Intracellular Delivery of Functional Proteins via Decoration with

ARTICLE

doi:10.1016/j.ymthe.2004.02.004

Intracellular Delivery of Functional Proteins via Decoration with Transporter Peptides Zurab Siprashvili, Jason A. Reuter, and Paul A. Khavari* VA Palo Alto Healthcare System, Palo Alto, CA 94025 and Program in Epithelial Biology, Stanford University, 269 Campus Drive, Room 2145, Stanford, CA 94305, USA

*To whom correspondence and reprint requests should be addressed. Fax: (650) 723-8762. E-mail: [email protected].

Available online 8 March 2004

Despite numerous attractive intracellular targets, protein therapeutics have been principally confined to the extracellular space due to the lack of a straightforward way to deliver functional polypeptides to the cell interior. Peptide sequences facilitating intracellular protein delivery have been identified; however, current strategies to apply them require problematic steps, such as generation of new in-frame fusion proteins, covalent chemical conjugation, and denaturation. We have developed a new approach to protein transfer into cells and tissues that relies on singlestep decoration by cysteine-flanked, arginine-rich transporter peptides. This approach facilitated cell and tissue delivery of a variety of functional proteins, including antibodies and enzymes. Decoration with transporter peptides thus provides an attractive general means of intracellular delivery of functional proteins in vitro and in tissue.

INTRODUCTION Development of therapeutic proteins to date has focused on secreted and cell surface proteins. This emphasis on the extracellular setting neglects the vast array of attractive intracellular targets that alter central cellular processes [1], which are not readily tractable to protein-based therapeutics currently due to difficulties in delivering functionally active polypeptides into cells. Recent efforts to overcome this problem have demonstrated the ability of specific peptide sequences, such as arginine-rich sequences from HIV Tat [2,3] and other proteins [4 – 6], to promote protein penetration into cells and tissues, facilitating design of a variety of synthetic peptide carriers encompassing protein transduction domain (PTD)like transporter peptide capacity [7 – 10]. PTD transporter sequences can mediate intracellular delivery of a wide variety of macromolecules [11 – 15], including polypeptides. While its mechanistic basis remains unclear, the protein transduction process can vary with the number and position of arginines [7], can be influenced by cellsurface heparan sulfate [16], and does not appear to occur through a classical receptor- or endosome-mediated fashion [9,17,18]. Although PTDs can enhance protein delivery, certain limitations exist, such as the necessity for generation of PTD – protein fusions [19 – 21], denaturation of the fusion proteins prior to transduction [13,22], or covalent cross-linking [18,23]. These labor-intensive processes require customization for each protein and do not

MOLECULAR THERAPY Vol. 9, No. 5, May 2004 Copyright B The American Society of Gene Therapy 1525-0016/$30.00

always result in polypeptides that retain biologic activity. Harnessing the capacity of PTD sequences to deliver functional proteins in an uncomplicated, efficient, and uniform manner represents a significant goal of current therapeutic protein delivery efforts.

RESULTS AND DISCUSSION To assess delivery of proteins via decoration with transporter peptides, we incubated h-galactosidase protein with the HGH6 transporter peptide at a series of molar ratios. HGH6 is a cysteine-flanked, internally spaced arginine-rich 25-amino-acid residue transporter peptide with the primary sequence CG(RHGH)5RGC [24]. Sedimentation analyses of the HGH6-decorated h-galactosidase displayed evidence of complex formation between 1:50 and 1:100 protein:peptide molar ratios (Fig. 1A, lanes 6 – 9) with formation of the large complexes sedimenting with the pellet at >1:100 molar ratios (Fig. 1A, lanes 12 – 15). Protein – peptide interaction was fully reversible with SDS and dithiothreitol (DTT) treatment, indicating contribution of both electrostatic and disulfide interactions in complex formation (Fig. 1A, lanes 17 – 20). We examined delivery of h-galactosidase protein decorated with HGH6 to COS cells in vitro both by enzymatic activity in cell lysates and by X-gal staining. After transfer, we treated cells with trypsin to cleave the peptidyl bond following the arginine residues and thus

