Characterization of a Class of Cationic Peptides Able to Facilitate ...

12 downloads 0 Views 3MB Size Report
able to function as protein transduction domains, 12-mer peptide sequences from an ... When tested in vivo, the PTDs were able to facilitate efficient and rapid.
doi:10.1006/mthe.2000.0137, available online at http://www.idealibrary.com on IDEAL

ARTICLE

Characterization of a Class of Cationic Peptides Able to Facilitate Efficient Protein Transduction in Vitro and in Vivo Zhibao Mi, Jeffrey Mai, Xiaoli Lu, and Paul D. Robbins1 Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 Received for publication June 28, 2000, and accepted in revised form August 23, 2000

Protein transduction domains (PTDs), such as the third helix of the Drosophila Antennapedia homeobox gene (Antp) and the HIV TAT PTD, posses a characteristic positive charge on the basis of their enrichment for arginine and lysine residues. To determine whether cationic peptides are able to function as protein transduction domains, 12-mer peptide sequences from an M13 phage library were selected for synthesis on the basis of their varying cationic charge content. In addition, polylysine and polyarginine peptides were synthesized in order to assess the effect of charge contribution in protein transduction. Coupling of the biotinylated peptides to avidin–β-galactosidase facilitated transduction in a wide variety of cell lines and primary cells, including islet β-cells, synovial cells, polarized airway epithelial cells, dendritic cells, myoblasts, and tumor cells. Two of the peptides, PTD-4 and PTD-5, mediated transduction nearly 600-fold more efficiently than a random control peptide, but with an efficiency similar to the TAT PTD and the 12 mers of polylysine and polyarginine. Furthermore, confocal analysis of biotinylated peptide–streptavidin–Cy3 conjugates demonstrated that the internalized PTDs are found in both the nuclei and the cytoplasm of treated cells. When tested in vivo, the PTDs were able to facilitate efficient and rapid protein delivery into rabbit synovium and mouse solid tumors following intraarticular and intratumoral administration, respectively. These novel PTDs can be used to transfer therapeutic proteins and DNA for the treatment of a wide variety of diseases, including arthritis and cancer. Key Words: protein transduction domains; internalization; cationic peptides; β-galactosidase; arthritis; cancer.

INTRODUCTION There have been many different approaches developed for the introduction of DNA and proteins into target cells. Multiple delivery systems have been used extensively for gene transfer including viral vectors such as retroviruses, lentiviruses, adenoviruses, and adeno-associated virus and nonviral vectors such as liposomes, bioballistics and even naked DNA. However, the low efficiency of the different nonviral vector systems, as well as the immunogenicity and transient expression of many of the viral vectors (1–3), have hindered their applications in vivo. In addition, many viral vectors are unable to infect certain cell types due to low levels of expression of specific receptors or coreceptors. Clearly, the develop-

1To whom correspondence and reprint requests should be addressed at Department of Molecular Genetics and Biochemistry, W1246 Biomedical Science Tower, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261. Fax: (412) 383-8837. E-mail: [email protected].

MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy 1525-0016/00 $35.00

ment of methods for improving uptake of DNA by cells or the rate of viral infection would facilitate the application of gene therapy for the treatment of a wide variety of acquired and genetic diseases. Different types of lipid particles have been used for delivery of therapeutic protein (4–10), but with relatively poor efficiency. Alternatively, peptides that bind to specific cell surface molecules have been used for targeted delivery of proteins and DNA to cells in vivo. Particularly, targeting of toxic proteins to tumor cells has been achieved with peptides able to bind to specific integrins or other tumor-specific cell surface proteins. However, the efficiency of internalization after peptide binding is not always high (11–17), and, in certain cases, the endocytic pathway may not permit the release of the protein to the appropriate compartment. Such internalized peptide complexes may even traffic to the lysosome where they are proteolytically degraded. Consequently, although receptor–ligand interactions are appropriate for certain applications, they may not be applicable for

