THE JOURNAL OF GENE MEDICINE
J Gene Med 2007; 9: 275–286.
Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/jgm.1014
Recombinant fusion proteins TAT-Mu, Mu and
Mu-Mu mediate efficient non-viral gene delivery
Rukkumani Rajagopalan
Jennifer Xavier
Nandini Rangaraj
Nalam Madhusudhana Rao
Vijaya Gopal*
Centre for Cellular and Molecular
Biology, Uppal Road, Hyderabad
500007, India
*Correspondence to: Vijaya Gopal,
Centre for Cellular and Molecular
Biology, Uppal Road, Hyderabad
500007, India.
E-mail: vijaya@ccmb.res.in
Received: 20 October 2006
Revised: 30 December 2006
Accepted: 15 January 2007
Copyright  2007 John Wiley & Sons, Ltd.
RESEARCH ARTICLE
Abstract
Background The inherent ability of certain peptides or proteins of viral,
prokaryotic and eukaryotic origin to bind DNA was used to generate novel
peptide-based DNA delivery protocols. We have developed a recombinant
approach to make fusion proteins with motifs for DNA-binding ability,
Mu and membrane transduction domains, TAT, and tested them for
their DNA-binding, uptake and transfection efficiencies. In one of the
constructs, the recombinant plasmid was designed to encode the Mu moiety
of sequence MRRAHHRRRRASHRRMRGG in-frame with TAT of sequence
YGRKKRRQRRR to generate TAT-Mu, while the other two constructs, Mu and
Mu-Mu, harbor a single copy or two copies of the Mu moiety.
Methods Recombinant his-tag fusion proteins TAT-Mu, Mu and Mu-Mu
were purified by overexpression of plasmid constructs using cobalt-based
affinity resins. The peptides were characterized for their size and interaction
with DNA, complexed with plasmid pCMVβ-gal, and shown to transfect
MCF-7, COS and CHOK-1 cells efficiently.
Results Recombinant fusion proteins TAT-Mu, Mu and Mu-Mu were cloned
and overexpressed in BL21(DE3)pLysS with greater than 95% purity. The
molecular weight of TAT-Mu was determined by matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) to
be 11.34 kDa while those of Mu and Mu-Mu were 7.78 and 9.83 kDa,
respectively. Live uptake analysis of TAT-Mu, Mu and Mu-Mu as DP
(DNA+peptide) or DPL (DNA+peptide+lipid) complexes into MCF-7 cells,
followed by immunostaining and laser scanning confocal microscopy,
demonstrated that the complexes are internalized very efficiently and
localized in the nucleus. DNA : peptide complexes (DP) transfect MCF-7,
COS and CHOK-1 cells. The addition of cationic liposomes enhances the
uptake of the ternary complexes (DPL) further and also brings about 3–7-fold
enhancement in reporter gene expression compared to DP alone.
Conclusions Recombinant proteins that are heterologous fusions, having
DNA-binding domains and nuclear localization epitopes, generated in this
study have considerable potential to facilitate DNA delivery and enhance
transfection. The domains in these fusion proteins would be promising in the
development of non-viral gene delivery vectors particularly in cells that do
not divide. Copyright  2007 John Wiley & Sons, Ltd.
Keywords gene delivery; non-viral gene delivery; cationic peptides; transfection;
cationic liposomes; transfection; TAT-Mu; recombinant DNA; gene expression
276
Introduction
The natural process of DNA uptake by eukaryotic cells
is inefficient owing to the chemical nature of the gene
and the cell membrane. Invariably, gene delivery has to
be facilitated by carriers both viral and non-viral with
examples of cationic lipids [1,2] and polymers [3,4] in
the non-viral category. These are known to interact with
DNA resulting in the formation of condensed, transfection-
competent complexes. Gene delivery by non-viral methods
is less efficient in cells where the mitotic rates are low.
Cells that divide slowly or differentiated cells are often
resistant to current DNA delivery methods. Understand-
ing the mechanism of nucleic acid uptake by cells coupled
with the development of methods to make this an efficient
process forms the major activity of research in the area of
gene delivery. Although gene transfer methods based on
viruses are successful, challenges such as immunogenicity
[5], toxicity [6] and their ability to integrate into the host
genome [7] have to be overcome. Cationic amphiphiles
have been widely employed for in vitro transfections as
these offer many advantages over that of viral systems in
terms of the ease of production, ability to carry larger DNA
transgenes, stability and safety in clinical trials [2,8]. In
the past decade, the optimization of conditions for trans-
fection has taken a rather long route as a large number
of these non-viral delivery agents and their derivatives
are initially tested empirically for their efficiency. A more
rational approach is to address the issues concerning vari-
ous barriers and pathways leading eventually to the entry
of DNA or transfection complexes of DNA into the nucleus,
which is the ultimate barrier in the transfection process.
This is elegantly discussed in recent reviews pertaining to
intracellular trafficking of nucleic acids, non-viral vectors
and polyplexes [9,10] that forms a major part of non-viral
gene delivery strategies.
Cationic peptides that condense DNA have a tremen-
dous potential to facilitate the passage of DNA through
various mechanisms [11–17] thereby circumventing
cell membrane barriers. The incorporation of pep-
tides in transfection protocols with DNA-binding poten-
tial/nuclear localization motifs has been shown to
improve the transfection efficiency [14,15,18–20] and
may offer solutions to deliver DNA more effectively.
Several adenoviral proteins such as the Mu, with pre-
dominantly basic amino acid residues, have the potential
to bind DNA and also bring about enhancement in cationic
lipid-mediated transfection [13,19]. Mu (µ) is based on
the adenoviral core peptide of 19 amino acids with a
sequence ‘MRRAHHRRRRASHRRMRGG’, twelve of which
are basic, conferring positive charge on the protein. This
peptide is found associated with the adenoviral core com-
plex [21,22]. Work in Miller’s laboratory has shed light
on using systems that are based on the DNA-condensing
ability of synthetic Mu peptide and has also shown its
ability to enhance cationic lipid-mediated transfections
in vitro [13].
The movement of macromolecules within cells is often
aided by nuclear localization sequences and factors that
Copyright  2007 John Wiley & Sons, Ltd.
R. Rajagopalan et al.
recognize and bind them. The use of such sequences in the
preparation of transfection complexes may therefore facil-
itate transport of the complexes efficiently to the nucleus
where gene expression must eventually occur [23], which
is the major rate-limiting step [10,24]. Transfection-
enhancing peptides, synthesized by solid-phase synthesis,
have reportedly shown reporter gene expression in various
cell lines.
Our strategy in this study aims at designing and pro-
ducing small proteins encoded by plasmid constructs by
conventional recombinant DNA strategies and bacterial
fermentation rather than the synthetic route. A basic
domain of TAT was earlier shown to be the minimal
sequence responsible for the cellular and nuclear uptake
which is contributed by potential nuclear localization
sequences (NLSs) in the 11-amino acid epitope YGRKKR-
RQRRR [25]. The TAT moiety has also been shown to
deliver functional fusion proteins in vivo [26]. By com-
bining the Mu motif with TAT, the inherent, desired
properties, such as DNA condensation of Mu and nuclear
localization conferred by TAT, facilitation of transport
across plasma and nuclear membrane, could be integrated
into one molecule. Here, we report the generation of the
plasmid constructs that overproduce TAT-Mu, Mu, and
Mu-Mu as heterologous, his-tag fusion proteins. We eluci-
date their DNA-binding, uptake and transfection potential
into cells. Recombinant fusion proteins TAT-Mu, Mu and
Mu-Mu have the potential to bind plasmid DNA, increase
internalization of transfection complexes into adherent
cells in culture, and thereby enhance transfection.
