Gene Therapy (1997) 4, 1313–1321
 1997 Stockton Press All rights reserved 0969-7128/97 $12.00
On the mechanism of DNA transfection: efficient gene
transfer without viruses
A Coonrod, F-Q Li and M Horwitz
Markey Molecular Medicine Center, Division of Medical Genetics, Department of Medicine, Box 357720, University of Washington,
Seattle, WA 98195, USA
From an investigation of how transfected DNA navigates
from the cell surface to the nucleus, we have developed
a transfection method for primary human fibroblasts that
approaches the efficiency of viruses. We have visually
tracked the subcellular routing of exogenous DNA and find
that all cells in an asynchronous population are surprisingly
competent in the nuclear uptake of DNA, but two steps
practically limit efficient transfection to a minority of cells.
First, regardless of the method used to traverse the cell
membrane – CaPO4 precipitation, lipofection or electroporation – it appears that nuclear transport of DNA requires
routing through endosomes and lysosomes. Apparent
abrogation of endosome–lysosome fusion or translocation
with microfilament or microtubule toxins, respectively,
inhibits the nuclear accumulation of transfected DNA, but
interruption of lysosomal function with protease inhibitors
promotes it. Second, in normal human fibroblasts, which
are refractory to transfection, the exogenous DNA is rapidly
excluded from the nucleus, but in HeLa cells, which are
readily transfected, there is prolonged nuclear stability of
the DNA, indicating the failure in HeLa cells of a mechanism
for the elimination of foreign DNA. These observations
imply strategies for optimizing gene transfer efficiency in
virus-independent approaches to gene therapy.
Keywords: transfection; lysosome; gene therapy
Introduction
Transfection is the process of experimental transfer of
DNA to cultured cells and arose from efforts to reproduce an intact virus from infection of cells with pure viral
DNA.1,2 Several transfection methods have since been
developed, including CaPO4 precipitation, electroporation and lipofection. 3–5 DNA must first cross the cell
membrane. It is thought that DNA entry with CaPO4 precipitation occurs through endocytic vesicles6 that later
fuse with lysosomes. With electroporation, an electric
field creates a transient pore,7 while with lipofection,
DNA–lipid complexes probably enter the cell through
endocytosis.8,9 How the DNA is then transported through
the cytoplasm to the nucleus is unknown. A refined
understanding of transfection could hasten development
of virus-independent gene therapy.
Several measures improve transfection efficiency.
Treatment of some cells with TPA modestly improves
efficiency through an unknown mechanism.10 Glycerol11
and DMSO shock12 slightly improve DNA uptake and are
speculatively attributable to effects upon membrane fluidity.13 Chloroquine modestly improves transfection
efficiency of some cells14 (although it is inhibitory in
others, as reviewed in Ref. 15), presumably by neutralizing lysosomal pH and inhibiting DNA degradation.15,16
Nevertheless, transfection remains unsuitable for gene
therapy applications because it is inefficient compared
with viral vectors and because primary human cells, in
Correspondence: M Horwitz
Received 5 May 1997; accepted 15 July 1997
contrast to transformed cell lines, are typically refractory
to transfection.3,4,6
Here we systematically investigate the mechanism of
transfection. We visually track the course of transfected
DNA and observe that all cells are surprisingly competent in the nuclear uptake of foreign DNA, but there
are two liabilities for DNA destruction that limit efficient
transfection to just a minority of cells: obligate routing
through the lysosome and nuclear purging of foreign
DNA. We have exploited these observations to devise a
transfection method for primary human fibroblasts
employing lysosomal cysteine protease inhibitors that
achieves efficiency approaching viral vectors and is
potentially applicable to in vivo gene therapy.
Results
Visualization of transfected DNA
It has not been possible sensitively and specifically to
stain transfected DNA.17 We therefore developed a protocol for the detection of transfected DNA by PCR incorporation of BrdU triphosphate followed by immunofluorescent staining with anti-BrdU. (A 1522 bp fragment of
the mouse MEF2A gene was arbitrarily selected as the
initial DNA.) The localization of DNA with respect to
lysosomes (counterstained with antiserum to the lysosomal cysteine protease cathepsin B) was observed at several time-points following CaPO4 transfection of primary
human fibroblasts (Figure 1). At 0.5 h after transfection,
the transfected DNA (green) appears in a faint granular
centripetal cytoplasmic pattern that is largely distinct
from the lysosome (red) and nucleus (blue). The timing
Mechanism of DNA transfection
A Coonrod et al
1314
Figure 1 Indirect immunofluorescent staining of transfected BrdU-labeled DNA. Primary human fibroblasts were transfected then fixed and stained at
the indicated time-points following the addition of DNA–CaPO4 precipitate. BrdU-DNA antibodies were detected with fluorescein-conjugated secondary
antibodies (appearing as green). The lysosome is stained by simultaneous indirect immunofluorescent detection of cathepsin B with rhodamine-conjugated
secondary antibodies (appearing as red); nuclei are DAPI counterstained (blue). The overlap of fluorescein and rhodamine appears in yellow. Also shown
is a negative control of cells treated with no primary or secondary antibodies.
