Bioscience Reports, Vol. 15, No.1, 1995
`
`Cholesterol Enhances Cationic Liposome(cid:173)
`Mediated DNA Transfection of Human
`Respiratory Epithelial Cells
`Michael J. Bennett/ Michael H. Nantz,1 Rajiv P.
`Balasubramaniam/ Dieter C. Gruenert/ and Robert W.
`Malone2
`4
`•
`
`Received December 14, 1994; accepted January 27, 1995
`
`· Cationic liposome transfection reagents are useful for transferring polynucleotides into cells, and have
`been proposed for human pulmonary gene therapy. The effect of adding cholesterol to cationic lipid
`preparations has been tested by first formulating the cationic lipid N-[1-(2,3-dioleoyloxy)propyl-N-[1-
`(2-hydroxy)ethyl]-N,N-dimethyl ammonium iodide (DORI) with varying amounts of dioleoylphos(cid:173)
`phatidylethanolamine (DOPE) and cholesterol. Cholesterol was found to enhance lipid-mediated
`transfection in both the respiratory epithelial cells and mouse fibroblasts. These findings will facilitate
`nucleic acid transfection of many cell types including differentiated epithelial cell monolayers, and
`therefore may be useful for examining gene regulation in various cell types and for developing
`pulmonary gene therapy.
`
`KEY WORDS: cationic lipcisome; cholester9l; transfection; human respiratory epithelial cells.
`
`ABBREVIATIONS: N-[1-(2,3-dioleoyloxy)propyl]-N-[1-(2-hydroxy)ethyl]-N,N-dimethyl ammonium
`iodide (DORI); dioleoylphosphatidylethanolamine (DOPE); N-[1-(2,3-dioleoyloxy) propyl]-N,N,N(cid:173)
`trimethyl ammonium chloride (DOTMA); Eagle's modified essential medium (Mem); Dulbecco's
`Modified Eagle's Medium (DMEM).
`
`INTRODUCTION
`
`Methods for polynucleotide delivery are essential for the correlation of gene
`expression with phenotype, investigation of the regulation of gene expression, and
`gene therapy. Commonly used gene delivery methods include the use of
`
`1 Gene Therapy Program, Department of Chemistry, University of California, Davis, California 95616.
`2 Medical Pathology, University of California, Davis, California 95616.
`3 Gene Therapy Core Center, Cystic Fibrosis Research Center and Cardiovascular Research Institute,
`Department of Laboratory Medicine, University of California, San Francisco, California 94143.
`4 To whom correspondence should be addressed.
`
`47
`0144-8463/95/0200-0047$07.50/0 © 1995 Plenum Publishing Corporation
`
`Moderna Ex 1010-p. 1
`Moderna v Arbutus
`
`