Materials and methods
Chemicals
1,2-Dioleoyl-3-trimethylammonium propane (DOTAP),
3 beta-(N(N
,N-dimethylaminoethane)carbamoyl) choles-
terol (DC-Chol) and dioleoylphosphatidylethanolamine
(DOPE) were purchased from Avanti Polar Lipids Inc.
N-(Lissamine rhodamine B sulfonyl)-1,2-dihexadecanoyl-
sn-glycero-3-phosphoethanolamine (Rh-DHPE) was from
Molecular Probes. Plasmid pTAT was a kind gift from
Steve Dowdy’s laboratory (UCSD). Plasmid pHind2 was
a kind gift from Akusjarvi of Uppsala University Sweden.
Plasmid DNA pCMV.SPORT-β-gal was from Invitrogen.
BD TALON was from BD Biosciences. The primary 6xhis
monoclonal antibody was from BD Biosciences. FITC
secondary antibodies were from Bangalore Genei. Flu-
orescein DNA-labeling kits for labeling plasmid DNA by
nick translation and random primer labeling were from
Jonaki, India. Ethidium bromide (EtBr) was from Sigma
Aldrich. All other chemicals used were of the highest
purity available.
PCR amplification, sub-cloning
The Mu moiety from the plasmid pHind2 was amplified
using forward and reverse primers with restriction
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Recombinant Fusion Proteins Mediate Efficient Gene Delivery
sites KpnI and EcoRI incorporated in them. This was
subsequently cloned in-frame into the pTAT construct
similarly restricted to generate pTAT-Mu that is inducible
in the presence of 1 mM IPTG when expressed in
an overexpression host BL21(DE3)pLysS. In the Mu
construct, a BamH I digestion of pTAT-Mu followed by
recircularization has deleted the TAT moiety. Inserting a
second copy of Mu amplified by polymerase chain reaction
(PCR) into the Mu construct generated the Mu-Mu
plasmid. These constructs, pTAT-Mu, pMu and pMu-Mu,
were verified for the integrity of the DNA insert sequence
prior to transformation in E. coli. The plasmids were then
transformed into competent E. coli BL21(DE3)pLysS cells,
grown and induced to express the desired protein.
Overexpression and purification of the
recombinant proteins
Recombinant TAT-Mu, Mu and Mu-Mu proteins were
purified from bacterial pellets by urea denaturation fol-
lowed by the use of cobalt affinity matrix (BD TALON;
BD Biosciences). This matrix eliminates the binding of
endogenous contaminating E. coli histidine-rich protein
that co-purifies with our protein in a nickel-affinity matrix,
thereby enriching the protein of our interest during elu-
tion. PD-10 desalting columns subsequently removed urea
employed as a denaturant in the purification protocols.
The final yield was typically 2 mg from a 1000 ml culture
where the protein was greater than 95% homogeneous,
as ascertained by sodium dodecyl sulfate/polyacrylamide
gel electrophoresis (SDS-PAGE). Western blotting of these
purified proteins was carried out using monoclonal anti-
his antibodies against the 6xhis-tag fusion proteins fol-
lowed by binding to the secondary antibody.
DNA binding and electrophoretic
mobility shift assays
Charge neutralization and DNA condensation of the
fusion peptides were observed by electrophoresis of
the complexes on agarose gels. Plasmid DNA pEGFPN3
(0.2–0.4 µg) was complexed with increasing amounts
of the TAT-Mu/Mu/Mu-Mu peptide, to obtain the corre-
sponding charge ratios, in Hepes buffered saline (150 mM
NaCl, 5 mM KCl, 0.75 mM Na2 HPO4, 20 mM Hepes pH
7.4) and incubated for 25 min at room temperature (RT).
The complexes were subsequently analyzed on a 1%
agarose gel [Tris-acetate buffer (TAE) system, pH 8.2]
and stained with EtBr post-electrophoresis.
EtBr exclusion assay
The binding of DNA with TAT-Mu, Mu and Mu-Mu was
studied by monitoring the fluorescence with the fluores-
cent probe EtBr. The intercalation of EtBr into DNA brings
about an increase in the fluorescence quantum yield. Upon
binding and condensing DNA, EtBr is expelled from the
Copyright  2007 John Wiley & Sons, Ltd.
277
DNA-EtBr complex and this displacement of EtBr by the
fusion proteins is reflected as a drop in the fluorescence
signal. EtBr from a stock solution of 1 mg/ml was added to
DNA in a cuvette and the fluorescence was measured using
a fluorescence spectrophotometer (Hitachi-F4000). The
excitation wavelength, λex , was 516 nm and the emission
wavelength was kept at 598 nm (slit width 5 × 5 nm).
Briefly, 2.3 µg of pCMV-β-gal plasmid DNA was added to
500 µl of sample in 20 mM Tris-HCl buffer (pH 7.4) in
a fluorescence cuvette. EtBr (0.23 mg) was added to the
DNA solution and the baseline fluorescence was deter-
mined. The fluorescence intensity obtained upon each
addition was normalized relative to the fluorescence sig-
nal of the DNA-EtBr complex in the absence of the fusion
proteins, which was taken as 100%. The binding of DNA
by the fusion peptides TAT-Mu, Mu and Mu-Mu, respec-
tively, was recorded after each addition at time intervals
of 5 min. The binding efficiency correlates with a drop in
the fluorescence intensity at various time points.
Preparation of complexes for
transfection
Cationic liposomes DOTAP and DC-Chol were prepared
by drying the appropriate amount from their chloroform
stock solutions. Plasmid DNA pCMVβ-gal was purified by
using the Qiagen kit (endotoxin free) using the manu-
facturer’s protocol. Complexes were prepared by taking
0.9 µg of plasmid DNA and varying amounts of the fusion
protein to obtain the desired charge ratio, in Hepes
buffered saline (pH 7.4), and incubated for 25 min at
RT. Transfection experiments were initially performed
with the TAT-Mu, Mu and Mu-Mu peptides and DNA
complexes. In liposome-mediated transfections, cationic
liposomes were added to the pre-incubated DNA : peptide
complexes and further incubated for 25 min at RT. This
ternary complex was then diluted with serum-free Dul-
becco’s modified Eagle’s medium (DMEM) to a final
volume of 300 µl before splitting them into triplicates
of 100 µl volume on a 96-well plate format containing
20 000 cells per well. The cells were incubated for at least
3 h following which complete medium with serum was
added and the reporter gene expression was assayed for
β-galactosidase 24–48 h post-transfection.
