CA2183647A1 - Process for treating eucaryotic cells - Google Patents

Abstract

In a method of treating eukaryotic cells, for applications in which the toxicity of the lipopolysaccharide has a detrimental effect, substances which bind lipopolysaccharide and neutralise its toxicity are used during transfection or in order to purify the foreign material. The process is of particular benefit when foreign material is introduced into the cells, particularly in gene transfer methods.
Compositions containing a lipopolysaccharide-binding substance may be used in cell culture applications as a medium and in therapeutic applications as a drug.

Description

218~64 1 S016573J.64 Methods of treating eukaryotic cells The present invention relates to methods of treating eukaryotic cells.

One problem which frequently occurs when working with eukaryotic cells in culture or in vivo is the presence of endotoxin.

Endotoxin (lipopolysaccharide, LPS), a major component of the cell walls of gram-negative bacteria, is frequently found, for example, as a cont~m;n~nt in plasmid DNA preparations, as up to 40~ of the surface LPS of E. coli are released with the methods conventionally used to prepare plasmid DNA. As a result of the negative charges, LPS behaves in a similar manner to DNA on anion exchange chromatography resins, but because of the size it has in its micellar form, LPS
behaves like a large DNA molecule on size exclusion resins. The density of LPS in CsCl is similar to that of plasmid/EtBr complexes, which means that the DNA in CsCl bands can easily become contaminated. When transfection with DNA is carried out the cells thus come into contact with LPS.

Since the LPS molecule is toxic and is a powerful stimulator of the m~mm~1ian imml~ne system, its presence during the treatment of cells is undesirable. For this reason there have already been numerous attempts at detected LPS and finding methods of eliminating this molecule or gram-negative bacteria or neutralising the harmful effects caused by th~m (Cordle et al., 1993;
Elsbach and Weiss, 1993; Golenbock et al., 1993; Haziot et al., 1993a; Lynn et al., 1991; Perera et al., 1993;
Rustici et al., 1993; Tobias et al., 1988; Tobias et 3~7 al., 1989; Ziegler-Heitbrock and Ulevitch, 1993).

The aim of the present invention is to overcome the toxicity problems connected with the presence of LPS in the treatment of eukaryotic cells, which arise in particular in connection with the introduction of foreign material, particularly DNA, into the cell.

In the course of the experiments carried out to find a solution to this problem, it was first established that toxicity problems occur when transferring plasmid DNA by means of adenovirus-aided receptor-mediated endocytosis, as described inter alia in WO 93/07283 and by Wagner et al., 1992; Cotten et al., 1992; Curiel et al., 1991;
Cotten et al., 1993; Cristiano et al., 1993a, and Cristiano et al., 1993b, into primary human skin fibroblasts or into primary human melanoma cells. Using experiments with LPS containing DNA or pure LPS, these problems have been traced back to the presence of LPS.
Toxicity occurred with amounts of 100 ng/ml of free LPS
or, if LPS were incorporated in the poly-lysine/adenovirus complexes, with amounts of 100 pg/ml.

Within the scope of the present invention, toxicity which could be traced back to LPS was also found in gene transfer methods independent of adenovirus; the possibility of the toxicity being connected with polylysine was also ruled out.

Starting from the findings obtained, first of all a method was developed for removing the LPS impurities from the DNA and thereby eliminating the toxicity. A
simple and effective method of removing LPS uses the detergent Triton X-114. At temperatures below 20C, Triton X-114 is miscible with aqueous solutions and at temperatures above 20C it separates into a distinct phase (Bordier, 1981). The phenomenon can be exploited 2183~47 in order to extract the lipophilic LPS molecule from aqueous protein solutions (Aida and Pabst, 1990) or from DNA preparations (Manthorpe et al., 1993).

An alternative method of separating off LPS uses polymyxin B, a cyclic mould peptide antibiotic (Storm et al., 1977), which binds with high affinity to the lipid A/ketodeoxyoctolonic acid component of LPS (Ka = 1.15 x 107 M 1; Rustici et al., 1993; Lynn and Golenbock, 1992;
Schindler and Osborn, 1979). Polymyxin has hitherto been used to remove LPS from proteins; to do this, polymyxin is used in a resin-bound form in chromatography columns.

The experiments with purified DNA showed that the removal of LPS from the DNA preparations brings about an increase in gene expression.

LPS is, however, not only an unwanted accompaniment to DNA preparations but a ubiquitous cont~min~nt which occurs, inter alia, in the cell culture media normally used. The LPS content of serum preparations may fluctuate over a wide range; unless special steps are taken to exclude contaminated preparations, the content may be up to 10 to 50 ng/ml. A problem also occurs in connection with the glass apparatus conventionally used which have come into contact with bacteria. Since it is extremely laborious to monitor the LPS content of all reagents and glass containers, in order to solve the problem set out hereinbefore, attempts have been made to find a method which combats, from the outset, the problems connected with the presence of LPS in the treatment of eukarotic cells, particularly when transporting foreign material into the cells.

The present invention relates to a method of treating eukaryotic cells, particularly for the incorporation of ~18:~647 foreign material, in which the cells are treated with a substance which binds lipopolysaccharide and blocks its toxicity to the cells, and/or the foreign material introduced into the cell is one which has been purified to remove LPS before the material is introduced into the cells.

Substances having this property are hereinafter referred to as "LPS binding substances" in the interests of simplicity; the substance may occur as a single substance or as a mixture.

The invention may be applied to all methods for introducing foreign material into the cell where the presence of LPS causes a toxicity which has a negative effect on the efficiency of the method. These include, inter alia, methods of importing drugs or drug conjugates or toxins into the cell.

Preferably, the method is applied to transfection and infection methods for introducing foreign DNA or RNA
into the cell, e.g. the calcium phosphate, microinjection and DEAE method, methods which operate with liposomes or cationic lipids, as well as methods which use recombinant viruses; it is particularly preferable to use the method in conjunction with gene transfer methods based on receptor-mediated endocytosis.
A summary of such methods is provided, for example, by Cotten and Wagner, 1993. The need to use the present invention during transfection of the cells can easily be determined by carrying out the method in the presence or absence of LPS, the other conditions remaining identical, and comparing the results obtained.

