Construction and biological activity of a full-length
molecular clone of human Torque teno virus (TTV)
genotype 6
Laura Kakkola
1
, Johanna Tommiska
1
, Linda C. L. Boele
2
, Simo Miettinen
1
, Tea Blom
1
,
Tuija Kekarainen
2
, Jianming Qiu
3
, David Pintel
3
, Rob C. Hoeben
2
, Klaus Hedman
1
and Maria So
¨derlund-Venermo
1
1 Department of Virology, Haartman Institute and Helsinki University Central Hospital, University of Helsinki, Finland
2 Department of Molecular Cell Biology, Leiden University Medical Center, the Netherlands
3 Department of Molecular Microbiology and Immunology, University of Missouri–Columbia, Life Sciences Center, Columbia, MO, USA
TT-virus (TTV) recently named Torque teno virus [1]
was found in 1997 in Japan from a patient with post-
transfusion hepatitis of unknown etiology [2]. The
virus is non-enveloped and contains a single-stranded
circular DNA genome of approximately 3.8 kb [3,4].
To date, five major phylogenetic groups have been
defined [1]. Due to its genome organization and struc-
ture, TTV resembles the chicken anemia virus (CAV)
of the Circoviridae family. This family of veterinary
viruses comprises the genus Gyrovirus, including CAV,
and the genus Circovirus, including the porcine circo-
virus (PCV) and the beak and feather disease virus of
birds. The human TT-virus is currently classified as a
member of a new, floating genus, Anellovirus [1].
The TTV genome consists of an approximately
2.6 kb coding and an approximately 1.2 kb noncoding
region. The latter contains a GC-rich area, a promoter
and transcriptional enhancer elements [3,5–8]. The
transcriptional capacity of the minute viral genome is
greatly expanded by splicing [9,10], resulting in six dis-
tinct yet partially overlapping viral proteins [11]. Little
is known of their functions. However, the longest gene,
Keywords
Anellovirus; replication; Torque teno virus;
transcription
Correspondence
L. Kakkola, Department of Virology,
Haartman Institute, Haartmaninkatu 3,
PO Box 21, University of Helsinki,
FIN-00014, Finland
Fax: +358 9 19126491
Tel: +358 9 19126676
E-mail: laura.kakkola@helsinki.fi
Website: http://www.hi.helsinki.fi/english
Note
Nucleotide sequence data are available in
the DDBJ ⁄EMBL ⁄GenBank databases under
the accession number AY666122
(Received 11 April 2007, revised 10 July
2007, accepted 11 July 2007)
doi:10.1111/j.1742-4658.2007.06020.x
Torque teno virus (TTV) is a non-enveloped human virus with a circular
negative-sense (approximately 3800 nucleotides) ssDNA genome. TTV
resembles in genome organization the chicken anemia virus, the animal
pathogen of the Circoviridae family, and is currently classified as a member
of a new, floating genus, Anellovirus. Molecular and cell biological research
on TTV has been restricted by the lack of permissive cell lines and func-
tional, replication-competent plasmid clones. In order to examine the key
biological activities (i.e. RNA transcription and DNA replication) of this
still poorly characterized ssDNA virus, we cloned the full-length genome of
TTV genotype 6 and transfected it into cells of several types. TTV mRNA
transcription was detected by RT-PCR in all the cell types: KU812Ep6,
Cos-1, 293, 293T, Chang liver, Huh7 and UT7 ⁄Epo-S1. Replicating
TTV DNA was detected in the latter five cell types by a DpnI-based restric-
tion enzyme method coupled with Southern analysis, a novel approach to
assess TTV DNA replication. The replicating full-length clone, the cell lines
found to support TTV replication, and the methods presented here will
facilitate the elucidation of the molecular biology and the life cycle of this
recently identified human virus.
Abbreviations
CAV, chicken anemia virus; DIG, digoxigenin; PBMC, peripheral blood mononuclear cells; PCV, porcine circovirus; TTV, Torque teno virus.
FEBS Journal 274 (2007) 4719–4730 ª2007 The Authors Journal compilation ª2007 FEBS 4719
ORF1, is assumed to encode a capsid protein that may
also participate in DNA replication [3,12] as is the case
with CAV [13].