721

ARTICLE

722

doi:10.1016/j.ymthe.2004.02.004

MOLECULAR THERAPY Vol. 9, No. 5, May 2004 Copyright B The American Society of Gene Therapy

doi:10.1016/j.ymthe.2004.02.004

remove any cell-surface-bound protein complexes, preventing measurement artifacts described earlier [25]. Consistent with sedimentation analyses, uptake of decorated h-galactosidase reached optimum levels at a 1:50 h-galactosidase:HGH6 transporter peptide molar ratio (Fig. 1B), with further decline at ratios >1:100 due to aggregate formation. Efficient uptake occurred within 30 min in a variety of cell types, including 293, NIH 3T3, and HeLa cell lines as well as primary human keratinocytes and fibroblasts (Figs. 1C and 1D). Enzymatically active h-galactosidase protein was present in all cells examined, in contrast to undecorated protein, which displayed minimal X-gal staining (Fig. 1C). Neither ATP depletion nor decreased temperature abolished protein transfer (Figs. 1E and 1F), consistent with prior studies of arginine-rich transporter peptide uptake, which argued against an energy-requiring, endocytic process [18]. Thus, decoration with arginine-rich transporter peptides facilitates delivery of functionally active protein to a variety of cell types in vitro. To determine if peptide decoration represents a general approach that can be applied to a variety of different proteins, we evaluated delivery of decorated placental alkaline phosphatase (PLAP), monoclonal antibodies to specific intracellular targets, and caspase-3. PLAP alone was unable to enter cells in culture; however, cells displayed universal uptake of PLAP protein decorated with HGH6 peptide after 30 min (Fig. 2A). For immunoglobulins, we assessed delivery of two different monoclonal antibodies: Cy3-conjugated anti-h-tubulin and FITC-conjugated antibodies to the Golgi matrix protein GM130. While confocal microscopy demonstrated that undecorated antibodies failed to penetrate cells, decorated antibodies entered living cells efficiently in all cells exposed to them and accumulated in the correct subcellular location. Tubulin antibodies localized to the microtubular network, while anti-Golgi antibodies stained the perinuclear Golgi region (Fig. 2B), subcellular localizations consistent with retention of functional binding for their specific targets. Recent studies have shown that fixation can cause artificial import of a protein into cells [26,27]. To examine this possibility, we compared the staining of COS cells with HGH6-decorated Cy3-conjugated anti-h-

ARTICLE

tubulin and FITC-conjugated anti-h-actin antibody before and after methanol fixation. As shown in Fig. 2B, the fluorescence patterns are essentially identical, indicating that intracellular localization seen with HGH6 decoration is not merely the result of a fixation artifact. We next examined induction of cell death by delivering one of the major intracellular effectors of apoptosis—caspase-3. Recombinant caspase-3 was either decorated or unmodified and then incubated with HeLa cells. Sixteen hours later, TUNEL analysis demonstrated marked increases in apoptosis with decorated caspase-3 compared to controls (Figs. 2C and 2D). Taken together, these in vitro results indicate that decoration with arginine-rich transporter peptides can be used to facilitate intracellular delivery of a variety of different proteins, while preserving their intact function. We next tested the capacity of transporter peptide decoration to achieve intracellular delivery of biologically active proteins in vivo. To do this, we administered proteins by intradermal (i.d.) injection to both murine skin and full-thickness human skin grafted to immunodeficient mice and then assayed uptake 13 h postinjection. Sequential X-gal-stained sections revealed that, in all cases, h-galactosidase detection was most prominent proximal to the injection site; however, HGH6-decorated protein displayed elevated tissue staining compared to hgalactosidase alone and established benchmark Tat – hgalactosidase fusion protein [13] controls (Fig. 3A). Furthermore, enzymatic activity measurements of tissue extracts were 5.4- and 73.8-fold higher compared to Tat – h-galactosidase and undecorated protein, respectively (Fig. 3B). We obtained analogous results using PLAP, establishing the general feasibility of this approach to multiple proteins both in vitro and in vivo (Fig. 3C). We also wished to examine the ability of transporter peptides to deliver proteins affecting integral cell biological processes in vivo. As a model, we delivered caspase-3 intratumorally and then assayed induction of apoptosis in tumor tissue in vivo. Specifically, we injected subcutaneous tumors formed by HeLa cell implantation into nude mice with decorated and undecorated caspase-3, as well as HGH6 peptide alone, into the central tumor mass for 4 consecutive days and