`

`48
`
`Bennett, Nantz, Balasubramaniam, Gruenert and Malone
`
`recombinant viral vectors, physical methods (such as direct DNA injection), or
`pharmaceutical reagents. One pharmaceutical gene delivery technique involves
`the use of DNA packaged with lipids (liposomes). Originally, neutral or weakly
`charged lipids were utilized for liposomal gene delivery (1, 2, 3). The discovery
`that cationic liposomes spontaneously associate with DNA, fuse with cell
`membranes, and deliver the DNA into cytoplasm (4) has greatly advanced the
`utility of liposomal polynucleotide transfection.
`Synthetic cationic transfection lipid preparations spontaneously interact with
`DNA in solution to form lipid-DNA complexes. When placed onto tissue culture
`cells, these lipid-DNA complexes interact with the plasma membrane (4). As a
`consequence of this process, a portion of the exogenous DNA becomes localized
`in the nucleus and is subsequently transcribed. There are two leading hypotheses
`defining the mechanism of cationic lipid-mediated transfection: 1) plasma
`membrane fusion and subsequent cytoplasmic delivery (4) or 2) a pathway
`involving endocytic uptake (5, 7, 8). These hypotheses are not mutually exclusive,
`and may be active to a greater or lesser extent in different cell types. Either
`transfection pathway may be facilitated by alterations in liposome formulation
`which effect the fluidity of lipid/polynucleotide/cell membrane complexes.
`Sterols are commonly used for modulating the fluidity of both natural and
`artificial membranes (9). Cationic cholesterol derivatives have also been shown to
`mediate effective DNA transfection of cells and tissues when formulated into
`sonicated vesicles with DOPE (10). Therefore, these experiments have been
`designed to investigate the effect of cholesterol, the predominant mammalian
`sterol, on the efficacy of cationic liposome transfection formulations.
`Since respiratory epithelial cells are a potential therapeutic target for cationic
`lipid-mediated gene therapy, experiments focusing on the transfection of the
`apical surface of polarized respiratory epithelial cells were performed. Human
`surface airway epithelial cells have been isolated and cultured in vitro, but even
`under optimal conditions such cells have a limited life span and senesce or
`terminally differentiate (11, 12). To circumvent this problem, cultures of polarized
`immortal human bronchial epithelial cells have recently been developed (13), and
`this cell line (16HBE14o-) was chosen as the principal target for the analysis of
`the effects of cholesterol on cationic lipid-mediated DNA transfection.
`
`MATERIAL AND MElliODS
`
`Chemicals
`
`Dioleoylphosphatidylethanolamine was purchased from A vanti Polar Lipids
`(Inc. (Birmingham, A1). Cholesterol was purchased from Sigma Chemical
`Company (St. Louis, MO). The cationic lipid N-[1-(2,3-dioleoyloxy)propyl]-N-[1-
`(2-hydroxy)ethyl]-N,N-dimethyl ammonium iodide (DORI) was prepared using a
`procedure developed in our laboratories (14).
`
`Moderna Ex 1010-p. 2
`Moderna v Arbutus
`
`

`

`Liposome-Mediated DNA Transfection
`
`49
`
`Cell Culture
`
`Human respiratory epithelial cells (16HBE14o-) were cultured by plating on
`24 well tissue culture plastic ware coated with fibronectin, vitrogen (collagen), and
`bovine serum albumen as previously described (11, 12). Growth media consisted
`of Eagle's modified essential medium (MEM) supplemented with 10% fetal
`bovine serum, and cells were transfected as subconfiuent monolayers.
`As a control for cell specific effects, NIH 3T3 cells were also tested for DNA
`transfection using the various liposome formulations. NIH 3T3 were obtained
`from ATCC (CRL 1658). 3T3 cells were cultured on standard 24 well tissue
`culture plates using Dulbecco's Modified Eagle's Medium and 10% calf serum.
`
`Liposome Formulation
`
`An appropriate mass of the cationic lipid DORI and a neutral lipid (DOPE
`and/or cholesterol), as solutions in chloroform, were added to 1.9 ml sample vials
`to yield a 50:50 mole ratio of DORI: neutral lipid. The chloroform was removed
`via rotary evaporation at 37°C. The resulting thin lipid films were placed under
`high vacuum overnight to insure that all traces of solvent have been removed.
`The lipid mixture was resuspended with vortex mixing using 1 ml sterile water for
`injection (American Reagent Laboratories Inc.), resulting in a total lipid
`concentration of 2 JLmole lipid/1 ml water. This solution was sonicated until clear
`using a Branson sonifer 450 sonicator equipped with a cup horn (35°C, 15
`minutes).
`
`Plasmid DNA
`
`The plasmid pCMVL consists of the P. pyralis luciferase eDNA subcloned
`into the plasmid pRc/CMV (Invitrogen). This plasmid was transformed into
`competent E. coli DH5-a cells, amplified in terrific broth, and prepared by
`alkaline lysis with the isolation of covalently closed circular plasmid DNA using
`two rounds of CsCl-EtBr gradient ultracentrifugation. The plasmid DNA was
`subsequently treated with DNAse-free RNAse, phenol/chloroform extracted, and
`purified by precipitation from an ethanol/sodium acetate solution. DNA purity
`was determined by agarose gel electrophoresis and optical density (OD 260/280
`greater than or equal to 1.8).
`
`Transfection of Cultured Cells
`
`Tissue culture well plates (24 well) were plated 24 hours prior to transfection
`at approximately 80% confluency with either NIH 3T3 cells or 16HBE14o- cells.
`The growth media was removed via aspiration and the cells were washed once
`with 0.5 ml PBS/well. The liposome/DNA complexes were formed through
`sequential addition of an appropriate amounts of DMEM (serum-free), plasmid
`pCMVL (4 micrograms), and liposome formulation into a 2 ml Eppendorf tube
`
`Moderna Ex 1010-p. 3
`Moderna v Arbutus
`
`