The infection mechanisms and pathogenicity of TTV
are unknown. Putative replicative forms of TTV DNA
have been found in peripheral blood mononuclear cells
(PBMC), bone marrow and liver [14–16], suggesting
replication at these sites. Low-level infectivity of TTV
has been shown in activated PBMC and in a few
human cell lines [17–19]. The TTV promoter has been
shown to be active in both human [5,8,11] and non-
human cells [8,9]. TTV mRNAs have been detected in
human PBMC [19], bone marrow [10] and several
other organs [20]. However, the main host cells and
the target organs of this virus are still undefined.
The study of the biological functions of TTV is par-
ticularly challenging. Replication does not appear to
be very efficient in primary cells [17–19], nor in the few
cell lines supporting virus growth [17]. On the other
hand, the veterinary circoviruses CAV and PCV have
been studied successfully with infectious plasmid clones
[21–24]. With a full-length TTV plasmid clone of geno-
type 1, RNA transcription and splicing was studied in
Cos-1 cells. However, neither DNA replication, nor
cell permissiveness was demonstrated [9].
The present study aimed to construct a full-length
TTV clone that can be used as a tool for exploring the
viral determinants important in virus–cell interactions,
and to find permissive cell lines for further molecular
and cell biological studies of this peculiar human virus.
For this purpose, we have cloned and sequenced the
full-length genome of TTV genotype 6, and used it to
detect the key biological functions (i.e. RNA transcrip-
tion and DNA replication) in a number of different
cell lines.
Results
Genotype 6 cloning and sequence analysis
A full length molecular clone, pTTV, in plasmid pST-
Blue-1 was constructed of TTV genotype 6 (Fig. 1A).
The cloned genome was sequenced (GenBank acces-
sion number AY666122; nucleotide numbering accord-
ingly), and found to be 3748 nucleotides in length. A
TATA-box (TATAA) was located at nucleotides 83–87
and a poly(A) sequence ATTAAA at nucleotides
2978–2983. A GC-rich area was 107 nucleotides in
length. In the present study, two forms of the full-
length clone were used for transfection experiments;
the excised linear genome (linTTV) and the intact plas-
mid pTTV (Fig. 1B,C). The full-length plasmid clone
contains at its left end the NG136 primer sequence [25]
in which two nucleotides differ from the in vivo geno-
type 6 sequence. However, the BspEI-excised linear
construct excludes these primer-derived nucleotides.
Production of TTV RNA
Prior to the studies, all the cells were tested and found
to be TTV DNA negative by generic UTR-PCR and
by genotype 6 specific PCR.
All seven cell lines were analyzed with RT-PCR and
were found to produce identical TTV RNA upon
transfection with either pTTV or linTTV. In RT-PCR
analysis of the TTV clone-transfected cells, two ampli-
cons were observed (Fig. 2): one from the spliced
TTV mRNA (454 bp) and the other from TTV DNA
(555 bp). RNase treatment prior to RT-PCR abolished
the 454 bp amplicon; and DNase treatment abolished
the 555 bp but not the 454 bp amplicon (Fig. 2D). In
addition, RT-PCR without the RT step, and RT-PCR
of the input DNA constructs, yielded only the 555 bp
amplicon (Fig. 2D). Furthermore, the sequence data of
both amplicons showed that in the 454 bp amplicon
the intron had, indeed, been spliced out. These experi-
ments substantiated that the 454 bp amplicon origi-
nated from the transcribed viral RNA and not from
DNA. The TTV RNA was shown by RT-PCR to per-
sist in subcultured cells for at least 11 days. Nontrans-
fected cells and cells transfected with the backbone
plasmid pSTBlue-1, remained negative for TTV RNA
(Fig. 2B,C) confirming the absence of endogenous,
transcriptionally active TTV. RT-PCR of retinoblas-
toma mRNA yielded the expected amplicons (data not
shown) demonstrating mRNA integrity. The results
were identical for all the cell lines.
