Structure-function correlation of chloroquine and analogues as transgene expression enhancers in nonviral gene delivery

Article abstract of DOI:10.1021/jm060736s

To understand how chloroquine (CQ) enhances transgene expression in polycation-based, nonviral gene delivery systems, a number of CQ analogues with variations in the aliphatic amino side chain or in the aromatic ring are synthesized and investigated. Our studies indicate that the aliphatic amino moiety of CQ is essential to provide increased gene expression. Further, the enhancements are more dramatically affected by changes to the aromatic ring and are positively correlated to the strength of intercalation between DNA and the CQ analogues. Quinacrine (QC), a CQ analogue with a fused acridinyl structure that can strongly intercalate DNA, enhances transfection similarly to CQ at a concentration 10 times lower, while N⁴-(4-pyridinyl)-N ¹,N¹-diethyl-1,4-pentanediamine (CP), a CQ analogue that has a weakly intercalating pyridinyl ring, shows no effect on gene expression. Subtle change on the 7-substituent of the chloroquine aromatic structure can also greatly affect the ability of the CQ analogues to enhance transgene expression. Transfection in the presence of N⁴-(7-trifluoromethyl-4- quinolinyl)-N¹,N¹-diethyl-1,4-pentanediamin e (CQ7a) shows expression efficiency 10 times higher than in the presence of CQ at-same concentration, while transfection in the presence of N⁴-(4- quinolinyl)-N¹,N¹-diethyl-1,4-pentanediamine (CQ7b) does not reveal any enhancing effects on expression. Through a number of comparative studies with CQ and its analogues, we conclude that there are at least three mechanistic features of CQ that lead to the enhancement in gene expression: (i) pH buffering in endocytic vesicles, (ii) displacement of polycations from the nucleic acids in polyplexes, and (iii) alteration of the biophysical properties of the released nucleic acid.

Full text of DOI:10.1021/jm060736s

6522
J. Med. Chem. 2006, 49, 6522-6531
Structure-Function Correlation of Chloroquine and Analogues as Transgene Expression
Enhancers in Nonviral Gene Delivery
Jianjun Cheng,†,‡ Ryan Zeidan,§ Swaroop Mishra,§,| Aijie Liu,† Suzie H. Pun,†,| Rajan P. Kulkarni, Gregory S. Jensen,†
Nathalie C. Bellocq,† and Mark E. Davis*,§
Insert Therapeutics, 2585 Nina Street, Pasadena, California 91107, Department of Materials Science and Engineering,
UniVersity of Illinois at Urbana-Champaign, Urbana, Illinois 61801, DiVision of Chemistry and Chemical Engineering, and
DiVision of Biology, California Institute of Technology, Pasadena, California 91125, and Department of Bioengineering,
UniVersity of Washington, Seattle, Washington 98195
ReceiVed June 19, 2006
To understand how chloroquine (CQ) enhances transgene expression in polycation-based, nonviral gene
delivery systems, a number of CQ analogues with variations in the aliphatic amino side chain or in the
aromatic ring are synthesized and investigated. Our studies indicate that the aliphatic amino moiety of CQ
is essential to provide increased gene expression. Further, the enhancements are more dramatically affected
by changes to the aromatic ring and are positively correlated to the strength of intercalation between DNA
and the CQ analogues. Quinacrine (QC), a CQ analogue with a fused acridinyl structure that can strongly
intercalate DNA, enhances transfection similarly to CQ at a concentration 10 times lower, while N4-(4-
pyridinyl)-N1,N1-diethyl-1,4-pentanediamine (CP), a CQ analogue that has a weakly intercalating pyridinyl
ring, shows no effect on gene expression. Subtle change on the 7-substituent of the chloroquine aromatic
structure can also greatly affect the ability of the CQ analogues to enhance transgene expression. Transfection
in the presence of N4-(7-trifluoromethyl-4-quinolinyl)-N1,N1-diethyl-1,4-pentanediamin e (CQ7a) shows
expression efficiency 10 times higher than in the presence of CQ at same concentration, while transfection
in the presence of N4-(4-quinolinyl)-N1,N1-diethyl-1,4-pentanediamine (CQ7b) does not reveal any enhancing
effects on expression. Through a number of comparative studies with CQ and its analogues, we conclude
that there are at least three mechanistic features of CQ that lead to the enhancement in gene expression: (i)
pH buffering in endocytic vesicles, (ii) displacement of polycations from the nucleic acids in polyplexes,
and (iii) alteration of the biophysical properties of the released nucleic acid.
Introduction
by inhibiting lysosomal enzyme degradation.19,20 Such hypoth-
eses are supported by the observation that polyethylenimine
(PEI), which is known to buffer the pH in endosomes, does not
benefit from the presence of CQ.8,21 Noting that CQ interacts
with DNA,22,23 others have suggested that CQ may increase
the transfection efficiency by facilitating dissociation of DNA
from polyplexes.8
Here, we test postulates of various mechanisms of CQ-
induced increases in gene expression by designing several CQ
analogues (Chart 1) and studying gene transfer in the presence
of these compounds. By correlating the chemical structures of
CQ analogues to their contributions in gene transfer, information
on the role of CQ is obtained. We use the cyclodextrin-
containing polycation developed in our laboratories as the main
transfection agent in these studies since it does not provide any
cellular toxicity at the conditions employed. However, selected
experiments are conducted with other transfection reagents in
order to confirm that the conclusions obtained are likely
applicable to other transfection media. Although similar structure-
property approaches have been previously applied to studying
the mechanism of CQ’s antimalarial activity,24-27 this is so far
the first structure-property investigation of CQ in gene transfer.
The intracellular transport of polycation/DNA complexes
(polyplexes) involves their cellular uptake, release from the
endocytic pathway into the cytoplasm, and delivery of DNA to
the cell nucleus.1 To achieve transgene expression, polyplexes
must overcome several extra- and intracellular barriers. Previous
studies suggest that entrapment of polyplexes in the endocytic
pathway, and their resulting degradation in lysosomes, represents
a major impediment to transfection efficiency.1-3
Significant increases in gene expression with many nonviral
gene delivery systems are observed when chloroquine (CQ) is
present during the transfection.4-8 We have also observed
increased transfection efficiency in the presence of CQ in our
studies using cyclodextrin-containing polycations (CDP) for the
delivery of nucleic acids.9-14 Although CQ has been widely
used as a transfection enhancement reagent, the mechanisms
by which CQ enhances gene expression remain unclear.
CQ is a weak base with pKas of 8.1 and 10.215 and is known
to buffer the luminal pH of endosomes.16,17 The buffering
activity of CQ could improve transfection efficiency by
facilitating DNA release from the endocytic pathway (as it may
generate swelling and destabilization of endosomes)18 and/or
Results
* To whom correspondence should be addressed. Phone: (626) 395-
4251. Fax: (626) 568-8743. E-mail: mdavis@cheme.caltech.edu.
† Insert Therapeutics.
Syntheses of CQ Analogues and Their Evaluation by In
Vitro Gene Transfection. Chloroquine (CQ) is an analogue
of quinoline that has a chloro group at the 7-position and an
aliphatic amino side chain at the 4-position (Chart 1). Because
of the aromatic ring-structure, CQ can intercalate DNA22,23 and
facilitate the unpackaging of DNA from polyplexes.8 CQ is also
‡ University of Illinois at Urbana-Champaign.
§ Division of Chemistry and Chemical Engineering, California Institute
of Technology.
| University of Washington.
Division of Biology, California Institute of Technology.
