97, 87–93.
Lanford, R. E., Sureau, C., Jacob, J. R., White, R. & Fuerst, T. R.
(1994). Demonstration of in vitro infection of chimpanzee
hepatocytes with hepatitis C virus using strand-specific RT/PCR.
Virology 202, 606–614.
Lanford, R. E., Chavez, D., Chisari, F. V. & Sureau, C. (1995).
Lack of detection of negative-strand hepatitis C virus RNA in
peripheral blood mononuclear cells and other extrahepatic tissues by
the highly strand-
specific rTth reverse transcriptase PCR. Journal of Virology 69, 8079–8083.
Laskus, T., Wang, L. F., Rakela, J., Demetris, A. S., Vargas, H., Pinna,
A. D., Tsamandas, A. C. & Fung, J. (1996). Dynamic behavior of
hepatitis C virus in chronically infected patients receiving liver graft
from infected donors. Virology 220, 171–176.
Laskus, T., Wang, L. F., Radkowski, M., Vargas, H., Cianciara, J.,
Poutous, A. & Rakela, J. (1997). Comparison of hepatitis B virus
core promoter sequences in peripheral blood mononuclear cells and
serum from patients with hepatitis B. Journal of General Virology 78,
649–653.
Lerat, H., Berby, F., Trabaud, M.-N., Vidalin, O., Major, M., Trepo,
C. & Inchauspe, G. (1996). Specific detection of hepatitis C virus
minus strand RNA in hematopoietic cells. Journal of Clinical
Investigation 97, 845–851.
Muller, H. M., Pfaff, E., Goeser, T., Kallinowski, B., Solbach, C. &
Theilmann, L. (1993). Peripheral blood leukocytes serve as a
possible extrahepatic site for hepatitis C virus replication. Journal
of General Virology 74, 669–76.
Saleh, M. G., Tibbs, C. J., Koskinas, J., Pereira, L. M. M. B., Bomford,
A. B., Portmann, B. C., McFarlane, I. G. & Williams, R. G. (1994).
Hepatic and extrahepatic hepatitis C virus replication in relation to
response to interferon therapy. Hepatology 20, 1399–1404.
Shimizu, Y. K., Purcell, R. H. & Yoshikura, H. (1993). Correlation
between the infectivity of hepatitis C virus in vivo and its infectivity
in vitro. Proceedings of the National Academy of Sciences, USA 90, 6037–
6041.
Simmonds, P., Holmes, E. C., Cha, T. A., Chan, S. W., McOmish, F.,
Irvine, B., Beall, E., Yap, P. L., Kolberg, J. & Urdea, M. S. (1993).
Classification of hepatitis C virus into six major genotypes and a
series of subtypes by phylogenetic analysis of the NS5 region. Journal
of General Virology 74, 2391–2399.
Wang, J. T., Sheu, J.-C., Lin, J.-T., Wang, T.-H. & Chen, D. S. (1992).
Detection of replicative form of hepatitis C virus RNA in peripheral
blood mononuclear cells. Journal of Infectious Diseases 166, 1167–
1169.
Received 10 April 1997 ; Accepted 24 June 1997
CHFA