Live uptake and immunolocalization of
peptide : DNA complexes by confocal
microscopy
MCF-7 cells were grown on either a chambered cover
glass (Lab-Tek, Nalgene, Nunc International Corp.) or
on cover slips placed on six-well plates. Cells plated to
∼70% confluence were washed with serum-free DMEM
(pH 7.4) and then incubated with the complexes at 1 : 4
DNA : peptide charge ratio, at 37 ◦ C and 5% CO2 . In the
case of fluorescent cationic lipid, DOTAP was labeled
with 5 mol% Rh-DHPE. For the live uptake experiments,
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
278
plasmid DNA was labeled (at 1 : 1 ratio of labeled vs.
unlabeled DNA) by either the random primer labeling
kit or by the nick translation kit using fluorescein-d UTP
from Jonaki, combined with the peptide and incubated
at 37 ◦ C. Subsequently, these were examined by confocal
microscopy to monitor the movement of the complexes
into cells. The image acquisition conditions for all the
experiments were identical.
The immunolocalization experiment was similar to the
live uptake experiment except that, after the incuba-
tion process, cells were exposed to anti-his antibodies
directed against the 6xhis-tag fusion proteins after uptake.
After incubation for 1 h, the cells were fixed with 2%
paraformaldehyde, permeabilized with 0.25% Triton X-
100 followed by immunostaining with anti-his antibodies,
at a dilution of 1 : 200, and incubated overnight at RT.
This was followed by the addition of propidium iodide
(PI) with the secondary antibody FITC-conjugated goat
anti-mouse IgG (anti-mouse used at a dilution o f 1 : 1000
and then analyzed with a confocal microscope (LSM 510
META, Carl Zeiss). Optical sections of cells were taken
using a 63× objective at 0.5 µm intervals. A 518 nm
laser line was used to excite the FITC fluorophore in the
secondary antibody and a 543 nm laser to visualize the
rhodamine fluorophore. Images were collected simulta-
neously for PI staining of the nuclei. The images were
analyzed by taking equal numbers of optical sections in
the central region and then merged.
Results
Construction and purification of
multifunctional DNA carrier proteins
TAT-Mu, Mu and Mu-Mu
The DNA-binding domain of Mu was derived as described
in the ‘Materials and methods’ section by restriction
digestion, PCR amplification and assembly in-frame in
the expression construct pTAT to obtain pTAT-Mu. The
pMu construct was obtained by BamH I excision of the
TAT moiety of pTAT-Mu plasmid. In the case of Mu-
Mu, an additional copy of the Mu moiety was cloned
in-frame into the pMu construct. All the three plasmid
constructs encoding the fusion protein are under the con-
trol of the T7 promoter with an N-terminal histidine tag.
Figure 1A shows the schematic of the fusion constructs
with the placement of 6xhis-tag at the N-terminus. The
presence of the his-tag in-frame with TAT-Mu, Mu and
Mu-Mu facilitates the purification of the recombinant pro-
teins using cobalt affinity matrix. The fusion proteins
were expressed in E. coli BL21(DE3)pLysS. Total bacterial
lysates were obtained with 8 M urea and the recombi-
nant protein purified by elution with 250 mM imidazole.
The fractions were analyzed on 15% SDS gels, pooled
and then the denaturant and imidazole were removed
with PD-10 columns. Figure 1B shows an SDS-PAGE anal-
ysis of the purified proteins appearing as single bands
Copyright  2007 John Wiley & Sons, Ltd.
R. Rajagopalan et al.
with an apparent molecular weight that is slightly larger
than that deduced, based on the DNA sequence of the
plasmid inserts. The recombinant TAT fusion proteins
migrated slower than expected with an apparent increase
of 7–10 kDa on SDS gels. This apparent difference could
be attributed to both the TAT leader sequences as well
as the highly cationic nature of the protein. That the
TAT leader sequences contribute to ∼5 kDa increase in
molecular weight has been reported for other TAT fusion
proteins [27]. The presence of the his-tag of the fusion
proteins was confirmed by immunoblotting using anti-his
antibodies that confirmed the authenticity of the proteins
(inset of Figure 1B, lanes 1–3). The corresponding molec-
ular masses for TAT-Mu, Mu and Mu-Mu, as derived by
matrix-assisted laser desorption/ionization time-of-flight
mass spectrometry (MALDI-TOFMS), are 11.34, 7.78 and
9.83 kDa, respectively.
DNA-binding analysis
Gel retardation assay
We next investigated the ability of recombinant proteins
TAT-Mu, Mu and Mu-Mu to bind plasmid DNA by an
agarose gel retardation assay. Plasmid pEGFPN3 , harbor-
ing the gene encoding the green fluorescent protein (GFP),
was used for the binding assay. The number of arginines
and lysines in the fusion protein were taken into account
in order to calculate the total positive charges on the
molecule. Complexes of DNA and protein were prepared
at various charge ratios and incubated with the plasmid
at 37 ◦ C for 15 min and electrophoresed. In the absence
of protein, which is the control, both relaxed and super-
coiled DNA migrated normally. At charge ratios of 1 : 8
and 1 : 16 (DNA : protein), the ability to bind DNA was
the highest for Mu-Mu while both TAT-Mu and Mu had
similar binding patterns, as seen from the shift in the
mobility of plasmid DNA. Both relaxed and supercoiled
forms of DNA were retarded by their interaction with
the fusion proteins (Figure 1C, panel a – TAT-Mu; panel
b – Mu; panel c – Mu-Mu). In the case of Mu-Mu at 1 : 8
charge ratio, the binding to EtBr is occluded as seen from
the decreased DNA band intensity. At 1 : 16 charge ratio,
the DNA appears to be inaccessible to EtBr binding. The
ability of the peptides to protect DNA in the complex from
nucleases was investigated by incubating the complexes
with DNase I. Protection from nucleases was complete at
charge ratios of 1 : 8 (data not shown).
Displacement of EtBr from DNA by TAT-Mu, Mu and
Mu-Mu: DNA-binding assay
Titrating DNA-EtBr complexes with the fusion proteins
further elucidated the differential DNA-binding charac-
teristics. In this experiment, the addition of the fusion
proteins to the DNA-EtBr complexes resulted in a rapid
decrease in the fluorescence intensity (Figure 1D). The
binding efficiency of the fusion proteins correlates with a
drop in the fluorescence intensity and this loss could be
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Recombinant Fusion Proteins Mediate Efficient Gene Delivery
interpreted as their ability to bind DNA that results in the
displacement of the intercalated EtBr. Although the three
fusion proteins were able to bind DNA, as seen from the
gel binding assays depicted in Figure 1C, the fluorescence
titrations clearly indicate that the efficiency with which
Copyright  2007 John Wiley & Sons, Ltd.
279
the DNA is condensed appears to be different in the case
of Mu-Mu, which is the most efficient, while TAT-Mu
and Mu displayed similar DNA-binding affinity. When we
monitored the binding properties of TAT, we observed
that TAT does not have any effect on EtBr exclusion even
at 1 : 8 charge ratio indicating its lack of binding to DNA
(not shown).