10.1021/jm060736s CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/07/2006

Chloroquine and Chloroquine Analogues in NonViral Gene DeliVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22 6523
Chart 1. Chemical Structures of Chloroquine and Chloroquine Analogues
Table 1. Chloroquine and Chloroquine Analogue-Mediated Transfection
of CDP/ p-DNA in HepG2 and HeLa Cellsa
widely recognized as an agent buffering endosomal pH16,17
which may reduce the degradation of polyplexes in acidic
CQ analogues
(concn, µM)
cell
line
cell viability
(%)
endolysosomal environments.19,20 The pH buffering capability
of CQ is largely due to its aliphatic amine side chains. To
thoroughly investigate the mechanism of action of CQ as an
enhancing agent in nonviral gene delivery, our strategy was to
study a family of CQ analogues having either side-chain
variations with fixed ring-structure (fixed intercalation capability
with variable pH buffering capability) or ring-structure variations
with fixed side chain (fixed pH buffering capability with variable
intercalation capability). CQ4a-f and CQO, CQ analogues with
fixed ring structure identical to CQ, were readily synthesized
in a one-step substitution reaction between 4,7-dichloroquinoline
and the corresponding aliphatic amines (see Supporting Infor-
mation).28 CQ analogues that contain 2-amino-5-diethylamino-
pentane as the side chain (identical to CQ) were prepared to
investigate CQ analogues with variable 7-substituents on the
quinoline ring (CQ7a,b) or with variable aromatic ring size
(CP). CQ7a,b and CP were synthesized via the substitution
reaction between 2-amino-5-diethylaminopentane and the cor-
responding 4-chloroquinoline derivatives or 4-bromopyridine,
respectively (see Supporting Information).
RLU/mg protein
none
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HepG2
HeLa
4.3 × 106
1.6 × 107
4.7 × 106
4.3 × 106
2.5 × 107
2.2 × 107
2.0 × 107
2.5 × 107
1.7 × 108
2.7 × 106
6.4 × 106
2.5 × 107
1.4 × 105
4.3 × 106
6.7 × 107
8.2 × 104
2.3 × 105
100
60
89
84
64
74
74
83
61
100
85
57
105
82
83
CQ (200)
CQ4a (150)
CQ4b (150)
CQ4c (150)
CQ4d (200)
CQ4e (100)
CQ4f (150)
CQ7a (200)
CQ7b (200)
CP (200)
QC (20)
none
CQ (200)
CQ7a (200)
CP (200)
HeLa
HeLa
HeLa
HeLa
101
92
CQO (200)
a HepG2 or HeLa cells were transfected for 4 h with CDP/pGL3-CV
polyplexes in the absence or presence of CQ or CQ analogues. Gene
expression was evaluated 48 h later by assaying the luciferase activity (in
relative light units, RLU) of cell lysates. The viability of the cells was
assessed by measuring the total amount of recovered protein. The CQ
analogue concentrations represented in the table are those which gave the
optimal increase in transfection efficiency. Transfection data with CQ or
CQ analogues at various concentrations using HepG2 cells is provided in
Supporting Information.
We utilized a linear, cyclodextrin-based polycation (CDP)
(Figure 1) to investigate the gene transfer effects of CQ and its
analogues. CDP is a new class of nontoxic and nonimmunogenic
polycation recently developed for use in nonviral gene
delivery.9-14,29-33 The plasmid pGL3-CV (pDNA), encoding the
luciferase gene under the control of the SV40 promoter, was
complexed with CDP at a charge ratio of 5 ( to give complete
pDNA binding.10 The polyplexes were added in the presence
of CQ analogues to cultured HepG2 or HeLa cells. Transfection
efficiencies were determined by assaying for luciferase protein
activity and reported in relative light units (RLU) per milligram
of total cellular protein. The concentrations of CQ and various
CQ analogues that resulted in optimal increase in transfection
efficiency in HepG2 cells are listed in Table 1. Transfection
efficiency increment at other concentrations of CQ or CQ
analogues in HepG2 cells can be found in the Supporting
Information.
The observed luciferase activity increases from 4.3 × 106
RLU/mg protein in the absence of CQ to 1.6 × 107 RLU/mg
protein in the presence of CQ. When the tertiary amine side
chain of CQ was replaced either with a hydrocarbon (CQ4a)
or with a negatively charged carboxyl group (CQ4b), the
luciferase activities of transfected cells were roughly the same
as that in the absence of CQ (Table 1). CQ4c, a CQ analogue
possessing a tertiary amine, enhances transfection to a degree
similar to CQ. These observations show the importance of the
terminal tertiary amine moiety of CQ for enhancing polyplex
transfection efficiency. However, increasing the basicity on the
side chain of CQ by increasing the number of amine groups
(three amine groups in CQ4d and four amine groups in CQ4e,f)
only results in slight increases in maximal luciferase activity
Figure 1. Cyclodextrin polycation (CDP) used as nonviral gene
delivery vehicle.

6524 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22
Cheng et al.
Table 2. Summary of Flow Cytometry Analysis of Polyplex Uptake in the Absence or Presence of CQ Analogues
30 min
2 h
pDNA or
CQ or
%
%
mean
%
%
mean
polyplexes
CQ analogue
negativ e
positiv e
fluores cence
negativ e
positiv e
fluoresce nce
-
-
-
-
99.2
99.3
82.8
85.5
82.2
85.4
87.4
0.8
0.7
4.1
4.0
12.9
10.0
15.0
9.2
99.7
98.9
20.9
32.5
24.7
35.0
24.0
0.3
1.1
3.45
5.48
125.7
79.3
88.3
78.1
pDNA
polyplexes
polyplexes
polyplexes
polyplexes
polyplexes
17.5
11.9
18.8
14.8
12.8
79.3
67.7
75.5
65.2
76.2
CQ
CQ7b
CQ7a
CP
7.6
109.0
when compared to CQ for HepG2 cells transfected in the
presence of those analogues (Table 1).
Dramatic changes in luciferase activity were observed for cells
transfected in the presence of CQ analogues with variable ring-
structures and with side chains identical to CQ (Table 1).
Replacement of the 7-chloro of CQ by a lipophilic trifluoro-
methyl group (CQ7a) results in an order of magnitude increase
in luciferase activity (1.6 × 107 RLU/mg protein for CQ to 1.7
× 108 RLU/mg protein for CQ7a). When the 7-substituent is
removed (CQ7b), the capacity for enhancing transgene expres-
sion is completely lost. Transfection of polyplexes in the
presence of CP, a CQ analogue with a pyridinyl structure, also
showed no increase in gene expression. Quinacrine (QC), a CQ
analogue having an acridine ring structure, enhances transfection
to a similar degree as CQ but at one-tenth the concentration.
QC is more toxic than CQ; a concentration of QC higher than
50 µM leads to complete cell death.
Figure 2. Uptake and intracellular accumulation of CQ and CQ
analogues in HepG2 cells (0.1 mM) in the presence or absence of
polyplexes. Fluorescence of cell lysates was used to evaluate the
intracellular concentration of CQ analogues at various time points after
initial exposure.
To study if there is cell-line dependency in the structure-
related gene-expression enhancement capability observed with
CQ and CQ analogues, we also investigated transfections by
CDP/pDNA polyplexes in other cells in the presence of CQ,
CQ7a, and CP. The variation of expression levels in the
presence of these reagents is very similar to that observed in
the transfection of HepG2 cells (HeLa cells, Table 1; BHK and
A2780 cells, Supporting Information). CQ7a gave an order of
magnitude increase in luciferase activity (6.7 × 107 RLU/mg
protein) compared to CQ (4.3 × 106 RLU/mg protein) in HeLa
cells (Table 1) as well as in BHK and A2780 cells (see
Supporting Information) and gave almost 3 orders of magnitude
increase in luciferase activity over CP (8.2 × 104 RLU/mg
protein) in HeLa cells (Table 1). Interestingly, reducing the
overall basicity on CQ’s side chain by replacing the aromatic
amino group with an oxygen (CQO) resulted in an order of
magnitude decrease of luciferase activity (2.3 × 105 RLU/mg
protein) (Table 1).
To confirm that increases in transfection efficiency were due
to interactions among CQ analogues, cells, and polyplexes, as
opposed to extracellular interactions between CQ analogues and
unpackaged pDNA, transfection experiments were carried out
with pDNA alone (naked pDNA). In the absence of a polycation
vector, cells transfected with pDNA in the presence of CQ
analogues (0.1 mM) did not show significant levels of gene
expression. In all cases, luciferase activity was below that
observed with polyplexes in the absence of CQ analogues (see
Supporting Information).
treated with naked pDNA is about the same as that of untreated
cells (Table 2). Cells exposed to complexes of labeled pDNA
and CDP displayed increases in mean cellular fluorescence,
indicating that polyplexes were internalized by cells. Similar
observations of internalization have been reported elsewhere.8,32
At both 30 min and 2 h, cells exposed to polyplexes revealed
significant uptake regardless of the presence or absence of CQ
or CQ analogues. 12-19% of cells at 30 min and 65-75% of
cells at 2 h contained detectable levels of fluorescently labeled
pDNA (% positive, Table 2; FACS plots are provided in
Supporting Information). The presence of CQ or CQ analogues
gave slightly reduced uptake relative to the case of polyplexes
alone. Also, the fluorescence distribution pattern was very
similar at both 30 min and 2 h for cells transfected in the
presence and absence of CQ or CQ analogues. Although cells
transfected in the presence of QC gave a similar percentage of
positive cells and similar mean fluorescence as cells transfected
in the presence of other CQ analogues, the values may not
reflect actual polyplex uptake because QC is itself fluorescent
at the wavelengths used for the FACS analysis (data not shown).
