Gene Therapy (2001) 8, 84–87
 2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00
www.nature.com/gt
BRIEF COMMUNICATION
In vivo nuclear delivery of oligonucleotides via
hybridizing bifunctional peptides
LJ Brandén1,2,3, B Christensson3 and CI Edvard Smith1,2,3
1
Clinical Research Center, Karolinska Institutet, Huddinge University Hospital; 2Department of Biosciences, Karolinska Institutet,
NOVUM, Huddinge; 3Department of Immunology, Microbiology, Pathology and Infectious Diseases, Huddinge University Hospital,
Huddinge, Sweden
Linking proteins directly to nucleic acids has been a complex
task. By hybridizing a bifunctional peptide nucleic acid (PNA)
consisting of a nucleic acid binding moiety and a nuclear
localization signal (NLS) we have previously demonstrated
that it is possible to link protein functions directly to nucleic
acids containing a PNA target site. By hybridizing fluor-
escently labeled oligonucleotides to PNA-NLS molecules
and subsequently transfecting different organs in vivo we
demonstrate an active nuclear translocation of the PNANLS/oligonucleotide complex in different mouse organs.
Gene Therapy (2001) 8, 84–87.
Keywords: PNA; NLS; in vivo; transfection; gene therapy
Nuclear entry has been a hurdle to overcome in transfection assays and in gene therapy protocols based on transfection. We have used the properties of PNA1 in combination with the SV40 NLS2 to create a hybrid PNA-NLS
peptide.3 By introducing a target site in the nucleic acid
to be transfected it is possible to hybridize the PNA-NLS
peptide to its cognate sequence thus creating a complex,
bioplex, (biological complexing), with an NLS function. In
the majority of experiments in this study we have not
used pure bioplexes, but instead polyplexes mixing PNANLS and oligonucleotides with the standard transfection
reagent polyethyleneimine (PEI). We have previously
demonstrated the feasibility of this technology under in
vitro conditions.4
In the present study we have investigated the possibility of using bifunctional PNA-NLS peptides to
enhance nuclear uptake in vivo, as well as using the PNANLS molecule as the active transfection reagent. In all
experiments two sets of oligonucleotides were used. The
fluorophore Cy-35 was coupled to an oligonucleotide
which was antisense to the PNA sequence, whereas the
Cy-5 fluorophore was coupled to the corresponding
sense oligonucleotide, serving as an internal control.
In a first set of experiments we investigated the effect
of intradermal/subcutaneous transfection. Following
intradermal/subcutaneous injection in mouse tail of the
PNA-NLS/Cy-3 complex and Cy-5 control oligonucleotide a pronounced nuclear translocation of the Cy-3 oligonucleotide could be detected (Figure 1). In the granular
cell layer an unspecific uptake of oligonucleotides can be
seen as compared with the spinous cell layer. This can
possibly be due to the stratification process of the skin.
Correspondence: L Brandén, Department of Biosciences, Karolinska Institutet, NOVUM, SE-14157, Huddinge, Sweden
Received 24 July 2000; accepted 14 September 2000
However, the nuclear uptake is more pronounced for the
PNA-NLS/Cy-3 oligonucleotide in both the granular and
spinous cell layer. For the spinous cell layer, indicated
with a white line, only the PNA-NLS/Cy-3 oligonucleotide shows a significant presence in the nuclei as can be
seen when comparing panel c and d/e. In panel a, the
spinous cell layer show a pink coloration due to the presence of the PNA-NLS/Cy-3 oligonucleotide and simultaneous DAPI staining of DNA in the nucleus. To show
the nuclear exclusion of non-targeted Cy-5 oligonucleotide, the image in panel e is enhanced. In this panel the
green fluorescence in the cytoplasm is clearly visible,
whereas nuclei are excluded. The intradermal/
subcutaneous transfection shows a clear nuclear translocation of the PNA-NLS/Cy-3 oligonucleotide whereas
the internal Cy-5 control was not present at significant
levels in the nuclei of Cy-3 positive cells. The densely
nucleated area in the tissue section is devoid of either
oligonucleotide, thus indicating that no detectable penetration of the transfection complexes has occurred in this
region. In total more than 300 nuclei were found to be
predominantly positive for Cy-3 fluorescence. We could
not detect any nuclei which were predominantly positive
for Cy-5. To illustrate the nuclear accumulation of the
PNA-NLS/Cy-3 oligonucleotide as compared with the
Cy-5 oligonucleotide we measured the fluorescence along
the indicated white line in Figure 1, panels a–e. The
arrows indicate representative nuclei both in the panels
and in the histogram.