FIG. 1. Intracellular delivery of functionally active h-galactosidase protein. (A) Sedimentation analyses of h-galactosidase – HGH6 peptide interaction using SDS – PAGE. Lanes 1 and 16, protein molecular weight markers; lanes 10 and 11, h-galactosidase control. Lanes 2 and 3, h-galactosidase in the absence of HGH6; lanes 4 and 5, h-galactosidase:HGH6 at ratio 1:5; lanes 6, 7, 19, and 20, ratio 1:50; lanes 8, 9, 17, and 18, ratio 1:100; lanes 12 and 13, ratio 1:200; lanes 14 and 15, ratio 1:300; S indicates the supernatant fraction; P indicates the pellet fraction. Note protein shift from the supernatant to the pellet fraction that occurs between 1:50 and 1:100 molar ratios (lanes 8 – 9, 12 – 15). Also, note h-galactosidase rescue from sedimentation by disruption of the protein – peptide interaction using SDS and DTT at protein:peptide ratios 1:50 and 1:100 (lanes 17 – 20). (B) 10 Ag of either h-galactosidase protein alone or protein decorated with HGH6 peptide at the molar ratios noted was applied to COS cells. h-Galactosidase enzymatic activity in mU/mg of total cell extract protein is shown (n = 3 independent experiments for each condition, error bars represent FSD). (C) X-gal staining of h-galactosidase protein transduction in a variety of cell lines, COS, 293, NIH 3T3, HeLa, primary human fibroblasts, and primary human keratinocytes (KC), after transduction of 10 Ag of h-galactosidase protein decorated with HGH6 at a 1:50 molar ratio. Scale bar, 40 Am. (D) Quantitation of relative h-galactosidase enzymatic activity in cell extracts compared to undecorated enzyme alone (n = 3, error bars represent FSD). (E) Effects of temperature and ATP depletion on h-galactosidase uptake. COS cells were incubated for 2 h with undecorated or HGH6decorated h-galactosidase at 37 or 4jC or in ATP-depleted medium and protein transduction was evaluated by enzymatic activity measurements in cell lysates (n = 3, error bars represent FSD). (F) Same as (E), but h-galactosidase uptake is visualized by X-gal staining. Scale bar, 20 Am.