Uptake and Intracellular Concentration of CQ Analogues.
CQ can penetrate the cell membrane and accumulate in acidic
cellular compartments, especially in lysosomes, in a time-
dependent manner.8 The concentration of CQ in HepG2 cells
(as measured by its fluorescence in cell lysates) increased as a
function of treatment time and reached nearly 12 mM following
4 h of treatment with 100 µM CQ (Figure 2). CQ accumulated
in HepG2 cells with a cell-to-medium concentration ratio of
∼100. Relative to CQ, CQ7a achieved a more rapid intracellular
accumulation and reached a concentration of 11 mM in cells
only 30 min after exposure to 100 µM CQ7a (Figure 2). Beyond
this time, the intracellular concentration of CQ7a remained
steady throughout the 4 h uptake study. The achieved intracel-
lular concentrations of both CQ and CQ7a noticeably exceeded
their concentrations in the surrounding medium.
Cellular Uptake Of Polyplexes in the Presence of CQ
Analogues. Flow cytometry analysis was used to give a relative
measure of pDNA uptake into cells (Table 2). pDNA was
labeled with Alexa-fluor-488, condensed into particles with
CDP, and exposed to HepG2 cells for 30 min or 2 h in the
presence and absence of CQ or CQ analogues. Cells were then
analyzed by fluorescence-activated cell sorting (FACS). Uptake
of naked pDNA was minimal, as the mean fluorescence of cells

Chloroquine and Chloroquine Analogues in NonViral Gene DeliVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22 6525
Figure 3. Intracellular buffering activity of CQ and CQ analogues
(0.1 mM) in the transfection of polyplexes containing of CDP and a
pH sensitive, SNARF-4F-labeled oligonucleotide in HeLa cells.
Figure 4. Electrophoretic mobility of pDNA in the presence of CQ,
CQ7a, or CQ7b on an agarose gel electrophoresis.
The uptake of CQ or CQ7a into HepG2 cells was also
investigated in the presence of CDP/pDNA polyplexes. Both
CQ and CQ7a exhibited strong uptake; after 4 h, intracellular
concentrations were 13-15 mM, slightly higher than their
concentrations in the absence of polyplexes (Figure 2). When
the substituent at the 7-position of CQ was removed, the
resulting CQ analogue (CQ7b) showed a lower tendency to
accumulate in cells. The maximum intracellular concentration
of CQ7b was around 2 mM after 4 h treatment, about five times
lower than that of CQ or CQ7a. These uptake analyses
demonstrate that the transmembrane capability of CQ and its
analogues is closely related to the substituent at the 7-position
of the quinoleic ring. The trifluoromethyl group is a strong
electron-withdrawing, lipophilic group, and CQ7a that contained
this group showed fast cell-membrane penetration and intra-
cellular accumulation relative to CQ.
The observed concentrations are averages for the entire cell
volume. When taken into cells, CQ analogues that accept
protons at the relevant pH are likely found at much greater
concentrations in vesicles of the endocytic pathway. CQ is
believed to accumulate in such vesicles as a result of its
intravesicular protonation17 and its interactions with vesicle
membranes.34
One of the motivations of our investigation on the uptake of
CQ and CQ analogues was to elucidate the role of endocytic
vesicle buffering by CQ in gene transfer. To assist in this study,
we synthesized an additional CQ analogue, CQO (where an
oxygen replaces the aniline amino group at the 4 position of
chloroquine), in order to examine a compound that has less
ability to buffer the pH of endocytic compartments.
To evaluate the actual intracellular buffering activity of CQO
and other CQ analogues, we employed the method of Kulkarni
et al.,35 where polyplexes labeled with the pH-sensitive fluo-
rophore SNARF-4F were administered to cells and used to
evaluate the local pH environment of polyplexes within
individual cells. The data collected provide a direct measurement
of the buffering activity of CQ analogues within live cells
(Figure 3). Using this technique, CQ and CQ7a (0.1 mM)
displayed strong buffering activity relative to the absence of
any CQ analogues. CQO displayed mild buffering activity,
while CP exhibited none.
electrostatic interaction between the cationic side chain and the
phosphate anions along the DNA backbone.22 For the examina-
tions of CQ analogue interaction with nucleic acids (and/or
polyplexes), we focused on a smaller set of molecules for further
study: CQ, CQ7a (shows strong effects on transfection and
strong interactions with DNA), and CP (not effective in
increasing transfection; a single-ring structure that should limit
interaction with DNA).
The interaction of pDNA and CQ or CQ analogues was
analyzed by studying the retardation of pDNA movement on a
gel in the presence of these small molecules. The pDNA was
incubated in the presence of CQ analogues at various concen-
trations and then subjected to gel electrophoresis (Figure 4).
The band from pDNA treated with CQ7a at 100 µM showed
essentially the same pattern as the untreated control, indicating
that CQ7a had little effect on the mobility of pDNA at this
concentration. The pDNA treated with CQ7a at a one-order-
of-magnitude higher concentration (1 mM) showed slightly
retarded movement while the migration of the pDNA was
completely prohibited when treated with CQ7a at 10 mM
(Figure 4). In contrast, CQ7b had little effect on pDNA
retardation across the same concentration range (0.1-10 mM).
CQ at 100 µM and 1 mM did not retard pDNA, but did so
slightly at 10 mM.
DNA Displacement from Polyplexes by CQ, CQ7a, or
CQ7b. Interactions between CQ and pDNA could destabilize
the association between the plasmid and cationic delivery
vectors.8 To assess the stability of CDP/pDNA polyplexes
against CQ analogues, solutions of polyplexes were allowed to
mix with CQ analogues for 30 min and then were filtered to
remove intact polyplexes. The unpackaged pDNA in the filtrate
was measured by a fluorescence method (Figure 5). CQ7a was
able to displace 40% of total condensed pDNA at a concentra-
tion as low as 500 µM. 40-50% of the pDNA was found in
the filtrate for CQ7a concentrations of 500 µM to 10mM, and
over 70% of pDNA was displaced with a CQ7a concentration
of 20 mM. CQ displayed weaker ability to displace pDNA than
CQ7a. The pDNA displaced from the polyplexes by CQ
gradually increased from 10% with 500 µM CQ to roughly 50%
with 20 mM CQ. With CQ7b concentrations of 2.5 mM or
lower, less than 3% of the pDNA was detected in the filtrate.
The displaced pDNA gradually increased from 10% to about
30% when polyplexes were treated with 5 mM to 20 mM of
CQ7b, respectively (Figure 5).
Interaction of CQ Analogues with pDNA. Electrophoretic
Mobility of pDNA in the Presence of CQ Analogues. It is
known that CQ can bind DNA.22,37 Previous work has indicated
that CQ exists as a doubly protonated cation in dilute aqueous
solutions at physiological pH,15 and its binding to DNA is related
to insertion of its quinoleic moiety into DNA base pairs and to
The comet assay is a standard method of evaluating DNA
damage.36 In this assay, cells are gently lysed, embedded in

6526 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22
Cheng et al.
Figure 5. DNA displacement from polyplexes by CQ, CQ7a, or CQ7b
after incubating CDP/pDNA polyplexes with various concentrations
of CQ, CQ7a, or CQ7b followed by the filtration to remove intact
polyplexes and the quantification of the displaced DNA in the filtrate.
Figure 7. PicoGreen dye exclusion from pDNA in the presence of
CQ analogues.