and pattern suggest an endosomal distribution.18,19 By
1 h after transfection this staining has intensified and has
grossly concentrated in the perinuclear region, somewhat
overlapping (yellow) with the cathepsin B staining. The
pattern and timing are consistent with central migration
of the endosomes and fusion with lysosomes. 19 At 2 h,
uniform nuclear staining begins to appear. The nuclear
staining reaches peak intensity 8 h following transfection.
At 16 h, the staining are less apparent and nearly absent
from the nucleus. By 40 h, staining is confined to large,
vacuolar cytoplasmic granules, typical in appearance for
residual bodies. This experiment reveals that transfected
Mechanism of DNA transfection
A Coonrod et al
DNA is possibly sorted through endosomes and lysosomes, enters the nuclei of all cells in an asynchronous
population, and is short-lived in the nucleus.
Inhibition of transfection with inhibitors of endosomal
fusion and lysosomal translocation
Endocytosis may be inhibited by blocking the intracellular movement of endosomes and lysosomes with drugs
that depolymerize microtubules such as vinblastine,
while the fusion of endosomes with lysosomes may be
inhibited with cytochalasins, which interrupt microfilament formation.16,19–22 In order to confirm the subcellular
distribution of transfected DNA, we transfected BrdUlabeled DNA in the presence of these inhibitors. In the
top row of Figure 2, the inhibitors were added to the culture media at the same time as the DNA precipitate, and
cells were stained at 8 h following transfection. Vinblastine inhibits intracellular DNA entry, adding support to
the contention that DNA uptake occurs through endocytosis. In the presence of cytochalasin B, the DNA is confined to peripherally distributed nonlysosomal granules
(versus the predominantly nuclear accumulation at 8 h in
the absence of inhibitors, as shown in Figure 1), suggesting that DNA transport is blocked at the point of fusion
of endosomes with lysosomes. In the bottom row of Figure 2, the inhibitors were added to the culture media
3 h following the addition of DNA, allowing time for
endocytosis of the DNA and fusion of endosomes with
lysosomes to be completed. Now in the presence of vinblastine, the DNA is visible only in the apparent lysosomes, and nuclear accumulation of DNA is inhibited.
Treatment with cytochalasin B at this time appears to
have no effect, presumably because the DNA has already
completed transit through the endosomes. These experiments support the possibility that endosomal and lysosomal transport is obligatory in the nuclear delivery of
transfected DNA, although we cannot exclude the possibility that the effects of filament disruption may extend
beyond trafficking.
Endosomal–lysosomal distribution with electroporation
and lipofection
We examined the distribution of DNA when it is introduced by lipofection or electroporation. At 3 h after transfection, the DNA is present in cytoplasmic granular structures that overlap to some degree with lysosomal staining
for cathepsin B (arrowheads, Figure 3). For lipofection, a
significant accumulation of DNA remains apparent at the
cell surface. (For both of these methods there is no
nuclear staining evident at 3 h. Nuclear accumulation
does occur eventually at later time-points, but is less
intense than that observed with CaPO4-mediated transfection (not shown) ). Electroporation of proteins and
dyes produces a uniform cytoplasmic staining, indicating
that the pores provide direct access to the soluble phase
of the cytosol,7 and it is thought that DNA transfected
Figure 2 DNA transfection in the presence of endocytosis inhibitors. Primary human fibroblasts were transfected with BrdU-labeled DNA and stained
8 h following transfection, as per Figure 1, except that 1 mm cytochalasin B or 2.2 mm vinblastine were added to the culture media with (top row) or
3 h after (bottom row) the DNA precipitate.
1315
Mechanism of DNA transfection
A Coonrod et al
1316
Figure 3 Transfection of BrdU-labeled DNA by electroporation and lipofection. Primary human fibroblasts were stained at 3 h following transfection,
as described for Figure 1. (Arrowheads highlight the fluorescein staining for electroporated DNA.)
by electroporation is also cytoplasmically soluble. 23 If the
labeled DNA is spread to the cytosol, we should expect
to detect it with indirect immunofluorescent staining. In
fact, our results indicate that, regardless of the route of
entry across the cell membrane – whether through endocytosis, lipid fusion or electrical pores – transfected DNA
follows a common route to the nucleus through vesicular
trafficking in endosomes and lysosomes.
