SUPPLEMENTARY EXPERIMENTAL PROCEDURES
Transposase and transposon constructs
The mouse codon-optimized PB transposase (mPB) was generated by codon usage
optimization of the wild-type PB transposase sequence using a proprietary algorithm (Bio
Basic Inc., Canada). The optimized sequence was then custom synthesized and cloned
into the EcoRI/XhoI sites of pcDNA3 (Life Technologies, USA) in which the
cytomegalovirus (CMV) promoter was replaced by the CMV b-globin intron from pAAVMCS (Stratagene) by EcoRI/MluI (blunt-end) cloning (Fig. 1a). The hyperactive PB
transposase (hyPB) containing seven amino acid substitutions (I30V, S103P, G165S,
M282V, S509G, N538K, N570S) was custom-synthesized (Bio Basic Inc., Canada) and
cloned into the ClaI/XhoI sites replacing the mPB cDNA (Fig. 1b). This plasmid was
restricted with ClaI/XhoI and ligated by blunt-end ligation to generate the control
transposase plasmid devoid of PB cDNA (‘empty’ control) (Fig. 1c). The terminal
inverted repeats (IRs) of the transposon-containing plasmids corresponded to truncated
wild-type IRs
56,57
(designated as IRwt) (3’ IRwt: 245 bp; 5’ IRwt: 313 bp) or contained
T53C and C136T point mutations in the 5’-minimal terminal repeat (designated as
IRmut16). This IRmut16 was selected because it was previously shown to increase
transposition efficiency with nearly 60% based on in vitro studies
18
. Alternatively, a
minimal IR was used that was further truncated (designated as IRmicro) 19. This IRmicro (5’
IRmicro: 40 bp; 3’ IRmicro: 67 bp) was selected because it resulted in a 1.5 fold increase in
transposition efficiency based solely on in vitro studies. These different IRs were custom-
synthesized (Integrated DNA Technologies, Leuven, Belgium) and then assembled to
generate the corresponding transposon backbone flanked by two IRs (Fig. 1).
The
resulting generic transposon constructs were specifically designed to also contain a
multiple cloning site and loxP sites. The three IRs were then isolated by BssHII digestion
and cloned into the correspond site of pBluescript II SK (+). The resulting plasmid was
then used for subsequent cloning of a
chimeric liver-specific promoter using the
NotI/NheI sites to ensure robust hepatocyte specific hFIX expression. This chimeric
promoter was composed of a liver-specific minimal transthyretin (TTRmin) promoter in
conjunction with a potent and small hepatocyte-specific cis-regulatory module
(designated as HS-CRM8), that was identified using an exhaustive in silico analysis,
described elsewhere
20
. Subsequently, we cloned the different human factor IX (hFIX)
cDNAs downstream of this potent liver-specific chimeric promoter by NheI/BglII
digestion. The PB-hFIXIA transposon (Fig. 1d) contained a hFIX cDNA harboring the
truncated 1.4kb hFIX intron 1 (intron IA) between the first hFIX exon (exon 1) and the
following exons (exons 2-8) including a partial 3’ hFIX untranslated region (3’-UTR). In
the PB-hFIXco transposons (Fig. 1e, f, g), the hFIXIA minigene was replaced by a
synthetic codon-optimized hFIX sequence (hFIXco) and a partial FIX 3’ UTR. The codon
usage optimization was carried out based on a proprietary algorithm (BaseClear, the
Netherlands). In addition, the minute virus of mice small intron (MVM) was introduced
downstream of the chimeric liver-specific HS-CRM8/TTRmin promoter, since introns are
known to boost FIX transgene expression levels
2,26
. The PB-hFIXco contained either
wild-type IR (Fig. 1e); IRmut16 (PB-hFIXco/IRmut16; Fig. 1g) or micro-IR (PBhFIXco/IRmicro; Fig. 1f, h). The PB-hFIXco-R338L/IRmicro transposon (Fig. 1h) contained
the hyper-functional co-hFIX-R338L (Padua) cDNA
23,
and was constructed by site-
directed mutagenesis (QuikChange II XL, Agilent Technologies, USA) using PBhFIXco/IRmicro as template. The PB-hFVIIIBco/IRmicro transposon (Fig. 1i) was
generated by replacing the hFIXco cDNA with the codon-optimized B-domain deleted
hFVIII sequence (hFVIIIBco).31 The PB-c-Myc transposon (Fig. 1l) was generated by
PCR amplification of the c-Myc cDNA that was subsequently cloned to substitute the
hFIXIA sequence in PB-hFIXIA using the corresponding NheI/BglII restriction sites. All
resulting transposon-containing plasmids were sequence-verified.