Replication of TTV DNA
All seven cell lines were transfected with linTTV and
with the intact pTTV. For detection of TTV replication,
total DNA from the transfected cells was treated with
the restriction enzymes BamHI and DpnI, and subjected
to Southern analysis. Two different probes were used
(Fig. 1): the one labelled with
32
P differentiates by size
the replicating TTV DNA from the input; and the other
labelled with digoxigenin (DIG) additionally documents
the susceptibility of the input DNA to DpnI.
In cells transfected with linTTV, the input DNA (after
BamHI digestion) was seen with the DIG-probe as a
3004-bp fragment (Figs 1B and 3A,B, marked with filled
circles), which was further digested by DpnI into a
fragment of 2162 bp (Figs 1B and 3A,B, marked with
filled squares). On day 3 post transfection, a full-
length TTV DNA of 3748 bp (after BamHI digestion,
Biological activity of a full-length TTV DNA clone L. Kakkola et al.
4720 FEBS Journal 274 (2007) 4719–4730 ª2007 The Authors Journal compilation ª2007 FEBS
detected with either probe) emerged in 293T, Huh7 and
UT7 ⁄Epo-S1 cells, and less pronounced also in 293 and
Chang liver cells (Fig. 3A,C, marked with an arrow).
However, no such bands appeared in KU812Ep6 and
Cos-1 cells. That BamHI digestion yielded a full-length
fragment indicates that circularization of the input lin-
ear construct had occurred. Furthermore, this 3748 bp
fragment was resistant to DpnI (Fig. 3A,C, marked with
an arrow), indicating TTV DNA replication. The linear-
ized backbone plasmid, not separated from the excised
linear TTV construct, did not replicate in these cells. As
an additional specificity control for the DpnI assay,
three (HindIII, EcoNI or ScaI) restriction enzymes,
other than BamHI, were used in Southern analysis,
resulting in identical DpnI-resistant bands on day 3. The
DpnI-resistant DNA progressively accumulated in the
transfected cells from day 0 to day 3, as shown for 293T
cells in Fig. 4. However, on days 3, 5 and 8–10 post
transfection upon cell passage, the amount of DpnI-
resistant (replicating) TTV DNA declined, and was
detectable in Southern analyses for up to day 5. Interest-
ingly, in those cells that permitted replication of the
excised linear construct, high molecular weight double
bands (sensitive to DpnI) were visible (Fig. 3). Single-
stranded DNA (ssDNA) could, however, not be visual-
ized by Southern analysis, suggesting that its production
remained below the detection limit. Of note, when com-
paring the 293 cells, with and without the SV40 large
T antigen, the amount of DpnI-resistant replicating
TTV DNA detected on day 3 post transfection (relative
to a standard amount of total cellular DNA) was invari-
ably much lower in 293 than in 293T cells (Fig. 3A, and
more pronounced in Fig. 3C), suggesting a possible
helper function for the SV40 large T antigen.
In cells transfected with the intact pTTV, the
input DNA (after BamHI digestion) was seen with
the DIG-probe as a 3165 bp fragment (Figs 1C and
3A,B, marked with filled circles), which was further
digested by DpnI into a fragment of 2162 bp
(Figs 1C and 3A,B, marked with filled squares). On
day 3 post transfection, the same 3165 bp fragment
(Fig. 3A, marked with an asterisk) was DpnI resis-
tant, indicating replication of the complete pTTV.
The results with the
32
P-probe verified this: the input
TTV genome
3748 bp
IEpsB
I
Ep
s
B
IHmaB
3004 bp
2162 bp
DIG
32P
pSTBlue-1
3851 bp
I
H
m
aB
32P
2162 bp
3165 bp
IH
ma
B
IHmaB
DIG
32P
32P
pTTV
total length
7775 bp
BC
3748/1
TTV
genotype6
•
NsiI
*
BspEI
pSTBlue-1
BspEI BspEI
tscr
tlat
•
TTVGCF NG136
A
Fig. 1. The full-length clone and the con-
structs for transfection. (A) The cloning
strategy of TTV into the pSTBlue-1 plasmid.
The GC-rich area is shown as a striped box
and the overlapping area in the clone as
spotted boxes. The three products used in
the construction of the TTV clone are indi-
cated with black lines. The key restriction
enzymes (see text for details), primers
(arrows: forward TTVGCF, reverse NG136),
the TATA-box (*), the poly(A) (d), the tran-
scription initiation (tscr) and the translation
initiation (tlat) sites are shown. Schematic
representations of (B) linear BspEI-excised
construct (linTTV), and (C) pTTV construct
used in transfection experiments. The viral
genome is represented by an empty bar and
the backbone plasmid by a thin black line.