The LPS binding substances are required both to bind LPS
with sufficient avidity and also block its interaction with cell components, which is responsible for the ~1836~

toxicity. (In contrast to substances which satisfy these requirements, there are proteins, e.g. the lipopolysaccharide binding proteins, an acute phase protein which does indeed bind to LPS with a high affinity but activates the response of the cell to the LPS toxicity; such an effect would be undesirable with the scope of the present invention).

In the embodiment of the invention in which the foreign material to be imported into the cell, especially DNA, is purified to remove LPS, the DNA is preferably treated with polymyxin, particularly by chromatography through a polymyxin resin, or by extraction with a suitable detergent such as Triton X-114.

In order to treat the cells during transfection the LPS
binding substances are used in amounts which saturate the LPS present at least to a degree such that its toxicity with regard to the intended use is neutralised and in which the substances themselves are not toxic.

In order to determine the suitability of the substance and its optimum concentration, transfection experiments are appropriately carried out, e.g. with a view to a particular transfection to be carried out with the cell type in question. By means of titrations, a suitable concentration is found at which the substance brings about an increase in gene expression without itself having a detrimental effect on the cells. In addition to the transfection system, the effect of the LPS
binding substance on the morphology of the cells may be investigated under the microscope. Furthermore, assays may be used to measure the release of cellular components as a measurement of the necrosis or apoptosis occurring as a reaction to the LPS toxicity. Example`s of this are the commercially obtainable lactate dehydrogenase assay which is used clinically on a large 2183~47 scale or the recently described epifluorescence method after staining with calcein-AM and propidium iodide (Lorenzo et al., 1994). Using such methods, it was found, within the scope of the present invention, that in the case of polymyxin B the medium should preferably contain no more than 100 ~g/ml and the lower limit may be 3 ~g/ml or below. The neutralisation achieved with polymyxin began to deteriorate at 1 ~g/ml and below.
This corresponds to a requirement for 250 ng of polymyxin to neutralise 10 to 100 ng of LPS. It should be borne in mind that the LPS may be incorporated in the transfection complexes and therefore need not absolutely be available for binding to the LPS binding substances.
It is also possible for neutralisation by means of the LPS binding substance to occur at an intracellular level if the LPS containing transfection complexes are released into the cytoplasm. Therefore, the excess antibiotic in the medium might be necessary to ensure sufficient intracellular concentrations of LPS binding substances to achieve neutralisation. The minimum concentration necessary to block the toxicity can be determined by titration.

In one embodiment of the invention the LPS binding substance is polymyxin B.

In another embodiment the substance is polymyxin E
(Storm et al., 1977), a derivative of polymyxin B which differs from it only by one amino acid. Like polymyxin B, it can bind LPS, but unlike polymyxin B it is not capable of blocking the activation of protein kinase C.

Other examples which are known to bind and neutralise LPS are: HDLs and LDLs (Levine et al., 1993; Flegel et al., 1993; Roth et al., 1993), apolipoprotein A1 (Flegel et al., 1993), BPI protein ("Bactericidal/Permeability Increasing Protein") or fragments thereof (Dentener et 2183~4~

al., 1993; Elsbach and Weiss, 1993) which have the properties required for the purposes of the invention, Limulus proteins (Roth and Tobias, 1993), derivatives of polymyxin which have a reduced cell toxicity, synthetic peptides having an affinity for LPS (Rustici et al., 1993), antibodies against LPS or lipid A (Burd et al., 1992). In addition, suitable substances may be designed, for example, by identifying the LPS binding domain of larger molecules known to have an LPS binding property and using the peptide, optionally in modified form. Alternatively, analogously to the proposal made by Rustici et al., 1993 for polymyxin B, the structure of lower molecular molecules with a known LPS binding property may be taken as a starting point for preparing new molecules optimally designed to bind and detoxify LPS.

The LPS binding substances for treating the cells are simply used by preferably making them part of the transfection medium. Generally, the LPS binding substances are only required to be present during transfection as the toxic effect of LPS manifests itself primarily during transfection; consequently, when the medium is changed after transfection, there is no need to add this substance. However, the presence of the substance even beyond transfection may be useful for the growth of the cell if, for example, the cell type used is sensitive to LPS; in this case, the fresh medium applied to the cells after transfection or infection will also contain the LPS binding substance.

In another embodiment of the process according to the invention, the cells are pretreated, before application of the transfection medium, with the LPS binding substances in order to bind any existing LPS and remove it if necessary.

2:183fi i7 The treatment of cells with LPS binding substances may be advantageous even irrespective of their treatment with foreign material which is to be introduced into the cell, e.g. during the culturing of cells which are very sensitive to LPS in their growth.

According to another aspect, the present invention relates to compositions for treating higher eukaryotic cells.

In the case of cell culture applications, this composition is a medium which contains one or more LPS
binding substances in addition to the usual components.
The usual components include nutrients for the cells, buffer substances, etc. If the LPS binding substance is a component of a transfection medium, this also contains the transfection components, i.e. the foreign material which is to be introduced into the cell and the components which mediate its transfer into the cell (e.g. recombinant viruses, cationic lipids, liposomes, as well as complexes for receptor-mediated gene transfer, optionally in conjunction with endosomolytic agents which increase the efficiency of gene transfer;
complexes of this kind are described, inter alia, in

3). When preparing a transfection medium of this kind it is recommended to add the LPS binding substance before the transfection components.

The present invention may be applied both to the treatment of cells in cell culture systems and also to therapeutic uses in vivo or ex vivo; in the latter case the medium containing the LPS binding substance acts as a drug.

The invention is beneficial, inter alia, in therapeutic applications in which the presence of LPS causes problems. One example of this in the field of gene ~183~7 therapy is the therapeutic use of recombinant adenovirus vectors in order to administer intact CFTR ("Cystic Fibrosis Transmembrane Regulator") genes to patients with cystic fibrosis (CF) in whom these genes are mutated. Since the lungs of CF patients may contain large quantities of Pseudomonas-LPS (Pseudomonas infections are an accompanying phenomenon of the disease), one restriction to the use of this method, which is carried out either through the nasal epithelium or by instillation directly into the lungs, might consist in toxicity to the pulmonary epithelium caused by the joint entrance of recombinant adenovirus and LPS.
Within the scope of the present invention, experiments were carried out with epithelial cells of the respiratory tract, which show that polymyxin B is capable of blocking the sharp decline in gene expression caused by LPS. The addition of an LPS binding substance to the medium containing adenovirus can thus achieve significant improvements in the therapeutic treatment of pulmonary epithelial cells.