`

`50
`
`Bennett, Nantz, Balasubramaniam, Gruenert and Malone
`
`gave a total volume of 800 ml. The addition of these substances was followed by
`thorough vortex mixing and incubation for 15 minutes at room temperature. A
`200 ml aliquot of the resultant transfection complex was added to each well and
`the cells were incubated for 4 hours at 37°C. At this time, 500 microliters of the
`appropriate growth media+ 10% calf serum/well was added and the cells were
`incubated for an additional 48 hours.
`
`Luciferase Assay
`
`Relative luciferase activity was determined by using the Enhanced Luciferase
`Assay Kit and Monolight 2010 luminometer (both from Analytical Luminescence
`Laboratories, San Diego, CA). This was accomplished by directly applying
`233.3 ml- of concentrated luciferase lysis buffer to each well and placing the cells
`on ice for 15 minutes. Removal of growth media was not necessary prior to the
`application of the lysis buffer. This technique proved to be more efficient and
`avoids the possibility of cell loss during media removal. An analogous experiment
`where the growth media was removed afforded similar results. Luciferase light
`emissions from 31.1 microliters of the lysate were measured over a 10 second
`period, and results expressed as a function of an assumed total lysate volume of
`933.3 JLL
`
`RESULTS
`
`To investigate the possibility that added cholesterol could enhance liposomal
`transfection ability, liposomes containing mixtures of the cationic lipid DORI and
`a neutral lipid component, DOPE and/ or cholesterol, were formulated and
`screened for their ability to transfect pCMVL DNA into human airway epithelial
`cells and NIH 3T3 murine fibroblasts. Optimal experimental conditions were
`obtained by varying the ratio of DOPE to cholesterol while keeping the amount
`of DORI (50:50 mole ratio of DORI:neutrallipid mixture) and the DORI:DNA
`phosphate charge ratio (2: 1) constant. These experimental conditions were
`specifically selected to allow direct comparison between the tri-component
`liposomes and 50:50 DOPE: DORI liposomes. Previous work by ourselves and
`others (6) has shown that a 50:50 DOPE:DORI formulation is optimal for
`hi-component transfection preparations composed of DOPE and DORI.
`The addition of cholesterol to the cationic lipid formulations enhances the
`transfection of differentiated human respiratory epithelial cells (Figure 1). This
`observation was consistent for all formulations, including those composed of only
`DORI and cholesterol as well as the tri-component liposomes. Furthermore,
`variable transfection activity was observed among the tri-component formula(cid:173)
`tions. The most active DNA transfection preparation was formulated using a
`molar ratio of 30:50:20 DOPE:DORI:cholesterol.
`To ascertain whether the enhancement of transfection which was observed
`with added cholesterol was limited to respiratory epithelial cells, NIH 3T3 cells
`
`Moderna Ex 1010-p. 4
`Moderna v Arbutus
`
`

