`(cid:211) 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
`www.nature.com/gt
`
`ACQUIREDDISEASES
`RESEARCHARTICLE
`Stabilized plasmid-lipidparticlesfor systemicgene
`therapy
`
`P Tam1*, M Monck1*, D Lee1, O Ludkovski1, EC Leng1, K Clow1, H Stark2, P Scherrer3,
`RW Graham1 and PR Cullis1,3
`1Inex Pharmaceuticals Corporation, Burnaby, Canada; 2Institute for Molekularbiologie und Tumorforschung, Marburg, Germany;
`and 3Department of Biochemistry and Molecular Biology, University of British Columbia, Vancouver, BC, Canada
`
`The structure of ‘stabilized plasmid-lipid particles’ (SPLP)
`and their properties as systemic gene therapy vectors has
`been investigated. We show that SPLP can be visualized
`employing cryo-electron microscopy to be homogeneous
`particles of diameter 72 – 5 nm consisting of a lipid bilayer
`surrounding a core of plasmid DNA. It is also shown that
`SPLPexhibitlongcirculationlifetimes(circulationhalf-life.6
`h) following intravenous (i.v.) injection in a murine tumor
`model resulting in accumulation of up to 3% of the total
`injected dose andconcomitant reporter gene expression at
`a distal (hind flank) tumor site. In contrast, i.v. injection of
`
`nakedplasmidDNAorplasmidDNA–cationicliposomecom-
`plexes did not result in significant plasmid delivery to the
`tumorsiteorgeneexpressionatthatsite.Furthermore,itis
`shown that high doses of SPLP corresponding to 175 mg
`plasmid per mouse are nontoxic as assayed by monitoring
`serum enzyme levels, whereas i.v. injection of complexes
`give rise to significant toxicity at dose levels above 20 mg
`plasmidpermouse.ItisconcludedthatSPLPexhibitproper-
`ties consistent with potential utility as a nontoxic systemic
`genetherapy vector. Gene Therapy (2000) 7, 1867–1874.
`
`Keywords: liposomes;cancergene therapy;intravenousgenetherapy; tumourtransfection
`
`Introduction
`Gene therapies for systemic diseases such as cancer or
`inflammatory disorders clearly require systemic vectors.
`However, currently available vectors for gene therapy
`have limited utility for systemic applications. Recombi-
`nant virus vectors, for example, are rapidly cleared from
`the circulation following intravenous injection, limiting
`potential transfection sites to ‘first pass’ organs such as
`the liver.1,2 Nonviral systems, such as plasmid DNA–
`cationic liposome complexes, are also rapidly cleared
`from the circulation, and the highest expression levels are
`again observed in first pass organs, particularly the
`lungs.3–8
`Intravenous administration of chemotherapeutic drugs
`encapsulated in small (diameter <100 nm), long-circulat-
`ing (circulation half-life t. >5 h in murine models) lipo-
`somes results in preferential delivery of encapsulated
`drug to distal tumors due to increased vascular per-
`meability in these regions.9–11 It therefore follows that
`intravenous injection of plasmid DNA encapsulated in
`small, long circulating lipid particles should give rise to
`preferential delivery of plasmid DNA to tumor sites.
`Recent work has shown that plasmid DNA can be encap-
`sulated in small (approximately 70 nm diameter) ‘stabil-
`ized plasmid-lipid particles’ (SPLP) that contain one plas-
`mid per particle.12 These particles contain the ‘fusogenic’
`
`Correspondence: P Tam, Inex Pharmaceuticals Corporation, 100 – 8900
`Glenlyon Parkway, Burnaby, BC V5J 5J8, Canada
`*These authors contributed equally to this work.
`Received 31 January 2000; accepted 18 July 2000
`
`low
`lipid dioleoylphosphatidylethanolamine (DOPE),
`levels of cationic lipid and are stabilized in aqueous
`media by the presence of a poly(ethylene glycol) (PEG)
`coating. Here, we show that the structure of SPLP can be
`directly visualized employing cryo-electron microscopy
`to reveal homogeneous particles consisting of plasmid
`DNA entrapped within a bilayer lipid vesicle. Further-
`more, we show that these SPLP exhibit long circulation
`lifetimes and no evidence of systemic toxicities following
`i.v. injection in a murine tumor model. Under the experi-
`mental conditions employed, approximately 3% of the
`total injected SPLP dose was delivered to a subcutaneous
`tumor site and 1.5 % of the total intact plasmid dose
`could be detected at the tumor site at 24 h. Significant
`levels of reporter gene expression were observed at the
`tumor site employing the SPLP system, whereas no
`expression was observed following i.v.