339

ARTICLE

1

2 FIG. 1. PTDs mediate internalization of a β-galactosidase marker protein into HIG-82 cells. PTD 1–6, a control peptide and the Ant and TAT PTDs, were coupled to a β-gal–avidin complex and the peptide–avidin–β-gal complexes added to the media of HIG-82 cell cultures at 37°C. Three hours post addition of the complexes, the cells were fixed and stained with X-gal. (A) 150 nM of PTDs 1–6 (1–6), Ant PTD (Ant), TAT PTD (TAT) and control peptide (Ran) avidin–β-gal complexes were added to HIG-82 cells. (B) Titration of PTDs 2–5, Ant PTD (Ant) and TAT PTD (TAT) avidin–β-gal complexes with the control peptide (Ran) and β-gal–avidin complex (β-gal) as controls using concentrations of the complexes ranging from 0.3 to 150 nM. FIG. 2. Inhibition of protein transduction by excess PTDs and Ant and TAT PTD. Nonbiotinylated peptide (100 µM) was added to HIG-82 cells at 37°C for 1 h, followed by addition of 1 µM of the biotinylated peptides conjugated to avidin–β-galactosidase. Three hours post addition of the complexes, the cells were fixed with 4% paraformaldehyde and stained with X-gal.

340

MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

ARTICLE

FIG. 3. PTD-5 facilitates protein transduction into a wide variety of cell types. PTD-5–avidin–β-gal complex (150 nM) was added to the cell culture media. Three hours following addition of the complexes, the indicated cells were fixed and stained with X-gal. (A) HIG-82 cells; (B) rabbit primary synovial cells; (C) human primary synovial cells; (D) primary human airway epithelial cells (HBE144); (E) polarized canine kidney cells (MDCK), (F) human islets; (G) murine myoblasts (C2C12); (H) murine fibrosarcoma tumor line (MCA205); (I) NIH/3T3 cells.

the efficient delivery of proteins or DNA to a wide variety of cells. Recently, protein transduction domains (PTDs) that are able to mediate receptor-independent delivery of protein complexes have been identified and characterized within several cellular and viral proteins. The first protein reported with transductional properties was the HIV transactivator protein, TAT (18, 19) in which the 11amino-acid protein transduction domain (PTD) was identified by virtue of its cationic content (20). Crosslinking of the TAT PTD to either β-galactosidase or horseradish peroxidase was able to confer efficient internalization of the marker proteins into cells. More recently,

chimeric, in-line fusion proteins containing the TAT PTD have been used to deliver proteins to a wide spectrum of cell types both in vitro and in vivo (21–25). The Drosophila Antennapedia homeodomain was also identified as enabling the homeoprotein to translocate across the plasma membrane and accumulate in the nuclei of cells in a receptor-independent manner. Its transduction domain, Antp or Ant PTD, was localized to a 16-amino-acid stretch from residues 43–58 comprising the third helix of the homeodomain (26–28). VP-22, a 38-kDa tegument protein from herpes simplex virus type 1 (HSV-1) also has been identified to possess the ability to transduce across cell membranes (29).

FIG. 4. PTDs-mediate both cytoplasmic and nuclear localization. PTD–streptavidin–Cy3 complexes were added to the media of HIG-82 cell cultures at 37°C. Three hours after addition of the complexes, the intracellular localization of the peptide–streptavidin–Cy3 complexes (red) was visualized by confocal microscopy. Nuclei were counterstained with Cytox green. (A) PTD-5 superimposed on differential interference contrast image, (B) PTD-5, (C) TAT PTD. MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

341

ARTICLE Since the peptide transduction domains previously identified are rich in positively charged amino acids, we have examined the ability of short, positively charged peptides from an M13 phage library to mediate efficient and rapid internalization of peptide-protein conjugates and peptide-protein chimeric fusions. We demonstrate that several cationic peptides are able to confer receptor independent protein transduction of a wide variety of cell types with an efficiency that is at least equivalent to the TAT PTD. Although several of the cationic peptides (PTDs) that are the most efficiently internalized do not have significant sequence homology, they share a similar positive charge. We also have demonstrated that the PTDs are able to confer efficient protein uptake by a wide variety of cell types in vitro, as well as by rabbit synovium and mouse tumors in vivo. Thus, these peptides should be highly useful for delivery of therapeutic proteins and potentially genes for the treatment of wide variety of diseases.