Figure 6. Agarose gel electrophoresis of the FITC-labeled nucleic acid
collected from HeLa cells transfected with CDP/FITC-oligonucleotide
polyplexes in the presence of CQ, CQ7a, and CQO.
an agarose gel on a glass slide or coverslip, and exposed to
electrophoresis. Only damaged cellular DNA is reduced to a
size that permits migration within the gel, and this damaged
DNA will generate a characteristic “comet” pattern upon
subsequent intragel labeling of nucleic acids by a fluorescent
marker. The size and fluorescence intensity of the comet reflects
the extent of DNA damage. By this method, individual cells
may be visually evaluated for DNA damage.
We modified this assay to produce an aggregate separation
and visualization of the delivered and unpackaged nucleic acids.
Rather than evaluating samples on an individual-cell basis, cells
are collected by centrifugation, lysed, embedded in an agarose
gel, and subjected to electrophoresis in order to generate a
relative measure of DNA unpackaging for a particular sample.
We transfected cells using nucleic acids conjugated with a
fluorescent marker rather than adding a fluorescent label after
electrophoresis. Thus, we are able to distinguish exogenous
(delivered) nucleic acids from cellular DNA. The unpacked
oligonucleotide measured by this assay is that which migrates
through the agarose gel to the same extent as in the case of
direct gel loading of the oligonucleotide.
HeLa cells were transfected in the presence of CQ analogues
and examined by the modified comet assay (Figure 6). Unlike
the behavior of CQ and CQ7a in displacing pDNA from
polyplexes (Figure 4), samples transfected with these CQ
analogues generate less unpackaged (migrating) DNA than
samples transfected without CQ analogues or samples trans-
fected in the presence of CQO.
PicoGreen experiences significant fluorescence enhancement
upon intercalation in nucleic acids, such that its exclusion
indicates inaccessibility of the nucleic acids to intercalation.37
This characteristic was used to evaluate the relative accessibility
of pDNA in the presence of CQ analogues. Polyplexes were
incubated with various concentrations of CQ analogues, com-
bined with solutions of PicoGreen, and evaluated for fluores-
cence intensity (Figure 7). CQ7a, CQ, and CQO demonstrated
Figure 8. In vitro transcription in the presence of CQ analogues.
Solutions of pDNA and CQ analogues were subjected to in vitro
transcription, and mRNA transcripts were quantified by absorption at
260 nm and normalized against the transcription in the absence of CQ.
enhancements in dye exclusion with increasing concentration.
Dye exclusion was ∼30% for 0.5 or 1 mM CQ7a and rose to
almost 100% at 4 mM. Dye exclusion was ∼23% for CQ (0.5
or 1 mM) and ∼17% for CQO (0.5 or 1 mM), with slightly
greater exclusion for CQ/CQO at 4 mM. PicoGreen was not
excluded from DNA by CP.
Interactions between pDNA and CQ analogues may render
pDNA less accessible to enzymes such as polymerases that
initiate gene expression. To evaluate this hypothesis, solutions
of pDNA and CQ analogues were subjected to in vitro
transcription. Relative to in vitro transcription in the absence
of CQ analogues, pDNA retained at least 75% of its transcript
production with CQ analogue concentrations of up to 2.5 mM
(Figure 8). However, small decreases in transcription were
observed as CQ analogue concentrations were increased. The
most pronounced reduction was observed with CQ7a, less of
an effect was shown with CQ and CQO, and the weakest effect
was recorded with CP. The correlation between dye exclusion
(Figure 7) and reduced mRNA production (Figure 8) suggests
that in vitro transcription is inhibited by interactions between
pDNA and CQ analogues.
Discussion
Chloroquine (CQ) can enhance the transfection efficiency
in many nonviral, polycation-based gene delivery vectors, but
the mechanisms by which this occurs are not well established.
In this study, we contribute to the understanding of CQ-related
enhancements in gene expression by synthesizing and investi-
gating a number of CQ analogues whose structures are designed
specifically to probe hypothesized mechanisms, i.e., pH buffer-

Chloroquine and Chloroquine Analogues in NonViral Gene DeliVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22 6527
ing of endocytic vesicles, displacement of polycations from
DNA, etc. This is so far the first systematic study to correlate
the CQ structure with its function during gene transfer.
As shown by the data in Table 1, a side chain with at least
one basic amino moiety is essential for the capability of these
small molecules to enhance transfection (CQ and CQ4c-f vs
CQ4a and CQ4b). CQ4a and CQ4b, the CQ analogues that
lack side-chain amino groups, show no effects on gene transfer.
However, the effects of the side-chain amino group alone are
not entirely responsible for CQ’s activity because transfection
is not enhanced by some of the CQ analogues containing this
group (CQ7b and CP) or by an aliphatic amine alone.8
CQ analogues bearing more than two amino moieties
(CQ4d-f) are more basic than CQ. Expression levels in the
presence of those CQ analogues are essentially the same as with
CQ or in the presence of a CQ analogue with only two amino
moieties (CQ4c). These results indicate that the level of basicity
exceeding that of CQ does not enhance the gene transfection.
The aromatic ring structure of CQ contributes to its function
in nonviral gene transfer (Table 1). CP, a pyridinyl-based CQ
analogue, does not increase gene expression at all, while QC,
an acridine-based CQ analogue with three fused aromatic rings,
enhances the transfection at an optimal concentration of 20 µM;
10 times lower than the optimal concentration of CQ.
Dramatic effects on gene expression occur with changes in
the 7-substituent of the CQ aromatic ring (while maintaining
the side chain of CQ). When the 7-Cl group of CQ is replaced
with a proton (CQ7b), the CQ analogue is not able to enhance
gene expression. If a trifluoromethyl group (CQ7a) is used in-
stead of the 7-Cl group, the transfection enhancement is mark-
edly improved relative to the enhancement by CQ (Table 1).
The presence of the CQ analogues does not significantly
affect the uptake of polyplexes by cells (Table 2), and the
intracellular concentrations of CQ and CQ7a are similar at 4 h
(Figure 2). Apparently, the differences in transfection enhance-
ment between CQ and CQ7a are not due to differences in
uptake of either the DNA or the CQ and CQ7a. However,
CQ7b demonstrates that the lipophilic or the electron-withdraw-
ing character of the 7-substituent can contribute to the uptake
of a CQ analogue, providing an explanation for the lack of effect
on transfection efficiency by CQ7b and possibly CP. Direct
uptake measurements of CP would be helpful, but the absor-
bance spectra of CP (and CQO) do not allow their cellular
uptake to be measured using the method employed for other
CQ analogues.
The results obtained with CQ, CQ7a, and CQO show a
correlation between enhancements in transfection efficiency and
buffering of the pH experienced by polyplexes within cells
(Figure 3). CQO interacts with nucleic acids in a manner similar
to CQ (Figure 7) but displays reduced buffering activity (Figure
3) and does not enhance transfection efficiency (Table 1). CQ
analogues without a side chain amino group (CQ4a,b) should
have reduced pH-buffering activity, and they do not reveal the
enhancement in gene expression that is seen with variants
containing one or more side chain amino groups (CQ, CQ4c-
f) (Table 1). Together, these results support the hypothesis that
CQ’s pH-buffering of endocytic vesicles is a necessary but not
sufficient condition for enhancing transfection efficiency.
While buffering of the endocytic vesicles appears to be
necessary for transfection enhancements, it does not provide a
complete explanation of how CQ increases transfection ef-
ficiency. The results from varying both the side chain and the
aromatic ring structures (Table 1) suggest that CQ achieves its
overall effect through multiple mechanisms. Also, the intracel-
lular buffering by CQ and CQ7a is functionally similar, so the
buffering hypothesis does not explain the significant increase
in transfection efficiency in all the four cell lines tested when
CQ is replaced with CQ7a (HepG2 and HeLa, Table 1; BHK
and A2780, see Supporting Information) When poly-lysine
(PLL)/pDNA complexes were transfected with CQ or CQ7a
in HepG2 cells, results similar to those of CDP were obtained
(see Supporting Information). The significantly enhanced gene
expression observed with CQ7a compared to CQ cannot be
simply ascribed to the buffering effect. Compounds such as
monensin, methylamine, spermine, and ammonium chloride are
known to buffer endocytic vesicles but do not contribute to
transfection efficiency,8 and these data provide additional
evidence that buffering activity is not a sufficient condition for
improved transfection.