Another example is the application to endothelial cells which are known to be difficult to transfect; one possible explanation for this difficulty might be in the tendency of these cells to respond to tiny amounts of LPS (Arditi et al., 1993; Haziot et al., 1993b; Pugin et al., 1993). Whilst the secretion of different cellular factors as a reaction of one cell to LPS influences other cells as well which have not themselves come into direct contact with LPS. Within the scope of the present invention, transfections of umbilical vein endothelial cells have shown that even very tiny amounts of LPS contAminAnts will produce toxicity; as LPS-cleansed DNA was used, these contAminAntS originated exclusively from tissue culture reagents. Polymyxin was able to neutralise even this toxicity caused by small amounts of LPS.

~i836g~7 The present invention may also be advantageous in the transfection of patient cells ex vivo when the cell population is contaminated with small amounts of LPS.
One example of such an application is the preparation of cancer vaccines in which tumour cells are taken from the patient, transformed ex vivo with a DNA coding for an immunostimulant polypeptide and given to the patient as a vaccination.

In therapeutic applications, the LPS-binding substance may be present as a component of the composition which acts as the transfection or infection medium and is thus applied to the cells together with the transfection components. It may also be present as an active component of a drug preparation which is added to the transfection medium before transfection or administered separately from the transfection composition, e.g.
before transfection. In its simplest form, a preparation of this kind is a solution of the LPS
binding substance, and the preparation may also contain conventional additives; methods of formulating pharmaceutical preparations are known to those skilled in the art. They may be found in the relevant textbooks, e.g. Remington's Pharmaceutical Sciences, 1980.

Thus, according to another aspect, the invention relates to pharmaceutical compositions containing an LPS binding substance for use in therapeutic treatment in which foreign material is introduced into the cell. The foreign material is preferably a nucleic acid, especially DNA.

According to another aspect, the present invention relates to the use of LPS binding substances for producing pharmaceutical compositions for use before and/or simultaneously with and/or after the treatment of 2183fi~'7 the human or animal body by transfection or infection with DNA or RNA.

These therapeutic treatments preferably include, in addition to viral gene transfer methods, gene therapy procedures as described in the summarising article by Cotten and Wagner 1993, including those methods in which inhibitory nucleic acid molecules such as antisense RNAs, ribozymes or DNA molecules coding for them are used specifically to inhibit cell function.

In another aspect the invention relates to a process for treating DNA for incorporation in human or animal cells, in which the DNA is cleansed of LPS.

Summary of Figures ig. 1: Influence of the endotoxin content of the DNA
on the expression of IL-2 in human melanoma cells Fig. 2: Reduction in the toxicity of LPS during the transfection of primary human fibroblasts by means of polymyxin B
Fig. 3: Mode of activity of polymyxin B at concentrations of 30 ~g/ml to 0.03 ~g/ml Fig. 4: Effect of polymyxin B during and after transfection Fig. 5: Correlation between the release of lactate dehydrogenase and transfection efficiency in the presence of polymyxin B
Fig. 6: Blocking the toxicity caused by natural LPS
contamination of plasmid DNA, using polymyxin B

ig. 7: Blocking the LPS toxicity in primary human epithelial cells of the respiratory tract, using polymyxin B

2 1~3647 ig. 8: Blocking the toxicity caused by LPS from tissue culture reagents, by means of polymyxin B in umbilical vein endothelial cells Fig. 9: Blocking the LPS-induced toxicity using polymyxin E
Fig. 10: Effect of polymyxin on the expression of DNA
when applying various methods of gene transfer In the Examples which follow, which illustrate the present invention, the following materials and methods were used unless otherwise specified:

a) Plasmid constructs i) pCMVL

The construction of the plasmid is described in WO 93/07283.

ii) pGShIL-2tet To obtain the plasmid which contains the sequence coding for human IL-2, the vector pWS2 was used as starting material: the plasmid pH~APr-1 (Gunning et al., 1987) was cut with BamHI and EcoRI. By agarose gel purification, a 2.5 kb fragment was isolated which contains the ampicillin resistance gene and the replication origin of pBR322 and the SV40 polyadenylation signal. This fragment was ligated with the CMV promoter/enhancer which had been amplified as a 0.7 kb PCR fragment from vector pAD-CMVI (described in EP-A 393 438) and digested with EcoRI/BamHI. The resulting plasmid was called pWS. The cDNA coding for human IL-2 was obtained as a PCR fragment from human pIL2-50A (Taniguchi et al., 1983) which contains the cDNA coding for human IL-2. The PCR fragment was ligated into the vector pWS opened by SalI/BamHI

21836~7 digestion and in this way pWS2 was obtained.

An IL-2 cassette containing the CMV enhancer/promoter, the sequence coding for IL-2 and the SV40-polyA
sequence, was obtained by PCR on the basis of pWS2. The PCR product was subjected to restriction enzyme digestion with EcoRI and cloned into the EcoRI/SmaI site of the plasmid pUC19 (Pharmacia). The resulting plasmid was known as pGShIL-2. The plasmid pBR327 (Soberon et al., 1980) which had been digested with SspI and AvaI
was used as the source for the tetracyclin resistance gene and parts of the "upstream" region of the ~-lactamase gene (ampicillin resistance gene). Together with an EcoRI/AvaI adaptor, the isolated tet sequence was cloned into the EcoRI/SspI site of pGShIL-2. The IL-2 cassette of the resulting clone pGShIL-2tet/amp was sequenced; then the amp sequence was excised with EamllO5I and SspI and the plasmid was religated. The resultant plasmid was termed pGShIL-2tet.