Animal husbandry
C57Bl/6JRj mice were purchased from Janvier (France). CB17/IcrTac/Prkdc
scid
mice
were purchased from Taconic (Denmark). FIX-deficient hemophilia B mice were kindly
provided by Dr. I. Verma & Dr. L. Wang (The Salk Institute for Biological Studies, CA,
USA) and Dr. M. Kay (Stanford University, CA, USA). All animal procedures were
approved by the institutional animal ethics committee of the Free University of Brussels
(VUB) (Brussels, Belgium) and the University of Leuven (Leuven, Belgium). Husbandry
was carried out in Thoren stages using individually ventilated cages with Hugenic
Animal Bedding from Lignocel. Temperature was maintained at approximately 21°C and
50-60% humidity. Animals were fed with SsniFF laboratory animal food ad libitum.
Hydrodynamic tail vein injection in mice
Plasmid DNA was produced using the PureLink® HiPure Plasmid Filter Maxiprep Kit
(Life Technologies, USA). DNA solutions containing different doses and ratio of
transposon/transposase plasmids were diluted in 2 ml of Dulbecco’s PBS and
hydrodynamically delivered to adult mice (6-7-week-old) by rapid tail vein injection.
Typically, the injection took less than 7 seconds for each mouse. At different time
intervals, whole blood was collected into buffered citrate by phlebotomy of the retroorbital plexus. The citrated plasma was stored at -80°C for future analysis.
DEN-induced hepatocarcinogenesis
C57BL/6 male mice were subjected to a preweaning protocol of diethylnitrosamine
(DEN) initiation 32 to induce hepatocellular carcinoma (HCC). Briefly, 12.5 mg/kg body
weight of DEN (Sigma-Aldrich, Diegem, Belgium) was injected intraperitoneally (i.p)
into 14-days-old mice. 4-weeks later, mice were hydrodynamically injected with the
transposon plasmids (Fig. 5a). A Kaplan-Meier survival analysis was conducted. Mice
were killed at 36 weeks after DEN initiation and examined for HCC nodules.
Analysis of hFIX and hFVIII antigen levels, hFIX activity levels and anti-hFIX
antibodies
hFIX and hFVIII antigen levels were assayed in citrated mouse plasma using a hFIX and
hFVIIIspecific enzyme-linked immunosorbent assay (ELISA) (Asserachrom® IX:Ag and
Asserachrom® VIII:Ag, Diagnostica Stago, France), as per the manufacturer’s
instructions. hFIX activity was determined in citrated hemophilia B mouse plasma using
a chromogenic hFIX activity assay, according to the manufacturer (Hyphen Biomed,
Neuville-sur-Oise, France). Standard curves were constructed with serial dilutions of
pooled normal human plasma corresponding to 100% FIX activity ( i.e. 5 mg/ml).
Antibodies directed against hFIX were detected using an indirect ELISA. Briefly, 96-well
microtiter plates (MaxiSorp, Nunc, Roskilde, Denmark) were coated with 1mg/ml
recombinant hFIX (BeneFIX, Wyeth Pharmaceuticals, Louvain-la-Neuve Belgium),
blocked with dilution buffer (PBS - 0.05% Tween-20 - 6% non-fat dried milk) and
incubated overnight at 4° with mouse plasma diluted in dilution buffer. After washing the
plates, they were
then incubated with polyclonal goat anti-mouse immunoglobulin
conjugated to horseradish peroxidase (HRP) (DAKO, Denmark) as secondary antibody
(diluted 1:2000). Anti-hFIX antibody levels were measured by 5 min incubation with the
3,3´, 5,5´-tetramethylbenzidine (TMB) substrate solution (Promega, USA). H2SO4 (1M)
was then added to terminate the reaction and spectrophotometric absorbance
determination at 450 nm. Plasma from PBS-injected mice was used as negative control
and antibody titers (in ng/ml) were determined by linear regression analysis using a
standard curve generated from wells coated with serially diluted purified murine IgG
proteins (Life Technologies, USA) spiked with negative control mouse plasma.