CQ, CQ7a, and QC all confer significant gains in transfection
efficiency. However, the dramatic variations in their perfor-
mance suggest that their effects on transfection efficiency could
be related to their interactions with DNA. Parker et al. have
demonstrated that CQ can bind to both DNA and RNA in vitro,
an interaction which can alter the biological and physical
properties of the DNA.38 Interactions between DNA and CQ
analogues are strongly associated with the aromatic structures
of those ligands. Because of its acridinyl fused aromatic
structure, QC exhibits strong interactions with DNA,39 and this
molecule is a potent enhancer of gene transfer at relatively low
concentration (Table 1). The single aromatic ring of CP renders
this molecule unable to bind DNA as effectively as its multiring
analogues (Figure 7), consistent with the fact that CP does not
enhance gene transfer (Table 1).
In addition to ring structure, the electron-withdrawing char-
acter of the substituent at the 7-position can affect the strength
of binding between CQ analogues and DNA.23 This effect is
demonstrated by gel retardation of pDNA at various concentra-
tions of CQ, CQ7a, and CQ7b (Figure 4). CQ7a shows the
strongest interaction with pDNA, displaying some retardation
of pDNA electrophoresis at 1 mM and essentially complete
retardation at 10 mM. CQ7b shows a weaker interaction with
pDNA, as pDNA mobility did not change in the presence of
CQ7b over a wide range of concentrations. The strength of
interaction between CQ and pDNA is likely intermediate to
those of CQ7a and CQ7b since CQ slightly retarded pDNA
electrophoresis at 10 mM.
pDNA must be dissociated from polycations in order to
achieve gene expression. We have shown that CDP/pDNA
polyplexes do not reach the nucleus of transfected cells but
deliver unpackaged pDNA to both the cytoplasm and cell
nucleus.32 It is likely that gene transfer is improved with
enhanced intracellular unpackaging.40 The presence of CQ
analogues can destabilize polyplexes, generating unpackaged
pDNA (Figure 5). The ability to displace pDNA from CDP
decreases in the order of CQ7a, CQ, and CQ7b, corresponding
to the order of pDNA binding strengths displayed in the gel
retardation assay.
Cells exposed to solutions of pDNA and CQ analogues do
not show gene expression (see Supporting Information), sug-
gesting that extracellular interactions between pDNA and CQ
analogues do not play a significant role in transfection. Although
cells may be exposed to relatively low concentrations of CQ
analogues during transfection, the concentrations within cells
are found to be much higher (Figure 2). For CQ and CQ7a,
these concentrations are sufficient to disrupt polyplexes (Figure
5). The average cellular concentrations of CQ analogues are
likely less than the concentrations found in vesicles of the

6528 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22
Cheng et al.
Figure 9. Proposed mechanism for the enhancement of transgene expression by CQ in nonviral gene delivery: accumulated intracellular high
concentrations of CQ in the endocytic pathway (1) leading to enhanced unpackaging of delivered DNA; (2) buffering the pH of endocytic vesicles;
(3) altering the trafficking, processing, or degradation of the DNA within cells.
endocytic pathway, as acidic intracellular compartments tend
to accumulate weakly basic compounds such as CQ.17,41-44 The
pH-partitioning may contribute to this effect,41,45 but data suggest
that such accumulation occurs to a greater extent than would
be predicted by pH-partitioning alone.46 Colombo and Bertini
have proposed that CQ accumulates in lysosomes due to pH-
partitioning as well as its interactions with the lysosomal
membrane.34 Thus, it is likely that CQ analogues accumulated
in the endocytic pathway at or above 10 mM can destabilize
polyplexes and unpack pDNA from polyplexes.
The biological and physical properties of DNA can be altered
when it is bound by CQ.38 CQ7a, which displays stronger
binding affinity for DNA than CQ, can be expected to have
similar or greater effects on altering the properties of DNA.
Beyond their activity in displacing DNA from polyplexes, CQ
and CQ7a appear to interact directly with nucleic acids, even
when those nucleic acids have been delivered to cells (Figures
4, 6-8). Results from the modified comet assay (Figure 6)
suggest that CQ and CQ7a may be associated with the DNA
that has been released from the polyplexes (lack of mobility in
the gel suggests that a cationic entity is bound to the DNA that
has been released from the polyplexes when CQ and CQ7a
are utilized in the transfection). Strong binding by numerous
CQ7a molecules could alter the physical properties of pDNA
and reduce the overall negative charge of its backbone (via
electrostatic interaction between anionic phosphate groups of
pDNA and cationic amino side chains of CQ7a) (Figure 4).
When bound by CQ, pDNA could achieve improved gene
transfer due to changes in its intracellular processing or its rate
of intracellular degradation.47,48 The pDNA can be transcribed
in the presence of CQ analogues, although increasing concen-
trations of CQ analogues lead to declining levels of mRNA
transcripts (Figure 8). Reduced accessibility of pDNA to
transcription-initiating polymerases may imply reduced pDNA
accessibility to degradative enzymes, but efforts to evaluate
DNAse protection by CQ analogues were inconclusive due to
difficulty in purifying pDNA from CQ analogues after DNAse
exposure (data not shown).
CQ7a greatly enhances gene expression over CQ in all the
four cell lines we studied (Table 1 and Supporting Information),
an effect that is not explained by its intracellular buffering effect
alone. The observed higher luciferase activity from transfection
in the presence of CQ7a is likely due to additional features
such as greater displacement of polycations (Figure 5) and
stronger interactions with intracellular nucleic acids that may
protect and traffic more of the intact nucleic acid into the nucleus
(Figures 4, 6-8).
In summary, a structure-function study among CQ and its
analogues in nonviral, polycation-based gene delivery is carried
out for the first time. It is demonstrated that the tertiary amino

Chloroquine and Chloroquine Analogues in NonViral Gene DeliVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22 6529
Essential Medium (HepG2) or Dulbecco’s Modified Eagle’s
Medium (HeLa), with 2 mM l-glutamine, 1.5 g/L sodium bicarbon-
ate, 0.1 mM nonessential amino acids, 1.0 mM sodium pyruvate,
and 10% fetal bovine serum. Media and supplements were
purchased from Gibco BRL (Gaithersburg, MD). Cells were
transfected in serum-reduced medium with CDP/pDNA polyplexes
as described previously, in the presence of CQ or CQ analogues
at various concentrations.
moiety of CQ is important to enhancing gene expression.
Further, changes in the aromatic ring of CQ significantly affect
its ability to increase gene expression. A CQ analogue with a
three fused aromatic ring structure (QC) shows much higher
cytotoxicity than CQ and can enhance transfection similarly to
CQ at a concentration 10 times lower. A CQ analogue with
one pyridinyl aromatic ring (CP) gives lower cytotoxicity than
CQ but has no effect on gene transfer. Variation at the 7-position
of CQ dramatically changes its ability to enhance transfection.
Transfection with a 7-trifluoromethyl CQ analogue (CQ7a)
shows luciferase activity an order of magnitude higher than
transfection with the same concentration (0.2 mM) of CQ, while
transfection is not affected in the presence of a 7-hydro CQ
analogue (CQ7b). The fused aromatic ring structure (acridinyl
ring) in QC and the strong electron-withdrawing group (triflu-
oromethyl) in CQ7a endow these molecules with higher binding
affinity to DNA. The stronger intercalator, CQ7a, has a higher
tendency than CQ to competitively displace the polycation from
CDP/pDNA polyplexes at a concentration that is achievable
within cells after several hours’ treatment. Finally, CQ and
CQ7a appear to interact directly with nucleic acids in cells.
Our study suggests that the mechanisms of CQ action are at
least 3-fold (Figure 9). One, intracellular accumulation of CQ
drives its concentration to a point where it is able to facilitate
unpackaging of nucleic acids from the polyplexes. CQ ac-
cumulates in cells to concentrations greatly exceeding that in
the extracellular medium; and it reaches still higher concentra-
tions in the vesicles of the endocytic pathway where internalized
polyplexes are also delivered. Second, consistent with numerous
literature reports, CQ buffers the vesicles of the endocytic
pathway, helping delivered nucleic acids escape from these
vesicles and/or avoid lysosomal degradation. Third, intracellular
interactions between CQ and unpackaged nucleic acids con-
tribute to enhanced gene expression. CQ can bind to the
delivered DNA via interaction with DNA bases and through
electrostatic interaction with anionic phosphate groups. These
interactions lead to changes in the biological and physical
properties of the delivered nucleic acids, possibly altering their
intracellular processing or slowing their degradation.