`injection of
`‘naked’ plasmid DNA or plasmid DNA–cationic lipo-
`some complexes.
`
`Results
`
`SPLPconsist ofa plasmidtrapped inside abilayer lipid
`vesicle
`Previous work has shown that plasmid DNA can be
`encapsulated (trapping efficiency approximately 70%) in
`SPLP by a detergent dialysis procedure employing octyl-
`glucopyranoside (OGP).12 These SPLP are composed of
`DOPE, 5–10 mol% of the cationic lipid dioleoyldimethyl-
`ammonium chloride (DODAC) and PEG attached to a
`ceramide anchor containing an arachidoyl acyl group
`
`PROTIVA - EXHIBIT 2004
`Moderna Therapeutics, Inc. v. Protiva Biotherapeautics, Inc.
`IPR2018-00739
`
`
`
`1868
`
`SPLP for systemic gene therapy
`P Tam etal
`
`(PEG-CerC20). SPLP can be separated from non-encapsu-
`lated plasmid by ion exchange chromatography and can
`then be further purified by density gradient centrifug-
`ation to remove empty vesicles produced during the
`dialysis procedure. On the basis of the size and plasmid-
`to-lipid ratio of these purified SPLP it was determined
`that each SPLP contained one plasmid molecule.12
`Here, we
`further
`characterize
`SPLP structure
`employing cryo-electron microscopy. Following the pro-
`cedures summarized in Materials and methods, purified
`SPLP were prepared from DOPE:DODAC:PEG-CerC20
`(83:7:10; mol:mol:mol) and pCMVluc, whereas large unil-
`amellar vesicles (LUV) with the same lipid composition
`were prepared by extrusion of the hydrated lipid mixture
`through 100 nm pore size filters. As shown in Figure 1a,
`the cryo-electron micrographs clearly reveal SPLP to con-
`
`Figure 1 Cryo-electron micrographs of (a) purified SPLP and (b) LUV
`prepared by extrusion. SPLP were prepared from DOPE:DODAC:PEG-
`CerC20 (83:7:10; mol:mol:mol) and pCMVluc and purified employing
`DEAE column chromatography and density gradient centrifugation. LUV
`were prepared from DOPE:DODAC:PEG-CerC20 (83:7:10; mol:mol:mol)
`by hydration and extrusion through filters with 100 nm diameter pore
`size. The arrows in panel (a) indicate the presence of residual ‘empty’
`vesicles formed during the detergent dialysis process that were not
`removed by the density centrifugation purification step. The bar in panel
`(b) indicates 100 nm. For details of sample preparation and cryo-electron
`microscopy see Materials and methods.
`
`Gene Therapy
`
`sist of a lipid bilayer surrounding an internal structure
`consistent with entrapped plasmid DNA molecules.
`Small (diameter approximately 30 nm), empty vesicles
`formed during the detergent dialysis process13 that were
`not removed by density centrifugation do not exhibit
`such internal structure (see arrows in Figure 1a). This
`internal structure is also not observed in the LUV pro-
`duced by extrusion (Figure 1b). It may also be noted that
`SPLP as detected by cryo-electron microscopy have a
`remarkably homogeneous size (diameter 72 – 5 nm), in
`close agreement with measurements of SPLP diameter
`employing freeze–fracture electron microscopy (diameter
`64 – 9 nm).12 The homogeneous size and morphology of
`SPLP contrasts with the irregular morphology and large
`size distribution of the extruded vesicles. The narrow size
`distribution of SPLP was also reflected by quasi-elastic
`light scattering (QELS) measurements (data not shown)
`which indicated a mean diameter of 83 – 4 nm. Plasmid
`DNA–cationic liposome complexes made from DOPE:
`DODAC (1:1; mol:mol) LUV exhibited a large, hetero-
`geneous
`size distribution as determined by QELS
`(diameter 220 – 85 nm, data not shown).