Phenotypic correction
Mice were anesthetized and tails were placed in prewarmed 37°C saline solution for 2
min and then cut at a diameter of 3 mm. The tails were re- immersed in prewarmed saline
solution and monitored for bleeding or clotting for 10 min. The blood-containing saline
was centrifuged at 520 g for 10 min at 4°C. Subsequently, 6 ml lysis buffer (10 mM
KHCO3, 150 mM NH4Cl, and 1 mM EDTA) was added to the red blood cell pellet. Lysis
proceeded for 10 min at room temperature after which the samples were centrifuged as
above. Hemogloblin content of the supernatants was detected by spectrophotometrically
at 575 nm. The survival rate was monitored andupon termination of each study mice were
subjected to euthanasia and different organs were harvested for further analysis.
Transposon genome copy number quantification
Genomic DNA was extracted from frozen liver samples of treated and control mice
according to DNeasy Blood & Tissue Kit protocol (Qiagen, Chatsworth, CA, USA).
RNase A (Qiagen, Chatsworth, CA, USA) treatment was carried out to eliminate carry
over RNA. To avoid possible variations due to primer amplification efficiency,
transposon copy number were quantified by qPCR using a primer set against a specific
region common to all PB constructs evaluated in this study using forward primer 5’AACAGGGGCTAAGTCCACAC
-3’
and
reverse
primer
5’-
GAGCGAGTGTTCCGATACTCT -3’. Briefly, 50 ng of genomic DNA from each
sample was subjected to qPCR in triplicate using an ABI Prism 7500 Fast Real-Time
PCR System (Applied Biosystems, Foster City/CA, USA) and Power SYBR® Green
PCR Master Mix (Applied Biosystems, Foster City/CA, USA). Copy number was
determined comparing the amplification signal with a standard curve consisting of serial
dilutions over a 6 log range of the corresponding linearized plasmid spiked with 50 ng of
liver genomic DNA from saline-injected mouse (slope ≈ -3,3, intercept ≈ 35, efficiency
% ≈ 100). Average copies per diploid genome were calculated taking into account that
one murine diploid genome = 5,92 pg.
hFIX mRNA expression analysis
Total RNA was extracted from frozen liver samples of treated and control mice using a
miRCURY™ RNA isolation kit (Exiqon, Denmark). DNase (Thermo Scientific, USA)
treatment was carried out to eliminate carry-over of any potential residual contaminating
genomic DNA. The reverse transcription reaction was performed starting from 1 mg of
total RNA from each sample using the SuperScript® III First Strand cDNA Synthesis Kit
(Life Technologies, USA). Next, a cDNA amount corresponding to 10 ng of total RNA
from each sample was analyzed in triplicate by quantitative (q)PCR using an ABI Prism
7500 Fast Real-Time PCR System (Applied Biosystems, Foster City/CA, USA) and
Power SYBR® Green PCR Master Mix (Applied Biosystems, Foster City/CA, USA).
The following primer set was used, that was complementary to the bGHpA sequence that
is
common
to
all
of
the
PB
transposons
used:
forward
primer
5’-
GCCTTCTAGTTGCCAGCCAT-3’, reverse primer 5’- GGCACCTTCCAGGGTCAAG3’. The hFIX mRNA levels were normalized using a primer set against the mRNA of the
housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (mGAPDH) which is
uniformly and constantly expressed in all samples (i.e. forward primer 5’ATCAAGAAGGTGGTGAAGCAGGCA
-3’
and
reverse
primer
5’-
TGGAAGAGTGGGAGTTGCTGTTGA -3’). RNA samples were amplified with and
without reverse transcriptase to exclude genomic DNA amplification. The 2-ΔΔCt relative
quantification method was used to determine the relative changes in hFIX mRNA
expression level. The ΔCt was calculated by subtracting the Ct of mGAPDH mRNA from
the Ct of the hFIX mRNA (CthFIX - CtGAPDH). The ΔΔCt was calculated by subtracting the
ΔCt of the reference sample (highest Ct) from the ΔCt of each sample (ΔCtsample ΔCtreference). Fold-change was determined by using the equation 2 − ΔΔCt.