Improvement of transfection efficiency with lysosomal
cysteine protease inhibitors
From Figure 1 it is apparent that transfected DNA
accumulates in the nucleus of every cell. Transfection is,
however, an inefficient process. For primary human
fibroblasts, less than 10−3 will express proteins from a
transiently transfected expression plasmid.3,4,6 Our
results, taken together with the observations that lysosomes readily hydrolyze DNA24 and that transfected
DNA is substantially damaged,25 suggest that a limiting
step in transfection may be lysosomal destruction or
denaturation.
The original rationale for testing chloroquine treatment
as a means of improving transfection efficiency14 was
based on the observation that chloroquine is an inhibitor
of the lysosomal cysteine protease cathepsin B.26 We have
found that cells genetically deficient in lysosomal protease function, including cathepsin B,27 the carboxypeptidase cathepsin A (absent in the disease galactosialidosis),
or the lysosomal targeting enzyme N-acetylglucosamine1-phosphotransferase (absent in I cell disease), are more
efficiently transfected (unpublished). We therefore tested
whether treatment of cells with lysosomal protease
inhibitors could improve transfection efficiency. Since
chloroquine is an amine, and other inhibitors of lysosomal function have an amine center as a common structural feature,16 we investigated tri-peptidyl aldehyde
cysteine protease inhibitors (LLnL, Z-LLL, LLM and E64) structurally related to chloroquine. These compounds
differ in lipid solubility and specific cysteine protease
inhibition.16,28
Brief treatment of primary human fibroblasts with
LLnL or Z-LLL markedly improves CaPO4 and lipofection-mediated transient transfection of a b-galactosidase
expression vector as judged by X-gal staining (Figure 4a),
activity assay (Figure 4b), Southern (Figure 4c) or Northern blot (Figure 4d) and stable transfection (Figure 4e) of
neo or hygromycin drug resistance markers. E-64 and
LLM are ineffective. The inhibitors must be present in the
tissue culture medium during at least the first 3 h following the addition of DNA (not shown), coincident with the
period during transit of the DNA through the lysosome.
The protease inhibitors likely protect DNA by preventing
the activation of lysosomal nucleases.
Stability of transfected DNA in HeLa nucleus
In contrast to primary human fibroblasts, HeLa cells may
be readily transfected at high efficiency (Figure 5a and
Ref. 3). To determine why, we performed time-course
studies. We found no evident difference in lysosomal
activity (not shown). However, transfected DNA in the
nuclei of HeLa cells is stable at prolonged periods following transfection (Figure 5b), in contrast to its absence at
40 h for primary human fibroblasts (compared with Figure 1). The transfected DNA persists in HeLa cells even
through cell division (note mitotic figures, arrowheads,
Figure 5b). We conclude that the efficient transfectability
of HeLa cells probably results from their incompetence in
the nuclear elimination of transfected DNA. HeLa cells,
however, appear to remain capable of differentiating
between endogenous and exogenous DNA, as the transfected DNA fails to condense into mitotic chromosomes
and uniformly stains in nucleoli (right panel, Figure 5b).
Discussion
Exogenous linear, double-stranded DNA is readily taken
up and accumulates in the nuclei of all cells in an asynchronous population, even though relatively few cells are
detectably transfected (as judged by the criteria of
expression of a foreign gene). Therefore, it does not
appear that delivery of DNA across the cell surface is a
limiting factor in transfection efficiency. Instead, we
hypothesize that it is the integrity of the DNA, once it
has reached the nucleus, that is significant. We identify
two intracellular steps that likely contribute to DNA
destruction. First, independent of the method of transfection, nuclear localization of transfected DNA appears to
require transport through vesicular compartments that
Mechanism of DNA transfection
A Coonrod et al
1317
Figure 4 Improved transfection efficiency in the presence of lysosomal protease inhibitors. Primary human fibroblasts were transiently transfected with
b-galactosidase expression vector in the presence of the indicated inhibitor and activity assayed by (a) X-gal staining, (b) cell extracts, (c) Southern
blot, or (d) Northern blot. Blots were probed with a fragment of the b-galactosidase gene. Primary human fibroblasts were also stably transfected (e)
in the presence of the protease inhibitors with neo and hygromycin drug resistance markers.
we provisionally have identified as endosomes and lysosomes. Our results are then in agreement with a previous
observation that, at least for CaPO4-mediated transfection, DNA moves directly to the nucleus through an
endosomal–lysosomal vesicular transport system.15,29
Second, the transfected DNA is rapidly eliminated from
the nuclei of primary human fibroblasts, through a process that apparently fails in HeLa cells. An implicit conclusion of our studies is that DNA can transit the nuclear
envelope, eliminating a hypothesis that a round of cell
division, with dissolution of the nuclear membrane, is
required for DNA entry into the nucleus.30 We caution
that our observations with labeled DNA have been made
exclusively with linear double-stranded DNA fragments,
and that these might behave differently to plasmid DNA.