`

`Liposome-Mediated DNA Transfection
`
`51
`
`::s
`
`... =
`0 u ..,
`= 0
`'-' "'
`[/J
`,...,
`<:>
`
`~
`
`'" = 0 ·;;:
`'" e
`.. ;:3
`..:
`'" "' ..
`"'
`~
`'-' ::s
`....:1
`
`2.1E+07
`
`1.8E+07
`
`1.5E+07
`
`1.2E+07
`
`9.0E+06
`
`6.0E+06
`
`3.0E+06
`
`O.OE+OO
`
`[J 50:50 DORI:Cho1estero1
`!lllJ 10:50:40 DOPE:DORI:Cho1estero1
`II 20:50:30 DOPE:DORI:Cholestero1
`IIIII 30:50:20 DOPE:DORI:Cholesterol
`II 40:50:10 DOPE:DORI:Cholesterol
`II 50:50 DOPE:DORI
`
`Fig. 1. Formulations containing 50 mole % of the cationic lipid DORI and varying amounts of
`DOPE and cholesterol were screened for their ability to functionally deliver the plasmid
`pCMVL (1 p.g) into human respiratory epithelial cells. Cell lysates obtained 48 hours after
`transfection were analyzed for luciferase specific activity. Each data point reflects the mean
`value of total light units derived from four transfections and the standard deviation from this
`mean.
`
`were transfected using the same formulations as were used for respiratory
`epithelial cell transfection (Figure 2). Again, it was determined that the
`tri-component formulation consisting of 30:50:20 DOPE:DORI:cholesterol was
`optimal for transfection of NIH 3T3 cells with DNA.
`
`4.0E+D7
`
`3.5E+D7
`
`3.0E+D7
`
`2.5E+07
`
`2.00+07
`
`1.5E+07
`
`!U5>:40DOPEDORI:Chollserol
`
`D 5U5l DORI:Cholesterol.
`lEI
`m 2U5>:30DOPEDORI:Chollserol
`Ill 3U5>:20DOPEDORI:Chollserol
`II 4U5>:1 ODOPEDORI:Chollserol
`IIIII 5U5l DOPE:DORI
`
`~
`00
`
`-
`
`~
`
`l.OE+D7
`
`;:j
`
`5.0E+06
`
`l!!
`8
`u
`"0
`§
`= ....
`"' §
`....
`"' .IQ
`~
`-a
`~ e 0.00+00
`~ !S
`.5
`Fig. 2. Optimization of NIH 3T3 fibroblast transfection mediated by liposomes containing the
`cationic lipid DORI (50 mole %) and variable amounts of DOPE and cholesterol. Analysis for·
`luciferase specific activity 48 hours after transfection indicates that !iposomes comprised of a
`30:50:20 mole ratio of DOPE:DORI:cholesterol are the most effective in mediating pCMVL
`delivery. These results compliment those obtained with the human respiratory epithelial cells.
`
`Moderna Ex 1010-p. 5
`Moderna v Arbutus
`
`