MATERIAL

AND

METHODS

Synthesis of protein transduction domains. To examine the ability of different cationic peptides to facilitate cell internalization, the six PTDs and a control peptide (Ran), Ant PTD, TAT PTD, a 12-mer polylysine (12-Lys) and a 12-mer polyarginine (12-Arg) were synthesized, biotinylated (Peptide Synthesis Facility, University of Pittsburgh) and conjugated to avidin-linked β-galactosidase and streptavidin–Cy3 (Sigma). Peptide internalization and localization. To determine if the peptides were able to facilitate internalization into a rabbit synovial cell line (HIG82), equimolar excess concentrations of the peptides were incubated with avidin–β-galactosidase for 2 h at room temperature in order to generate the complexes. Peptide–avidin–β-galactosidase complexes (0.3 to 150 nM) were then added to the media of HIG-82 cell cultures at 37°C. Three hours post addition of the complexes, the cells were washed extensively with TBS buffer and fixed with 4% paraformaldehyde and stained with Xgal. To examine the specificity of internalization, nine different types of cells, HIG-82 cells, rabbit primary synovial cells, human primary synovial cells, primary human airway epithelial cells (HBE144), human islets, murine myoblasts (C2C12), polarized canine kidney cells (MDCK), NIH 3T3 cells, and a murine fibrosarcoma tumor line (MCA205) were used. To examine if the peptides were able to facilitate internalization into different locations of HIG-82 cells, PTD streptavidin–Cy3 (Sigma) conjugates were made as described previously (16, 17). The PTD–Cy3 complexes were added to the media of HIG-82 cell cultures at 37°C. Three hours post addition of the complexes, the cells were washed extensively with Tris–NaCl buffer and examined by confocal microscopy. Quantitation of β-galactosidase activity in Hig-82 cells. To compare the transduction efficiency among the characterized PTDs, a quantitative βgalactosidase activity assay was utilized. Peptide–avidin–β-galactosidase complexes (1.5 to 150 nM) were added to the media of HIG-82 cell cultures at 37°C. Three hours following addition of the complexes, the cells were washed extensively in TBS and cell lysates were collected for quantitation using 1,2-dioxetane-based light emission (Tropix, Inc., Bedford, MA), displayed as relative light units (RLUs). Peptide competition assay. For peptide competition assays, 100 µM of nonbiotinylated peptides (100-fold excess) was added to HIG-82 cells at 37°C for 1 h followed by addition of 1 µM biotinylated peptide coupled to β-galactosidase. Three hours following addition of the complexes, the cells were fixed with 4% paraformaldehyde and stained with X-gal for 12 h. In vivo experiments. For analysis of peptide mediated internalization in vivo, 0.5 ml of a 150 nM solution of the various peptide–β-galactosidase complexes was injected intraarticularly into New Zealand white rabbit

342

knee joints and 50 µl of a 150 nM solution of the complexes was administered by intratumoral injection into C57BL/6 mice bearing day 10 MCA 205 tumors. Three hours post injection, the rabbits or mice were sacrificed, the joint capsules or the solid tumors isolated and the tissue stained by X-gal. 1 × 1010 particles of adenoviral vector encoding LacZ gene and a control viral vector Ψ5 were injected into the rabbit joints or mouse tumors three days before sacrifice of the rabbits or mice. Construction of peptide–eGFP fusion protein. The construction of the peptide–eGFP fusion protein was performed by PCR. The 12 amino acids of PTD-5 were inserted at the amino terminus of eGFP, whereas a 6 histidine amino acid tag was inserted at the carboxy-terminus. The fusion protein was expressed in the pET3b plasmid in Escherichia coli, purified on a nickel column and added to HIG-82 cells, human islets and human dendritic cells. Two hours post addition of the protein, the cells were washed extensively with Tris–NaCl buffer and the presence of eGFP examined using fluorescent microscope. Histology. Rabbit knee synovium was harvested from dissected rabbit knees, fixed in 4% paraformaldehyde, stained with X-gal, and then fixed in 10% formalin. The formalin fixed tissues were embedded in paraffin, sectioned at 5 µm, and stained with X-gal and eosin counterstain. The tumor tissues were frozen at −80°C, sectioned at 5 µm, fixed in 4% paraformaldehyde, stained with X-gal and counterstained with eosin and hematoxylin. Statistical analysis. Quantitative data collected were expressed as mean ± SD and statistical significance was analyzed by a paired, two-sample Student’s t test of the means.