In a next set of experiments we investigated the effect
of bifunctional peptides in intramuscular injections of oligonucleotides. As can be seen in Figure 2, panel c there
is a pronounced fluorescence in the perinuclear/nuclear
region derived from the PNA-NLS/Cy-3 oligonucleotide.
Subsequently, we studied injections into the liver. The
liver is a more difficult organ to analyze due to the background fluorescence in the Cy-3 spectrum. Even though
Nuclear delivery of oligonucleotides
LJ Brandén et al
85
Figure 1 Subcutaneous/intracutaneous injection of oligonucleotides, skin
section: In all animal experiments CBA mouse strain was used and the
animals were killed 8 h after injection. In panel a, a composite image can
be seen consisting of gray scale Differential Interference Contrast (DIC),
blue DAPI-stain, red PNA-NLS/Cy-3 oligonucleotide and green Cy-5
conjugated sense oligonucleotide. The nuclei are shown in panel (b) and
the PNA-NLS hybridized oligonucleotide is shown in red (Cy-3), panel
c, while the sense oligonucleotide in green (Cy-5) is shown in panels d
and e. The fluorescence along the white line was measured and the different color channels (red, green and blue) were plotted against each other.
The colors from panel a to e correspond to the colors indicated in the
histogram. The arrows in panels a-e correspond to the arrows in the histogram indicating representative nuclei. The PNA was synthesized at Oswel
Research Products Ltd (University of Southampton, Southampton, UK).
The complete sequence is GCGCTCGGCCCTTCC-linker-PKKKRKV. The
two fluorochrome-labeled oligonucleotides were synthesized at DNA Technology (Science Park Aarhus, Aarhus, Denmark); antisense (AS) Cy-3labeled, antisense to the PNA-NLS dual-function peptide; and sense (S)
Cy-5-labeled, sense to the PNA–NLS dual-function peptide. The PNANLS peptide, 33 ␮m, and the corresponding complementary DNA oligonucleotide, 33 ␮m, were mixed in 1:1 relationship in water. An equal
amount of competitor DNA oligonucleotide, 33 ␮m, was added to the
hybridization mix after 4 h incubation at room temperature followed by
incubation at room temperature for 12 h. The final volume was 50 ␮l.
The hybridization mix was complexed with 25 kDa PEI at a charge ratio
of 3:1 taking in account the charge of the PNA-NLS. The transfection
mixture was incubated for 20 min to allow for complex formation. The
transfection mixture was supplemented with phosphate buffered NaCl to
physiological levels. For the intradermal/subcutaneous and intrahepatic
injection the volume was 10 ␮l corresponding to 4 ␮g of oligonucleotide.
The color separated images from the subcutaneous/intracutaneous injection were imported into NIH image software and the histograms along
the indicated line of measurement were assembled thus showing the Cy3, Cy-5 and DAPI fluorescence intensities plotted against each other. All
experiments were approved by the animal ethical committee.
the background in the Cy-3 emission spectra is high in
this organ it is possible to separate the fluorescence
derived from the PNA-NLS/Cy-3 oligonucleotide. In Figure 3b we can see that a number of nuclei immediately
surrounding the injection point of the liver are positive
for the PNA-NLS/Cy-3 oligonucleotide. Distal from the
point of injection, the number of positive nuclei decreases
as well as the cells which are positive for the Cy-5
labeled oligonucleotide.