`
`SPLP exhibitextended circulation lifetimes,preferential
`accumulationat tumorsites,andlowsystemic toxicities
`following intravenous injection
`The next set of experiments was aimed at characterizing
`the pharmacokinetics and biodistribution of SPLP follow-
`ing i.v. injection into tumor-bearing mice. SPLP were pre-
`pared with trace amounts of the lipid label, 3H-choles-
`teryl hexadecylether (3H-CHE) and were injected at a
`dose level equivalent to 100 mg plasmid DNA per mouse
`into C57Bl/6 mice bearing a subcutaneous Lewis lung
`carcinoma (approximately 200 mg) in the hind flank. The
`clearance of SPLP from the circulation as assayed by the
`lipid label (Figure 2a) corresponds to a first order process
`with a t. of 6.4 – 1.1 h. Relatively low levels of uptake by
`the lung and liver are observed (Figure 2b and c) whereas
`approximately 3% of the injected SPLP dose accumulates
`at the tumor site over 24 h (Figure 2d). Such tumor
`accumulation levels are comparable with those achieved
`for small, long-circulating liposomes containing conven-
`tional drugs such as doxorubicin, where approximately
`5% of the injected dose can be found at 24 h in larger
`(.0.5 g) tumors.14 In contrast to the behavior of the SPLP
`system,
`3H-CHE-labeled plasmid DNA–cationic lipo-
`some complexes were rapidly cleared from the circu-
`lation (t. ¿ 15 min), appearing predominantly in the lung
`and liver, and less than 0.2% of the injected dose was
`found at the tumor site at 24 h. The biodistribution of 3H-
`CHE labeled SPLP and complexes 4 and 24 h following
`injection are summarized in Table 1. Only trace amounts
`were detected in kidney, heart and lymph nodes.
`The levels of intact plasmid DNA in the circulation and
`tumor tissue following i.v. injection of naked plasmid
`DNA, plasmid DNA–cationic lipid complexes and SPLP
`were analyzed by Southern blot hybridization (Figure 3a,
`b and c, respectively) and quantified by phosphor-
`imaging analysis (Figure 3d and e). For naked plasmid,
`less than 0.01% of the injected dose remained intact in
`the circulation at 15 min, and no intact tumor-associated
`plasmid could be observed at any time. For plasmid
`administered in complexes, only a small fraction (,2%)
`was still intact in the circulation at 15 min and less than
`0.2% was found to be intact in tumor tissue at 1 h. In
`
`
`
`1869
`
`SPLP for systemic gene therapy
`P Tam etal
`
`(ALT) or aspartate aminotransferase (AST) were assayed
`for evidence of toxicity following i.v. administration of
`SPLP and plasmid DNA–cationic liposome complexes.
`Elevated ALT and AST levels are usually associated with
`liver damage, although elevated AST levels can also indi-
`cate systemic tissue damage. Mice receiving SPLP at dose
`levels as high as 175 mg plasmid DNA per mouse did not
`have significantly elevated serum levels of ALT and AST
`(Figure 4a). However, mice receiving doses of plasmid
`DNA–cationic liposome complexes corresponding to
`plasmid doses above 20 mg per mouse exhibited pro-
`gressively higher serum levels of ALT and AST, reaching
`levels 100-fold above normal levels at plasmid doses of
`75 mg (Figure 4b).
`
`Intravenously administeredSPLPpromotegene
`expression in adistaltumor
`It is of obvious interest to determine whether SPLP-
`mediated delivery of intact plasmid to the tumor site
`results in transgene expression at that site. Luciferase
`gene expression in tumor tissue was therefore monitored
`following i.v. injection of SPLP, naked plasmid DNA and
`plasmid DNA–cationic liposome complexes at dose levels
`corresponding to 100 mg plasmid DNA per mouse. This
`dose level corresponded to the maximum tolerated dose
`of complexes as evidenced by animal morbidity and mor-
`tality. As shown in Figure 5, administration of SPLP
`results in reporter gene expression at the tumor site, with
`maximum levels corresponding to 32 pg luciferase per
`gram of tumor tissue at the 48 h time-point and signifi-
`cant gene expression extending to 96 h after injection.
`Injection of free plasmid DNA or plasmid DNA–cationic
`liposome complexes, on the other hand, resulted in no
`detectable gene expression at the tumor site. It is of inter-
`est to note that i.v. administration of complexes did result
`in transfection in the lung, liver and spleen, whereas
`administration of SPLP did not result in detectable levels
`of gene expression in these organs (data not shown). In
`an attempt to understand why SPLP did not give rise to
`significant gene expression in the liver, the levels of intact
`plasmid in the liver 24 h after injection of SPLP into
`C57BI/6 mice (100 mg plasmid per mouse) bearing a sub-
`cutaneous Lewis lung carcinoma were analyzed by
`Southern blot hybridization. No intact plasmid could be
`detected in the liver whereas intact plasmid was readily
`detected at the tumor site (results not shown). This sug-
`gests that the ability of SPLP to transfect cells at the
`tumor site but not in the liver may reflect relatively rapid
`breakdown of SPLP and associated plasmid following
`uptake into liver phagocytes (Kuppfer cells), which play
`a dominant role in clearing liposomal systems from the
`circulation.15 Lower gene expression in the liver may also
`reflect the finding that nonviral vectors such as SPLP
`transfect dividing cells much more efficiently than non-
`dividing cells16 or that Kuppfer cells are less readily
`transfected than tumor cells.