RESULTS

AND

DISCUSSION

Cationic peptides can mediate internalization of a β-galactosidase marker protein. Although both of the well characterized protein transduction domains, TAT PTD and Ant PTD, do not have sequence homology, they both have a high percentage of positively charged amino acids. To examine the ability of different cationic peptides to facilitate cell internalization, six 12-mer peptide sequences were selected for analysis on the basis of their varying cationic charge content from an M13 phage random peptide display library (New England Biolabs). The six peptide selected sequences are all positively-charged, as they are enriched for lysine and arginine residues (Table 1). The six PTDs, two previously characterized transduction peptides, the Ant and TAT PTDs, and a control peptide (Ran) were synthesized, biotinylated and coupled to avidin -β-gal (30, 31). Initially, to determine whether the peptides were able to facilitate internalization into rabbit synovial fibroblasts, 150 nm of the peptide–avidin–β-gal complexes were added to the media of cells. Three hours following addition of the complexes, the cells were fixed and stained for the presence of β-gal activity. As shown in Fig. 1A, nearly 100% of the cells were positive for β-gal protein for PTD-2, 3, 4, and 5 as well as for the Ant and TAT PTDs. Interestingly, the peptide–β-gal complexes for the positive peptides were found predominantly concentrated in the nuclei. Moreover, PTD-4 and PTD-5 were able to deliver the β-gal complex to cells more efficiently than PTD-2 and PTD-3 whereas PTD-1 and PTD-6 were only able to deliver the β-gal complex at low levels. Furthermore, efficient internalization of the peptide–avidin–β-gal complex occurred at 4, 25 (data not MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

ARTICLE TABLE 1 Peptide Sequences characterized for Protein Transduction Length

MW

Lys + Arg/L

PTD-1

KRIIQRILSRNS

12

1482.81

0.333

PTD-2

KRIHPRLTRSIR

12

1531.88

0.417

PTD-3

PPRLRKRRQLNM

12

1563.94

0.417

PTD-4

PIRRRKKLRRLK

12

1629.81

0.667

PTD-5

RRQRRTSKLMKR

12

1614.98

0.583

PTD-6

MHKRPTTPSRKM

12

1468.80

0.333

Ran

ARPLEHGSDKAT

12

1280.40

0.167

TATPTD

YGRKKRRQRRR

11

1558.63

0.727

Ant PTD

RQIKIWFQNRRMKWKK

16

2245.78

0.438

Note. The six synthesized PTDs, a control peptide (Ran), the 11-amino-acid protein transduction domain from the TAT HIV transactivator protein (TAT PTD), and the 16amino-acid sequence comprising the third helix of the Drosophila Antennapedia homeodomain (Ant PTD) are listed. Molecular weights for these peptides (MW) are shown. The ratios of lysine and arginine content to peptide length (Lys + Arg/L) have been calculated since the PTDs are enriched for these residues. These two positively charged amino acids, arginine (R) and lysine (K), have been boldfaced in the sequences.

shown), and 37°C, suggesting a receptor and endosomeindependent mechanism of internalization. A titration assay on rabbit synovial cells was performed to determine further the efficiency of the different peptides to facilitate uptake of avidin–β-gal (Fig. 1B). The Ant PTD was approximately 50-fold less efficient in delivering the protein complex into the cells than PTD-4, PTD-5 or the TAT PTD that had a similar efficiency. Interestingly, intracellular β-gal was observed over a large range of peptide concentrations, with the levels of X-gal staining correlating with the concentration of the added complex. Inhibition of peptide-mediated internalization by nonbiotinylated peptides. To determine whether the different peptides are able to enter cells through a similar pathway, an inhibition experiment using nonbiotinylated peptides as competitors was performed (Fig. 2). A 100fold excess of nonbiotinylated PTD-5 was able to inhibit uptake of the different PTD–β-gal complexes into rabbit fibroblasts. Addition of nonbiotinylated PTD-2 and PTD3 was able to block transduction of the PTD-2 and PTD3 complexes, but not PTD-4 and PTD-5 complexes. Moreover, nonbiotinylated PTD-6 was able to block uptake of PTD-2 and PTD-3 complexes, but not PTD-4 and PTD-5 complexes (data not shown), even though PTD-6 was only weakly able to transduce cells. PTD-4 and 5 were also able to inhibit uptake mediated by both the Ant and TAT PTDs. The Ant PTD was able to block for uptake mediated by PTD-3, but only weakly inhibited internalization by PTD-4, PTD-5 and TAT PTD. The TAT PTD also was able to inhibit uptake mediated by the Ant PTD, PTD-4 and PTD-5. Polylysine was able to inhibit internalization mediated by PTD-3, PTD-5 and Ant PTD, but only partially competed with PTD-4 and the TAT PTD. These results suggest that the identified peptides (PTDs) were internalized through a mechanism similar to the Ant and TAT PTDs and that PTD-4 and 5 were able to facilitate protein uptake at least as efficiently as the TAT PTD. The ability of the polylysine to inhibit uptake suggests that the charge of the peptide is important for MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