When comparing the transfection efficacy of different
combinations of oligonucleotides, PEI6 and PNA-NLS,
some of the differences are notable. Transfection with oligonucleotides alone gave poor cellular uptake compared
with transfection of oligonucleotides condensed with PEI
or oligonucleotides hybridized with PNA-NLS. PEI condensed oligonucleotides, polyplex, can be detected inside
the cells as large complexes with a non-homogeneous
cellular distribution. Oligonucleotides hybridized with
PNA-NLS, in the absence of PEI, bioplex, condense into
complexes that also are taken up by cells efficiently.
These complexes are dispersed intracellularly in a less
aggregated way than the polyplexes and translocate more
efficiently to the nuclei (Figure 4). In Figure 4, panel a
and c, transfection has been made with PEI or with pure
bioplex respectively. Panel b shows a section of muscle
transfected with bioplex. Panel d is a nucleus from a
Figure 2 Intramuscular injection of oligonucleotides: The PNA-NLS/Cy3 oligonucleotide is shown in red and the sense Cy-5 oligonucleotide in
green. A white box is drawn to indicate where the nucleus is located. In
panel a the phase contrast image is merged with the different color channels for the fluorescent labels. Panel b indicates the nuclei in the tissue
section. Panel c shows the location of the PNA-NLS/Cy-3 oligonucleotide.
Panel d shows the unspecific cellular distribution of the control
oligonucleotide in green. For technical details, see legend of Figure 1. For
the intramuscular transfection, a total volume of 100 ␮l was injected
intramuscularly, corresponding to a total amount of 40 ␮g of oligonucleotides.
Gene Therapy
Nuclear delivery of oligonucleotides
LJ Brandén et al
86
Figure 3 Intrahepatic transfection of oligonucleotides: The arrows indicate the point of injection. (a) Nuclear staining with DAPI indicating
where the nuclei are located; (b) the Cy-3-labeled oligonucleotides in red
hybridized to PNA-NLS molecules; (c) the Cy-5-labeled oligonucleotides
that have not been hybridized to the PNA-NLS molecule. For technical
details, see legend of Figure 1.
Figure 4 Comparison between polyplex- and bioplex-mediated transfection. The panels show the PNA-NLS/Cy-3 oligonucleotide fluorescence.
(a) PEI has been used as transfection reagent and in both panels a and c
HeLa cells were used as target cells. In panels b, c and d the pure bioplex
has been used as the transfection reagent in intramuscular injections.
Panels a and c show the transfected HeLa cells in × 40 magnification.
Panel b is a section of muscle tissue at × 10 magnification indicating the
tissue distribution and intracellular location of the PNA-NLS/Cy-3 oligonucleotide. Panel d is a × 63 magnification of a myofiber nucleus transfected with PNA-NLS/Cy-3 oligonucleotide. All complex formations were
performed according to the technical details in the legend to Figure 1. The
transfection was subsequently performed with or without the addition of
PEI. In the in vitro experiments HeLa cells were used. The cells were
plated 24 h before transfection in six-well dishes and transfected at 100%
confluency. The cells were transfected by adding transfection mix to the
complete cell culture media and mixing by pipetting up and down gently.
The Petri dishes were then swirled momentarily for optimal distribution
of the transfection mixture.
Gene Therapy
myofiber at high magnification. The intranuclear distribution of fluorescently labeled oligonucleotide is very
similar to the intranuclear distribution in Figure 2. The
overall amount of fluorescence does not differ from PEItransfected cells as compared with bioplex-transfected
cells. Thus, this demonstrates that bioplexes themselves
can enter cells efficiently and may also disperse more easily once inside the cells. The data from our bioplex
experiments indicate that transfection by bioplex alone
could be as efficient as polyplex transfection.
