Vol. 262 No. 10 fasue of April 5 pp. 4429-4432 1987 THE JOURNAL OF BIOLOGICAL CHEHrSTRY

Q 1987 by The kmerican Society of Biciogical Chemis&. Ine. Printed in U.S.A.

Receptor-mediated in Vitro Gene Transformation by a Soluble DNA Carrier System*

(Received for publication, December 22,1986) George Y. Wu$ and Catherine H. W u From the Division of Gastroenterobgy-Liver Disease, Department of Medicine, University of Connecticut School of Medicine, Farmington, Connecticut 06032

We present, here, evidence that foreign DNA can be specifically delivered to cells by a soluble carrier sys- tem that takes advantage of receptor-mediated endo- cytosis. Our experiments were based on the following concepts: 1) hepatocytes possess a unique receptor that binds and internalizes g a l a c ~ - t e r m i n a l (asialo-)gIy- coproteins; 2) DNA can bind to polycations in a strong but noncovaIent manner forming soluble complexes; and 3) the gene for chloramphenicol acetyltransferase, a bacterial enzyme that acetylates chloramphenicol, is not present in mammalian cells.

We coupled asialoorosomucoid (ASOR) to poly-L-ly- sine to form an asialoorosomucoid-poly-L-lysine con- jugate. The plasmid, pSV2 CAT, was complexed to the conjugate in a molar ratio of 1:2. To test this complex, a model system was used consisting of hepatoma cell lines, Hep 62, asiaIoglycopro~in receptor (+), and SK- Hep 1, receptor (-). Each cell line was incubated with filtered ASOR*poly-L-lysine*DNA complex, or con- trols consisting of DNA plus ASOR, DNA plus poly-L- lysine, or DNA alone. Cells were assayed for the pres- ence of chloramphenicol acetyltransferase activity as a measure of gene transformation. SK-Hep 1, receptor (-) cells, produced no detectable acetylated chloram- phenicol derivatives under any condition. However, Hep 62, receptor (+) cells, incubated with the ASOR*~Iy-L-Iysine*DNA complex were transformed as indicated by the presence of chloramphenicol ace- tyltransferase activity (0.028 chloramphenicol acetyl- transferase units/106 cells). Mixtures of individual components of the complex failed to transform these cells. Competition by a 10-fold excess of ASOR pre- vented gene transformation by the ASOR*poly-~-ly- sinemDNA complex.

Introduction of foreign genes into m a m m ~ i a n cells in vitro is a common procedure for studying gene regulation, The most popular technique employs a precipitation method in which DNA is coprecipitated with calcium phosphate to form insoluble particles (1). Some of these particles become inter-

* This work was supported, in part, by an American Gastroenter- ological Association/Industry Research Scholar Award (to G. Y. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “ ~ v e r t i s e ~ n ~ ” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed Division of Gas- troenterology-Liver Disease, Dept. of Medicine, University of Con- necticut Health Center, Rm. AM-044, 263 Farmington Ave., Far- mington, CT 06032.

nalized within host cells by phagocytosis (2). Following inter- nalization, a fraction of the DNA avoids digestion and even- tually enters the nucleus resulting in expression of new genes (1). Although the precipitation technique is widely applicable in vitro, its potential for delivery of genes in vivo is limited by a lack of cell specificity and especially by the ~ ~ o L ~ ~ nature of the precipitates. We wondered whether a simple, soluble, DNA carrier system could be developed to target DNA spe- cifically to hepatocytes using receptor-mediated endocytosis. We based our experiments on the following concepts. I) Hepatocytes possess unique receptors that bind and internal- ize galactose-terminal (asialo-)glycoproteins (3). Coupling of DNA to an asialoglycoprotein could permit internalization of DNA via asialoglycoprotein receptors. However, covalent COU- pling might alter the DNA and preclude proper gene expres- sion. 2) Alternatively, DNA can be bound by polycations, e.g. poly-L-lysine in a strong, electrostatic interaction forming soluble complexes (4). 3) Chloramphenicol acetyltransferase, a bacterial enzyme, catalyzes acetylation of the antibiotic, chloramphenicol (5 ) . The gene for this enzyme is not present in mammalian cells.