The DIG- and the
32
P-labelled probes are
indicated. BamHI (vertical bars) and DpnI
(r) restriction enzymes were used for the
analysis of DNA replication. The predicted
TTV DNA products in replication analyses:
linTTV-derived 3004 bp fragment and pTTV-
derived 3165 bp fragment after BamHI
digestion are marked with d; linTTV and
pTTV derived 2162 bp fragments after
BamHI ⁄DpnI-digestion are marked with j.
L. Kakkola et al.Biological activity of a full-length TTV DNA clone
FEBS Journal 274 (2007) 4719–4730 ª2007 The Authors Journal compilation ª2007 FEBS 4721
pTTV was seen (after BamHI digestion) as two
DpnI-sensitive restriction fragments of 3165 bp and
4610 bp (Fig. 3C) and, on day 3 post transfection,
these fragments had become DpnI resistant (Fig. 3C).
In all pTTV-transfected cells, as detected with either
probe, a full-length DpnI-resistant 3748 bp fragment
(which would indicate rescue and replication of the
TTV genome from the backbone plasmid) remained
absent.
As opposed to the five other cell lines, in KU812Ep6
and Cos-1 cells, no replicating DNA (or, in some
experiments with the latter cells, barely exceeding
detection threshold) were detected upon transfection
with either construct (Fig. 3A). No apparent cyto-
pathological changes were microscopically detected in
the cells supporting TTV DNA replication.
The results were the same regardless of the DNA iso-
lation method (total cellular; Hirt extraction) and of the
detection probes (DIG- and
32
P-labelled probe) (data
shown for 293 and 293T cells in Fig. 3A,C). The non-
transfected cells and the backbone plasmid-transfected
cells were always negative for TTV DNA. In Southern
analysis, the input DNA served as an internal control to
verify restriction enzyme activity: the input DNA was
sensitive to DpnI (Fig. 3A,B, marked with filled squares)
whereas the newly synthesized DNA was resistant
(Fig. 3A, marked with an arrow) but remained digest-
ible with other restriction enzymes (data not shown).
Circularization of the linear construct
In the linTTV-transfected cells on day 3, the emer-
gence of the 3748 bp TTV DNA after BamHI diges-
tion (Fig. 3A) indicated that the input linear
TTV DNA had circularized. To confirm this, two
other restriction enzymes (HindIII and ScaI, Fig. 5A)
that, like BamHI, cut the TTV genome only once, were
used with identical results (i.e. a single product of
approximately 3.7 kb was detected; data not shown).
The circularization was further confirmed by digestion
of the DNA samples with pairs of restriction enzymes
that cut on both sides of the linearization breakpoint
(Fig. 5). BamHI ⁄SalIorXhoI⁄SalI double-digestions
of the input linear construct yielded three restriction
fragments of approximately 2100, 740 and 890 bp with
the first enzyme pair, and of 2200, 630 and 890 bp
with the latter. However, the same double-digestions
of replicating TTV DNA from 293T cells on day 3
post transfection yielded additional fragments of
approximately 1630 bp and 1520 bp, respectively
(Fig. 5B), indicating fusion of the linearization break-
point ends. Taken together, these results show that, in
the linTTV-transfected 293T cells, circular forms of the
TTV genome had been formed.
Effect of aphidicolin on TTV DNA replication
To reconfirm that TTV DNA replication had occurred
and to investigate whether it utilizes the cellular replica-
tion machinery, aphidicolin (an inhibitor of eukaryotic
C
B
D
Ant 112
DNA
nt 112 nt 182 nt 284
RT1F
nt 129-147
RT1R
nt 683-660
RNA
nt 1 nt 3748
poly-A
DDDDD
non-transf.