For example, microinjection of oligonucleotides indicates
that short, linear, single-stranded DNA rapidly diffuses
to the nucleus whereas large supercoiled plasmid DNA
does not.31–34 A second limitation of our methods is that
BrdU-labeled DNA may be relatively resistant to lysosomal degradation and that the kinetics of nuclear accumulation may not be representative of native DNA; this
possibility cannot be formally excluded.
These observations invite the question of what serves
as the lysosomal and nuclear localization signals for
DNA. Except during mitosis, at which time the DNA
is highly condensed into nucleoprotein chromatin
structures, DNA is never found within the cytoplasm
or outside of membrane-bound organelles (nuclei,
mitochondria or chloroplasts). In contrast, since both
single- and double-stranded RNA is freely soluble in the
cytoplasm (as a component of, or when complexed with,
free ribosomes), it is reasonable to propose that the
chemical signals for the subcellular localization of DNA
reside on the deoxyribose moiety. For the case of polypeptides, the lysosomal localization signal is post-translational modification with mannose-6-phosphate 35 and,
so with DNA, similar recognition of a carbohydrate unit
might also occur. We are currently systematically chemically substituting DNA in order to define the positions
on the molecule relevant for subcellular routing.
Primary human fibroblasts are particularly refractory
to transfection,36–38 but because these cells are not transformed, they are useful in the study of normal cell function37 and may be attractive targets for DNA transfer in
gene therapy.38 Brief treatment of primary human fibroblasts with either of two lysosomal cysteine protease
inhibitors, LLnL or Z-LLL, markedly improves transfection efficiency. The presumed effect of the protease
inhibitors on transfection efficiency is to prevent the
Mechanism of DNA transfection
A Coonrod et al
1318
Figure 5 (a) Efficient CaPO4 transfection of HeLa cells in the absence of the inhibitor. Cells were transiently transfected with b-galactosidase marker
and X-gal stained. (b) Prolonged stability of transfected DNA in the nucleus of HeLa cells 40 h following transfection. BrdU-labeled DNA was detected
as described for Figure 1. Areas staining with both fluorescein and DAPI (merge) are depicted in shades of red and pink. Arrowheads indicate two (of
several) mitotic figures.
maturation of lysosomal nucleases and thereby the
hydrolysis of DNA. Although it is not known whether
lysosomal nucleases are proteolytically activated from
zymogen precursors, there is evidence for a proteolytic
cascade in the maturation of hydrolases in the yeast vacuole, which is the equivalent of the lysosome.39
In contrast to primary human fibroblasts, HeLa cells
are readily transfectable at high efficiency. This distinction cannot be accounted for by obvious differences in
DNA uptake or lysosomal processing. Time-course studies in primary human fibroblasts suggest that the bulk of
transfected DNA is rapidly eliminated from the nucleus
following a period of brief accumulation. This purge
takes place before mitosis resumes following transfection.
A significant finding in HeLa cells is that the transfected
DNA persists in the nucleus for a prolonged period of
time. A possible explanation for the high efficiency transfection of HeLa cells is that they are defective in a second
step involving the nuclear elimination of exogenous
DNA that normally functions in primary cells. In primary
fibroblasts, the transfected DNA is persistent for a longer
period of time in the nucleus following treatment with
the protease inhibitors (not shown), conceivably suggesting that DNA damage serves as the signal promoting the
removal of the exogenous DNA. In fact, it is possible that
HeLa cells have a particular defect in at least the recognition of the BrdU-labeled DNA used in our experiments
or, more generally, damaged DNA. Indeed, transfection
causes cells to transiently arrest in G1 and induces p53.40
In HeLa cells, p53 is inactive as a consequence of complex
formation from the papillomavirus E6 and E7 proteins.41
Interruption of the DNA damage inducible G1 arrest
with drugs, such as caffeine, that override this checkpoint, may be a potential secondary strategy for further
improvements in transfection efficiency that we are currently testing. A supportive observation is that
expression of papillomavirus E6 or E7 proteins promotes
the genomic integration of foreign DNA through just
such a mechanism.42
There is some practical import of our results with
respect to the refinement of strategies for human therapy
by gene replacement or antisense. The pharmacological
properties of tri-peptidyl aldehyde cysteine protease
inhibitors have been investigated as potential treatments
for muscular dystrophy, cancer and cataracts; in vivo inhibition of the lysosome in animals can be achieved with
systemic administration of nontoxic doses.43 For patients
undergoing gene therapy, it is conceivable that treatment
with a pulse of these drugs could improve gene transfer
rates. We would further suggest that approaches aimed
at increasing gene delivery to the cell surface through,
for example, the manipulation of cell receptors, are less
likely to result in substantial gains in efficiency than are
methods aimed at facilitating the intracellular routing of
DNA to the nucleus. Since transit through the lysosome
is an unavoidable prerequisite for ultimate nuclear delivery, hyper-glycosylation of DNA to increase its lysosomal
uptake might also improve gene transfer efficiency. Continued refinement of high efficiency transfection protocols may offer an eventual means to gene therapy independent of the limitations and complications of viral
vectors.