at least part of the process of peptide-mediated internalization. PTD-5 facilitates internalization into a wide variety of cell types. To determine if the peptides are able to facilitate uptake into cell types other than rabbit synovial fibroblasts, the ability of PTD-5 to deliver β-gal protein to a variety of cell types was examined. As shown in Fig. 3, PTD-5 was able to deliver the avidin–β-gal complex to HIG-82 cells (A), rabbit primary synovial cells (B), human primary synovial cells(C), primary human airway epithelial cells (HBE144, D), polarized canine kidney cells (MDCK, E), human islets (F), murine myoblasts (C2C12, G), murine fibrosarcoma tumor line (MCA205, H) and NIH 3T3 cells (I). Indeed, PTD-5 was able to mediate efficient protein transduction in all cell types tested. The ability of the peptide to deliver β-gal complex to differentiated airway cells, HBE144, that are resistant to viral infection suggests that these peptides could potentially be used for protein and viral transfer to airway epithelium. PTDs mediate cytoplasmic and nuclear localization following internalization. To confirm that the PTDs are able to facilitate internalization and transport of protein complexes to the nucleus, the biotinylated peptides were coupled to streptavidin–Cy3 (32, 33) for confocal microscopy. PTD-5, TAT PTD, and PTD-2, 3, and 4 (data not shown) were able to mediate both cytoplasmic and nuclear localization following internalization of the streptavidin–Cy3 complexes (Fig. 4). PTDs facilitate efficient internalization in vivo. We have been developing methods for intraarticular delivery of genes for the treatment of pathologies associated with arthritis. In particular, we have been examining the ability of transfer of genes encoding apoptotic proteins to induce apoptosis of the hyperplastic synovium (34). To determine whether the PTDs can facilitate protein uptake by synovium, the peptide–avidin–β-gal complexes were injected into rabbit knee joints. The treated rabbits were sacrificed 3 h post injection and tissue from the joint capsules was stained for the presence of β-gal. As shown in Fig. 5A, significant X-gal staining was observed in the treated knees, but not in control joints. As expected, the level of β-gal internalized following intraarticular injection was highest for PTD-4, PTD-5, and TAT PTD with weaker staining observed for PTD-2, PTD-3, and Ant PTD. No staining was observed with the random peptide or saline. Interestedly, the level of staining observed was significantly higher than that observed following intraarticular injection of 1 × 1010 particles of an adenoviral vector carrying the gene for β-gal (Ad-lacZ), injected 3 days prior to injection of the peptides. Histological analysis demonstrated that the β-gal staining was intracellular and largely restricted to the synovial lining, although some superficial vessel and connective tissue staining was observed in the synovium transduced with the PTD-5 complexes (Fig. 5B). Similarly, injection of the PTD–avidin–β-gal complexes into established day 10 murine tumors resulted in extensive intratumoral β-gal

343

ARTICLE

FIG. 5. PTDs facilitate uptake of protein complexes into rabbit synovial lining and mouse fibrosarcomas. 0.5 ml of a 150 nM solution of the peptide–β-galactosidase complexes was injected intraarticularly into rabbit knee joints or intratumorally into day 10 MCA 205 tumors in C57BL/6 mice. Three hours post injection, the rabbits or mice were sacrificed, the joint capsules or tumors isolated, and the tissues stained by X-gal. As a control, an adenoviral vector encoding the LacZ gene was injected into the rabbit joints three days before sacrifice. (A) X-gal stained rabbit knee joint capsules. (B) Histology of X-gal stained sections from the rabbit synovial lining with eosin used as a counterstain. (C) X-gal staining of tumor tissue with a hematoxylin and eosin counterstain.