`
`Discussion
`This study demonstrates that SPLP consist of plasmid
`DNA encapsulated in a bilayer vesicle, and that systemic
`administration of SPLP results in significant accumu-
`lation and transfection at a distal tumor site. There are
`three important features of these results. The first con-
`cerns the structure of SPLP, which represents a major
`
`Gene Therapy
`
`Figure 2 Pharmacokinetics, tissue distribution and tumor accumulation
`of SPLP and plasmid DNA–cationic liposome complexes following intra-
`venous administration in tumor-bearing mice as reported by the 3H-CHE
`lipid marker. The levels of complexes (s) and SPLP (d) in the circulation,
`the lung, the liver and in Lewis lung tumor tissue are shown in panels
`(a), (b), (c) and (d), respectively. The accumulations in liver, lung and
`tumor were corrected for plasma contributions29 and are expressed as a
`percentage of the total injected dose.
`
`Table 1 Biodistribution of SPLP and plasmid DNA–cationic lipo-
`some complexes in mice 4 and 24 h following i.v. injection
`
`Tissue
`
`% Injected dose (s.e.m.)
`
`SPLP
`
`Complexes
`
`4 h
`
`24 h
`
`4 h
`
`24 h
`
`Plasma
`Liver
`Lung
`Spleen
`Tumor
`
`55.0 (1.7)
`7.0 (0.6)
`0.0 (0.1)
`0.4 (0.1)
`0.2 (0.0)
`
`6.4 (1.0)
`23.0 (4.3)
`0.2 (0.1)
`1.6 (0.1)
`2.8 (0.5)
`
`1.7 (0.2)
`35.2 (2.3)
`1.8 (0.8)
`0.2 (0.2)
`0.2 (0.1)
`
`1.4 (0.3)
`35.1 (3.5)
`0.5 (0.0)
`0.1 (0.3)
`0.3 (0.2)
`
`Both SPLP and complexes contained pCMVLuc as well as trace
`levels of the 14C-labeled CHE lipid marker and were administered
`at a dose level of 100 mg plasmid per mouse. The biodistribution
`was measured employing the CHE lipid marker.
`s.e.m., standard error of the mean.
`
`contrast, following i.v. injection of SPLP, approximately
`85% of the injected plasmid DNA remained in intact form
`in the circulation at 15 min, and progressively higher lev-
`els of intact plasmid accumulated at the tumor site over
`the time-course of the experiment. The levels achieved at
`24 h correspond to approximately 1.5% of the total
`injected plasmid DNA dose. The circulation half-life of
`intact plasmid DNA following injection of SPLP was cal-
`culated to be 7.2 –
`1.6 h, in good agreement with the
`circulation half-life of 3H-CHE-labeled SPLP, confirming
`the highly stable nature of SPLP in the circulation.
`Serum enzyme levels of alanine aminotransferase
`
`
`
`1870
`
`SPLP for systemic gene therapy
`P Tam etal
`
`Figure 3 Pharmacokinetics and tumor accumulation of plasmid DNA following intravenous administration of naked plasmid, plasmid DNA–cationic
`liposome complexes and SPLP as reported by a Southern blot analysis. The Southern blot hybridizations shown in panels (a), (b) and (c) result from
`plasmid DNA isolated from blood and tumor tissue of mice injected with naked plasmid DNA, plasmid DNA–cationic lipid complexes and SPLP,
`respectively. Each panel shows pCMVluc (2 ng) to indicate the position of intact plasmid DNA. The levels of intact plasmid resulting from i.v. injection
`of naked plasmid DNA (l), plasmid DNA–cationic liposome complexes (s) and SPLP (d) were quantified for plasma (panel d) and tumor tissue (panel
`e) by phospor-imaging analysis and converted to mass quantities of plasmid DNA by comparison to a standard curve made from known amounts of
`plasmid DNA. Tumor accumulations of plasmid were corrected for plasma contributions and expressed as a percentage of the total injected plasmid
`DNA dose.29
`
`advance for plasmid encapsulation in liposomal delivery
`systems. Second, it is of interest to compare the proper-
`ties of the SPLP system for systemic gene delivery and
`distal tumor transfection with the properties of other
`viral or nonviral gene delivery systems. Finally, the possi-
`
`bilities for further optimization of the SPLP system are
`of interest. We discuss these areas in turn.