723

ARTICLE

724

doi:10.1016/j.ymthe.2004.02.004


ARTICLE

doi:10.1016/j.ymthe.2004.02.004

then analyzed them by TUNEL assay. While HGH6 transporter peptide alone produced minimal effects, decorated caspase-3 boosted the frequency of TUNEL+ cells >184-fold over untreated tumors and >5-fold beyond that observed with undecorated caspase-3 alone (Figs. 4A and 4B). Thus, decoration of caspase-3 with arginine-rich transporter peptides facilitated enhanced tumor cell death in vivo. In this study, we have developed a straightforward approach for the delivery of functionally active globular proteins to the cellular interior that is effective in both cultured cells and intact tissue. This approach relies on the use of short arginine-rich transporter peptides flanked with cysteine residues to achieve both electrostatic interactions and reversible association with cargo proteins via disulfide bonding. This strategy is effective for globular proteins displaying a range of sizes and multimer subunit arrangement, including hgalactosidase [28] (tetramer, 465 kDa), PLAP [29] (dimer, 116 kDa), caspase-3 [30] (17 and 12 kDa, heterodimer), and immunoglobulins (IgG, 150 kDa). Protein delivery was not cell-type specific and approached 100% efficiency in COS, 293, HeLa, and NIH 3T3 cell lines as well as in primary human fibroblasts and primary human keratinocytes. Moreover, protein uptake occurred in both normal skin and malignant tumor tissue. With the exception of the deliberately cytotoxic caspase-3 experiments, the current strategy lacked significant cell or tissue toxicity, as judged by cell and tissue morphology, trypan blue staining of cultured cells, and TUNEL assay of cells and tissue, consistent with our previous studies of arginine-rich peptide-mediated DNA transfection, which used up to 800-fold greater amounts of peptide [24]. Therefore, protein delivery based on decoration with transporter peptides lacks substantial cytotoxic effects and functions in a variety of cells and tissues. Protein transfer displayed a number of distinct technical features. First, transfer was dependent on the peptide:protein molar ratio used and, to a lesser extent, on cargo/transporter concentration. Second, replacement of L-arginine residues with D-isomer had no effect on protein transfer, whereas lysine substitution, as well as cysteine omission diminished it significantly (data not shown). Regarding molar ratio, high uptake levels occurred at only a narrow range of ratios, with a significant reduction in protein internalization observed at peptide:protein

molar ratios >1:100. Reduced uptake at high ratios may result from the formation of large complex aggregates initiated by excess of the cationic charge or intermolecular disulfide bridging, as has been described in DNA condensation [31]. Based on previous reports with arginine-rich peptides [9,17], we analyzed protein delivery in settings of low temperature and ATP depletion. Our studies indicate that HGH6-decorated proteins enter cells at low temperature in an energy-independent manner. This suggests that such transport occurs with minimal input from classical endocytic mechanisms. The ability to decorate presynthesized, commercially available proteins with a standard transporter peptide described here offers the advantages of rapidity, simplicity, and general applicability over currently available approaches that require customized fusion protein generation and/or denaturation. Additionally, the clear preservation of intracellular function observed in each case is attractive because PTD – protein fusions or proteins tagged by covalent chemical cross-linking may suffer from loss or alterations of function. Delivery efficiency of decorated proteins in all the cell types examined in this study was effectively 100% in culture, comparable to fusion protein-based PTD approaches; however, delivery in vivo was actually more effective with protein decoration. Therefore, protein delivery by the present simplified approach appears comparable, and in some cases superior, to the best available current approaches. Decoration with arginine-rich transporter peptides thus represents a straightforward general platform for delivery of functionally active proteins into the cell interior, both in vitro and in living tissue.

MATERIALS AND METHODS Transporter peptide and functional proteins. The HGH6 peptide (CG(RHGH)5RGC) [24] was synthesized as described [32]. Commercially available functional proteins h-galactosidase, Cy3-conjugated monoclonal Tub 2.1 anti-h-tubulin, and FITC-conjugated monoclonal AC-15 antih-actin antibody were from Sigma (St. Louis, MO, USA); human PLAP was from EMD Biosciences (San Diego, CA, USA); caspase-3 was from Upstate Biotechnology (Lake Placid, NY, USA); and FITC-conjugated monoclonal anti-GM130 antibody was from BD Biosciences (San Jose, CA, USA). pTAT bacterial expression vector (generous gift from S. Dowdy) was used to produce the recombinant Tat – h-galactosidase fusion protein, which was further purified to homogeneity as described [12]. Protein decoration. Proteins were decorated with the HGH6 transporter peptide in 100 Al bicine buffer (N,N-bis(2-hydroxyethyl)glycine; Sigma), pH 7.5, for 10 min at 25jC. HGH6/h-galactosidase complexes were