Materials and methods
Cell culture
Primary human foreskin fibroblasts were obtained at
passage six and were a gift from A Saulewicz (University
of Washington, Seattle, WA, USA). (Identical results were
obtained with primary fibroblasts obtained through dermal biopsy from three different individuals.) They were
cultured in MEM medium with Earle’s BSS, 2 × concentration of essential and nonessential amino acids and vit-
Mechanism of DNA transfection
A Coonrod et al
amins, and 20% fetal bovine serum in the presence of 1%
penicillin and streptomycin, with the final pH of the
media adjusted to 7.40 with NaOH. HeLa cells were
obtained from ATCC (Rockville, MD, USA). These cells
were grown in DMEM with 1% penicillin and streptomycin, supplemented with 10% fetal bovine serum. Cells
were grown in a humidified 37°C incubator with 5% CO2
as adherent cultures on plastic dishes and passaged at
confluence by trypsin–EDTA treatment. All cell culture
reagents
were purchased
from
GIBCO
BRL
(Gaithersburg, MD, USA).
Transfection
Lipofectin was obtained from GIBCO BRL and used
according to the manufacturer’s instructions with OptiMEM I reduced serum media. Calcium phosphate transfection was performed by adding 5 mg of DNA (in a volume of TE buffer no greater than 20 ml) to 125 ml of 250
mm CaCl 2. To this was added dropwise an equal volume
of 2 × HBS (containing 1.6% NaCl, 0.021% Na2HPO4 , 1.3%
Hepes, pH 7.0). The mixture was allowed to stand for
approximately 1 min and then added to a 35-mm tissue
culture plate. Cells were approximately 75% confluent
(105 cells) and were seeded approximately 15 h before
transfection. The mixture remained undisturbed on the
cells for 15 h, and then cells were fed on the following
day with fresh media. Twenty-four hours later the cells
were stained or harvested for b-galactosidase assay or
nucleic acid extraction. For stable transfection of human
fibroblasts, plates of cells were transfected as above with
15 mg of pSV2neo44 or ptgCMVHy/TK,45 and 24 h following transfection, trypsinized and replated in a 250-mm
plate in the presence of 0.25 mg/ml active concentration
of G418 (GIBCO BRL) or 7.5 U/ml hygromycin B
(Calbiochem, La Jolla, CA, USA), respectively. After 7
days, the concentration of G418 and hygromycin B was
reduced by one half. Cells were cultured for a total of 21
days, then stained with crystal violet in order to count
resistant colonies. For electroporation, 107 trypsinized
human fibroblasts were resuspended in 400 ml PBS containing approximately 6 mg DNA (including 5 mg
pCS2+b-gal plasmid carrier DNA plus approximately
1 mg BrdU PCR products (see below)), chilled on ice for
10 min then discharged in a 2-mm gap cuvette with a
BTX Electro Cell Manipulator 600 (San Diego, CA, USA)
with field strength of 3.0 kV/cm and pulse length of 0.5
ms, chilled again on ice for 10 min then immediately
replated. The inhibitors were added to the tissue culture
media with the CaPO4–DNA or lipofectin–DNA mixture,
and were not replaced after the cells were later fed with
fresh media. Z-LLL28 was provided by MyoGenics
(Cambridge, MA, USA). LLnL, LLM, E-64, cytochalsin B
and vinblastine sulfate were from Calbiochem. Stock solutions were dissolved in DMSO or, for vinblastine, methanol. All other reagents were from Sigma (St Louis,
MO, USA). All plasmids were prepared on Qiagen
(Chatsworth, CA, USA) columns.
X-gal staining and b-galactosidase assay
The b-galactosidase test vector was pCS2+b-gal, containing a simian CMV promoter–enhancer driving the E. coli
lacZ gene.46,47 Cells were fixed on plates by incubation
for 5 min in 4% paraformaldehyde in phosphate-buffered
saline (PBS) and b-galactosidase activity was detected by
staining for 2 h in PBS to which was added 5 mm
K4Fe(CN)6, 5 mm K3Fe(CN)6, 1 mm MgCl2 , and 1 mg/ml
X-gal. Quantitative b-galactosidase assay on ONPG substrate was performed by freeze–thaw lysis of cells collected by scraping from tissue culture plates as
described.48 All quantitative assays were performed at
least in triplicate and confidence intervals representing
the standard error of the mean are shown.