`The cryo-electron microscopy results presented here
`establish the structure of SPLP as a plasmid surrounded
`by a lipid bilayer envelope. This represents the first direct
`
`Gene Therapy
`
`
`
`SPLP for systemic gene therapy
`P Tam etal
`
`1871
`
`Figure 5 Transgene expression at a distal tumor site following intra-
`venous injection of naked plasmid DNA (l), plasmid DNA–cationic lipid
`complexes (s) and SPLP (d). Mice bearing subcutaneous Lewis lung
`tumors were injected i.v. with doses containing 100 mg of pCMVluc.
`Tumors were harvested at the indicated times and assayed for luciferase
`activity. The level of transgene expression reported is normalized for the
`weight of the tumor tissue.
`
`enable transgene expression corresponding to 30 pg
`luciferase per gram tumor at that site with no evidence
`of toxicity (as indicated by serum enzyme levels). Deliv-
`ery of approximately 3% of the injected dose of SPLP at
`a 200 mg tumor site corresponds to more than 1000 plas-
`mid copies per tumor cell, assuming a cell density of 1 ·
`109 per milliliter. It is of interest to compare these proper-
`ties with the behavior of other gene delivery systems. In
`the case of viral vectors, there have been three reports of
`transgene expression in liver metastases and in a distal
`tumor following systemic administration of a recombi-
`nant vaccinia virus18,19 and a selectively replicating
`adenovirus.20 These viral vectors are replication incom-
`petent in normal nondividing cells but can selectively
`replicate in tumor cells resulting in transgene expression
`in tumors and antitumoral efficacy. The major drawback
`of these viral vectors is the immune response, which
`occurs within 6 days. In the case of nonviral vectors such
`as plasmid DNA–cationic polymer ‘polyplexes’, there is
`only one report showing transfection of distal tumors fol-
`lowing i.v. injection.21 This work utilized a PEG-contain-
`ing polyplex that exhibits plasmid circulation half-lives
`of less than 0.5 h following intravenous injection and
`gave rise to transfection at a distal tumor site, achieving
`transfection levels corresponding to approximately 250
`pg/g tumor, approximately eight-fold higher than the
`levels reported here.
`liposome
`With regard to plasmid DNA–cationic
`complexes (‘lipoplexes’), a number of studies have
`characterized transfection properties
`following i.v.
`administration,3–7 however, only two studies by Xu and
`co-workers have demonstrated transfection at a distal
`tumor site.22,23 In the initial study,22 less than 5% of the
`cells at the tumor site were transfected as indicated by
`immunohistochemical staining, whereas in the second
`study using transferrin targeted complexes 20–30% of the
`cells were transfected. The levels of gene expression
`could not be related to the levels observed here. Issues
`related to circulation lifetimes, plasmid tumor accumu-
`lation and toxicity were not addressed. An additional
`study24 has demonstrated the presence of complexes at a
`distal tumor site following i.v. injection but the levels of
`
`Gene Therapy
`
`Figure 4 Toxicity resulting from i.v.
`injection into mice of varying
`amounts of SPLP (panel a) and plasmid DNA–cationic lipid complexes
`(panel b) as assayed by determining serum levels of the hepatic enzymes
`alanine aminotransferase (ALT) and aspartate aminotransferase (ASP).
`The serum levels of AST (l) and ALT (L) were measured 24 h after
`injection.