b) DNA preparation The plasmids were first cultured in bacterial strain E. coli DH5~ in the presence of 100 ~g/ml ampicillin (pCMVL) or tetracyclin (pGShIL-2tet) in LB medium. The overnight culture was centrifuged and from it the DNA
was prepared as follows: the CsCl density gradient centrifugation was carried out using the method described by Cotten et al., 1993. In order to do this the bacterial deposit from 1 litre of culture was incubated in 10 ml of 20~ (weight/volume) sucrose, 10 mM
EDTA, 50 mM Tris, pH 7.5 (solution 1) on ice for 10 minutes. Then 2.2 ml of lysozyme (10 mg/ml in solution 1) were added for a further 10 minutes on ice, then 5 ml of 0.2 M EDTA, pH 7, were added, the sample was incubated on ice for 10 minutes and finally 10 ml of 2~
(volume/volume) Triton X-114, 60 mM EDTA and 40 mM Tris, 3~47 pH 7.5, were added followed by 15 minutes incubation on ice. This lysate was then centrifuged for 30 minutes (Sorvall SS34, 17K) and 28.5 g CsCl and 400 ~l of ethidium bromide (10 ~g/ml) were added to the supernatant (26 ml initial volume). This material was centrifuged for 18 hours in a Beckman VTi50 rotor at 49,000 rpm at 20C. The lower one of the two ethidium-rich bands was collected and centrifuged again, directly in a Beckman VTi65 rotor, for 4 hours at 63,000 rpm and at 20C. The ethidium-rich band was harvested again, extracted with CsCl-saturated isopropanol until the pink colour had disappeared, exhaustively dialysed against TE
(10 mM Tris, 0.1 mM EDTA, pH 7.4), mixed with 1/10 volume 3 M sodium acetate, pH 5, and precipitated with 3 volumes of ethanol at -20C. The DNA precipitate obtained was further treated with RNase A, proteinase K, phenol/chloroform and chloroform, precipitated once more and the final DNA pellet was taken up in TE and quantitatively measured by optical absorption, working on the assumption that 0.05 mg/ml DNA has an absorption of 1 at 260 nm.

c) DNA purification of LPS

i) Triton X-114 extraction In order to obtain a homogeneous preparation of the detergent, Triton X-114 (Sigma) was subjected to three 0C/30C temperature cycles as described by Bordier, 1981. The extraction of the lipopolysaccharides from the DNA sample was carried out, in a modified version of published methods (Aida and Pabst, 1990; Manthorpe et al., 1993) as follows: the DNA sample (0.5 - 1.5 mg/ml in 10 mM Tris, 0.1 mM EDTA, pH 7.4 (TE)) was applied to 0.3 M sodium acetate (pH 7.5). Then 3 ~1 of Triton X-114 were added per 100 ~1 of DNA solution, the samples were vigorously mixed in a vortex and incubated on ice ~1836~7 for 10 minutes. To allow the two phases to separate, the samples were stored for 5 minutes at 30C, centrifuged in a preheated Eppendorf centrifuge at 2,000 rpm for 2 minutes and the aqueous phase was placed in a fresh Eppendorf test tube. This extraction was carried out twice more and the aqueous phase finally obtained was precipitated with 0.6 volumes of isopropanol at ambient temperature, the precipitate was obtained by centrifuging, washed twice with 80~ ethanol, air-dried, taken up in TE again and the quantity was measured. In order to do this, the sample was treated with RNase A, proteinase K, phenol/chloroform and chloroform, re-precipitated and the final DNA pellet was suspended in TE and the absorption at 260 nm was determined, working on the assumption that a concentration of 0.05 mg/ml of DNA has an absorption value of 1. (This method was used for the DNA used in the Examples).

ii) Polymyxin chromatography One volume of polymyxin resin slurry (Affi-Prep-Polymyxin, Biorad) corresponding to the volume of the DNA sample was briefly mixed with three volumes of 0.1 N
NaOH, then washed three times with five resin volumes of TE. The pelleted resin was taken up again with the DNA
samples (in TE 0.8 - 1.2 mg/ml) and the mixture was stirred overnight at 4C. Then the sample was placed on a disposable column pretreated with 0.1 NaOH, and washed with TE. The eluate was collected, the resin was washed with another volume of TE and the eluate was combined with the washing liquid. The DNA of this pooled sample was precipitated with 1/10 volume of 3 M sodium acetate, pH 5, and 2 volumes of ethanol. Further treatment of the precipitate and DNA measurement were carried out as described above. (This method was used in the preliminary test.) ~1836'17 d) LPS preparation A commercially available LPS preparation from Escherichia coli was used (0111:B4, Sigma). The preparation was dissolved in LPS-free water at the rate of 10 mg/ml and before the preparation of serial dilutions in LPS-free water it was sonicated for 5 minutes (SONOREX bath, 360 W). The final dilutions were sonicated for 5 minutes before use.

The LPS assays were carried out using the BioWhittaker assay (chromogenic Limulus assay; Iwanaga, 1993), and it was found that all the reagents used were LPS-free (<0.1 endotoxin units/50 ~l of solutlon).

e) Adenovirus preparation The E4-deficient adenovirus 5, dll014 (Bridge and Ketner, 1989) was cultured in the complementary cell line W162 (Weinberg and Ketner, 1983). Pellets of infected cells were suspended in amounts of 2 ml/2 x 107 cells in 20 mM HEPES, pH 7.4, 1 mM PMSF (phenylmethyl-sulphonyl fluoride) and subjected to three freeze/thaw cycles (liquid nitrogen, 37C). The suspension was then mixed with an equal volume of freon in the vortex and centrifuged for 10 minutes at 3,000 rpm (Heraeus Sepatech, 2705 Rotor). The aqueous (upper) phase was removed and the freon phase was treated with 1/5 volume 20 mM HEPES, pH 7.4 in the vortex and centrifuged again.
The aqueous phases were combined, transferred into a Beckman VTi50 centrifugal test tube (15 ml/test tube) and underlayered with 15 ml of 1.2 g/cm3 CsCl/20 mM
HEPES, pH 7.4 and 7 ml of 1.45 g/cm3 CsCl, 20 mM HEPES
pH 7.4. The samples were centrifuged for 40 minutes at 20C in a Beckman VTi50 rotor at 49,000 rpm. The lowèr opalescent band of mature virus particles at 1.34 to 1.35 g/cm3 (measured by means of the refractive index) 2183~47 and the upper band (immature particles at 1.31 to 1.32 g/cm3) were collected separately and subjected to equilibrium centrifugation (more than 4 hours) at 63,000 rpm in a VTi65 rotor. The opalescent virus bands (either 1.31 g/cm3 immature or 1.34 g/cm3 mature) were harvested. Biotinylation of the resulting virus particles with N-hydroxysuccinimide biotin (Pierce), inactivation with 8-methoxypsoralene/WA and purification by gel filtration using a Pharmacia PD10 column, equilibrated with HBS/40~ glycerol, were carried out as described in WO 93/07283 or by Wagner et al., 1992 and Cotten et al., 1992. The virus samples were quantitatively determined by means of the protein concentration (Biorad Bradford Assay using BSA as standard) and analysed, using the equation 1 mg/ml of protein = 3.4 x 1012 adenovirus particles/ml (Lemay et al., 1980).