Our hypothesis was: If poly-L-lysine were coupled to an asialoglycoprotein, subsequent addition of DNA to form a soluble asialoglycoprotein.po1y-L-lysine . DNA complex could permit targeted delivery of DNA specifically to asialoglyco- protein receptor-bearing cells. Our strategy to test this system was to deliver DNA in the form of a bacterial plasmid, pSV2 CAT, which contains the CAT gene. Because mammalian cells lack this gene, the appearance of chloramphenicol ace- tyltransferase enzyme activity in target cells could be used as a convenient marker for gene transformation.

EXPERIMENTAL PROCEDURES

Celfs and Cell Culture-Hep G2 (from B. Knowles, Wistar Insti- tute, Philadelphia, PA) and SK-Hep l (from D. Shafritz, Albert Einstein College of Medicine, Bronx, NY) were maintained in plastic dishes con~in ing minimum Eagle’s medium plus 10% fetal calf serum at 37 “C and 5% COz (6).

As~~glycoprotein Uptake Activity of Hep G2 and SK-Hep 1 Cells- Orosomucoid was prepared from pooled human serum (American Red Cross, Farmington, CT) according to the method of Whitehead and Sammons (7) and desialylated to expose galactose residues using neuraminidase (8). The asialoorosomucoid (ASOR),’ thus formed, was determined to have no residual sialic acid (9). For studies on DNA. protein complex formation, ASOR was iodinated with carrier- free N a T (IO).

Specific uptake of [12JI]ASOR by Hep G2 and SK-Hep 1 cell l i e s was determined according to the method of Schwartz et aL (11). For Hep G2, uptake was 140 ngfhfmg cell protein: for SK-Hep 1, uptake was less than 10 ng/hfmg cell protein. On this basis, Hep G2 was designated receptor (+) and SK-Hep 1 was designated receptor (-).

Preparation of a Targetable DNA Carrier System-To prepare a targetable carrier system for DNA, [lZ5I]ASOR was coupled to poly- L-lysine (mean M, = 69,000, Sigma) in a 5:l molar ratio using N - succinimidyl 3-(2-pyridyldithio) propionate (Pierce Chemical Co.) according to the method of Jung et al. (12). The conjugate was separated from noncoupled [‘Z51]ASOR and poly-L-lysine on a Bio- Gel A-1.5m column (Bio-Rad) eluted with 0.01 M HEPES, 2 M guanidine HC1, pH 7.4. The conjugate peak was identified as that containing both lZ6I radioactivity and a lysine content in excess of that contributed by ASOR alone as determined by amino acid analysis after hydrolysis in 6 N HCI at 100 “C for 24 h. From the specific

’ The abbreviations used are: ASOR, asialoorosomucoid HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

4429

4430 Targeted Gene Delivery in Vitro

FIG. 1. The effect of increasing proportions of ASOR-poly-L-lysine conjugate on plasmid DNA electrophoretic migration through a 0.8% agarose gel. Panel A, increasing amounts of ASOR-poly-L-lysine conjugate in 2 M NaCl were added to DNA samples in 2 M NaCl to make each 87 nM with respect to DNA. After incubation a t 25 "C for 1 h, each sample was dialyzed and filtered as described under "Experimental Procedures." Samples were electrophoresed through an 0.8% agarose gel using a Tris-phosphate buffer system as described previously (15) and stained with ethidium bromide to visualize DNA. Lane I , DNA alone; lanes 2 through 9, DNA with progressively increasing proportions of conjugate; lane 10, saline; lane 11, ['251]ASOR-poly-~-lysine conjugate. Panel R, the agarose gel was dried and autoradiographed as described under "Experimental Procedures."

activity of ['2sII)ASOR and the lysine content of the conjugate, the molar ratio of ASOR to poly-L-lysine in the conjugate was calculated to be 5:l. The conjugate was found to be stable at -20 "C for a t least 4 months. A 10% sodium dodecyl sulfate-polyacrylamide gel of the conjugate after electrophoresis (13), stained with Coomassie Blue, revealed a single band which, in the presence of mercaptoethanol, completely resolved the conjugate into its asialoorosomucoid and poly-L-lysine components.