`
`demonstration that plasmid can be entrapped in small
`(diameter approximately 70 nm), well defined vesicular
`systems containing a single plasmid per vesicle. Entrap-
`ment of a plasmid such as pCMVluc, which contains 5650
`bp, in a supercoiled configuration in a 70 nm diameter
`vesicle represents a solution for a difficult packing prob-
`lem. For example, electron micrographs of supercoiled
`4.4 kbp plasmids reveals extended lengths of approxi-
`mately 500 nm and average (two dimensional) diameters
`in the range of 350 nm, suggesting an average diameter
`for free supercoiled pCMVluc of approximately 400 nm.17
`The detergent dialysis process clearly involves a partial
`condensation of entrapped plasmid to allow encapsul-
`ation in a 70 nm diameter vesicle. The mechanism of
`entrapment is not understood in detail, but appears to
`proceed via association of plasmid with lipid structures
`formed as
`intermediates
`in the detergent dialysis
`process.12
`SPLP exhibit extended circulation lifetimes (t. approxi-
`mately 7 h) following i.v. injection, can deliver significant
`amounts of intact plasmid to a distal tumor site and
`
`
`
`1872
`
`SPLP for systemic gene therapy
`P Tam etal
`
`gene transfer were not measured. In general, i.v. injection
`of complexes gives rise to high levels of
`transgene
`expression in the lungs, with lower levels of expression in
`the spleen, liver, heart and kidneys. Similar results were
`observed for the complexes employed in this investi-
`gation. The lung expression appears to arise from depo-
`sition in lung microvasculature and reflects the rapid
`clearance of plasmid DNA–cationic lipid complexes from
`the circulation due to their large size (.200 nm diameter)
`and high cationic lipid content.8 This is consistent with
`the observation that murine B16 tumors seeded in the
`pulmonary vascular compartment can be transfected by
`i.v. administered complexes.3 Finally, as clearly shown in
`this study, administration of complexes is often associa-
`ted with significant toxicity.
`The final point of discussion concerns the utility of
`SPLP as a systemic gene therapy vector and the potential
`for further optimization. As indicated above, despite the
`delivery of large amounts of intact plasmid to the tumor
`site, the levels of gene expression observed for the SPLP
`system are modest, albeit comparable with or superior
`than can be achieved with other vectors. It is likely that
`the low levels of transfection reflect low levels of uptake
`of SPLP into cells at the tumor site due to inhibition of
`cell association and uptake by the PEG coating.25 In vitro
`studies have shown that SPLP containing PEG-CerC20 are
`accumulated into cells to a very limited extent, however,
`the SPLP that are taken up are highly transfection
`potent.26 The challenge that faces the next stage of SPLP
`development is, therefore, to devise methods of enhanc-
`ing intracellular delivery of SPLP following arrival at the
`tumor site. There are a number of avenues to explore.
`First, the dissociation rate of the PEG coating from the
`SPLP can be modulated by varying the acyl chain length
`of the ceramide anchor,12 suggesting the possibility of
`developing PEG-Cer molecules that remain associated
`with the SPLP long enough to promote passive targeting
`to the tumor, but which dissociate quickly enough to
`allow transfection after arriving at the tumor site. Alter-
`natively, improvements may be expected from inclusion
`of cell-specific targeting ligands in SPLP to promote cell
`association and uptake. Finally, the nontoxic properties
`of SPLP allow the possibility of higher doses. A dose of
`100 mg plasmid DNA per mouse corresponds to a dose
`of approximately 5 mg plasmid DNA per kilogram body
`weight. This is a relatively low dose level in comparison
`to small molecules used for cancer therapy, which typi-
`cally are used at dose levels of 10 to 50 mg per kg
`body weight.
`In summary, we have shown that SPLP consist of plas-
`mid encapsulated in a lipid vesicle. Furthermore, we
`have demonstrated that, in contrast to naked plasmid or
`complexes, SPLP exhibit extended circulation lifetimes
`following intravenous injection, resulting in plasmid
`accumulation and transgene expression at a distal tumor
`site in a murine model. The levels of transgene expression
`achieved are modest, but are comparable or superior to
`distal tumor expression levels achieved employing other
`vectors. Further improvements can be expected due to
`the low toxicity and flexible nature of the SPLP system.
`Materials andmethods
`Lipids andplasmid
`1,2-Dioleoyl-3-phosphatidylethanolamine (DOPE) was
`obtained from Northern Lipids (Vancouver, BC, Canada).
`
`The cationic lipid N,N-dioleyl-N,N-dimethylammonium
`chloride (DODAC) and 1-O-[29-(v-methoxypolyethylene-
`glycol)succinoyl]-2-N-arachidoylsphingosine (PEG-CerC20)
`were prepared at Inex Pharmaceuticals (Burnaby, BC,
`Canada) using previously described methods.27 3H-lab-
`eled cholesteryl hexadecyl ether (CHE) was purchased
`from Dupont NEN Products (Boston, MA, USA). The
`pCMVluc plasmid encodes the Photinus pyralis luciferase
`gene under the control of the human CMV immediate–
`early promoter. Plasmid DNA was propagated in E. coli
`(DH5a) and purified using the alkaline lysis method fol-
`lowed by two rounds of CsCl/ethidium bromide density
`equilibrium centrifugation.