TAT-Mu, Mu and Mu-Mu mediate
transfer of DNA into cells
Live uptake
Having ascertained the DNA-binding characteristics of the
fusion proteins, the internalization of DNA : peptide (DP)
or DNA : peptide : lipid (DPL) complexes into the cells
was investigated by monitoring the live uptake of the
complexes by confocal laser scanning microscopy. After
the application of complexes of DP or DPL, cells were
incubated in the absence of serum for 1 h and replaced
with Hepes/Hanks buffer containing 10% serum. In the
case of the DP complexes, nick translated plasmid DNA
was combined with the fusion proteins and the uptake
into MCF-7 cells was visualized live. The localization of
DP complexes appeared as punctate particles in both the
cytoplasm and the nucleus. It is evident that the DP
complexes, D + TM, D + Mu and D + Mu-Mu, prepared
with TAT-Mu, Mu and Mu-Mu, respectively, facilitate the
Figure 1. Construction and gel analysis of DNA-binding carrier
proteins. (A) Schematic representation of chimeric fusion
constructs encoding (a) TAT-Mu, (b) Mu, and (c) Mu-Mu. The
motifs harboring the TAT sequence YGRKKRRQRRR and/or Mu
sequence MRRAHHRRRRASHRRNRGG are cloned in-frame in a
pTAT expression vector and expressed as N-terminal his-tag
fusion proteins in E. coli. The plasmids are designated as
pTAT-Mu, pMu and pMu-Mu. The Mu construct is created by
a deletion of the TAT moiety of the plasmid construct pTAT-Mu
as described in the ‘Materials and methods’ section. (B) Sodium
dodecyl sulfate/polyacrylamide gel electrophoretic (SDS-PAGE)
analysis of the purified fusion proteins. Lane 1, 2, 3 correspond
to TAT-Mu (2 µg), Mu (∼2 µg) and Mu-Mu (1 µg) after desalting
from PD-10 columns. The molecular weights of the expressed
proteins TAT-Mu, Mu and Mu-Mu are 11.34, 7.78 and 9.23 kDa,
respectively. Inset shows the immunoblot analysis of purified
proteins TAT-Mu, Mu and Mu-Mu after electrophoresis and
Western transfer. The 6xhis-tagged proteins were detected with
monoclonal anti-his antibodies followed by secondary antibody
goat anti-mouse IgG AP conjugate. Detection was by NBT/BCIP
treatment. M, molecular weight standards in kDa. (C) Gel
retardation assay of the fusion proteins TAT-Mu, Mu and Mu-Mu.
In each panel, i.e. (a)–(c), plasmid pEGFPN3 (0.4 µg in (a) and
0.2 µg in (b) and (c)) was incubated in 1× Hepes buffered
saline (pH 7.4) with increasing amounts of the purified protein
corresponding to the charge ratio indicated above each lane.
Samples were electrophoresed on 1% agarose gel in TAE buffer
and visualized by staining with ethidium bromide (EtBr). Control
in each panel pertains to plasmid DNA alone. (D) DNA binding
of TAT-Mu, Mu and Mu-Mu to DNA: Titration curves depicting
the release of EtBr from pCMVβ-gal plasmid DNA upon binding
TAT-Mu (solid squares), Mu (solid triangles) and Mu-Mu (open
circles) fusion proteins in a buffer containing 10 mM TrisHCl
(pH 7.4). 8 µg of protein corresponds to approximately 1 : 2.5
charge ratio (DNA : protein). Details as in the ‘Materials and
methods’ section
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
280 R. Rajagopalan et al.

Figure 2. Live uptake of transfection-competent complexes into MCF-7 cells. (A) Confocal microscope analysis of uptake of
FITC-labeled DNA + fusion protein complexes into MCF-7 cells. Panels from left to right: Control: DNA alone; DNA : TAT-Mu;
DNA : Mu; DNA : Mu-Mu all at 1 : 4 charge ratio DNA : protein. 0.9 µg of DNA was used to make the transfection complexes with
the corresponding protein and applied to cells in the absence of serum. Uptake of the complexes is for a duration of 3 h before
replacement with complete medium containing 10% serum. Thirty minutes before the confocal analyses, the cover slips were placed
in Hepes/Hanks buffer. (B) Confocal microscope analysis of uptake of DNA + fusion protein + lipid complexes into MCF-7 cells.
Complexes were prepared using rhodamine-labeled DOTAP liposomes (5 mol% Rh-DHPE) and the uptake experiment was carried out
as described in the ‘Materials and methods’ section. Panels on the left show the images in the fluorescence channel and those on the
right show the images merged with the transmission channel. Eight optical sections in the middle region, each of 0.5 µm thickness,
were combined to generate the images. D + L = DNA:liposomes at 1 : 1 charge ratio, D + TM + L = DNA + TAT-Mu + liposomes
(1 : 4 : 1), D + Mu + L = DNA + Mu + liposomes (1 : 4 : 1 charge ratio) and D + Mu-Mu + L = DNA + Mu-Mu + liposomes (1 : 4 : 1)

uptake of DNA into the cytoplasm and the nucleus when
compared to the control DNA (extreme panel on the
left) that is labeled DNA alone (Figure 2A). Both nick
translation and random primer labeling methods to label
plasmid DNA yielded identical end results.
In the case of live uptake of DNA : peptide : lipid
complexes (DPL), cationic lipid DOTAP was labeled with
5 mol% Rh-DHPE. The punctate red fluorescence (seen
in all panels) was observed upon uptake and entry
of the particles into the cytoplasm. When compared
to DL (DNA + lipid) complexes, the uptake of DPL
complexes prepared with the three fusion proteins
was enhanced as seen from the increased fluorescence
intensity (Figure 2B). In a situation where both labeled
DNA and rhodamine-labeled lipid were used to prepare
the DPL complexes, with the respective proteins, the
presence of a dual label, originating from the FITC-labeled
plasmid DNA as well as the rhodamine-labeled lipid,
within cells indicated that the DNA and lipid co-localized
(not shown).
Immunocytochemistry of TAT-Mu, Mu and Mu-Mu fusion
proteins
In yet another set of experiments, we also monitored
the intracellular localization of the complexes after an
incubation period of 1 h followed by immunolocalization

Copyright  2007 John Wiley & Sons, Ltd. J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Recombinant Fusion Proteins Mediate Efficient Gene Delivery 281

Figure 3. Immunolocalization and confocal microscopy of fusion proteins. Fusion proteins : plasmid pCMVβ-gal DNA complexes
(DP) prepared with TAT-Mu (a), Mu (b), and Mu-Mu (c) at 1 : 4 charge ratio and incubated for 1 h with cells at 37 ◦ C. After
incubation, cells were fixed followed by immunostaining with anti-his antibodies followed by propidium iodide (PI) addition with
the FITC-conjugated secondary antibody goat anti-mouse IgG. The localization of the his-tagged proteins in the cytoplasm and the
nucleus is visualized as green fluorescence while PI staining of the nucleus is in red. Equal numbers of optical sections in the
middle region, each of 0.5 µm thickness, were combined to generate panels in (a)–(f). Top panels correspond to DP complexes
alone incubated with MCF-7 cells for 1 h while the bottom panels (d–f) correspond to TAT-Mu, Mu and Mu-Mu, respectively, to
which DPL complexes prepared with DC-Chol : DOPE at 1 : 4:1 charge ratios were applied to cells and incubated for 1 h. Details as
described in the ‘Materials and methods’ section

with FITC-labeled secondary antibodies (Figure 3). We
observed that the entry of complexes into the cytoplasm
and nucleus occurs rapidly within 1 h, as seen from the
green fluorescence in the cytoplasm and the nucleus
in the case of DP complexes prepared with TAT-Mu,
Mu or Mu-Mu (Figures 3a–3c). The nuclear staining is
very intense in all the panels where cells were treated
with DPL complexes prepared with DC-Chol : DOPE
(Figures 3d–3f). Panels a and b in the Supplemental
Section S1 correspond to without-primary antibody
control and cell control, respectively (see Supplementary
Material).