f) Transfection complexes The modified adenovirus particles (8 ~l, 1 x 10l2 particles/ml) were diluted in 150 ~l of HBS and mixed at room temperature with 1 ~g of StrpL (streptavidin-modifed polylysine, prepared as described in WO 93/07283) in 150 ~1 of HBS for 30 minutes. Aliquots of 6 ~g of plasmid DNA were mixed with increasing amounts of LPS in 100 ~l. The DNA solutions were then mixed for 30 minutes with the adenovirus/StrpL solution at room temperature. Finally, a 100 ~l aliquot of HBS
containing 5 ~g of TfpL, prepared as described in WO 93/07283, was added to each sample, followed by 30 minutes at ambient temperature. Of the 500 ~l of transfection medium thus obtained, 1/10 was used in each well of a cell culture plate containing 20,000 cells.

~183~4~

g) Cell cultures i) Human fibroblasts After surgical removal, skin biopsies were placed in 4C
DMEM, containing 10~ FCS, 2 mM glutamine and gentamycin.
The biopsies were finely comminuted in a tissue culture device with forceps and a surgical blade in a l~m' n~r air current in sterile 6 cm plastic dishes. Then 3 ml of DMEM containing 20~ FCS, 2 mM glutamine and antibiotics were added and the culture was placed in a 37C incubator. After 10 days the medium was replaced by DMEM containing 10~ FCS. Then the medium was changed twice a week. 4 weeks after the beginning of the culture the cells which had grown out of the tissue fragments were trypsinised and plated out into new culture dishes for transfection. Primary cultures from passage 5-10 were used. The fibroblasts were transfected as described by Wagner et al., 1992.

ii) Human melanoma cells Primary human melanoma cells were isolated and cultivated in RPMI 1640 medium (Gibco/BRL) supplemented with 100 I.U./ml of penicillin, 100 ~g/ml of streptomycin, 2 mM L-glutamine, 1~ sodium pyruvate and 10~ heat inactivated FCS.

iii) Human respiratory tract epithelial cells Human respiratory tract epithelial cells were isolated from nasal polyps as described by Van Scott et al., 1986. The cells were cultured on standard cell culture dishes coated with human placental collagen (Sigma, Catalogue No. C 7521) in bronchial epithelial cell growth medium (BEGM, Promocell, Catalogue No. C-2106).

2183fi~ '7 iv) Human umbilical vein endothelial cells Human umbilical vein epithelial cells (HWECs) were obtained from cell systems (Kirklandt, Washington) and cultured on 0.1~ gelatine-coated cell culture dishes.

h) Measurement of expression i) Luciferase assay The preparation of cell extracts, the standardisation of the protein content and determination of the luciferase activity were carried out as described by Zenke et al., 1990 and Cotten et al., 1990 and in EP 388 758.

ii) IL-2 assay The expression of interleukin-2 was determined using a bioassay as described by Karasuyama and Melchers, 1988.
In addition, the IL-2 production was carried out using the IL-2 ELISA kit made by Becton Dickinson (Catalogue No. 30032) in accordance with the manufacturer's instructions.

Example 1 Influence of the endotoxin content of DNA on the expression of IL-2 in primary human melanoma cells Primary human melanoma cells (2 x 105 cells/6 cm culture dish) were transfected with 6 ~g of the plasmid purified by various methods, according to the data in Fig. 1.
The endotoxin content of the plasmid preparation before purification is shown in the drawing as a dark shaded bar. After purification by means of polymyxin resin or extraction with Triton X-114 all the preparations were ~I 83~ 7 given less than 0.1 EU lipopolysaccharide/6 ~g of DNA.
The IL-2 content was measured by ELISA in the cell supernatant; the values shown in Fig. 1 indicate units/106 cells and 24 hours.

Example 2 Reduction in the toxicity of LPS during transfection of primary human fibroblasts by means of polymyxin B

First of all the toxicity of LPS during gene transfer by means of adenovirus-aided receptor-mediated endocytosis into primary human fibroblasts was investigated. The content of LPS, based on 6 ~g of DNA, is shown in Fig. 2. 24 hours after transfection the cells were harvested and the luciferase measurement was carried out. As the LPS content of the DNA increased, the gene expression fell by nearly 2 powers of ten. The disrupted morphology of the cells which had been exposed to a high LPS content accorded with the low expression values (Fig. 2, upper Table).

Next, the cells were subjected to the same DNA/adenovirus/LPS complexes in the presence of increasing amounts of polymyxin B (polymyxin B sulphate, Sigma, Catalogue No. P 4932). Polymyxin was present in the medium both during and after transfection in the concentrations specified, i.e. the fresh medium changed 3 hours after transfection contained the specified concentration of polymyxin. A typical antibiotic concentration of polymyxin B is 1,000 units/ml. At a specific activity of 7,500 units/mg this corresponds to a concentration of 133 ~g/ml of polymyxin B. It was found that there were virtually no differences in expression in the presence of 3, 10 and 30 ~g/ml of polymyxin B. This shows that the antibiotic is tolerated by these cells. In all three concentrations 2183~7 used, polymyxin B allows successful transfections with virus/DNA complexes containing LPS contamination, with complete protection being achieved against 100 ng LPS/6 ~g DNA (cf. samples 2, 5, 8 and 11) and with a virtually total protective effect against 1,000 ng LPS/6 ~g DNA (cf. samples 3, 6, 9 and 12). (The values shown in the Figure are the averages of two transfections.) Example 3 Analysis of the method of activity of polymyxin at concentrations of 30 ~g/ml to 0.03 ~g/ml The minimum concentration of polymyxin necessary to neutralise LPS toxicity was determined by titrations similar to those in Example 1. It was found that total neutralisation of a content of 100 or 1,000 ng of LPS/6 ~g of DNA in the DNA complexes is achieved with 10 ~g/ml of polymyxin B. With 3 to 0.3 ~g/ml of polymyxin B only partial neutralisation of the toxicity was obtained and at lower concentrations only a very slight neutralisation was observed. A content of 0.6 ~g of DNA in the transfection mixtures shows that DNA
complexes containing LPS in concentrations of 100 or 1,000 ng/6 ~g DNA contain 10 to 100 ng of LPS in a volume of 0.25 ml. The neutralisation brought about by polymyxin began to deteriorate at 1 ~g/ml and below.
This corresponds to a requirement for 250 ng of polymyxin to neutralise 10 to 100 ng of LPS. The excess antibiotic in the medium would appear to be necessary in order to ensure adequate intracellular polymyxin concentrations for neutralisation. The results of the titration are shown in Fig. 3: where shown in the Fig., polymyxin B was present, both during and after transfection. In each case, double transfections were carried out. The luciferase values (measured 24 hours ~J836~

after transfection) are expressed for each polymyxin concentration as a percentage of the value of the control sample.