`
`SPLP andplasmidDNA–cationic liposome complex
`preparation
`Plasmid DNA was encapsulated in SPLP composed of
`DOPE/DODAC/PEG-CerC20 (83:7:10; mol:mol:mol) by
`the detergent dialysis method.12 Lipids were dissolved in
`ethanol and dried to a lipid film in a round-bottom flask.
`The lipid mixture was resuspended in HBS (5 mm Hepes,
`150 mm NaCl, pH 7.5) containing 200 mm OGP and 0.4
`mg/ml pCMVluc. The final lipid concentration was 10
`mg/ml. When required, 3H-CHE was added to a specific
`activity of 1.0 mCi/mg total lipid. The mixture of lipid,
`plasmid and OGP was dialyzed against 4 l of HBS for 2
`days with three changes. Untrapped plasmid was
`removed by DEAE-Sepharose CL-6B chromatography,
`and plasmid DNA-containing SPLP were purified by
`sucrose gradient centrifugation (2.5%/5%/10%)
`in a
`Beckman SW 28 rotor (16 h at 107000 g) (Beckman, Fuller-
`ton, CA, USA). DNA-containing particles banding at the
`5%/10% sucrose interface were collected and concen-
`trated by ultrafiltration before the DNA concentration
`was adjusted to 500 mg/ml. The final lipid composition
`was determined by HPLC analysis. DNA was quantified
`by picogreen (Molecular Probes, Eugene, OR, USA) flu-
`orescence of TX-100-solubilized SPLP preparations. Plas-
`mid DNA–cationic liposome complexes were prepared
`by adding pCMVluc to large unilamellar vesicles (LUV)
`composed of DOPE:DODAC (1:1; mol:mol) to a final
`charge ratio (+/- ) of 3.0 in 5% glucose. The LUV were
`prepared by extrusion through 100 nm pore size filters
`according to standard procedures.28
`
`Cryo-electron microscopy
`A drop of buffer containing SPLP was applied to a stan-
`dard electron microscopy grid with a perforated carbon
`film. Excess liquid was removed by blotting leaving a
`thin layer of water covering the holes of the carbon film.
`The grid was rapidly frozen in liquid ethane, resulting in
`vesicles embedded in a thin film of amorphous ice.
`Images of the vesicles in ice were obtained under cryo-
`genic conditions at a magnification of 66 000 and a defo-
`cus of - 1.5 micron using a Gatan cryo-holder in a Philips
`CM200 FEG electron microscope (Eindhoven, The
`Netherlands).
`
`Quasi-elastic lightscattering
`The mean diameter of SPLP was measured by quasi-elas-
`tic light scattering (QELS) using a Nicomp Model 370
`Sub-Micron particle sizer (Santa Barbara, CA, USA)
`operated in the particle mode.
`
`Gene Therapy
`
`
`
`Clearance, biodistribution andtumor accumulationof
`SPLP
`Lewis lung carcinoma cells (300000; ATCC CRL-1642)
`were implanted subcutaneously in the hind flank of 6-
`week-old female C57BL/6 mice (Harlan, Indianapolis,
`IN, USA) and the tumor allowed to grow to approxi-
`mately 200 mg (12–14 days). Injected materials were then
`administered intravenously (lateral tail vein injection).
`All injected doses are reported in micrograms of plasmid
`DNA per mouse. Blood from animals was collected at
`the appropriate time-points into blood collection tubes by
`cardiac puncture. Tumors and organs were quickly
`removed and frozen at - 70(cid:176) C. Aliquots of plasma separ-
`ated from blood were analyzed for 3H-CHE by liquid
`scintillation counting. Plasmid DNA was purified by tre-
`ating 50 ml plasma with 1· proteinase K buffer (1.0
`mg/ml proteinase K, 100 mm NaCl, 10 mm Tris-HCl, 25
`mm EDTA, 0.5% SDS, pH 8.0). After incubation at 37(cid:176) C
`for 3 h, the samples were purified by phenol/chloroform
`extraction followed by ethanol precipitation. Tumors
`were homogenized in PBS containing 100 mm EDTA pH
`8.0 by using the Fast-Prep 120 homogenizer system (Bio
`101, Vista, CA, USA). DNA was purified from an aliquot
`of the tumor homogenates using the DNAzol reagent
`according to the manufacturer’s guidelines (Life Techno-
`logies, Bethesda, MD, USA). The DNA preparations from
`tumor homogenates were digested with EcoRI. DNA
`samples were subject to electrophoresis through 1.0%
`agarose gels, transferred to nylon membranes and sub-
`jected to Southern blot hybridization using a random
`primed 32P-labeled restriction fragment from the lucifer-
`ase gene. Hybridization intensity was quantified using a
`STORM840 phosphor-imager (Molecular Dynamics, Sun-
`nyvale, CA, USA) and converted to mass of DNA using
`a standard curve constructed with known amounts of
`plasmid DNA.