TAT-Mu, Mu and Mu-Mu transfection in MCF-7 cells
Having ensured the DNA-binding and uptake properties
of DP and DPL complexes prepared with the recombinant
fusion proteins TAT-Mu, Mu and Mu-Mu, we went on
to determine the relative transfection efficiencies of
fusion proteins. Transient transfections during the initial
characterization of the fusion proteins TAT-Mu, Mu and
Mu-Mu were tested at various charge ratios varying from
1 : 2 to 1 : 16. We have consistently obtained maximum
transfection when cells were transfected with complexes
at 1 : 8 charge ratio of DNA : fusion protein with TAT-Mu,
Mu or Mu-Mu (not shown).
After the initial characterization, transient transfections
were carried out in MCF-7 cells with plasmid pCMVβ-
gal : lipid DL or DPL complexes prepared with TAT-
Mu, Mu or Mu-Mu fusion proteins and reporter gene
expression was assayed 48 h post-transfection, as detailed
in the ‘Materials and methods’ section. We compared
the transfection efficiency of the transfection complex
at the optimized charge ratio of 1 : 8. We chose the
optimized charge ratio of 1 : 1 DNA : lipid as in our
hands DDAB : DOPE, Lipofectin and also DC.Chol : DOPE
gave the best results. Where complexes (DPL) were
prepared with the fusion proteins and the cationic lipid,
we obtained enhanced transfection at 1 : 8:1 charge
ratio of D : P : L. Figures 4a–4c show a representative
set of the results obtained in MCF-7 cells. Complexes
of pCMVβ-gal : fusion protein DP, i.e. D + TM, D + Mu or
D + Mu-Mu, at 1 : 8 charge ratio, were very effective at
transfecting cells when compared to plasmid pCMVβ-gal
alone. Lipopolyplex mixtures DPL prepared with TAT-
Mu (D + TM + L) resulted in ∼7.5-fold enhancement
in transfection when compared to D + TM and 3.5-
fold increase was observed when compared to D + L
(Figure 4a). Similar fold increases were obtained for TAT-
Mu with the cationic lipid Lipofectin or DDAB : DOPE (not
shown).
In the case of the fusion protein Mu, ∼4-fold
increase in the transfection efficiency with D + Mu + L
(L = DC-Chol : DOPE) was obtained when compared with
D + Mu complexes. The fold increase in transfection
efficiency obtained with D + Mu + L prepared with DC-
Chol : DOPE is ∼3-fold when compared with that of
D + L (Figure 4b). We also obtained good transfection
efficiencies with Mu-Mu fusion protein where D +
Mu-Mu + L complexes resulted in ∼7-fold enhancement
in transfection when compared to D + Mu-Mu alone
and ∼3-fold when compared with D + L. Complexes

Copyright  2007 John Wiley & Sons, Ltd. J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
282
Figure 4. Transient transfection of MCF-7 cells with (a) TAT-Mu
(TM), (b) Mu, and (c) Mu-Mu. Complexes prepared with the
fusion proteins and the lipids are at a charge ratio of 1 : 8:1 and
prepared as indicated in the ‘Materials and methods’ section.
Protein estimation of the cell lysates was done to express
β-galactosidase activity in mU after normalizing the protein
values. D-plasmid DNA alone (control); L-lipid DC-Chol : DOPE.
The lipids used were DC-Chol : DOPE in all panels at 1 : 1
charge ratio. The reporter gene assay was carried out 48 h
post-transfection
of the fusion proteins and plasmid DNA harboring the
β-galactosidase or GFP gene also transfect COS, HeLa
and CHOK1 cells efficiently both in the absence and
presence of cationic liposomes (data not shown). Purified
TAT as such did not result in any enhancement in
transfection either in the absence or presence of Lipofectin
Copyright  2007 John Wiley & Sons, Ltd.
R. Rajagopalan et al.
or DC.Chol : DOPE under the conditions tested (not
shown).
A charge ratio of 1 : 1 of DNA to lipid was generally
found to be optimal for several cationic lipids developed
in our laboratory and with those reported in the literature.
Unlike cationic lipids, higher charge ratios greater than
1 : 8, resulting from peptides in a polyplex, were non-toxic.
The effect of charge ratio in a lipoplex in transfection is
not similar. In the case of DPL complexes, a charge ratio
higher than 1 : 8:1 or lower than 1 : 4:1 always resulted in
less efficient transfection (not shown).
We also examined the transfection efficiency of a
control peptide such as poly-l-lysine in the presence and
absence of Lipofectin. Poly-l-lysine 3 kDa and 30 kDa
were both tested in MCF-7 cells and compared to a lipid-
DNA control under the same conditions. The transfection
efficiency with poly-l-lysine in the absence of lipid was
similar to the fusion peptides. The transfection efficiency
with the three fusion proteins was ∼5–8 times higher
when compared to poly-l-lysine in the presence of the
lipid (not shown).
Discussion
Non-viral vectors are comparatively much safer although
not as efficient as viral vectors [8]. It is vital to
develop non-viral vector systems and make them clinically
efficient. The use of peptides to overcome the cellular
barriers of DNA entry into the cell and then into the
nucleus is gaining momentum. When testing out gene
delivery systems, the cell membrane is the first major
barrier. An approach to overcome this problem is to
incorporate motifs with peptides that are capable of
penetrating the cell membrane. Subsequently, the delivery
of DNA from the cytoplasm into the nucleus, that is also an
inefficient process and considered to be the major barrier
in non-dividing cells, is challenging. A strategy to improve
nuclear uptake of DNA is to exploit the cellular nuclear
import machinery where peptides containing an NLS can
be conjugated and thereby facilitate the translocation
of macromolecules across the nuclear envelope inside
cells. Few synthetic peptides with DNA-binding ability or
nuclear import capabilities have been shown to enhance
transfection efficiency. In this report we have presented a
strategy to make combinations of various peptide motifs
using recombinant genetic approaches and to produce
them by heterologous expression. The approach allows
the testing of various combinations of the peptides that
are efficient DNA carriers. Based on this, novel peptide-
based gene carriers could be designed for gene therapy
applications.