Example 4 Investigating the effect of polymyxin B during and after transfection In the preceding Examples, polymyxin was present in the culture medium both during transfection and afterwards, until the cells were harvested for the luciferase assay.
Since it was assumed on the basis of preliminary tests that the event responsible for the toxicity of LPS is its entry into the cell together with the adenovirus, the protective function of polymyxin B should only be necessary during the period of contact of the cells with the transfection complexes. To confirm this assumption, primary fibroblast cultures were exposed to the LPS-containing transfection complexes in the absence of polymyxin (samples 1-4) and in the presence of polymyxin B, on the one hand, only during transfection (in order to do this, after transfection the samples were mixed with fresh medium free from polymyxin B; samples 9-12) and, on the other hand, also after incubation with the transfection complexes (in order to do this, the samples were mixed with fresh medium which contained identical quantities of polymyxin B to the transfection medium;
samples 5-8). The luciferase expression obtained was measured 24 hours after transfection. The results of these experiments are shown in Fig. 4: it was found that the presence of 30 ~g/ml of polymyxin B solely during transfection has virtually the same protective effect as the continued presence of the antibiotic beyond the time after transfection. (The values shown in the Figure are the averages of two transfections.) ~ 1 83~ 7 Example 5 Correlation between the release of lactate dehydrogenase and transfection efficiency in the presence of polymyxin B

In the tests carried out in Example 1 it was established that the morphology of cells which had been transfected with DNA complexes containing LPS accorded with the cell toxicity which causes a reduction in gene expression.
Since the cytoplasmic enzyme lactate dehydrogenase (LDH) is released into the surrounding medium when cells undergo necrosis or apoptosis, simply measuring the LDH
activity in the cell medium is an indicator of cell toxicity. An inverse correlation has been found between the release of LDH into the medium and the successful transfer of the luciferase gene (Fig. 5, samples 1-4), whilst increases in LDH in the medium coincided with an appreciable drop in luciferase gene expression. In accordance with the results of Examples 1 and 2, the presence of 10 or 30 ~g/ml of polymyxin B (polymyxin B
was present in the medium only during transfection) blocked the LPS-induced reduction in gene transfer (Fig. 5, samples 5-12). In accordance with the blocking of the toxicity by polymyxin, the presence of polymyxin blocks the LPS-induced release of LDH (Fig. 5, samples 5-12). (The values shown in the Figure are the averages of two transfections.) The LDH assay was carried out by removing aliquots of the cell culture medium (5 to 50 ~l) at the specified times after transfection and mixing them with 500 ~l of LD-L reagent (Sigma, Catalogue No. 228-10). The samples were incubated at 37C for 45 minutes and the absorption at 340 nm was measured by comparison with an LD-L blind value.

~1~.3~4 ~

Example 6 Blocking of the toxicity caused by natural LPS
contamination of plasmid DNA by means of polymyxin B

In the preceding Examples, LPS-free DNA was used to which known quantities of defined LPS preparations had deliberately been added. In this Example, an investigation was carried out to find out whether the toxicity of LPS contamination as typically found in DNA
preparations can be blocked by polymyxin B. For this purpose, LPS-free luciferase plasmid was mixed with an excess of DNA plasmid which had been obtained by the Qiagen method with a plasmid DNA resin (Diagen) without treatment to remove LPS. This DNA thus contains the molecular form of LPS which is typically found in a plasmid DNA preparation. The DNA sample used contained about 16 ng LPS/6 ~g DNA; this corresponds to 6.4 ng/ml during transfection. As can be seen from Fig. 6, only weak DNA transfer was obtained with this preparation (Fig. 6, sample 1), but an almost 10-fold increase was achieved when 3 ~g/ml of polymyxin B was present during the transfection (Fig. 6, sample 2). At higher concentrations of antibiotic, no further improvement was observed. The 3 ~g/ml of polymyxin represent an almost 500-fold excess by mass over the LPS concentration obtained during transfection. (The values shown in the Figure are the averages from two transfections.) Example 7 Blocking the LPS toxicity in primary human respiratory tract epithelial cells using polymyxin B

Primary human respiratory tract epithelial cells were treated with LPS-containing transfection complexes as described in the preceding Examples, on the one hand in T~ 1 ~ 3 5 ~ 7 the absence of polymyxin B and on the other hand in the presence of 10 ~g/ml of polymyxin B during transfection.
48 hours after transfection the cells were harvested and the luciferase expression was measured. It was found that the presence of LPS in the DNA complexes causes a sharp drop in gene expression (Fig. 7, samples 1-4).
The toxicity caused by the LPS concentrations used can largely be blocked by means of polymyxin B (Fig. 7, samples 5-8). (The values shown in the Figure are the averages from three transfections.) Example 8 Blocking the toxicity caused by LPS from tissue culture reagents by means of polymyxin B in umbilical vein endothelial cells Human umbilical vein endothelial cells were transfected using the standard procedure. It was found that these cells are sensitive to the presence of LPS in the complexes; cf. Fig. 8, sample 1, in which the complex contains no LPS, with samples 2, 3 and 4 in which LPS
was present in amounts of 10, 100 and 1,000 ng/6 ~g of DNA. However, the transfection efficiency was low even when LPS-free DNA was used. In order to establish that this is caused by a toxicity which originates from the tissue culture reagents, the quantities of polymyxin B
specified in the Figure were used, which were present in the medium both during and after transfection. It was found that the presence of 10 ~g/ml of polymyxin B
causes a 5-6-fold increase in gene expression (cf.
samples 1 and 5). Higher concentrations of polymyxin B
were less effective.