`
`Luciferase assays
`Tumor tissue was homogenized in 1· Cell Culture Lysis
`Reagent (CCLR) (Promega, Madison, WI, USA) using the
`Fast-Prep 120 homogenizer system (Bio 101). The homo-
`genates were centrifuged at 10000 g for 2 min before 20
`ml of the supernatant was assayed for luciferase activity
`using the Luciferase Assay System Kit (Promega) on an
`ML3000 microtiter plate luminometer (Dynex Techno-
`logies, Chantilly, VA, USA). Luciferase activities were
`converted to mass quantities of purified luciferase by
`comparison with a standard curve generated by assaying
`known amounts of purified Photinus pyralis luciferase
`enzyme (Boehringer-Mannheim, Laval, PQ, Canada)
`diluted into untreated tumor extract.
`
`Hepaticrelease enzyme assays
`Plasma from normal C57BL/6 mice injected with SPLP,
`plasmid DNA–cationic lipid complexes or HBS was reco-
`vered 24 h after injection by centrifugation and assayed
`immediately for ALT or ASP using commercially avail-
`able kits (Sigma, St Louis, MO, USA).
`
`Acknowledgements
`We thank T Nolan, N Turcotte and J Johnson for excellent
`technical assistance, Dr P Joshi for assistance in preparing
`the manuscript and Dr S Ansell and Dr Z Wang for
`supplying the DODAC and PEG-CerC20 respectively.
`
`SPLP for systemic gene therapy
`P Tam etal
`
`Plasmid DNA was prepared by C Giesbrecht and J
`Thompson.
`
`1873
`
`References
`1 Huard J et al. The route of administration is a major determinant
`of the transduction efficiency of rat tissues by adenoviral recom-
`binants. Gene Therapy 1995; 2: 107–115.
`2 Worgall S, Wolff G, Falck PE, Crystal RG. Innate immune mech-
`anisms dominate elimination of adenoviral vectors following in
`vivo administration. Hum Gene Ther 1997; 8: 37–44.
`3 Zhu N, Liggitt D, Liu Y, Debs R. Systemic gene expression after
`intravenous DNA delivery into adult mice. Science 1993; 261:
`209–211.
`4 Thierry AR et al. Systemic gene therapy: biodistribution and
`long-term expression of a transgene in mice. Proc Natl Acad Sci
`USA 1995; 92: 9742–9746.
`5 Li S, Huang L. In vivo gene transfer via intravenous adminis-
`tration of cationic lipid-protamine-DNA (LPD) complexes. Gene
`Therapy 1997; 4: 891–900.
`6 Liu Y et al. Factors influencing the efficiency of cationic lipo-
`some-mediated intravenous gene delivery. Nat Biotechnol 1997;
`15: 167–173.
`7 Templeton NS et al. Improved DNA: liposome complexes for
`increased systemic delivery and gene expression. Nat Biotechnol
`1997; 15: 647–652.
`8 Huang L, Li S. Liposomal gene delivery: a complex package.
`Nat Biotechnol 1997; 15: 620–621.
`9 Proffitt RT et al. Liposomal blockade of the reticuloendothelial
`system: improved tumor imaging with small unilamellar ves-
`icles. Science 1983; 220: 502–505.
`10 Gabizon A, Papahadjopoulos D. Liposome formulations with
`prolonged circulation time in blood and enhanced uptake by
`tumors. Proc Natl Acad Sci USA 1988; 85: 6949–6953.
`11 Chonn A, Cullis PR. Recent advances in liposomal drug-
`delivery systems. Curr Opin Biotechnol 1995; 6: 698–708.
`12 Wheeler JJ et al. Stabilized plasmid-lipid particles: construction
`and characterization. Gene Therapy 1999; 6: 271–281.
`13 Mimms LT et al. Phospholipid vesicle format