FIG. 2. In vitro protein delivery via decoration with transporter peptides. (A) PLAP uptake in COS cells. 35 Ag PLAP and HGH6 peptide (1:100 molar ratio) was incubated with COS cells and protein uptake analyzed 30 min posttransduction. Scale bar, 40 Am. (B) HGH6-mediated immunoglobulin delivery in mammalian cells. Monoclonal Cy3-conjugated anti-h-tubulin (3 Ag) or FITC – anti-GM130 (2.5 Ag) and FITC – anti-h-actin (5.6 Ag) antibodies were incubated with HGH6 (1:10 molar ratio) for 10 min and overlaid onto cultured COS cells for 30 min. Cells were washed and methanol-fixed, nuclei were counterstained with Hoechst 33342, and antibody internalization was observed by laser confocal fluorescence microscopy. Images show double staining of cells with Cy3 (red) or fluorescein (green) and Hoechst 33342 (blue). The right shows comparison of FITC – anti-h-actin/HGH6 and Cy3 – anti-h-tubulin/HGH6 uptake in COS cells before and after methanol fixation. Scale bar, 10 Am. (C) Caspase-3 transduction in HeLa cells. Apoptosis was evaluated by TUNEL staining 16 h after protein addition in untreated HeLa or caspase-3- or caspase-3/HGH6-transduced cells. The bright green dots correspond to representative TUNEL-positive fluorescent nuclei. Scale bar, 40 Am. (D) Same as (C) with apoptosis quantified and reflected as an enrichment index given as the mean level of three experiments F SD.


725

ARTICLE

doi:10.1016/j.ymthe.2004.02.004

FIG. 3. In vivo delivery of decorated proteins. (A) h-Galactosidase transduction in human skin regenerated on immune-deficient mice. 35 U of h-galactosidase/HGH6 (50 Ag, 1:50 molar ratio), undecorated h-galactosidase, or a previously characterized Tat – h-galactosidase fusion protein was injected i.d. into full-thickness human skin and biopsy obtained 13 h later. Sequential X-gal-stained cryosections are presented. Scale bar, 300 Am. Higher magnification images of eosincounterstained sections on the right demonstrate h-galactosidase protein in dermal and epidermal tissue. (B) Quantitation of hgalactosidase enzymatic activity in human skin from (A) (n = 3 independent injected grafts analyzed, error bars represent FSD). (C) PLAP transduction in murine skin. 52.5 Ag of undecorated PLAP protein or PLAP/ HGH6 complex (1:100 molar ratio) was injected i.d. into mouse skin and alkaline phosphatase enzymatic activity visualized (dark staining) in tissue sections 15 h later.

formed in 5 mM buffer using 1 – 50 Ag (700 U/mg) h-galactosidase and HGH6 at 1:1 – 300 protein:peptide molar ratios. Sedimentation analyses of h-galactosidase/HGH6 peptide interaction were performed using 20 Ag h-

726

galactosidase protein. Decorated protein was centrifuged 20 min at 10,000g and supernatant removed. The pellet was washed two times in 5 mM bicine buffer (pH 7.5), lysed in reducing sample buffer, and then


doi:10.1016/j.ymthe.2004.02.004

subjected to SDS – PAGE. Protein – peptide complexes formed at 1:50 and 1:100 molar ratios were disrupted by treatment with 1% SDS and 1.5% DTT for 15 min before centrifugation. PLAP was decorated in 50 mM bicine buffer using 35 Ag (103 U/mg) PLAP and HGH6 at a 1:100 molar