Plasmid Preparation-The plasmid pSV2 CAT (5) (from G. Car- michael, University of Connecticut, Farmington, CT) was grown in Escherichia coli, isolated and purified according to the method of Birnboim and Doly (14). Purity was confirmed by 0.8% agarose gel electrophoresis, demonstrating an absence of bacterial cellular DNA.

Conditions for Formation of a Targetable DNA Complex-To de- termine the proportion of ASOR-poly-L-lysine conjugate that should be mixed with plasmid DNA to optimize the DNA content of a soluble complex, increasing amounts of conjugate in 2 M NaCl were added dropwise to contain 87 nM DNA in 2 M NaCI. Samples were incubated for 1 hr at 25 "C and then dialyzed for 24 h against 0.15 M saline through membranes with an M , limit of 3500 (Spectrum Medical Industries, Los Angeles, CA). Although clear solutions were obtained after dialysis, all samples were filtered through 0.2-pm membranes (Millipore Corp., Bedford, MA) to ensure that complexes to be used did not contain precipitates. An aliquot of each sample was heat- denatured and the DNA content was measured fluorometrically (16). There was no significant loss of DNA due to the filtration step. Samples with equal concentrations of DNA (87 nM) were electropho- resed on an 0.8% agarose gel a t 50 V for 3 h as described previously (15) and stained with ethidium bromide to visualize DNA. T o detect the location of the [1251]ASOR-poly-~-lysine conjugate, the gel was dried and an autoradiogram was obtained using XAR film (Eastman) at -70 "C for 8 h.

Assay for Targeted DNA Deliuery-Hep G2 and SK-Hep 1 cells were grown separately to one-quarter confluence in 100-mm dishes and then switched to minimum Eagle's medium, 10% fetal calf serum, 5 mM Ca2' containing ASOR.poly-L-lysine.DNA complex (1.04 nM ASOR, 200 PM poly-L-lysine, 520 pM DNA) alone or complex plus

lysine plus 520 p~ DNA; or 520 PM DNA alone. Each sample was filtered through 0.2-pm membranes prior to incubation with cells. After incubation for 48 h a t 37 "C under 5% COP, medium was aspirated and the cells were washed with ice-cold phosphate-buffered saline scraped from dishes in that same solution and centrifuged a t 1,500 rpm for 5 min at 4 "C. The pellets were washed twice and resuspended in 1 ml of 0.2 M Tris, pH 7.5, sonicated for 1 min and then centrifuged a t 10,000 rpm for 10 min a t 4 "C. Aliquots of the supernatants were assayed for protein content (17) and chloramphen- icol acetyltransferase activity as described by Gorman et al. (5). In brief, aliquots of supernatant containing equal amounts of protein were incubated a t 25 "C with ['4C]chloramphenicol (Du Pont New England Nuclear) in 0.25 M Tris, pH 7.5, to which was added acetyl coenzyme A (Sigma). The samples were extracted with ice-cold ethyl