For cell transfection and luciferase assay, cells were plated at 5
× 104 cells/well in 24-well plates 24 h in advance. Immediately
prior to transfection, cells were rinsed once with PBS (pH 7.4),
and 200 µL of Opti-MEM (Gibco) was added to each well. pGL3-
CV (1 µg, 5 µL of a 0.2 µg/µL solution in DNAse-free water) was
mixed with an equal volume of polymer solution (5 µL of 24.9
mg/mL freshly prepared CDP solution in DNAse-free water) to
give a charge ratio 5 +/-. An Opti-MEM solution (185 µL) and
an appropriate amount of CQ (or CQ analogues) in 5 µL of Opti-
MEM solution were mixed with CDP/pDNA complexes (10 µL)
to give a final volume of 200 µL. These solutions were immediately
transferred to each well. After 4 h of incubation (37 °C, 5% CO2),
the media in each well was replaced with 1 mL of culture media.
After another 44 h, the media was removed by aspiration. Cells
were washed twice with PBS (pH 7.4) before addition of 100 µL
of 1× cell culture lysis buffer (Promega). Cell lysates were analyzed
for luciferase activity with luciferase assay reagent (Promega). Light
units were measured in duplicate with a luminometer (Monolight
2010, Analytical Luminescence Laboratory, San Diego, CA).
Determination of cell viability and total protein concentration
was conducted by a modified Lowry protein assay. The amount of
protein in cell lysates obtained 48 h after transfection was used as
a measure of cell viability. Protein levels of transfected cells were
determined by the DC Protein Assay (Bio-Rad, Hercules, CA) and
normalized with protein levels of cells transfected with naked DNA.
A protein standard curve was run with various concentrations of
bovine IgG (Bio-Rad) in cell culture lysis buffer.
Uptake of Polyplexes in the Absence or Presence of CQ
Analogues. pDNA was prepared for fluorophore binding by reacting
with Label-IT Amine (Mirus, Madison, WI). After 1 h at 37 °C,
the amine-terminated DNA was recovered by precipitation with
ethanol and then incubated overnight with Alexa-fluor-488 NHS
ester (Molecular Probes, Eugene, OR). The labeled DNA was
purified by precipitation with ethanol. The percentage of pDNA
labeling was determined to be 3.3 wt % by measuring the
fluorescence intensity of labeled DNA (λexcitation 488 nm; λemission
535 nm) using a Spectrafluor Plus plate reader (Tecan, Durham,
NC) and comparing with a standard curve.
Experimental Section
Chloroquine (CQ), quinocrine (QC), and other chemicals
reagents used in this study were purchased from Sigma-Aldrich
(St. Louis, MO) and used without further purification unless
otherwise indicated. NMR spectra were collected on Varian 300
MHz and Inova 500 MHz spectrometers. â-Cyclodextrin polymer
(CDP) containing two positively charged amidine groups in each
repeat with weight average molecular weight (Mw) of 6000 Da was
synthesized and used for condensing DNA or oligonucleotide in
this study.10,11 Each repeat unit contains two positively charged
amidine groups that can efficiently bind and condense DNA in
plasmid (pDNA) or oligonucleotide form.10
General Procedure for the Synthesis of CQ Analogues. CQ
analogues with alkylamino side chains and 7-chloroquinoline
structure (CQ4a-f and CQO) were synthesized via substitution
reaction of an appropriate amine and 4,7-dichloroquinoline (see
Supporting Information). CQ analogues with aromatic ring-size and
ring-substituent variations were synthesized similarly using an
appropriate quinoline or pyridine derivative as starting materials
(see Supporting Information).
Labeled pDNA was condensed with CDP prior to being
administered to HepG2 cells at a concentration 5 µg DNA/mL
culture medium (same concentration as in gene transfection study).
Uptake was terminated at 30 min and at 2 h. Cells were treated
with 0.25% trypsin, washed with phosphate-buffered saline and
Hanks balanced salt solution, and analyzed by fluorescence-
activated cell sorting (FACS).
Uptake and Intracellular Accumulation of CQ, CQ7a, and
CQ7b. HepG2 cells (2 × 106 cells) were incubated at 37 °C in 2
mL of Opti-MEM (pH 7.4) with or without CDP/pDNA (20 nmol/3
pmol) in the presence of 100 µM of CQ, CQ7a, or CQ7b. After
incubation in a humidified atmosphere (37 °C, 5% CO2) for
scheduled time, the medium was removed; cells were washed three
times with cold PBS containing 100 µM of corresponding CQ
analogue, washed three more times with cold PBS, and then lysed
by 20 min incubation in 1 mL of 1× cell culture lysis buffer. The
lysates was diluted with 1 mL of NaOH (0.1 N), and 80 µL of
such solution was analyzed on a Spectrafluor Plus plate reader
(Tecan) at excitation wavelength of 360 nm and emission wave-
length of 465 nm. The resulting fluorescence results were converted
to concentrations of CQ or CQ analogues using a standard curve.
The cellular concentration of CQ or CQ analogues was calculated
assuming a cell volume of 3 × 10-6 µL.8
Cell Culture and Transfection Experiments. The DNA plasmid
pGL3-CV (Promega, Madison, WI) containing the firefly luciferase
gene under the control of the SV40 promoter was amplified by E.
coli strain DH5R and was then purified using the Ultramobius 1000
plasmid kit (Novagen, San Diego, CA). HepG2 and HeLa cells
were purchased from the ATCC (Rockville, MD). Cells were
cultured according to recommended procedures using Minimum
Intracellular Buffering Activity of CQ Analogues. The activity
of CDP in buffering the pH environment of intracellular polyplexes

6530 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22
Cheng et al.
was evaluated using the method of Kulkarni et al.35 Polyplexes were
prepared by mixing CDP and a 21-base RNA phosphorothioate
oligonucleotide (5′-GACGUAAACGGCCACAAGUUC-3′) labeled
at the 5′ position with the pH-sensitive fluorophore SNARF-4F
(Molecular Probes, Carlsbad, CA).35 HeLa cells were treated with
the polyplexes and imaged using an inverted Zeiss LSM 510 META
confocal microscope with a 63× oil objective (NA 1.4). Excitation
of the SNARF fluorophore was achieved with the 543 nm line of
the He-Ne laser line supplied with the microscope. After passing
through a variable confocal pinhole, the SNARF-4F emission was
collected by the META detector in distinct wavelength bins of 569-
601 nm and 633-687 nm. Using a standard curve, the relative
intensity of individual signals across these bins was used to evaluate
the pH environment of the corresponding intracellular polyplexes.
Electrophoretic Mobility of pDNA in the Presence of CQ
Analogues. Each polycation was examined for its ability to bind
pDNA through gel electrophoresis experiments as previously
described.9,10 pGL3-CV (5 µL of a 0.2 µg/µL solution in DNAse-
free water) was mixed with an appropriate amount of CQ, CQ7a,
or CQ7b. Each solution was incubated for approximately 10 min.
Loading buffer (5 µL) was added to each sample, followed by an
appropriate volume of Borax buffer (0.025 M, pH 8.4) to give a
final volume of 25 µL for each sample. 10 µL of such mixture was
transferred into each well of a 0.5% agarose gel (6 µg of ethidium
bromide/100 mL TAE buffer (40 mM Tris-acetate, 1 mM EDTA))
and electrophoresed.
Displacement of DNA from Polyplexes by CQ Analogues.
pGL3-CV (2 µg, 0.6 pmol) in 10 µL of DI water was complexed
with CDP (24 µg, 4 nmol) in 10 µL. The solution was diluted by
adding 80 µL of PBS and kept for 30 min at 25 °C. After a further
30 min at 25 °C, each solution was passed through a Pall Life
Science Supor Membrane (100 nm) presoaked and rinsed with a
solution containing the equivalent concentration of the CQ ana-
logues. An amount of 40 µL of the filtrate was mixed with 40 µL
of 5.6 µM 4,6-diamidino-2-phenylindole, dihydrochloride (DAPI).