`

`52
`
`Bennett, Nantz, Balasubramaniam, Gruenert and Malone
`
`DISCUSSION
`
`The inclusion of cholesterol in cationic lipid transfection preparations
`enhances DNA transfection of differentiated human respiratory epithelial cells.
`This finding is not limited to epithelial cells, but is also applicable to a commonly
`used murine fibroblastic cell line. Respiratory epithelial cell transfection is
`important for investigations of gene expression, transport and secretion, and for
`the development of human gene therapies. Furthermore, methods for DNA
`transfection of cultured mammalian cells are critical to a wide range of
`experimentation concerning the biochemistry and molecular biology of nucleic
`acids. Therefore, this observation may be of interest to investigators who do not
`routinely transfect respiratory epithelial cells.
`The theoretical basis for the observation that the addition of cholesterol to
`cationic lipid preparations enhances DNA transfection of mammalian cells is
`speculative. It is known that cholesterol and some of it's derivatives modulate the
`physical state of both natural and artificial membranes. In general, at tempera(cid:173)
`tures above the phase transition temperature of phospholipid-containing membr(cid:173)
`anes (e.g. 37°C), cholesterol decreases membrane fluidity by restricting the
`mobility of fatty acyl chains (15). Thus the optimized tri-component formulation
`which is described (30:50:20 DOPE:DORI:cholesterol) is likely to be more rigid
`than the corresponding formulations which lack cholesterol. If this is the case, one
`explanation for cholesterol enhancement of transfection would be that a more
`rigid lipid component of a cationic lipid: DNA complex facilitates endocytic
`uptake rather than direct membrane fusion. Therefore, these results may support
`the hypothesis that endocytosis is the principal mechanism for cationic liposome(cid:173)
`mediated polynucleotide transfection.
`
`ACKNOWLEDGEMENTS
`
`We thank Jill G. Malone and Amy Sawyer for assistance with cell culture, Jill
`G. Malone and Karl Bassler for assistance with manuscript preparation and
`Phillip Montbriand for plasmid DNA preparation. This work was supported by
`grants from the Cystic Fibrosis Foundation (S884) and the California Tobacco
`Related Disease Research Program (4KT-0205) (RWM), and NIH grants
`DK47766 and DK46002 (DCG).
`
`REFERENCES
`
`1. Fraley, R. and Papahadjopoulos, D. (1982) Curr. Top. Microbial. Immuno/. 96:171-91.
`2. Uchida, T. (1988) Exp. Cell Res. 178:1-17.
`3. Gould-Fogerite, S., Mazurkiewicz, J. E., Raska Jr., K., Voelkerding, K., Lehman, J. M. and
`Mannino, R. J. (1989) Gene 84:429-38.
`4. Feigner, P. L., Gadek, T. R., Holm, M., Roman, R., Chan, H. W., Wenz, M., Northrop, J. P.,
`Ringold, G. M. and Danielsen, M. (1987) Proc. Nat/. Acad. Sci USA 84:7413-7.
`
`Moderna Ex 1010-p. 6
`Moderna v Arbutus
`
`

`

`Liposome-Mediated DNA Transfection
`
`53
`
`5. Farhood, H., Bottega, R., Epand, R. M. and Huang, L. (1992). Biochim. Biophys. Acta
`1111:239-46.
`6. Feigner, 1. N., Kummar, R., Sridhar, C. N., Wheeler, C., Tsai, Y. 1., Border, R., Ramsay, P.,
`Martin, M. and Feigner, P. (1994) J. Biol. Chern. 269:2550-2561.
`7. Pinnaduwage, P., Schmitt, L. and Huang, L. (1989) Biochim. Biophys. Acta 985:33-8.
`8. Debs, R., Pian, M., Gaensler, K., Clements, J., Friend, D. S. and Dobbs, L. (1992) Am. J. Respir.
`Cell Mol. Biol. 7:406-13.
`9. New, R. R. C. (1990) In: Liposomes-A Practical Approach. (New, R. R. C. ed), IRL Press,
`Oxford, UK, pp. 19-22.
`10. Gao, X. and Huang, L. (1991). Biochem. Biophys. Res. Commun. 179:280-5.
`11. Gruenert, D. C., Basbaum, C. B. and Widdicombe, 1. H. (1990) In: Vitro Cell Dev. Biol.
`26:411-8.
`12. Lechner, 1. LaVeck (1985) J. Tissue Culture Methods 9:43-48.
`13. Cozens, A. L., Yezzi, M. 1., Kunzelmann, K., Ohrui, T., Chin, L., Eng, K., Finkbeiner, W. E.,
`Widdicombe, 1. H. and Gruenert, D. C. (1994) Am. J. Respir. Cell Mol. Biol. 10:38-47.
`14. Bennett, M. 1., Malone, R. W. and Nantz, M. H. (Submitted). Tetrahedron Lett.
`15. Bloch, K. (1991) In: Biochemistry of Lipids, Lipoproteins and Membranes. (Vance, D. E. and
`Vance, 1. eds), Elsevier, New York, NY. pp. 368-369.
`
`Moderna Ex 1010-p. 7
`Moderna v Arbutus
`
`