10.4 nM ASOR 1.04 nM ASOR PIUS 520 PM DNA; 200 PM poly-L-

FIG. 2. Assay fo r gene transformation in SK-Hep 1 cells. Cells were grown to one-quarter confluence and then incubated for 48 h a t 37 "C under 5% COz in the presence of filtered ASOR.poly- L . lysine complex or components of the complex. After harvesting and sonication, cells were assayed for chloramphenicol acetyltrans- ferase activity by thin layer chromatography as described under "Experimental Procedures" and originally by Gorman et al. (5). The effects of incubation with ASOR.poly-L-lysine complex (1.04 nM ASOR, 200 PM poly-L-lysine, 520 pM DNA) alone is shown in lane 2; complex plus 10.4 nM ASOR, lane I ; 520 pM DNA, lane 3; 200 PM poly-L-lysine plus 520 PM DNA, lane 4 ; and 1.04 nM ASOR plus 520 PM DNA, lane 5. ['*C]Chloramphenicol alone is shown in lane 6; chloramphenicol acetyltransferase enzyme standard (std. CAT) (0.001 units), lane 7; and chloramphenicol acetyltransferase (0.01 units), lane 8. CA, chloramphenicol; PL, poly-L-lysine.

acetate, spotted onto silica gel thin layer chromatography plates (E. Merck, Darmstadt, West Germany), and developed using a 95:5 mixture of chloroform and methanol (v/v) (5). After drying, the plates were autoradiographed using XAR film a t -70 "C for 4 days. Varying amounts of standard chloramphenicol acetyltransferase enzyme (P-L Biochemicals) incubated with ['4C]chloramphenicol were spotted onto each plate for quantitation of chloramphenicol acetyltransferase activity in cell samples by densitometric scanning of triplicate thin layer chromatographic assays of which Figs. 2 and 3 are representa- tive.

Targeted Gene Delivery in Vitro 4431

HeD G 2

@@ 3-Acetate

I - A c e t a t e

, A

7 8

FIG. 3 . Assay for gene transformation in Hep G 2 cells. Cells were grown to one-quarter confluence and then incubated for 48 h at 37 "C under 5% C 0 2 in the presence of filtered ASOR.poly-r,-lysine complex or components of the complex. After harvesting and soni- cation, cells were assayed for chloramphenicol acetyltransferase ac- tivity hy thin layer chromatography as described under "Experimental Procedures" and originally by Gorman et a/. (5). The effect of incu- bation with ASOR.polv-L-lysine complex (1.04 nM ASOR, 200 pM poly-I,-lysine, 520 PM DNA) alone is shown in /one 2; complex plus 10.4 nM ASOR, Ianr I : 520 pM DNA, lane 3; 200 pM poly-L-lysine plus 520 PM DNA, lane 4 ; 1.04 nM ASOR plus 520 PM DNA. lane 5. ["'C](:hloramphenicoI alone is shown in lanr 6; chloramphenicol acetyltransferase enzyme standard ( S t d . CAT) (0.05 units), lane 7; and chloramphenicol acetyltransferase (0.1 units), inn^ 8. CA, chlor- ampllenicol; PI,, poly-[.-lysine.

RESIJLTS AND DISCUSSION

Increasing t,he proport,ion of ['"51]ASOR-poly-~,-lysine con- jugate and it,s effect on electrophoretic migration of plasmid DNA in an agarose gel is shown in Fig. 1. Panel A shows this gel stained wi th ethidium bromide to visualize DNA. Panel I3 is an autoradiogram of' the same gel to detect [12511]ASOR- poly-I,-lysine conjugate. DNA alone is shown in lanes 1 and 1 '; and blanks (saline), lanes 10 and I O ' . Lane 1 reveals four DNA bands corresponding to the four forms of the plasmid DNA: super-coil, nicked circular, and dimers of each of these forms (18). Lanes I 1 and I I ' contain only ['"'IIASOR-poly- r,-lysine conjugate. LanP 11' shows t,hat the conjugate ap- peared as a single radioactive band that did not migrate from the top of the gel. As the proportion of ['"II]ASOR-~OI~-L- lvsine conjugate in the samples increased (evident in lanes 2' through 9 ' ) , there was a decrease in DNA that entered the gel as seen in lanes 2 through 9. There was a corresponding increase in DNA that remained at the top of the gel with the conjugate. With increasing proportions of conjugate in the sample, more DNA was retained by the ASOR-poly-L-lysine conjugate in the wells. A 2:1 molar ratio of the conjugate (based on ASOR content) to DNA optimized the complex formation and this molar ratio was used in all subsequent experiments.