218~

Example 9 Blocking the LPS-induced toxicity using polymyxin E

The tests carried out in this Example were performed chiefly as described in Example 2, except that polymyxin E (colistin-methanesulphonate, Sigma, Catalogue No. C 1511) was used instead of polymyxin B and it was present only during the transfection. The results of these experiments are shown in Fig. 9. It was found that polymyxin E has the ability to block LPS in an amount of 100 ng/6 ~g of DNA in a similar manner to polymyxin B. However, unlike polymyxin B, polymyxin E
was unable to neutralise LPS contamination of 1,000 ng/6 ~g DNA (the apparent stimulation at 30 ~g polymyxin E/ml (samples 4-6) can be put down to a significant inhibition of the control sample (30 ~g polymyxin E/ml, no LPS) rather than to stimulation of the sample which contained 100 ng LPS/6 ~g DNA).

Example 10 Effect of polymyxin on the expression of DNA when using different gene transfer methods The experiments carried out in this Example serve as a comparison in order to establish whether the toxicity observed during gene transfer by means of adenovirus-aided receptor-mediated endocytosis can be put down to components of the transfection complexes.

a) Gene transfer with recombinant adenovirus These experiments were intended to show whether the toxicity which occurs in the presence of adenovirus and LPS can be put down to polylysine. For this purpose, a recombinant adenovirus which carries a genomic ~-~1~3~

galactosidase gene (Stratford-Perricaudet et al., 1992) was used as the gene transfer vehicle. The adenovirus was applied to the cells (2 x lOq per well) in the presence of LPS. After 24 hours the ~-galactosidase expression was determined as a measurement of the survival of cells which had taken up the adenovirus particles. These experiments are shown in Fig. 10: an LPS-dosage dependent reduction in ~-galactosidase activity was found (Fig. lOA). Measurement of the release of LDH (cf. Example 5) into the medium showed that cell death and lysis could be responsible for the reduction in ~-galactoside expression. It is thus found that the presence of polylysine is not necessary in order for the toxic reaction to occur, and this supports the simple statement that toxicity is a consequence of the internalisation of LPS. The toxic effect caused by LPS was able to be counteracted by the addition of polymyxin (Fig. 10, samples 5-8).

b) Gene transfer by means of non-viral systems Next, investigations were carried out to discover whether adenovirus might play a part in signal transmission after the entry of LPS. If interactions between adenovirus and the cell play a part in toxicity, the transfer of DNA into the cells by non-viral methods should not involve any LPS-induced toxicity. For this purpose, glycerol on the one hand and cationic lipids on the other hand were used as non-viral gene transfer methods.

i) During the transfection of the fibroblasts (0.4 ~g pL
or StrpL in 75 ~l HBS were incubated with 3 ~g of pCMVLuc in 75 ~l of HBS for 30 minutes, then 3 ~g of TfpL or 2 ~g of pL in 75 ~l of HBS were added and incubation was continued for a further 30 minutes.
After the addition of the medium and 13~ glycerol to the 21~3~ll7 complexes, the transfection medium was applied to the cells for 4 hours) as a result of the addition of LPS in various concentrations (see Fig. 10B, samples 5-8) a reduction in luciferase expression was observed which was almost as great as in the method based on adenovirus-aided endocytosis.

ii) During transfection with the cationic lipid Transfectam (DOGS; a commercially obtainable preparation was used according to standard procedure) a reduction in the luciferase activity was observed as a function of the LPS content (Fig. 10B, samples 9-11). The presence of 1000 ng LPS/6 ~g DNA reduced the luciferase expression to about 10~ of the control value obtained with LPS-free DNA.

~1~36~7 Bibliography Aida, Y. and Pabst, M.J., 1990, J. Immunol. Methods 132, 191-195.
Arditi, M., Zhou, J., Dorio, R., Rong, G.W., Goyert, S.M. and Kim, K.S., 1993, Infection and Immunity 61, 3149-3156.
Bordier, C., 1981, J. Biol. Chem. 256, 1604-1607.
Bridge, E. and Ketner, G., 1989, J. Virol. 63, 631-638.
Burd, R.S., Cody, C.S. and Dunn, D.L., 1992, Immunotherapy of gram-negative bacterial sepsis.
R.G. Landes Company, Austin, Texas, USA.
Cordle, S.R., Donald, R., Read, M.A. and Hawiger, J., 1993, J. Biol. Chem. 268, 11803-11810.
Cotten, M., Laengle-Rouault, F., Kirlappos, H., Wagner, E., Mechtler, K., Zenke, M., Beug, H. and Birnstiel., M.L., 1990, Proc.Natl.Acad.Sci. USA 87, 4033-4037.
Cotten, M., Wagner, E., Zatloukal, K., Phillips, S., Curiel, D. and Birnstiel, M.L., 1992, Proc.Natl.Acad.Sci. USA 89, 6094-6098.
Cotten, M., Wagner, E., Zatloukal, K. and Birnstiel, M.L., 1993, J. Virol. 67, 3777-3785.
Cotten, M. and Wagner, E., 1993, Current Opinion in Biotechnology 4, 705-710.
Cristiano, R.J., Smith, L.C. and Woo, S.L.C., 1993a, Proc.Natl.Acad.Sci. USA 90, 2122-2126.
Cristiano, J.R., Smith, L.C., Kay, M.A., Brinkley, B.R.
and Woo, S.L.C., 1993b, Proc.Nat.Acad.Sci. USA 90, 11548-11552.
Curiel, D.T., Agarwal, S., Wagner, E. and Cotten, M., 1991, Proc.Natl.Acad.Sci. USA 88, 8850-8854.
Dentener, M.A., Von Asmuth, E., Francot, G., Marra, M.
and Buurman, W., 1993, J. Immunol. 151, 4258-4265.
Elsbach, P. and Weiss, J., 1993, Current Opinion in Immunology 5, 103-107.
Flegel, W.A., Baumstark, M.W., Weinstock, C., Berg, A.