Our efforts to identify potential cationic proteins for
transfection led to the construction of plasmids that
encode the fusion proteins TAT-Mu, Mu and Mu-Mu that
are cationic in nature. Mu is characterized by the 19-amino
acid epitope, rich in arginines, and potential to condense
the viral DNA [21]. While Mu helps condense DNA,
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Recombinant Fusion Proteins Mediate Efficient Gene Delivery
the short, basic peptide sequence from the HIV-1 TAT
protein known as TAT or protein transduction domain
(PTD) (residue 47–57) confers the property of cellular
uptake across membranes and nuclear localization. As
evident from our results, the single and double copies of
the Mu moiety in the recombinant Mu and Mu-Mu are
as efficient as TAT-Mu in bringing about reporter gene
expression although the DNA-binding pattern of Mu-Mu
is different from that of TAT-Mu and Mu. Interestingly
the transfection efficiency of these DP complexes was
further enhanced when the transfection complexes were
formulated with cationic liposomes, i.e. DPL. These fusion
proteins have enhanced the transfection of efficient
and well-studied cationic lipid formulations of which
only results obtained with DC-Chol : DOPE have been
presented here.
Although several factors such as the primary sequence,
net charge and sequence context may all be important,
a combination of these could possibly influence gene
transfer in a positive manner. The presence of the TAT
motif in TAT-Mu may thereby synergistically facilitate
the import of plasmid DNA into the nucleus, as it
has better DNA-binding ability than TAT alone. These
properties finally relate to the enhanced transfection
observed for TAT-Mu both in the absence and presence
of cationic lipids. Hyndman et al. have observed the
unexpected negative effect of also tagging TAT with
poly-l-lysine [19] which indicates that the primary
sequence of TAT may play a role in determining peptide-
mediated enhancements in the transfection efficiency in
the presence of liposomes. From the results with poly-
l-lysine, we reasoned that a stretch of positive charges
in the cationic polymer may not be as efficient as
peptides described in this study. The presence of multiple
arginines in the double copy of the Mu-Mu construct
may also explain the stronger binding to DNA as seen by
immunolocalization. Nuclear import of nucleic acids may
be inefficient during cationic lipid-mediated transfection.
A number of attempts to improve the transfection
efficiency by attempting to enhance nuclear uptake of
nucleic acids by using polycationic polymers bearing
peptide sequences with nuclear localizing capabilities
have met with success. The presence of NLSs does
not promote transfection without adequate nucleic acid
charge neutralization and condensation as well. For
example, synthetic TAT alone complexed with DNA was
unable to transfect cells but facilitated transfection in
the presence of cationic liposomes [19]. The domains in
TAT-Mu provide the attributes important in contributing
the individual functions and may lay the foundation for
designing next-generation constructs where issues such
as the binding strength and nuclear localization potential
need to be considered. In the early studies, with cationic
peptides in transfection, the peptides were generated by
solid-phase synthesis. Such labor-intensive methods of
peptide synthesis limit testing of larger peptides and also
peptide combinations. To circumvent these issues, few
laboratories have attempted to synthesize peptides using
recombinant DNA approaches. Heterologous production
Copyright  2007 John Wiley & Sons, Ltd.
283
of peptides rich in basic amino acids, in bacterial systems,
usually results in low yields and in our hands required
the standardization of protocols to further improve
the yields. In our experience, the fusion proteins are
present as insoluble fractions and require denaturation
followed by renaturation to obtain pure preparations.
To aid in purification of the peptides, DNA sequences
corresponding to the peptides were cloned downstream
of the his-tag present at the N-terminus. The presence of
the his-tag at the N-terminus enabled us to confirm the
purity of the peptide preparation by Western blotting.
In addition, the presence of histidine residues in a
transfection complex is known to bring about osmotic lysis
of endosomes upon protonation due to acidification of
endosomes [28] conferring a broad utility of these fusions
in gene therapy applications. All the three fusion proteins,
TAT-Mu, Mu and Mu-Mu, have DNA-binding properties
and also the ability to protect the DNA from nucleases. The
charge ratio required for retardation of the plasmid DNA
and protection from nuclease occurs at 1 : 8 DNA : protein,
indicating complete occlusion of the plasmid by peptides.
Within 1 h of addition of DP, extensive internalization of
peptides along with the plasmid DNA was observed. The
rapid entry of DP and particulate fluorescence suggests
that the endocytic pathway mediates the uptake of the
complexes. The cellular entry of DP or DPL complexes
was confirmed by using fluorescent probes where both
DNA and the lipid were shown to co-localize. Based on
intensity of fluorescence as shown by immunostaining and
a fluorescence-based DNA binding assay, it is apparent
that Mu-Mu shows maximum internalization even in the
absence of lipid. A significant amount of fluorescence was
also seen inside the nucleus suggesting that in 1 h DP or
DPL complexes have entered the nucleus. In comparison
to this, Mu apparently shows more internal fluorescence.
When the internal fluorescence was monitored using the
his-tag antibodies, the results showed much stronger
fluorescence inside the nucleus of MCF-7 cells. The
intensity was more pronounced with DPL than DP. Though
significant particulate fluorescence is seen with DP or
DPL in fixed samples strong fluorescence seen inside the
nucleus may be partially attributed to the process of
fixation per se. During fixation, the cationic peptides may
access the nuclear DNA and bind avidly thereby resulting
in strong fluorescence. However, the results obtained
with live imaging of DP uptake using FITC-labeled DNA
and co-localization experiments clearly indicated that
these peptides have facilitated rapid entry of the DP
or DPL into the MCF-7 cells. Efficient internalization of
plasmid DNA was supported by our data that the fusion
proteins are able to bring about reporter gene expression
in MCF-7 cells (∼100-fold) when compared to plasmid
pCMVβ-gal DNA alone. Between the three fusions, the
efficiency of transfection was enhanced in the presence
of DC-Chol : DOPE contributing to 3–7-fold differences
between DP and DPL for TAT-Mu, Mu and Mu-Mu
fusion proteins. It has been demonstrated that cationic
polymers such as poly-l-lysine and protamine enhance
lipid-mediated transfections in several cell lines in vitro
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
284
[29]. Similarly, several DNA-condensing natural peptide
sequences including Mu [13], HMG [30], and histones
[31], etc., have been demonstrated to enhance reporter
gene expression when included in the transfection
protocols. A recent study from Wels’ laboratory has shown
that recombinant derivatives of human HMGB2 facilitate
non-viral gene delivery [15]. Yet another study used a
similar approach by generating a recombinant polymer-
fusion protein (KH)-FGF2 that had the ability to condense
DNA and mediate transgene expression in mammalian
cells [32].
Synthetic Mu was extensively investigated for its ability
to enhance transfection efficiency in post-mitotic neuronal
cells and also in vivo lung transfection. These studies
have led to LMD formulations containing lipids and Mu
peptide which are efficient transfection formulations at
very low amounts of DNA and in the presence of 100%
serum [13]. Examination of peptide sequences of the
fusion proteins with PSORT II [33] predicts the nuclear
localization scores of the TAT-Mu, Mu and Mu-Mu to be
similar. The contribution of arginines as a key feature in
nuclear membrane translocation and localization has been
reported previously [34,35]. Recently, Miller’s laboratory
has shown that Mu functions as an NLS, since a fusion
of Mu with β-galactosidase was shown to efficiently
accumulate in the nucleus [36]. This is also supported
by our results where all the three fusion proteins have
been shown to internalize efficiently and localize in the
nucleus thereby transfecting various cell lines.