The amount of unpackaged DNA was determined from the
fluorescence intensities (λexcitation 350 nm; λemission 465 nm) of the
DAPI-containing filtrate solution.8 The percentage of pDNA
dissociated from polyplexes was calculated according to (If - I0)/(
It - I0) × 100, where It is the fluorescence intensity of DAPI in
the presence of the total full of pDNA, I0 is the fluorescence
intensity of DAPI in the absence of pDNA, and If is the fluorescence
intensity of the filtrate containing dissociated pDNA.
Measure of Unpackaged Intracellular Nucleic Acids. HeLa
cells in 24-well plates were transfected with polyplexes of CDP
and a 25-base DNA oligonucleotide (5′-ACTGCTTACCAGG-
GATTTCAGTGCA-3) labeled at the 5′ position with FITC (FITC-
oligo). Fourteen hours after transfection, cells were collected by
trypsinization and pelleted by centrifugation (6 min at 2400 rpm).
Each sample was resuspended in 10 µL of 1× cell culture lysis
buffer (Promega) and stored at 4 °C. Controls consisted of known
amounts of FITC-oligo in 1× cell culture lysis buffer. After 30
min incubation, each sample was mixed with 25 µL of 37 °C low-
melting-point agarose (1% in TAE buffer) and transferred im-
mediately to a well of an agarose gel (0.5% in 60 mL TAE buffer).
The agarose gel was subjected to gel electrophoresis and imaged
under UV illumination. Detection of FITC-oligo was achieved
through its fluorescent signal. Migration of FITC-oligo from control
samples was used to indicate the expected migration of FITC-oligo
unpackaged from cells.
Dye Exclusion Assay To Evaluate Accessibility of Condensed
pDNA. The accessibility of polyplexes in the presence of CQ
analogues was evaluated by dye exclusion through adaptation of a
published procedure.37 After formulation, polyplexes were diluted
2.5× in dH2O. For each sample, 25 µL of polyplex solution was
transferred to a well of an opaque, black 96-well plate and combined
with 25 µL of CQ analogue solution at twice the concentration of
interest. Separately, a 200× dilution of the supplied PicoGreen stock
solution (Molecular Probes) was prepared in 10 mM HEPES buffer.
After polyplex-CQ analogue solutions were incubated for 5 min
at room temperature, each 50 µL sample was combined with 50
µL of PicoGreen solution. The fluorescence (λexcitation 488 nm,
λemission 535 nm) of the resulting solutions was evaluated with a
Spectrafluor Plus plate reader (Tecan, Durham, NC). The percentage
dye exclusion was calculated from the ratio ((FDNA - Fsample)/(FDNA
- FH2O)), where FDNA is the fluorescence of a sample of DNA
alone (no polycation), Fsample is the sample fluorescence, and FH2O
is the fluorescence of a blank (control) sample.
In Vitro Transcription in the Presence of Chloroquine
Analogues. An in vitro transcription assay was used to evaluate
the transcription accessibility pDNA in the presence of CQ
analogues. The plasmid pT7-luc contains the firefly luciferase gene
under the control of the T7 promoter. This plasmid was amplified
by E. coli strain DH5R and purified using the Ultramobius 1000
plasmid kit. The MAXIscript T7 kit (Ambion, Austin, TX) was
used to produce firefly luciferase mRNA from mixtures of pT7-
luc and CQ analogues. Following the reaction, pT7-luc was digested
using the TURBO DNase supplied with the kit. The RNA reaction
product was purified by ammonium acetate/ethanol precipitation,
resuspended in dH2O, and quantified by absorbance at 260 nm.
Yields were normalized using reactions conducted in the absence
of CQ analogues.
Acknowledgment. We thank Insert Therapeutics for finan-
cial support.
Supporting Information Available: Detailed synthesis and
characterization of chloroquine analogues, in vitro transfection of
CDP/pDNA in the presence of chloroquine and chloroquine
analogues in HepG2, HeLa, and A2780 and BHK cells, transfection
of naked pDNA in the presence of CQ and CQ analogues,
transfection of polylysine/pDNA in the presence of chloroquine
and chloroquine analogue, fluorescence-activated cell sorting
(FACS) histogram of the time-dependent uptake of polyplexes in
the absence or presence of CQ analogues, and UV analysis of the
interaction between DNA and CQ or CQ analogues. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Hwang, S. J.; Davis, M. E. Cationic polymers for gene delivery:
Designs for overcoming barriers to systemic administration. Curr.
Opin. Mol. Ther. 2001, 3, 183-191.
(2) Davis, M. E. Nonviral gene delivery systems. Curr. Opin. Biotechnol.
2002, 13, 128-131.
(3) Lechardeur, D.; Verkman, A. S.; Lukacs, G. L. Intracellular routing
of plasmid DNA during non-viral gene transfer. AdV. Drug DeliVery
ReV. 2005, 57, 755-767.
(4) Wagner, E.; Zenke, M.; Cotten, M.; Beug, H.; Birnstiel, M. L.
Transferrin-Polycation Conjugates as Carriers for DNA Uptake into
Cells. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 3410-3414.
(5) Zenke, M.; Steinlein, P.; Wagner, E.; Cotten, M.; Beug, H.; Birnstiel,
M. L. Receptor-Mediated Endocytosis of Transferrin Polycation
Conjugatess an Efficient Way to Introduce DNA into Hematopoietic-
Cells. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 3655-3659.
(6) Midoux, P.; Mendes, C.; Legrand, A.; Raimond, J.; Mayer, R.;
Monsigny, M.; Roche, A. C. Specific Gene-Transfer Mediated by
Lactosylated Poly-l-Lysine into Hepatoma-Cells. Nucleic Acids Res.
1993, 21, 871-878.
(7) Monsigny, M.; Roche, A. C.; Midoux, P.; Mayer, R. Glycoconjugates
as Carriers for Specific Delivery of Therapeutic Drugs and Genes.
AdV. Drug DeliVery ReV. 1994, 14, 1-24.
(8) Erbacher, P.; Roche, A. C.; Monsigny, M.; Midoux, P. Putative role
of chloroquine in gene transfer into a human hepatoma cell line by
DNA lactosylated polylysine complexes. Exp. Cell Res. 1996, 225,
186-194.
(9) Gonzalez, H.; Hwang, S. J.; Davis, M. E. New class of polymers for
the delivery of macromolecular therapeutics. Bioconjugate Chem.
1999, 10, 1068-1074.
(10) Hwang, S. J.; Bellocq, N. C.; Davis, M. E. Effects of structure of
beta-cyclodextrin-containing polymers on gene delivery. Bioconjugate
Chem. 2001, 12, 280-290.
(11) Pun, S. H.; Davis, M. E. Development of a nonviral gene delivery
vehicle for systemic application. Bioconjugate Chem. 2002, 13, 630-
639.
(12) Reineke, T. M.; Davis, M. E. Structural effects of carbohydrate-
containing polycations on gene delivery. 1. Carbohydrate size and
its distance from charge centers. Bioconjugate Chem. 2003, 14, 247-
254.

Chloroquine and Chloroquine Analogues in NonViral Gene DeliVery
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 22 6531
(13) Reineke, T. M.; Davis, M. E. Structural effects of carbohydrate-
containing polycations on gene delivery. 2. Charge center type.
Bioconjugate Chem. 2003, 14, 255-261.
enzyme (DNAzyme) to tumor tissue by transferrin-modified, cyclo-
dextrin-based particles. Cancer Biol. Ther. 2004, 3, 641-650.
(32) Mishra, S.; Webster, P.; Davis, M. E. PEGylation significantly affects
cellular uptake and intracellular trafficking of non-viral gene delivery
particles. Eur. J. Cell Biol. 2004, 83, 97-111.
(33) Pun, S. H.; Bellocq, N. C.; Liu, A. J.; Jensen, G.; Machemer, T.;
Quijano, E.; Schluep, T.; Wen, S. F.; Engler, H.; Heidel, J.; Davis,
M. E. Cyclodextrin-modified polyethylenimine polymers for gene
delivery. Bioconjugate Chem. 2004, 15, 831-840.