5-Bromo-2 ′deoxyuridine (BrdU) labeling of DNA
PCR with a 1:1 ratio of dTTP and BrdUTP was used to
label DNA specifically.49 The 1522 bp DNA was amplified from the mouse MEF2A gene (GenBank X63381)
using primers corresponding to nucleotides 415–434 and
1937–1919. (Similar results were found using a 1340 bp
fragment of the mouse MCK gene.) The reaction was performed in a total volume of 50 ml containing 1 × PCR
buffer (50 mm KCl, 10 mm Tris-HCl, pH 8.3, 1.5 mm
MgCl2, 0.01% (w/v) gelatin), 200 mm of each of the four
dNTPs, 200 mm BrdUTP (Sigma), 10 ng/ml of each
primer, 2 ng/ml of template DNA and 2 U of AmpliTaq
DNA polymerase (Perkin Elmer, Norwalk, CT, USA)
using the following conditions: denaturation (94°C,
1 min) annealing (55°C, 1 min), and extension (72°C,
3 min) for 36 cycles. The ethanol precipitated products of
one PCR reaction combined with 5 mg pCS2+b-gal plasmid DNA as a carrier were used as a source of labeled
DNA for transfection per each 35-mm plate of cells.
Immunofluorescent staining
Transfected BrdU-labeled DNA was detected by
immunofluorescent staining with BrdU-DNA antibody
and simultaneous costaining with cathepsin B antiserum
to counterstain lysosomes. Cells were fixed on the plastic
dish for 3 min at room temperature with 50% methanol–
50% acetone, washed three times in PBS, and incubated
for 60 min at room temperature with 0.4 mg/ml of mouse
monoclonal IU-4 BrdU-DNA antibody (Caltag, South San
Francisco, CA, USA) and 1:200 of rabbit cathepsin B antiserum (Athens Research, Athens, GA, USA) in PBS and
secondary detection by incubation for 30 min at room
temperature with 20 mg/ml fluorescein- or rhodamineconjugated, noncrossreactive, donkey anti-mouse or antirabbit, respectively, antibodies (Jackson Labs, West
Grove, PA, USA) in PBS. (We found acid treatment to
denature DNA to be an unnecessary step with the IU-4
antibody.) Nuclei were counterstained by incubation for
3 min in 0.5 mg/ml DAPI in PBS. Epifluorescence and
phase contrast photomicroscopy were performed on a
Zeiss Photomicroscope III (Thornwood, NY, USA) with
25 × plan-neo objective using 45 s exposure on Kodak
Ektachrome P1600 film (Rochester, NY, USA), processed
by one stop push of E6 protocol, scanning of transparencies on to a Kodak Photo CD, and image compositing with Adobe Photoshop (to crop, resize, adjust
brightness/contrast and overlay images) and Illustrator
software (Mountain View, CA, USA) for Macintosh with
images printed on a Tektronix IIsd dye sublimation
printer (Beaverton, OR, USA). Images were collected serially with fluorescein, rhodamine and phase/DAPI bandwidth filters.
Nucleic acid blotting
DNA and RNA were purified 40 h after CaPO4 transient
transfection of primary human fibroblasts with 5 mg
pCS2+b-gal using the TRIzol reagent (GIBCO BRL)
1319
Mechanism of DNA transfection
A Coonrod et al
1320
according to the manufacturer’s instructions. The total
cellular DNA was digested with EcoRI and BamHI. Four
micrograms of each DNA sample and 5 mg of each RNA
sample were electrophoresed in 1% agarose gels without
and with, respectively, formaldehyde, and blotted on to
Hybond N membranes (Amersham, Arlington Heights,
IL, USA). Consistency of loading for Southern blot was
verified by ethidium bromide staining and consistency of
loading and transfer for Northern blot by methylene blue
staining. The probe was a random-prime 32P-labeled
(using NEBlot kit; New England BioLabs, Beverly, MA,
USA) BamHI–EcoRI fragment of lacZ from pCS2+b-gal.
Hybridization was performed in a Denhardt’s solutionbased buffer at 65°C followed by high-stringency wash
in 0.1 × SSC buffer at 65°C.48
15
16
17
18
19
20
Acknowledgements
We thank B Clurman (Fred Hutchinson Cancer Research
Center, Seattle, WA) for discussion, A Saulewicz
(University of Washington, Seattle) for a gift of primary
human fibroblasts, and V Palombella and Y-T Ma
(MyoGenics, Cambridge, MA) for a gift of Z-LLL. This
work was supported by the Markey Foundation, Howard
Hughes Medical Institute pilot projects grant, and PHS
grants NICHD HD01080–03 and NHLBI SP01 HL53750–02.