TAT is probably the most investigated among PTDs and
extensive literature is available on its ability to deliver
proteins and genes into cells [37–39]. The domain that
confers this unique property to TAT is the stretch of
YGRKKRRQRRR, rich in arginine and lysine, responsible
for membrane transduction and also nuclear localization
[38]. Mu-Mu alone was more efficient than Tat-Mu
or Mu in binding with the DNA as seen from the
DNA-binding assays and EtBr-exclusion experiments. The
binding ability of Mu is also reflected in its ability to
bring about transfection. Reporter gene expressions with
the three peptides indicated that expression levels are
more or less comparable and correlate with the increased
uptake as seen from the confocal experiments. Although
all the three peptides show efficient internalization of
plasmid DNA, these have been demonstrated to be more
efficient when combined with cationic lipids. Peptides
when combined with lipids have been shown to be more
efficient in bringing about reporter gene expression than
peptides alone.
For more than a decade, linear poly-l-lysine has
been used in gene delivery. Modifications or variations
in the length of the positively charged polymeric
scaffold can have a varying effect on transfection
efficiency thereby contributing to the development of
the oligolysine molecules also as gene delivery agents.
Several studies have employed poly-l-lysine of different
sizes for binding and condensation. These complexes are
effective at concentrations which are usually toxic to
cells [40–42]. Although poly-l-lysine served as a good
Copyright  2007 John Wiley & Sons, Ltd.
R. Rajagopalan et al.
example, subsequent modification was very essential
to increase its utility in gene delivery. A stretch of
positive charges alone may not be sufficient for enhancing
transfection. Also, the presence of nuclear localization
does not promote transfection without adequate nucleic
acid charge neutralization and condensation as well.
In all the reported instances, the inclusion of peptides or
cationic polymers has resulted in more uniform complexes
of 100–200 nm size compared to much larger particles
that result from complexes prepared with cationic lipids.
We have also observed that the size of DPL is ∼150 nm
compared to a few hundreds of nanometers in the
case of DL prepared with DC-Chol : DOPE and detailed
investigations are underway. Though it is expected
that the cellular uptake of the DPL is mediated by
endocytosis, it is not very clear how various components
of DPL interact with various organelles inside the cell.
Preliminary transfection data obtained by the addition of
chloroquine, a lysosomal disrupting agent, enhanced the
reporter gene expression significantly (data not reported)
indicating the involvement of endocytotic pathways in
the uptake of DPL. These results demonstrate that the
internalization of the DNA is enhanced in the presence
of peptide. Although we observed clear enhancements
in transfection efficiencies with DP complexes, the exact
contributions of TAT-Mu, Mu and Mu-Mu in enhancing
DNA delivery needs to be investigated. Nevertheless, the
observed uptake, co-localization and immunolocalization
data when combined would explain the efficiency of
transfection observed by us.
Cationic peptides, although less commonly used, are
however gaining attention due to their versatility and
advantages. By combining cues from viral proteins
with that of current transfection protocols that employ
cationic lipids, novelty combined with multifunctionality
is introduced making them more stable, potential, next-
generation formulations. Cationic lipids do not offer all
the features that are required for a unique delivery system
and most current methods seek to generate formulations
not confined to cationic lipids alone. The benefit of using
recombinant peptides is the versatility that can be added
to the recombinant molecule by genetic engineering
methods to suit various cell types. The addition of cell-
specific epitopes would make these heterologous fusions
therapeutically significant particularly in differentiated
cells as NLS-like features characterize these constructs.
This would also advance the development of vectors
where a combination vector system could be reconstituted
depending on desired protocol.
Conclusions
Recombinant proteins that are heterologous fusions with
DNA binding and nuclear localization epitopes of viral ori-
gin, generated in this study, have considerable potential
to facilitate DNA delivery and enhance transfection. By
using this strategy with cues from viral proteins, versatility
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Recombinant Fusion Proteins Mediate Efficient Gene Delivery
to the molecule can be added to the modules rapidly in a
combinatorial fashion. Unique cell-targeting moieties can
be added to these basic constructs thereby making them a
more specific-potential transfecting agent to suit all gene
delivery strategies. An important aspect that emerges from
this study is that although the recombinant proteins have
been shown to efficiently transfect actively dividing cells,
the presence of NLS domains in these fusion proteins may
confer the potential to transfect non-dividing/resting cells.
This would impact strategies using ex vivo cell therapies.
In situations where the nuclear membrane dissolution
does not take place, the complexes could gain entry into
the nucleus and subsequently become available to the
transcriptional machinery. This would have tremendous
therapeutic significance and would advance studies in the
direction of gene medicine.
Acknowledgements
The authors would like to thank the Department of Biotechnology
for a DBT fellowship to RR. We would also like to thank
Prof. David Dean, Northwestern University, Chicago, for going
through the initial draft of the manuscript. We acknowledge
Soumya Sudhakar, summer student, in the preparation of the
pMu construct.
Supplementary Material
The supplementary electronic material for this paper
is available in Wiley InterScience at: http://www.
interscience.wiley.com/jpages/1099-498X/suppmat/.
References
1. Felgner PL, Gadek TR, Holm M, et al. Lipofection: a highly
efficient, lipid-mediated DNA-transfection procedure. Proc Natl
Acad Sci U S A 1987; 84: 7413–7417.
2. Miller AD. Cationic liposomes for gene therapy. Angew Chem Int
Ed Engl 1998; 37: 1768–1785.
3. Lollo CP, Banaszczyk MG, Mullen PM, et al. Poly-L-lysine-based
gene delivery systems. Synthesis, purification, and application.
Methods Mol Med 2002; 69: 1–13.
4. Zaric V, Weltin D, Erbacher P, Remy JS, Behr JP, Stephan D.
Effective polyethylenimine-mediated gene transfer into human
endothelial cells. J Gene Med 2004; 6: 176–184.
5. Herz J, Gerard RD. Adenovirus-mediated transfer of low
density lipoprotein receptor gene acutely accelerates cholesterol
clearance in normal mice. Proc Natl Acad Sci U S A 1993; 90:
2812–2816.
6. Simon RH, Engelhardt JF, Yang Y, et al. Adenovirus-mediated
transfer of the CFTR gene to lung of nonhuman primates:
toxicity study. Hum Gene Ther 1993; 4: 771–780.
7. Ali M, Lemoine NR, Ring CJ. The use of DNA viruses as vectors
for gene therapy. Gene Ther 1994; 1: 367–384.
8. Kostarelos K, Miller AD. Synthetic, self-assembly ABCD
nanoparticles; a structural paradigm for viable synthetic non-
viral vectors. Chem Soc Rev 2005; 34: 970–994.
9. Zhou R, Geiger RC, Dean DA. Intracellular trafficking of nucleic
acids. Expert Opin Drug Deliv 2004; 1: 127–140.