218364~

and Northoff, H., 1993, Infection and Immunity 61, 5140-5146.
Golenbock, D.T., Liu, Y., Millham, F.H., Freeman, M.W.
and Zoeller, R.A., 1993, J. Biol. Chem. 268, 22055-22059.
Gunning, P., Leavit, J., Muscat, G., Ng, S.Y. and Kedes, L., 1987, Proc.Natl.Acad.Sci. USA 84, 4831-4835.
Haziot, A., Tsuberi, B. and Goyert, S.M., 1993a, J.
Immunology 150, 5556-5565.
Haziot, A., Rong, G.W., Silver, J. and Goyert, S.M., 1993b, J. Immunol. 151, 1500-1507.
Iwanaga, S., Curr. Opin. Immunol. 5, 74-82.
Karasuyama, H. and Melchers, F., 1988, Eur. J. Immunol.
18, 97-104.
Lemay, P., Boudin, M., Milleville, M. and Boulanger, P., 1980, Virology 101, 131-143.
Levine, D.M., Parker, T.S., Donnelly, T.M., Walsh, A.
and Rubin, A.L., 1993, Proc.Natl.Acad.Sci. USA 90, 12040-12044.
Lorenzo, A., Razzaboni, B., Weir, G.C. and Yankner, B.A., 1994, Nature 368, 756-760.
Lynn, W.A., Raetz, C.R.H., Qureshi, N. and Golenbock, D.T., 1991, J. Immunology 147, 3072-3079.
Lynn, W.A. and Golenbock, D.T., 1992, Immunology Today 13, 271-276.
Manthorpe, M., Cornefert-Jensen, F., Hartikka, J., Felgner, J., Rundell, A., Margalith, M. and Dwarki, V., 1993, Human Gene Therapy 4, 419-431.
Perera, P.Y., Manthey, C.L., Stutz, P.L., Hildebrandt, J. and Vogel, S.N., 1993, Infection and Immunity 61, 2015-2023.
Pugin, J., Ulevitch, R.J. and Tobias, P.S., 1993, J.
Exp. Med. 178, 2193-2200.
Roth, R.I. and Tobias, P.S., 1993, Infection and Immunity 61, 1033-1039.
Roth, R.I., Levin, F.C. and Levin, J., 1993, Infection and Immunity 61, 3209-3215.

~183~;~7 Rustici, A., Velucchi, M. Faggioni, R., Sironi, M., Ghezzi, P., Quataert, S., Green, B. and Porro, M., 1993, Science 259, 361-365.
Schindler, J. and Osborn, M.J., 1979, Biochem. 18, 4425-4430.
Soberon, X., Covarrubias, L. and Boliviar, F., 1980, Gene 9, 287.
Stratford-Perricaudet, L., Makeh, I., Perricaudet, M.
and Briand, P., 1992, J. Clin. Invest. 90, 626-630.
Storm, D.R., Rosenthal, K.S. and Swanson, P.E., 1977, Am. Rev. Biochem. 46, 723-763.
Taniguchi, T., Matsui, H., Fujita, T., Takaoka, C., Kashima, N., Yoshimoto, R. and Hamuro, J., 1983, Nature 302, 305-310.
Tobias, P.S., Mathison, J.C. and Ulevitch, R.J., 1988, J. Biol. Chem. 263, 13479-13481.
Tobias, P.S., Soldau, K. und Ulevitch, R.J., 1989, J.
Biol. Chem. 264, 10867-10871.
Van Scott, M.R., Yankaskas, J.R. and Boucher, R.C., 1986, Exp. Lung. Res. 11, 75-94.
Wagner, E., Zatloukal, K., Cotten, M., Kirlappos, H., Mechtler, K., Curiel, D. and Birnstiel, M.L., 1992, Proc.Natl.Acad.Sci. USA 89, 6099-6103.
Weinberg, D.H. and Ketner, G., 1983, Proc.Natl.Acad.
Sci. USA 80, 5383-5386.
Zenke, M., Steinlein, P., Wagner, E., Cotten, M., Beug, H. and Birnstiel, M.L., 1990, Proc.Natl.Acad.Sci.
USA 87, 3655-3659.
Ziegler-Heitbrock, H.W.L., and Ulevitch, R.J., 1993, Immunology Today 14, 121-125.
Remington's Pharmaceutical Sciences, 1980, Mack Publ.
Co., Easton, PA, Osol (ed.).

Claims (20)

Patent Claims

1. Process for treating eukaryotic cells, particularly for the incorporation of foreign material, characterised in that the cells are treated with a substance which binds lipopolysaccharide and blocks its toxicity to the cells and/or the foreign material to be introduced into the cell is one which has been purified to remove LPS
before being introduced into the cells.

2. Process according to claim 1, characterised in that the foreign material is a nucleic acid, particularly DNA.

3. Process according to claim 1 or 2, characterised in that the cells are treated with the lipopolysaccharide-binding substance while the foreign material is introduced into the cell.

4. Process according to claim 3, characterised in that the lipopolysaccharide-binding substance is applied to the cells as a component of a transfection or infection medium.

5. Process according to one of claims 1 to 4, characterised in that the lipopolysaccharide-binding substance is polymyxin B.

6. Process according to one of claims 1 to 4, characterised in that the lipopolysaccharide-binding substance is polymyxin E.

7. Process according to claim 2, characterised in that the DNA is treated by extraction with a detergent.

8. Process according to claim 7, characterised in that Triton X-114 is used as the detergent.

9. Process according to claim 2, characterised in that the DNA is treated with a polymyxin resin.

10. Process according to claim 9, characterised in that polymyxin B is used.

11. Composition for treating cells for the incorporation of foreign material, characterised in that it contains a lipopolysaccharide-binding substance.

12. Composition according to claim 11, characterised in that it also contains the foreign material to be incorporated in the cell as well as the components which mediate the uptake of the foreign material into cells.

13. Composition according to claim 11 or 12, characterised in that the foreign material is DNA.

14. Composition according to claims 12 and 13, characterised in that it contains, as components for the uptake of DNA into the cell, a conjugate of a cellular ligand and a substance binding to DNA as well as an endosomolytically active agent.

15. Composition according to claim 14, characterised in that the conjugate is transferrin-polylysine and the endosomolytically active agent is an adenovirus.

16. Pharmaceutical composition, characterised in that it contains a composition according to one of claims 11 to 15.

17. Cell culture medium, characterised in that it contains a composition according to one of claims 11 to 15.

18. Use of lipopolysaccharide-binding substances for preparing drugs for administration before and/or simultaneously with and/or after the treatment of human or animal cells by transfection or infection with a nucleic acid, particularly DNA.

19. Use according to claim 18, characterised in that the lipopolysaccharide-binding substance is polymyxin B.

20. Method of treating DNA for uptake into human or animal cells, characterised in that the DNA is purified to remove LPS.