344

MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

ARTICLE

FIG. 6. Efficient internalization of a PTD-5–eGFP fusion protein. A PTD-5–eGFP fusion protein was constructed by PCR, expressed in bacteria, and purified using a nickel column. The presence of eGFP was visualized by fluorescent microscopy (C, E, and G). (A) Diagram of the PTD-5–eGFP fusion protein. (B and C) Transduced rabbit synovial cells; (D and E) human dendritic cells; (F and G) human islets.

staining (Fig. 5C). Taken together, these in vivo results suggest that the PTDs can facilitate efficient internalization of protein complexes into joints and tumors and could thus be useful for delivery of therapeutic proteins for arthritis and cancer, as well as other diseases. PTD-5 is able to mediate internalization of eGFP when fused to the amino terminus. To determine if the PTDs are able to deliver proteins when fused directly to a marker protein, a PTD-5–eGFP fusion protein was constructed by PCR. The chimeric eGFP protein was generated with the 12 amino acids of PTD-5 located at the N-terminus of the eGFP and a 6 histidine amino acid tag located at the Cterminus (Fig. 6A). The fusion protein was expressed in E. coli, purified on a nickel column and added to rabbit synovial cells (Figs. 6B and 6C), dendritic cells (Figs. 6D and 6E) and human islets (Figs. 6F and 6G). The fusion eGFP protein was internalized efficiently by the different cell types, localizing predominantly to the nuclei (Figs. 6C, 6E, and 6G). These results demonstrate that PTD-5 can be used to facilitate internalization when used as a fusion protein. Given that a previous report suggested that TAT PTD works more efficiently when the fusion protein is denatured, it is important to note that in all of the experiments described above, the proteins complexes or fusions were in their native conformation. Comparison of transduction efficiencies among PTDs. To compare the transduction efficiencies among the various MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

peptides, the quantitation of β-galactosidase activity in Hig-82 cells was conducted on all six PTDs and the TAT and Ant PTDs. In addition, 12-mer polylysine (Lys) and polyarginine (Arg) were also synthesized and tested for efficiencies of internalization. PTDs 1, 6, and Ant PTD demonstrated relatively weak transduction (results not shown), whereas PTDs 4 and 5, TAT, Arg and Lys all displayed significantly higher β-galactosidase internalization than the control peptide complexes (up to 600-fold greater transduction, Fig. 7). These results indicate that the content of the positive charge plays an important role in the protein transduction. However, the fact that PTD-5 and PTD-4 worked as efficiently as polylysine and polyarginine suggests that beyond a certain level, increased charge may yield diminishing improvements in transductional efficiency and that other factors, such as conformation of the PTDs, may play a role in transduction. In this report, we have described the identification and characterization of a group of cationic peptides able to facilitate the efficient internalization of large protein complexes. All peptides able to facilitate internalization were positively charged and have been termed protein transduction domains (PTDs). PTD-5, PTD-4, and TAT have similar CD spectra in solution and are highly fusogenic in a lipid environment (M. Cascio, Z.M., J.M., and P.D.R., manuscript in preparation). Mutational analysis of

345

ARTICLE

FIG. 7. Protein transduction efficiencies of cationic PTDs. The six 12-mer PTDs from the phage library (PTDs 1–6), a control peptide (Ran), Ant and TAT PTDs, and the 12 mers of polylysine (Lys) and polyarginine (Arg) were synthesized and biotinylated. The peptides were conjugated to avidin-linked β-galactosidase and were added to the media of HIG-82 cells at 37°C. Three hours following addition of the complexes, the cells were washed extensively in TBS and cell lysates were collected for quantitation using 1,2-dioxetane-based light emission, displayed as relative light units (RLUs). PTDs 1 and 6 and Ant demonstrated relatively weak transduction (results not shown).

the various PTDs may serve to identify the optimal peptide length, structure, and composition, and may lead to the identification of more efficient transduction domains. Confocal analysis demonstrated that the PTDs, including polylysine and polyarginine, had both cytoplasmic and nuclear subcellular localization. PTD-4 and 5 were able to facilitate efficient transduction into a wide variety of cell types in vitro including human β-cells, dendritic cells, polarized airway epithelial cells, tumor cells and synovial fibroblasts. Similar to the biotin–avidin complexes, the PTD-5–eGFP in-line fusion was able to efficiently transduce a number of cell types. The observation that PTDs 4 and 5, as well as polyarginine and polylysine, are able to confer efficient protein transduction demonstrates that the transductional properties found in the TAT and Antp sequences are not unique. We also have demonstrated that PTD-4 and PTD-5 can efficiently transduce rabbit synovium and murine fibrosar-