The effect of ASOR.poly-r,-lysine. DNA complex on gene transformation of SK-Hep I. receptor (-) cells, is shown in

Fig. 2. I,anes 7 and 8 show that chloramphenicol acetyltrans- ferase enzyme activity can be detected by the presence of 1- and .?-acetylchloramphenicol derivatives. The effects of in- cubation of SK-Hep 1 cells with ASOR.poly-L-lysine.DNA complex is shown in lane 2; incubation with DNA alone, lane 3; poly-L-lysine plus DNA, lane 4 ; and ASOR plus DNA, lane 5. Control samples contained amounts of components identi- cal to those present in the sample containing the complete ASOR .poly-L-lysine. DNA complex. There was no detectable chloramphenicol acetyltransferase activity in SK-Hep 1 cells, indicating a lack of gene transformation in this cell line under any of the conditions.

However, Fig. 3, lane 2, shows that incubation of Hep G2, receptor (+) cells, with ASOR.poly-~-lysine. DNA complex, lane 2, did result in gene transformation as shown by the formation of [14C]acetylchoramphenicol derivatives, densito- metrically quantitated to be 0.028 * 0.003 chloramphenicol acetyltransferase units/1O6 cells. The transformation effi- ciency of the complex in Hep G2 was approximately double that of the calcium phosphate method in our hands (data not shown). Incubation of Hep G2 cells with controls is shown as follows: DNA alone, lane 3; DNA plus poly-L-lysine, lane 4 ; or DNA plus ASOR, lane 5 did not result in gene transfor- mation. Furthermore, competition by a 10-fold excess of ASOR, lane I , inhibited transformation and expression of chloramphenicol acetyltransferase in the cells indicating that recognition of the complex by Hep G2 is directed by the asialoglycoprotein component of the complex.

We conclude that asialoglycoprotein-poly-L-lysine conju- gates can be used to target genes in a soluble form resulting in specific delivery t,o hepatocytes and expression of foreign genes by these cells.

Previous studies have shown that the presence of polycat- ions in culture medium can increase i n vitro cellular uptake of substances, e.g. albumin (19) and nucleic acids (20, 21), in a manner dependent on the size (19) and concentration (19, 20) ofthe polycation. Polylysines of a size comparable to that used in our conjugate required concentrations approximately 2000 times greater than present in our complex. The lack of gene transformation in our control experiments using a com- bination of poly-L-lysine plus DNA in concentrations identi- cal to that provided by the complete complex indicate that enhancement of DNA uptake by the polycation alone was not responsible for the transforming effect of the ASOR.poly-L- Iysine'DNA complex. In addition, our finding of inhibition of transformat.ion of Hep G2 by a 10-fold excess of ASOR supports the notion that the recognition of the complex by Hep G2 cells is directed by the asialoglycoprotein component of the complex.

Asialoglycoproteins internalized by their receptor have been shown to be ultimabely delivered via membrane-limited vesi- cles to lysosomes with subsequent glycoprotein degradation (22). Because recognition of our carrier system is directed by an asialoglycoprotein, our complex would be expected to be subject to the same fate. However, the efficiency of the deg- radative pathway can be less than complete. For example, Chang and Kullberg (23) showed that a conjugate consisting of an asialoglycoprotein bound to diphtheria toxin A-chain was internalized by the asialoglycoprotein receptor and was lethal for targeted cells, indicating that escape from the nor- mal degradative pathway is possible for some proteins.

Similarly, expression of foreign DNA introduced by phag- ocytosis of calcium phosphate precipitates ( 5 ) or polycation complexes (20) also suggests that DNA under appropriate condit,ions can enter the cytoplasm avoiding nuclease diges- tion.