21
22
23
24
References
25
1 Ostro MJ, Giacomoni D. Liposomes as a tool in molecular
biology: a comparison to other methodologies. In: Ostro MJ (ed).
Liposomes. Marcel Dekker: New York, 1983, pp 145–208.
2 Wolff JA, Lederberg J. A history of gene transfer and therapy.
In: Wolff JA (ed). Gene Therapeutics. Methods and Applications of
Direct Gene Transfer. Birkhauser: Boston, 1994, pp 3–25.
3 Guy J, Dubravka D, Antoniou M. Delivery of DNA into mammalian cells by receptor-mediated endocytosis and gene therapy. Mol Biotechnol 1995; 3: 237–248.
4 Kriegler M. Gene Transfer and Expression: a Laboratory Manual.
WH Freeman and Company: New York, 1990.
5 Keown WA, Campbell CR, Kucherlapati RS. Methods for introducing DNA into mammalian cells. Meth Enzymol 1990; 185:
527–537.
6 Cotten M et al. High-efficiency receptor-mediated delivery of
small and large (48 kilobase gene constructs using the endosome-disruption activity of defective or chemically inactivated
adenovirus particles. Proc Natl Acad Sci USA 1992; 89: 6094–6098.
7 Weaver JC. Electroporation: a general phenomenon for manipulating cells and tissues. J Cell Biochem 1993; 51: 426–435.
8 Singhal A, Huang L. Gene transfer in mammalian cells using
liposomes as carriers. In: Wolff JA (ed). Gene Therapeutics.
Methods and Applications of Direct Gene Transfer. Birkhauser:
Boston, 1994, pp 118–142.
9 Gao X, Huang L. Cationic liposome-mediated gene transfer.
Gene Therapy 1995; 2: 710–722.
10 Reston JT, Gould-Fogerite S, Mannino RJ. Differential effects of
the phorbol ester TPA on DNA-mediated transfection in a variety of cell lines. Biochim Biophys Acta 1993; 1177: 49–53.
11 Frost E, Williams J. Mapping temperature-sensitive and hostrange mutations of adenovirus type 5 by marker rescue. Virology
1978; 91: 39–50.
12 Stow ND, Wilkie NM. An improved technique for obtaining
enhanced infectivity with herpes simplex virus type 1 DNA.
J Gen Virol 1976; 33: 447–458.
13 Sabelnikov AG. Nucleic acid transfer through cell membranes:
towards the underlying mechanisms. Prog Biophys Mol Biol 1994;
62: 119–152.
14 Luthman H, Magnusson G. High efficiency polyoma DNA
26
27
28
29
30
31
32
33
34
35
36
37
transfection of chloroquine treated cells. Nucleic Acids Res 1983;
11: 1295–1308.
Chang PL. Calcium phosphate-mediated DNA transfection. In:
Wolff JA (ed). Gene Therapeutics. Methods and Applications of
Direct Gene Transfer. Birkhauser: Boston, 1994, pp 157–179.
Seglen PO. Inhibitors of lysosomal function. Meth Enzymol 1983;
96: 737–764.
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 USA 1982; 79:
422–426.
Pastan IH, Willingham MC. Journey to the center of the cell:
role of the receptosome. Science 1981; 214: 504–509.
van Deurs B, Holm PK, Kayser L, Sandvig K. Delivery to lysosomes in the human carcinoma cell line HEp-2 involves an actin
filament-facilitated fusion between mature endosomes and preexisting lysosomes. Eur J Cell Biol 1995; 66: 309–323.
Matteoni R, Kreis TE. Translocation and clustering of endosomes and lysosomes depends on microtubules. J Cell Biol 1987;
105: 1253–1265.
Davies P, Allison AC. Effects of cytochalasin B on endocytosis
and exocytosis. Front Biol 1978; 46: 143–160.
Cremaschi D et al. Endocytosis inhibitors abolish the active
transport of polypeptides in the mucosa of the nasal upper concha of the rabbit. Biochim Biophys Acta 1996; 1280: 27–33.
Bertling W, Hunger-Bertling K, Cline MJ. Intranuclear uptake
and persistence of biologically active DNA after electroporation
of mammalian cells. J Biochem Biophys Meth 1987; 14: 223–232.
De Duve C et al. Tissue fractionation studies. 6. Intracellular distribution patterns of enzymes in rat-liver tissue. Biochem J 1955;
60: 605–617.
Wake CT et al. How damaged is the biologically active subpopulation of transfected DNA? Mol Cell Biol 1984; 4: 387–398.