346

comas in vivo, suggesting that such PTDs could be adapted for the regional delivery of a number of other protein and/or DNA cargoes for therapeutic or investigational purposes. Indeed, PTD-5 is able to deliver a marker protein to rabbit synovium and established murine tumors far more efficiently than adenoviral vectors. We have used PTD-5 to deliver an anti-bacterial, mitochondrial disruption peptide to cells, resulting in rapid and efficient caspase-mediated apoptosis. Intraarticular or intratumoral administration of this peptide results in rapid and extensive apoptosis of synovial and tumor cells, respectively (J.M., Z.M., S.H.K., and P.D.R., manuscript in preparation). Those results demonstrate that these cationic protein transduction peptides could be used to deliver apoptotic or tumor suppressor proteins to tumors or to hyperplastic synovium. Although there are limitations to the applications of peptide mediated protein delivery, the efficient and tranMOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

ARTICLE sient delivery of proteins has certain advantages over delivery of therapeutic genes. For instance, peptide-mediated delivery of apoptotic proteins could allow for more efficient killing of proliferating tumor cells and synovial tissue without long-term toxicity. It also may be possible to generate viral vectors that use PTDs for internalization or to increase the uptake of plasmids and oligonucleotides through the use of PTD-nucleic acid conjugates. In addition to possible in vivo applications for protein and DNA delivery, the use of peptides to facilitate protein uptake clearly will be an important tool for delivery of transcription and growth regulatory factors, for the purpose of modulating cell growth and differentiation of cell in culture. ACKNOWLEDGMENTS The authors thank Helga Georgescu, Seon-Hee Kim, Andrea Gambotto, Nick Giannoukakis, and Joseph Pilewski for providing cell lines, as well as Eric Lechman and Bruce Baldwin for their technical assistance. The authors also thank Simon Watkins and the Structural Biology and Imaging Center for assistance. This work was partially supported by a grant from the Cystic Fibrosis Foundation and by Contract AR-6-2225 from the National Institutes of Health.

REFFERENCES 1

Robbins, P. D., and Ghivizzani, S. C. (1998). Viral vectors for gene therapy. Pharmacol. Ther. 80: 35–47. 2 Robbins, P. D., Tahara, H., and Ghivizzani, S. C. (1998). Viral vectors for gene therapy. Trends Biotechnol. 16: 35–40. 3 Anderson, W. F. (1998). Human gene therapy. Nature 392: 25–30. 4 Pillai, O., Nair, V., Poduri, R., and Panchagnula, R. (1999). Transdermal iontophoresis. Part II: Peptide and protein delivery. Methods Fundam. Exp. Clin. Pharmacol. 21: 229–240. 5 Witschi, C., and Mrsny, R. J. (1999). In vitro evaluation of microparticles and polymer gels for use as nasal platforms for protein delivery. Pharmacol. Res. 16: 382–390. 6 Galan, J. E., and Collmer, A. (1999). Type III secretion machines: bacterial devices for protein delivery into host cells. Science 284: 1322–1328. 7 Crotts, G., and Park, T. G. (1998). Protein delivery from poly(lactic-co-glycolic acid) biodegradable microspheres: Release kinetics and stability issues. J. Microencapsul. 15: 699–713. 8 Tracy, M. A. (1998). Development and scale-up of a microsphere protein delivery system. Biotechnol. Prog. 14: 108–115. 9 Yeh, M. K., Davis, S. S., and Coombes, A. G. (1996). Improving protein delivery from microparticles using blends of poly(DL lactide co-glycolide) and poly(ethylene oxide)poly(propylene oxide) copolymers. Pharmacol. Res. 13: 1693–1698. 10 Cevc, G. (1996). Transfersomes, liposomes and other lipid suspensions on the skin: Permeation enhancement, vesicle penetration, and transdermal drug delivery. Crit. Rev. Ther. Drug Carrier Syst. 13: 257–388. 11 Kato, Y., and Sugiyama, Y. (1997). Targeted delivery of peptides, proteins, and genes by receptor-mediated endocytosis. Crit. Rev. Ther. Drug Carrier Syst. 14: 287–331.