(34) Colombo, M. I.; Bertini, F. Properties of Binding-Sites for Chloro-
quine in Liver Lysosomal Membranes. J. Cell. Physiol. 1988, 137,
598-602.
(35) Kulkarni, R. P.; Mishra, S.; Fraser, S. E.; Davis, M. E. Single cell
kinetics of intracellular, nonviral, nucleic acid delivery vehicle
acidification and trafficking. Bioconjugate Chem. 2005, 16, 986-
994.
(36) Singh, N. P.; McCoy, M. T.; Tice, R. R.; Schneider, E. L. A simple
technique for quantitation of low levels of DNA damage in individual
cells. Exp. Cell Res. 1988, 175, 184-191.
(37) O’Brien, R.; Allison, J.; Hahn, F. Evidence for intercalation of
chloroquine into DNA. Biochim. Biophys. Acta 1966, 129, 622-
624.
(38) Parker, F. S.; Irvin, J. L. The Interaction of Chloroquine with Nucleic
Acids and Nucleoproteins. J. Biol. Chem. 1952, 199, 897-909.
(39) Doglia, S. M.; Albinsson, B.; Hiort, C.; Norden, B.; Graslund, A.
QuinacrinesSpectroscopic Properties and Interactions with Poly-
nucleotides. Biopolymers 1993, 33, 1431-1442.
(40) Schaffer, D. V.; Fidelman, N. A.; Dan, N.; Lauffenburger, D. A.
Vector unpacking as a potential barrier for receptor-mediated polyplex
gene delivery. Biotechnol. Bioeng. 2000, 67, 598-606.
(41) Lloyd, J. B. Lysosome membrane permeability: implications for drug
delivery. AdV. Drug DeliVery ReV. 2000, 41, 189-200.
(42) Rashid, F.; Horobin, R. W.; Williams, M. A. Predicting the behaviour
and selectivity of fluorescent probes for lysosomes and related
structures by means of structure-activity models. Histochem. J. 1991,
23, 450-459.
(43) Diwu, Z.; Chen, C. S.; Zhang, C.; Klaubert, D. H.; Haugland, R. P.
A novel acidotropic pH indicator and its potential application in
labeling acidic organelles of live cells. Chem. Biol. 1999, 6, 411-
418.
(44) Hurwitz, S. J.; Terashima, M.; Mizunuma, N.; Slapak, C. A. Vesicular
anthracycline accumulation in doxorubicin-selected U-937 cells:
participation of lysosomes. Blood 1997, 89, 3745-3754.
(45) de Duve, C.; de Barsy, T.; Poole, B.; Trouet, A.; Tulkens, P.; Van
Hoof, F. Commentary. Lysosomotropic agents. Biochem. Pharmacol.
1974, 23, 2495-2531.
(46) Duvvuri, M.; Krise, J. P. A novel assay reveals that weakly basic
model compounds concentrate in lysosomes to an extent greater than
pH-partitioning theory would predict. Mol. Pharmacol. 2005, 2, 440-
448.
(47) Lukacs, G. L.; Haggie, P.; Seksek, O.; Lechardeur, D.; Freedman,
N.; Verkman, A. S. Size-dependent DNA mobility in cytoplasm and
nucleus. J. Biol. Chem. 2000, 275, 1625-1629.
(14) Popielarski, S. R.; Mishra, S.; Davis, M. E. Structural effects of
carbohydrate-containing polycations on gene delivery. 3. Cyclodextrin
type and functionalization. Bioconjugate Chem. 2003, 14, 672-678.
(15) Irvin, J. L.; Irvin, E. M. Spectrophotometric and Potentiometric
Evaluation of Apparent Acid Dissociation Exponents of Various
4-Aminoquinolines. J. Am. Chem. Soc. 1947, 69, 1091-1099.
(16) Maxfield, F. R. Weak Bases and Ionophores Rapidly and Reversibly
Raise the Ph of Endocytic Vesicles in Cultured Mouse Fibroblasts.
J. Cell Biol. 1982, 95, 676-681.
(17) Poole, B.; Ohkuma, S. Effect of Weak Bases on the Intralysosomal
Ph in Mouse Peritoneal-Macrophages. J. Cell Biol. 1981, 90, 665-
669.
(18) Sonawane, N. D.; Szoka, F. C.; Verkman, A. S. Chloride accumula-
tion and swelling in endosomes enhances DNA transfer by polyamine-
DNA polyplexes. J. Biol. Chem. 2003, 278, 44826-44831.
(19) Wibo, M.; Poole, B. Protein Degradation in Cultured-Cells .2. Uptake
of Chloroquine by Rat Fibroblasts and Inhibition of Cellular Protein
Degradation and Cathepsin-B1. J. Cell Biol. 1974, 63, 430-440.
(20) Poole, B.; Ohkuma, S.; Warburton, M. J. Accumulation of Weakly
Basic Substances in Lysosomes and Inhibition of Intracellular Protein
Degradation. Acta Biol. Med. Ger. 1977, 36, 1777-1788.
(21) Behr, J. P. The proton sponge: A trick to enter cells the viruses did
not exploit. Chimia 1997, 51, 34-36.
(22) Allison, J. L.; Obrien, R. L.; Hahn, F. E. DNAsReaction with
Chloroquine. Science 1965, 149, 1111-1113.
(23) Cohen, S. N.; Yielding, K. L. Spectrophotometric studies of the
interaction of chloroquine with deoxyribonucleic acid. J. Biol. Chem.
1965, 240, 3123-3131.
(24) Egan, T. Structure-Function Relationships in Chloroquine and
Related 4-Aminoquinoline Antimalarials. Mini-ReV. Med. Chem.
2001, 1, 113-123.
(25) Egan, T. J.; Hunter, R.; Kaschula, C. H.; Marques, H. M.; Misplon,
A.; Walden, J. Structure-function relationships in aminoquino-
lines: Effect of amino and chloro groups on quinoline-hematin
complex formation, inhibition of beta-hematin formation, and anti-
plasmodial activity. J. Med. Chem. 2000, 43, 283-291.
(26) Kaschula, C. H.; Egan, T. J.; Hunter, R.; Basilico, N.; Parapini, S.;
Taramelli, D.; Pasini, E.; Monti, D. Structure-activity relationships
in 4-aminoquinoline antiplasmodials. The role of the group at the
7-position. J. Med. Chem. 2002, 45, 3531-3539.
(27) De, D. Y. D.; Krogstad, F. M.; Byers, L. D.; Krogstad, D. J.
Structure-activity relationships for antiplasmodial activity among
7-substituted 4-aminoquinolines. J. Med. Chem. 1998, 41, 4918-
4926.
(28) De, D. Y.; Byers, L. D.; Krogstad, D. J. Antimalarials: Synthesis of
4-aminoquinolines that circumvent drug resistance in malaria para-
sites. J. Heterocycl. Chem. 1997, 34, 315-320.
(29) Bellocq, N. C.; Pun, S. H.; Jensen, G. S.; Davis, M. E. Transferrin-
containing, cyclodextrin polymer-based particles for tumor-targeted
gene delivery. Bioconjugate Chem. 2003, 14, 1122-1132.
(30) Bellocq, N. C.; Kang, D. W.; Wang, X. H.; Jensen, G. S.; Pun, S.
H.; Schluep, T.; Zepeda, M. L.; Davis, M. E. Synthetic biocompatible
cyclodextrin-based constructs for local gene delivery to improve
cutaneous wound healing. Bioconjugate Chem. 2004, 15, 1201-1211.
(31) Pun, S. H.; Tack, F.; Bellocq, N. C.; Cheng, J. J.; Grubbs, B. H.;
Jensen, G. S.; Davis, M. E.; Brewster, M.; Janicot, M.; Janssens, B.;
Floren, W.; Bakker, A. Targeted delivery of RNA-cleaving DNA
(48) Lechardeur, D.; Sohn, K. J.; Haardt, M.; Joshi, P. B.; Monck, M.;
Graham, R. W.; Beatty, B.; Squire, J.; O’Brodovich, H.; Lukacs, G.
L. Metabolic instability of plasmid DNA in the cytosol: a potential
barrier to gene transfer. Gene Ther. 1999, 6, 482-497.
JM060736S