ARTICLE ratio. Caspase-3 (4 104 U/mg) was decorated in 50 Al 10 mM bicine buffer using 1 Ag of protein and peptide at a 1:50 molar ratio and immunoglobulins were decorated using 3 Ag anti-h-tubulin (20 pmol), 2.5 Ag anti-GM130 (16.7 pmol), and 5.6 Ag anti-h-actin (37 pmol) at a 1:10 molar ratio using 250 mM bicine buffer. Decorated proteins were prepared immediately before use. Protein delivery in vitro. COS-7, NIH 3T3, HeLa, 293T, and primary human fibroblast cells were cultured in DMEM supplemented with 10% FBS. Protein complexes were added to cells at 65% confluency in FBScontaining medium for at least 30 min prior to analyses. For antibody transduction, cells were kept in 250 mM bicine buffer throughout protein incubation. After incubation, cells were washed twice with PBS and then analyzed with fluorescence microscopy before fixation. Duplicate chambers were fixed 2 min with 100% ice-cold methanol, rehydrated in PBST (0.1% Triton X-100 in PBS), stained 1 min with 1 Ag Hoechst 333342, and, after being washed with PBST, mounted in Vectashield mounting medium (Vector Laboratories, Burlingame, CA, USA). h-Galactosidase activity was determined in triplicate independent experiments using a h-galactosidase enzyme assay kit (Promega, Madison, WI, USA). For in vitro protein uptake studies, cells were treated with trypsin to remove any cell surfacebound protein complexes and prevent measurement artifacts. Duplicate wells of h-galactosidase+ cells were visualized with X-gal staining after 2 min fixation in PBS with 1% formaldehyde. Endogenous alkaline phosphatases were heat-inactivated by incubation in PBS at 70jC for 30 min and then PLAP delivery was analyzed using BM purple AP substrate according to the manufacturer’s instructions (Roche, Indianapolis, IN, USA). Apoptosis was evaluated 16 h after caspase-3 transduction using an in situ cell death detection kit according to the manufacturer’s instructions (Roche). The apoptotic enrichment index was calculated relative to untreated control by dividing the number of apoptotic cells in five captured optical fields by the total number of cells, using OpenLab software (Improvision, Lexington, MA, USA). For cellular uptake assays 10 Ag (700 U/mg) of h-galactosidase decorated with HGH6 peptide (1:50 molar ratio) was added to 3 105 COS cells in 35-mm plates and incubated at 37 or 4jC for 2 h. Cells were washed three times and trypsinized, and h-galactosidase activity was measured in the lysates or visualized in duplicate wells as above. For low-temperature uptake experiments, all buffers and solutions were equilibrated at 4jC and incubations, washes, and spins conducted at the same temperature. For uptake under ATP-depleted conditions, COS cells were preincubated at 37jC for 1 h in ATP-depleted medium (glucose-free MEM, 10% FBS, 10 mM sodium azide) and, after addition of the complexes, incubated at 37jC in the same medium for 2 h. Protein uptake in vivo. Animal and human tissue studies were approved by the appropriate institutional review boards. Full-thickness human skin xenografts were placed on immune-deficient mice and studied 6 to 8 weeks postgrafting. Wholly murine skin studies were performed using adult C57BL/6 mice. Proteins were decorated as described, using 35 U of both h-galactosidase/HGH6 (50 Ag, 1:50 molar ratio) and Tat – h-galactosidase (62 Ag, 565 U/mg), 5.4 U of PLAP/HGH6 (52.5 Ag, 1:100 molar ratio), and 40 U of caspase-3/HGH6 (1 Ag, 1:50 molar ratio). h-Galactosidase, PLAP, and corresponding protein – peptide complexes were injected intradermally using a 28-gauge needle. For caspase-3 uptake studies, 5 106 HeLa cells in 150 Al PBS were injected subcutaneously into 4-week old female athymic nu/nu mice. Fourteen days later, when tumors were established, decorated

FIG. 4. Assessment of apoptosis after intratumoral administration of HGH6decorated caspase-3 in vivo. (A) HeLa tumors developed by subcutaneous implantation of 5 106 cells in athymic nude mice were treated with intratumoral injections of 1 Ag decorated caspase-3 (1:50 molar ratio), undecorated caspase-3, or HGH6 peptide alone. TUNEL-stained (green) tumor tissue cryosections were counterstained with Hoechst 33342 (blue). Scale bar, 100 Am. (B) Apoptosis levels from (A) are quantified and reflected as an enrichment index, calculated by dividing the number of apoptotic cells in experimental samples by untreated control.