10. Medina-Kauwe LK, Xie J, Hamm-Alvarez S. Intracellular
trafficking of nonviral vectors. Gene Ther 2005; 12: 1734–1751.
11. Graham FL, van der Eb AJ. A new technique for the assay of
infectivity of human adenovirus 5 DNA. Virology 1973; 52:
456–467.
Copyright  2007 John Wiley & Sons, Ltd.
285
12. Loyter A, Scangos GA, Ruddle FH. Mechanisms of DNA uptake
by mammalian cells: fate of exogenously added DNA monitored
by the use of fluorescent dyes. Proc Natl Acad Sci U S A 1982;
79: 422–426.
13. Murray KD, Etheridge CJ, Shah SI, et al. Enhanced cationic
liposome-mediated transfection using the DNA-binding peptide
mu (mu) from the adenovirus core. Gene Ther 2001; 8: 453–460.
14. Sorgi FL, Bhattacharya S, Huang L. Protamine sulfate enhances
lipid-mediated gene transfer. Gene Ther 1997; 4: 961–968.
15. Sloots A, Wels WS. Recombinant derivatives of the human high-
mobility group protein HMGB2 mediate efficient nonviral gene
delivery. FEBS J 2005; 272: 4221–4236.
16. Demirhan I, Hasselmayer O, Chandra A, Ehemann M, Chan-
dra P. Histone-mediated transfer and expression of the HIV-1
tat gene in Jurkat cells. J Hum Virol 1998; 1: 430–440.
17. Balicki D, Reisfeld RA, Pertl U, Beutler E, Lode HN. Histone
H2A-mediated transient cytokine gene delivery induces efficient
antitumor responses in murine neuroblastoma. Proc Natl Acad
Sci U S A 2000; 97: 11500–11504.
18. Li S, Huang L. In vivo gene transfer via intravenous
administration of cationic lipid-protamine-DNA (LPD)
complexes. Gene Ther 1997; 4: 891–900.
19. Hyndman L, Lemoine JL, Huang L, Porteous DJ, Boyd AC,
Nan X. HIV-1 Tat protein transduction domain peptide facilitates
gene transfer in combination with cationic liposomes. J Control
Release 2004; 99: 435–444.
20. Akita H, Tanimoto M, Masuda T, et al. Evaluation of the
nuclear delivery and intra-nuclear transcription of plasmid DNA
condensed with micro (mu) and NLS-micro by cytoplasmic and
nuclear microinjection: a comparative study with poly-L-lysine.
J Gene Med 2006; 8: 198–206.
21. Anderson CW, Young ME, Flint SJ. Characterization of the
adenovirus 2 virion protein, mu. Virology 1989; 172: 506–512.
22. Hosokawa K, Sung MT. Isolation and characterization of an
extremely basic protein from adenovirus type 5. J Virol 1976;
17: 924–934.
23. Varga CM, Wickham TJ, Lauffenburger DA. Receptor-mediated
targeting of gene delivery vectors: insights from molecular
mechanisms for improved vehicle design. Biotechnol Bioeng
2000; 70: 593–605.
24. Elouahabi A, Ruysschaert JM. Formation and intracellular
trafficking of lipoplexes and polyplexes. Mol Ther 2005; 11:
336–347.
25. Wadia JS, Dowdy SF. Modulation of cellular function by TAT
mediated transduction of full length proteins. Curr Protein Pept
Sci 2003; 4: 97–104.
26. Nagahara H, Vocero-Akbani AM, Snyder EL, et al. Transduction
of full-length TAT fusion proteins into mammalian cells: TAT-
p27Kip1 induces cell migration. Nat Med 1998; 4: 1449–1452.
27. Becker-Hapak M, McAllister SS, Dowdy SF. TAT-mediated
protein transduction into mammalian cells. Methods 2001; 24:
247–256.
28. Kichler A, Leborgne C, März J, Danos O, Bechinger B. Histidine-
rich amphipathic peptide antibiotics promote efficient delivery
of DNA into mammalian cells. Proc Natl Acad Sci U S A 2003;
100: 1564–1568.
29. Gao X, Huang L. Potentiation of cationic liposome-mediated
gene delivery by polycations. Biochemistry 1996; 35:
1027–1036.
30. Zaitsev SV, Haberland A, Otto A, Vorob’ev VI, Haller H,
Bottger M. H1 and HMG17 extracted from calf thymus nuclei
are efficient DNA carriers in gene transfer. Gene Ther 1997; 4:
586–592.
31. Balicki D, Beutler E. Histone H2A significantly enhances in vitro
DNA transfection. Mol Med 1997; 3: 782–787.
32. Hatefi A, Megeed Z, Ghandehari H. Recombinant polymer-
protein fusion: a promising approach towards efficient and
targeted gene delivery. J Gene Med 2006; 8: 468–476.
33. Paul Horton, K.-J.P.T.O.&.K.N. Protein Subcellular Localization
Prediction with WoLF PSORT. Proceedings of the 4th Annual
Asia Pacific Bioinformatics Conference APBC06 39–48, 2006;
Available: http://wolfpsort.seq.cbrc.jp.
34. Futaki S. Arginine-rich peptides: potential for intracellular
delivery of macromolecules and the mystery of the translocation
mechanisms. Int J Pharm 2002; 245: 1–7.
35. Siprashvili Z, Scholl FA, Oliver SF, et al. Gene transfer
via reversible plasmid condensation with cysteine-flanked,
internally spaced arginine-rich peptides. Hum Gene Ther 2003;
14: 1225–1233.
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
286
36. Keller M, Harbottle RP, Perouzel E, et al. Nuclear localisation
sequence templated nonviral gene delivery vectors: investigation
of intracellular trafficking events of LMD and LD vector systems.
Chembiochem 2003; 4: 286–298.
37. Frankel AD, Pabo CO. Cellular uptake of the tat protein from
human immunodeficiency virus. Cell 1988; 55: 1189–1193.
38. Schwarze SR, Ho A, Vocero-Akbani A, Dowdy SF. In vivo protein
transduction: delivery of a biologically active protein into the
mouse. Science 1999; 285: 1569–1572.
39. Jeang KT, Xiao H, Rich EA. Multifaceted activities of the HIV-
1 transactivator of transcription, Tat. J Biol Chem 1999; 274:
28837–28840.
Copyright  2007 John Wiley & Sons, Ltd.
R. Rajagopalan et al.
40. Wu GY, Wu CH. Receptor-mediated in vitro gene transformation
b y a soluble DNA carrier system. J Biol Chem 1987; 262:
4429–4432.
41. Wagner E, Zenke M, Cotton M, Beug H, Birnstiel ML.
Transferrin-polycation conjugates as carriers for DNA uptake
into cells. Proc Natl Acad Sci U S A 1990; 87: 3410–3414.
42. Kircheis R, Wagner E. Polycation/DNA complexes for in vivo
gene delivery. Gene Ther Regul 2000; 1: 95–114.
J Gene Med 2007; 9: 275–286.
DOI: 10.1002/jgm
Supplementary information for the online version
of the paper ‘Recombinant fusion proteins TAT-Mu,
Mu and Mu-Mu mediate efficient non-viral gene
delivery’

Figure S1.