Wibo M, Poole B. Protein degradation in cultured cells. II. The
uptake of chloroquine by rat fibroblasts and the inhibition of
cellular protein degradation and cathepsin B1. J Cell Biol 1974;
63: 430–440.
Gogos JA et al. Gene trapping in differentiating cell lines: regulation of the lysosomal protease cathepsin B in skeletal myoblast
growth and fusion. J Cell Biol 1996; 134: 837–847.
Rock KL et al. Inhibitors of the proteasome block the degradation of most cell proteins and the generation of peptides
presented on MHC class I molecules. Cell 1994; 78: 761–771.
Orrantie E, Chang PL. Intracellular distribution of DNA
internalized through calcium phosphate precipitation. Exp Cell
Res 1990; 190: 170–174.
Takeshita S et al. Increased gene expression after liposomemediated arterial gene transfer associated with intimal smooth
muscle cell proliferation. In vitro and in vivo findings in a rabbit
model of vascular injury. J Clin Invest 1994; 93: 652–661.
Leonetti JP et al. Intracellular distribution of microinjected antisense oligonucleotides. Proc Natl Acad Sci USA 1991; 88: 2702–
2706.
Chin DJ et al. Rapid nuclear accumulation of injected
oligodeoxyribonucleotides. New Biol 1990; 2: 1091–1100.
Zabner J et al. Cellular and molecular barriers to gene transfer
by a cationic lipid. J Biol Chem 1995; 270: 18997–19007.
Clarenc J-P, Lebleu B, Leonett J-P. Characterization of the
nuclear binding sites of oligodeoxyribonucleotides and their
analogs. J Biol Chem 1993; 268: 5600–5604.
Kornfeld S, Sly WS. I. Cell disease and pseudo-Hurler polydystrophy: disorders of lysosomal enzyme phosphorylation and
localization. In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds).
The Metabolic and Molecular Bases of Inherited Disease, 7th edn.
McGraw-Hill: New York, 1995, pp 2495–2508.
Baum C, Forster P, Hegewish-Becker S, Harbers K. An optimized electroporation protocol applicable to a wide range of cell
lines. BioTechniques 1994; 17: 1058–1060.
Green MHL, Lowe JE, Teufel R, Petit-Frere C. Reducing the concentration of selected marker improves efficiency of cotransfer
of unselected DNA into SV40-transformed human fibroblasts.
Exp Cell Res 1991; 192: 298–301.
Mechanism of DNA transfection
A Coonrod et al
38 Veelken H et al. Primary fibroblasts from human adults as target
cells for ex vivo transfection and gene therapy. Hum Gene Ther
1994; 5: 1203–1210.
39 Rupps S, Wolf DH. Biogenesis of the yeast vacuole (lysosome).
The use of active-site mutants of proteinase yscA to determine
the necessity of the enzyme for vacuolar proteinase maturation
and proteinase yscB stability. Eur J Biochem 1995; 231: 115–125.
40 Renzing J, Lane DP. p53-Dependent growth arrest following calcium phosphate-mediated transfection of murine fibroblasts.
Oncogene 1995; 10: 1865–1868.
41 Scheffner M, Munger K, Byrne JC, Howley PM. The state of the
p53 and retinoblastoma genes in human cervical carcinoma cell
lines. Proc Natl Acad Sci USA 1991; 88: 5523–5527.
42 Kessis TD, Connolly DC, Hedrick L, Cho KR. Expression of
HPV16 E6 or E7 increases integration of foreign DNA. Oncogene
1996; 13: 427–431.
43 Tamai M et al. In vitro and in vivo inhibition of cysteine proteinases by EST, a new analog of E-64. J Pharmaco Dyn 1986; 9:
672–677.
44 Southern PJ, Berg P. Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40
early region promoter. J Mol Appl Genet 1982; 1: 327–341.
45 Lupton SD, Brunton LL, Kalberg VA, Overell RW. Dominant
positive and negative selection using a hygromycin phosphotransferase-thymidine kinase fusion gene. Mol Cell Biol 1991; 11:
3374–3378.
46 Turner DL, Weintraub H. Expression of achaete-scute homolog
3 in Xenopus embryos converts ectodermal cells to a neural fate.
Gene Devel 1994; 8: 1434–1447.
47 Rupp RA, Weintraub H. Xenopus embryos regulate the nuclear
localization of XMyoD. Gene Devel 1994; 8: 1311–1323.
48 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual 2nd edn. Cold Spring Harbor Laboratory Press:
Cold Spring Harbor, New York, 1989.
49 Muhlmann-Diaz M, Dullea RG, Bedford JS. Application of 5bromo-2′ deoxyuridine as a label for in situ hybridization.
BioTechniques 1996; 21: 82–86.
1321