727

ARTICLE

caspase-3 protein and controls were administered intratumorally once daily for 4 consecutive days and apoptotic index was analyzed as noted above.

ACKNOWLEDGMENTS We thank P. Wender and S. Oliver for peptide reagent support, S. Dowdy for the Tat – b-galactosidase plasmid, J. Rothbard and P. Wender for helpful comments, and N. Griffiths and P. Bernstein for administrative help. This work was supported by the USVA Office of Research and Development and by NIH AR44012 to P.A.K. and NIH/NIAMS T32 AR07422 to Z.S. RECEIVED FOR PUBLICATION NOVEMBER 5, 2003; ACCEPTED FEBRUARY 3, 2004.

REFERENCES 1. Wheeler, Y. Y., Chen, S. Y., and Sane, D. C. (2003). Intrabody and intrakine strategies for molecular therapy. Mol. Ther. 8: 355 – 366. 2. Green, M., and Loewenstein, P. M. (1988). Autonomous functional domains of chemically synthesized human immunodeficiency virus tat trans-activator protein. Cell 55: 1179 – 1188. 3. Frankel, A. D., and Pabo, C. O. (1988). Cellular uptake of the tat protein from human immunodeficiency virus. Cell 55: 1189 – 1193. 4. Elliott, G., and O’Hare, P. (1997). Intercellular trafficking and protein delivery by a herpesvirus structural protein. Cell 88: 223 – 233. 5. Joliot, A., Pernelle, C., Deagostini-Bazin, H., and Prochiantz, A. (1991). Antennapedia homeobox peptide regulates neural morphogenesis. Proc. Natl. Acad. Sci. USA 88: 1864 – 1868. 6. Chatelin, L., Volovitch, M., Joliot, A. H., Perez, F., and Prochiantz, A. (1996). Transcription factor hoxa-5 is taken up by cells in culture and conveyed to their nuclei. Mech. Dev. 55: 111 – 117. 7. Wender, P. A., et al. (2000). The design, synthesis, and evaluation of molecules that enable or enhance cellular uptake: peptoid molecular transporters. Proc. Natl. Acad. Sci. USA 97: 13003 – 13008. 8. Ho, A., et al. (2001). Synthetic protein transduction domains: enhanced transduction potential in vitro and in vivo. Cancer Res. 61: 474 – 477. 9. Futaki, S., et al. (2001). Arginine-rich peptides: an abundant source of membranepermeable peptides having potential as carriers for intracellular protein delivery. J. Biol. Chem. 276: 5836 – 5840. 10. Mi, Z., et al. (2000). Characterization of a class of cationic peptides able to facilitate efficient protein transduction in vitro and in vivo. Mol. Ther. 2: 339 – 347. 11. Morris, M. C., Chaloin, L., Heitz, F., and Divita, G. (2000). Translocating peptides and proteins and their use for gene delivery. Curr. Opin. Biotechnol. 11: 461 – 466. 12. Schwarze, S. R., Hruska, K. A., and Dowdy, S. F. (2000). Protein transduction: unrestricted delivery into all cells? Trends Cell Biol. 10: 290 – 295. 13. Schwarze, S. R., Ho, A., Vocero-Akbani, A., and Dowdy, S. F. (1999). In vivo protein transduction: delivery of a biologically active protein into the mouse. Science 285: 1569 – 1572.