MOLECULAR THERAPY Vol. 2, No. 4, October 2000 Copyright  The American Society of Gene Therapy

12

Arap, W., Pasqualini, R., and Ruoslahti, E. (1998). Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 279: 377–380. 13 Koivunen, E., Arap,W., Rajotte, D., Lahdenranta, J., and Pasqualini, R. (1999). Identification of receptor ligands with phage display peptide libraries. J. Nucl. Med. 40: 883–888. 14 Pasqualini, R., Koivunen, E., Ruoslahti, E., and Alpha, V. (1997). Integrins as receptors for tumor targeting by circulating ligands Nat. Biotechnol. 15: 542–546. 15 Pasqualini, R., and Ruoslahti, E. (1996). Organ targeting in vivo using phage display peptide libraries. Nature 380: 364–366. 16 Ivanenkov, V. V., Felici, F., and Menon, A. G. (1999). Targeted delivery of multivalent phage display vectors into mammalian cells. Biochim. Biophys. Acta 1448: 463–472. 17 Leopold, P. L., Ferris, B., Grinberg, I., Worgall, S., Hackett, N. R., and Crystal, R. G. (1998). Fluorescent virions: Dynamic tracking of the pathway of adenoviral gene transfer vectors in living cells. Hum. Gene Ther. 9: 367–378. 18 Green, M., and Loewenstein, P. M. (1988). Autonomous functional domains of chemically synthesized human immunodeficiency virus tat trans-activator protein. Cell 55: 1179–1188. 19 Frankel, A. D., and Pabo, C. O. (1988). Cellular uptake of the tat protein from human immunodeficiency virus. Cell 55: 1189–1193. 20 Fawell, S., et al. (1994). Tat-mediated delivery of heterologous proteins into cells. Proc. Natl. Acad. Sci. USA 9: 664–668. 21 Nagahara, H., et al. (1998). Transduction of full-length TAT fusion proteins into mammalian cells: TAT-p27Kip1 induces cell migration. Nat. Med. 4: 1449–1452. 22 Vives, E., Brodin, P., and Lebleu B.(1997). A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J. Biol. Chem. 272: 16010–16017. 23 Schwarze, R. S., Ho, A., Vocero-Akbani, A., and Dowdy, F. S. (1999). In vivo protein transduction: Delivery of a biologically active protein into the mouse. Science 285: 1569–1572. 24 Vocero-Akbani, A. M., Heyden, N. V., Lissy, N. A., Ratner, L., and Dowdy, S. F. (1999). Killing HIV-infected cells by transduction with an HIV protease-activated caspase-3 protein. Nat. Med. 5: 29–33. 25 Moy, P., Daikh, Y., Pepinsky, B., Thomas, D., Fawell, S., and Barsoum, J. (1996). Tatmediated protein delivery can facilitate MHC class I presentation of antigens. Mol. Biotechnol. 6: 105–113. 26 Derossi, D., Joliot, A. H., Chassaings, G., and Prochiantz, A. (1994). The third helix of the Antennapedia homeodomain translocates through biological membranes. J. Biol. Chem. 269: 10444–10450. 27 Derossi, D., Chassaing, G., and Prochiantz, A. (1998). Trojan peptides: The penetratin system for intracellular delivery. Trends Cell Biol. 8: 84–87. 28 Derossi, D., Calvet, S., Trembleau, A., Brunissen, A., Chassaing, G., and Prochiantz, A. (1996). Cell internalization of the third helix of the Antennapedia homeodomain is receptor-independent. J. Biol. Chem. 271: 18188–18193. 29 Elliott, G., and O’Hare, P. (1997). Intercellular trafficking and protein delivery by a Herpes virus structural protein. Cell 188: 223–233. 30 O’Sullivan, M. J., et al. (1978). A simple method for the preparation of enzyme-antibody conjugates. FEBS Lett. 95: 311–313. 31 O’Sullivan, M. J., et al. (1979). Comparison of two methods of preparing enzyme–antibody conjugates: Application of these conjugates for enzyme immunoassay. Anal. Biochem. 100: 100–108. 32 Bayer, E. A., and Wilchek, M. (1980). The use of the avidin–biotin complex as a tool in molecular biology. Methods Biochem. Anal. 26: 1–45. 33 Bayer, E. A., Skutelsky, E., Wynne, D., and Wilchek, M. (1976). Preparation of ferritin–avidin conjugates by reductive alkylation for use in electron microscopic cytochemistry. J. Histochem. Cytochem. 24: 933–939. 34 Yao, Q., et al. (2000) Adenoviral mediated delivery of Fas ligand to rheumatoid synovium results in extensive apoptosis in the synovial lining J. Gene Med. 2: 147–220.

347