Moreover, our data suggest that the physical properties
of the NLS signal are the cause of the efficient transfection of the bioplex. When using a PNA molecule alone
lacking the NLS portion no detectable transfection took
place (unpublished data). The same is true when adding
extended PNA molecules lacking NLS signals. The effect
of the NLS portion is therefore likely to be two-fold: first
it affects the physical properties of the hybridized PNANLS/Cy-3 oligonucleotide enabling intracellular uptake
and second, once internalized, it provides a specific signal for nuclear translocation. Our in vivo findings demonstrate that a PNA molecule conjugated to an SV40 NLS
peptide can work as a nuclear targeting signal when
hybridized to a fluorescently labeled oligonucleotide. The
NLS-mediated nuclear translocation of nucleic acids was
observed in all cell types tested. The differences between
cell types may mainly be related to their ability to take up
the transfection complexes. This issue is being addressed
presently by linking cellular receptor ligands7 to the
nucleic acid via a PNA adaptor. Furthermore, the levels
of karyopherins/importins could vary between different
tissues.8,9 The potential influence of such a variability on
nuclear translocation efficacy could be overcome by linking multiple types of NLS signals to the plasmid. The
variation of NLS receptor species in different tissues
might also be exploited to achieve a higher degree of selectivity in transfection of targeted organs potentially in
combination with tissue-specific promotors. In the intramuscular injection, the oligonucleotide appears to be
located to the perinuclear area, as well as to the nucleus.
This could reflect that different levels of nuclear import
proteins may be present in various tissues. As has been
reported, the different NLS receptor subunits (␣/␤) are
expressed at varying levels in the tissue of mice whereas
the ␤-subunit is maintained at relatively constant levels.
It is hypothesized that the ␣-subunit is regulating the
nuclear transport of proteins containing NLS signals of
different types. In our experiments we have established
the feasibility of an active in vivo targeting of nucleic acid
complexes to the nuclei of cells in vivo. This platform
technology is of potential value for transfections in
general and may also be applied in the context of gene
therapy.10
Acknowledgements
This research has been supported by the Swedish
Research Council.
References
1 Nielsen PE. Peptide nucleic acids: on the road to new gene
therapeutic drugs. Pharmacol Toxicol 2000; 86: 3–7.
Nuclear delivery of oligonucleotides
LJ Brandén et al
2 Yoneda Y et al. A long synthetic peptide containing a nuclear
localization signal and its flanking sequences of SV40 T-antigen
directs the transport of IgM into the nucleus efficiently. Exp Cell
Res 1992; 201: 313–320.
3 Oswel Research Products Ltd, Lab 5005, Medical and Biological
Sciences Building, University of Southampton, UK,
http://www.oswel.com
4 Brandén LJ, Mohamed AJ, Smith CIE. A peptide nucleic acidnuclear localization signal fusion that mediates nuclear transport of DNA. Nature Biotechnol 1999; 17: 784–787.
5 Mujumdar RB et al. Cyanine dye labeling reagents: sulfoindocyanine succinimidyl esters. Bioconj Chem 1993; 4: 105–111.
6 Boussif O et al. A versatile vector for gene and oligonucleotide
7
8
9
10
transfer into cells in culture and in vivo: polyethylenimine. Proc
Natl Acad Sci USA 1995; 92: 7297–7301.
Kircheis R et al. Coupling of cell-binding ligands to polyethylenimine for targeted gene delivery. Gene Therapy 1997; 4: 409–418.
Miyamoto Y et al. Differential modes of nuclear localization signal (NLS) recognition by three distinct classes of NLS receptors.
J Biol Chem 1997; 272: 26375–26381.
Köhler M et al. Cloning of two novel human importin-␣ subunits
and analysis of the expression pattern of the importin-␣ protein
family. FEBS Lett 1997; 417: 104–108.
Liu F, Song YK, Liu D. Hydrodynamics-based transfection in
animals by systemic administration of plasmid DNA. Gene Therapy 1999; 6: 1258–1266.
87
Gene Therapy