728

doi:10.1016/j.ymthe.2004.02.004

14. Torchilin, V. P., Rammohan, R., Weissig, V., and Levchenko, T. S. (2001). TAT peptide on the surface of liposomes affords their efficient intracellular delivery even at low temperature and in the presence of metabolic inhibitors. Proc. Natl. Acad. Sci. USA 98: 8786 – 8791. 15. Lewin, M., et al. (2000). Tat peptide-derivatized magnetic nanoparticles allow in vivo tracking and recovery of progenitor cells. Nat. Biotechnol. 18: 410 – 414. 16. Tyagi, M., et al. (2001). Internalization of HIV-1 tat requires cell surface heparan sulfate proteoglycans. J. Biol. Chem. 276: 3254 – 3261. 17. Suzuki, T., et al. (2002). Possible existence of common internalization mechanisms among arginine-rich peptides. J. Biol. Chem. 277: 2437 – 2443. 18. Mai, J. C., Shen, H., Watkins, S. C., Cheng, T., and Robbins, P. D. (2002). Efficiency of protein transduction is cell type-dependent and is enhanced by dextran sulfate. J. Biol. Chem. 277: 30208 – 30218. 19. Peitz, M., Pfannkuche, K., Rajewsky, K., and Edenhofer, F. (2002). Ability of the hydrophobic FGF and basic TAT peptides to promote cellular uptake of recombinant Cre recombinase: a tool for efficient genetic engineering of mammalian genomes. Proc. Natl. Acad. Sci. USA 99: 4489 – 4494. 20. Park, J., et al. (2002). 9-Polylysine protein transduction domain: enhanced penetration efficiency of superoxide dismutase into mammalian cells and skin. Mol. Cells 13: 202 – 208. 21. Cao, G., et al. (2002). In vivo delivery of a Bcl-xL fusion protein containing the TAT protein transduction domain protects against ischemic brain injury and neuronal apoptosis. J. Neurosci. 22: 5423 – 5431. 22. Schwarze, S. R., and Dowdy, S. F. (2000). In vivo protein transduction: intracellular delivery of biologically active proteins, compounds and DNA. Trends Pharmacol. Sci. 21: 45 – 48. 23. Chen, B. X., and Erlanger, B. F. (2002). Intracellular delivery of monoclonal antibodies. Immunol. Lett. 84: 63 – 67. 24. Siprashvili, Z., et al. (2003). Gene transfer via reversible plasmid condensation with cysteine-flanked, internally spaced arginine-rich peptides. Hum. Gene Ther. 14: 1225 – 1233. 25. Lundberg, M., and Johansson, M. (2002). Positively charged DNA-binding proteins cause apparent cell membrane translocation. Biochem. Biophys. Res. Commun. 291: 367 – 371. 26. Leifert, J. A., Harkins, S., and Whitton, J. L. (2002). Full-length proteins attached to the HIV tat protein transduction domain are neither transduced between cells, nor exhibit enhanced immunogenicity. Gene Ther. 9: 1422 – 1428. 27. Richard, J. P., et al. (2003). Cell-penetrating peptides: a reevaluation of the mechanism of cellular uptake. J. Biol. Chem. 278: 585 – 590. 28. Fowler, A. V., and Zabin, I. (1978). Amino acid sequence of beta-galactosidase. J. Biol. Chem. 253: 5521 – 5525. 29. Le Du, M. H., Stigbrand, T., Taussig, M. J., Menez, A., and Stura, E. A. (2001). Crystal structure of alkaline phosphatase from human placenta at 1.8 A˚ resolution. J. Biol. Chem. 276: 9158 – 9165. 30. Rotonda, J., et al. (1996). The three-dimensional structure of apopain/CPP32, a key mediator of apoptosis. Nat. Struct. Biol. 3: 619 – 625. 31. Ledley, F. D. (1995). Nonviral gene therapy: the promise of genes as pharmaceutical products. Hum. Gene Ther. 6: 1129 – 1144. 32. Robbins, P.B., et al. (2002). Peptide delivery to tissues via reversibly linked protein transduction sequences. Biotechniques 33: 190 – 192, 194.