4432 Turgeted Gene Delivery in Vitro

Gene transformation of mammalian cells in vitro by a variety of techniques including delivery by viruses, liposomes, electroporation, microinjection, cell fusion, DEAE-dextran, and calcium phosphate precipitation has been achieved pre- viously and the methods have been reviewed (24). We have demonstrated here, for the first time, that targeted gene transformation can be accomplished using a soluble DNA carrier system based on a cell-specific receptor-mediated en- docytotic process. The use of this type of DNA complex can permit targeted gene delivery that, by virtue of its solubility and specificity, may be useful in the study of gene regulation in vivo.

A c k ~ ~ ~ ~ g ~ n t s - W e thank Dr. Gordon Carmichael for his kind and considerable help and advice in the preparation of the pSV2 CAT plasmid. The excellent assistance of Rosemary Pavlick in the prepa- ration of this manuscript is gratefully acknowledged.

REFERENCES 1. Graham, F. L., and Van der Eb, A. J. (1973) Virology 52, 456-

2. Loyter, A., Scangos, G. A., and Ruddle, F. H. (1982) Proc. Natl.

3. Ashwell, G., and Morell, A. G. (1977) Trends Biochem. Sci. 2:

4. Li, H. J., Chang-Wu, C., and Weiskopf, M. (1973) Biochem. J.

5. Gorman, C. M., Moffat, L. F., and Howard, B. H. (1982) MOL

467

Acod. Sci. U. S. A. 79,422-426

76-78

12,1763-1771

Cell. Bwl. 2 , 1044-1051

6.

7.

8.

9. 10.

11.

12.

13. 14.

15.

16.

17. 18.

19. 20.

21. 22.

23.

24.

Knowles, B. B., Howe, C. C., and Aden, D. P. (1980) Science

Whitehead, D. H., and Sammons, H. G. (1966) Biochim. Biophys.

Kawasaki, T., and Ashwell, G. (1977) J. Bioi. Chem. 252,6536-

Warren, L. (1959) J. Biol. Chem. 234,1971-1975 Greenwood, F. C., Hunter, W. M., and Glover, J. S. (1963)

Schwartz, A. L., Fridovich, S. E., Knowles, B. B., and Lodish, H.

Jung, G., Kohnlein, W., and Fiders, G. (1981) Biochem. Biophys.

Laemmli, U. K. (1970) Nature 277 , 680-685 Bimboim, H. C., and Doly, J. (1979) NrccCic Acids Res. 7,1513-

1518 Maniatis, T., Fritsch, E. F., and Sambrook, G. (1982) in ~ o ~ c u ~ r

Cloning, A Laboratory Manual, pp. 150-161, Cold Spring Har- bor Laboratory, CoId Spring Harbor, NY

Kapuscinski, J., and Skoczylas, B. (1977) Anal. Biochem. 83,

Bradford, M. A. (1976) Anal. Biochem. 72,248-254 Breitman, M. L., Tsui, L.-C., Buchwald, M., and Siminovitch, L.

Ryser, H. J.-P. (1967) Nature 215,934-936 Farber, F. E., Melnick, J. L., and Butel, J. S. (1975) Biochim

Amos, H. (1981) Biochem. Biophys. Res. Commun. 5 , l - 4 Wall, D. A., Wilson, G., and Hubbard, A. L. (1980) Cell 21,”

Chang, T.-M., and Kullberg, D. J. (1982) J. Biol. Chem. 257,

Kucherlapati, R., and Skoultchi, A. (1984) Crit. Rev. Biochem.

209,497-499

Acta 124,209-211

6543

Biochem. J. 114-119

F. (1981) J. Biol. Chem. 256,8878-8881

Res. Commun. 101,599-606

252-257

(1982) Mol. Cell. Biol. 2, 966-976

Biophys. Acta 390,298-311

93

12563-12572

16,349-379