cells
Mw
Mw
day1 day3 day5 day10
linTTV
500
400
300
bp
day3
Mw
pSTBlue-1pTTV
Mw
500
400
bp DDDDD
day1 day3 day6 day10
600
day3
Mw -RT D
linTTV day3
400
500
600
bp +RT
400
500
600
bp Mw -RT R D-RT RD
input
linTTV
input
pTTV
+RT+RT
R
Fig. 2. RT-PCR of TTV RNA. (A) A schematic drawing of TTV
RT-PCR. Transcription initiates at nucleotides 112, and splicing
removes nucleotides 182–284 [11]. RT-PCR primers are shown
with arrows. Amplicon sizes are 555 bp for DNA and 454 bp for
spliced mRNA. RT-PCR results of representative 293T cells trans-
fected with (B) the linear excised TTV (linTTV) and (C) intact pTTV.
Nontransfected cells and cells transfected with the backbone plas-
mid (pSTBlue-1) were included as controls. (D) RT-PCR controls
from 293T cells transfected with linTTV (day 3) and from the input
constructs. +RT, normal RT-PCR; –RT, without the RT-step; R,
RNase-treated; D, DNase-treated.
Biological activity of a full-length TTV DNA clone L. Kakkola et al.
4722 FEBS Journal 274 (2007) 4719–4730 ª2007 The Authors Journal compilation ª2007 FEBS
3639
C
A
B
B B/DU
bp
4899
3639
2799
Mw
input
linTTV
BB/DU
bp
3639
2799
Mw
input
pTTV
4899
4899
3639
2799
293T
BB/DBB/D B B/D B B/D B B/D
day1 day3 day1 day3 day3
linTTV pTTV
bp
pSTBlue-1
Mw
293
B B/D B B/D B B/D B B/D B B/D
day1 day3 day1 day3 day3
linTTV pTTV
bp
pSTBlue-1
4899
3639
2799
Mw
4899
2799
bp B B/D B B/D B B/D B B/D B B/D
day1 day3 day1 day3 day3
Mw
Cos-1
linTTV pTTV pSTBlue-1
KU812Ep6
BB/D B B/D B B/D B B/D B
day1 day3 day1 day3
linTTV pTTV
cells
bp B/D
4899
3639
2799
Mw
3639
UT7/Epo-S1
day1 day4 day1 day4
linTTV pTTV
cells
bp B B/D B B/D B B/D BB/D B B/D
4899
2799
Mw
BB/D B B/D B B/D B B/D B B/D
day1 day3 day1 day3 day3
bp
4899
3639
2799
Mw
Huh7
linTTV pTTV pSTBlue-1
Chang liver
BB/D B B/D B B/D B B/D B B/D
day1 day3 day1 day3 day3
linTTV pTTV pSTBlue-1
bp
4899
3639
2799
Mw
linTTV
293T 293
input
pTTV pTTV linTTV pTTV
bp
4625
3165
Fig. 3. Southern analysis of TTV DNA replication. (A) 293T, 293, KU812Ep6, UT ⁄Epo-S1, Huh7, Cos-1 and Chang liver cells transfected with
the excised linear (linTTV) or the intact pTTV construct. The key products of the replication assay are marked in the 293T-cell figure: the input
linTTV yielding a 3004 bp fragment and the input pTTV yielding a 3165 bp fragment after BamHI digestion are marked with d; the input
linTTV and pTTV yielding 2162 bp fragments after BamHI ⁄DpnI digestion are marked with j; the DpnI-resistant circularized full-length
TTV DNA is marked with an arrow; DpnI-resistant pTTV is marked with *. (B) Southern analysis of the input constructs. The products of the
restriction enzyme digestions are marked as those in the 293T-cell figure. (Note the absence of a 3748 bp product.) U, undigested. (C) South-
ern analysis of Hirt-extracted (BamHI ⁄DpnI-digested) DNA from the 293T and 293 cells (with and without T antigen, respectively) transfected
with pTTV or linTTV. Input pTTV digested with BamHI as a control. Arrows indicate DpnI-resistant full-length TTV DNA. For detection of
TTV DNA, either a DIG- (A,B) or a
32
P- (C) labelled probe was used.
L. Kakkola et al.Biological activity of a full-length TTV DNA clone
FEBS Journal 274 (2007) 4719–4730 ª2007 The Authors Journal compilation ª2007 FEBS 4723