Cell penetration by transportan

^a Department of Neurochemistry and Neurotoxicology, Arrhenius Laboratories, Stockholm University, S-10691 Stockholm, Sweden
^b Estonian Biocentre, EE-2400, Tartu, Estonia
^c Institute of Biochemistry, Medical Faculty, University of Ljubljana, 1000 Ljubljana, Slovenia

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Transportan is a 27 amino acid-long peptide containing 12 functional amino acids from the amino terminus of the neuropeptide galanin and mastoparan in the carboxyl terminus, connected via a lysine. Transportan is a cell-penetrating peptide as judged by indirect immunofluorescence using N¹³-biotinyl-transportan. The internalization of biotinyl-transportan is energy independent and takes place efficiently at 37°, 4°, and 0°C. Cellular uptake of transportan is probably not mediated by endocytosis, since it cannot be blocked by treating the cells with phenylarsine oxide or hyperosmolar sucrose solution and is nonsaturable. The kinetics of internalization was studied with the aid of the ¹²⁵I-labeled peptide. At 37°C, the maximal intracellular concentration is reached in about 20 min. The internalized transportan is protected from trypsin. The cell-penetrating ability of transportan is not restricted by cell type, but seems to be a general feature of this peptide. In Bowes' melanoma cells, transportan first localizes in the outer membrane and cytoplasmatic membrane structures. This is followed by a redistribution into the nuclear membrane and uptake into the nuclei where transportan concentrates in distinct substructures, probably the nucleoli.—Pooga, M., Hällbrink, M., Zorko, M., Langel, Ü. Cell penetration of transportan FASEB J. 12, 67–77 (1998)

Key Words: peptide synthesis • Bowes' melanoma cells • endocytosis • GTPase activity

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
GALPARAN, A 27 AMINO ACID-LONG member of a class of galanin-based chimeric peptides, was recently designed and synthesized (1). It contains the first 13 amino acids from the highly conserved amino-terminal part of galanin and the 14 amino acid-long wasp venom peptide toxin, mastoparan, in the carboxyl terminus. The amino-terminal fragment 1–13 is the smallest highly active galanin receptor ligand with agonist properties (2, 3). The other component of galparan, mastoparan, is known to activate G-proteins (4). At high concentrations, mastoparan is cytotoxic, perturbs membranes, and shows lytic effects (5, 6). Mastoparan is penetrating the cell membrane where it creates short-living pores by translocating into the inner leaflet (7); it inhibits Golgi transport (8) and facilitates the opening of mitochondrial permeability transition pores (9).

Galparan possesses properties apart from those of its components. The chimeric peptide activates the Na⁺,K⁺-ATPase in rat cortical membranes, whereas galanin has no effect and mastoparan shows inhibitory properties (1). Galanin inhibits glucose-induced insulin release (10). Galparan, on the other hand, is a potent stimulator of insulin secretion, whereas unsubstituted mastoparan is a low efficacy stimulator (11). Galanin inhibits acetylcholine release in the ventral hippocampus, whereas galparan induces the release in frontal cortex (12). The neurotransmitter and insulin-releasing activity of galapran is not exerted through the galanin receptors or the sites where mastoparan acts (11, 12). These findings promoted the study of the mechanism of action of this galanin-mastoparan chimera and especially its cell-penetrating properties.

To study the possible cellular uptake of galparan as well as the localization on the cell surface or in the cell interior, we have synthesized a galparan analog, [Lys¹³]galparan, and characterized the internalization and intracellular localization of this peptide. We have used the biotinylated and radioactively labeled analogs of the peptide in this study. Because of the efficient cellular uptake of the peptide and the availability of a reactive amino group at Lys¹³, which is a suitable handle to connect to cargo molecules, it would be possible to apply the peptide as a carrier vector for hydrophilic macromolecules; hence we have taken the liberty to apply the name transportan to this novel peptide.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Materials
8-Anilino-1-naphthalenesulphonic acid, MgCl₂, NaCl, H₃PO₄, and EDTA were from Merck, Darmstadt, Germany; [-³²P]GTP, ¹²⁵I-Tyr²⁶-galanin (2200 Ci/mmol) and Na¹²⁵I were from NEN Research Products (Hounslow, U.K.); tert-butyloxycarbonyl amino acids were from Bachem (Bubendorf, Switzerland) and Chemimpex (Wood Dale, Ill.). Streptavidin conjugates with fluorescein, Texas red, and alkaline phosphatase were purchased from Amersham (Amersham, U.K.). Cell culture media and reagents were from Gibco (Stockholm, Sweden). All other chemicals were from Sigma Chemical Co., St. Louis.

Peptide synthesis
Peptides were synthesized in a stepwise manner in a 0.1 mmol scale on an Applied Biosystem Model 431A peptide synthesizer on solid support using dicyclohexyl-carbodiimide/hydroxybenzotriazole activation strategy. tert-Butyloxycarbonyl amino acids were coupled as hydroxybenzotriazole esters to a p-methylbenzylhydrylamine (MBHA)² resin (1.1 mmol of amino groups/g; Bachem) to obtain carboxy-terminally amidated peptides. Deprotection of the side chains from formyl and benzyl groups was carried out using the "low TFMSA" method (13). The protecting groups on histidine (DNP) were removed by treatment for 1 h at room temperature with 20% (v/v) thiophenol/DMF. For the synthesis of biotin containing transportan, N¹³-Fmoc-transportan-MBHA resin was synthesized in the conventional way. The N-Fmoc protecting groups were removed with 20% piperidine in DMF. Biotin was coupled manually by adding a threefold excess of HOBt and DCC-activated biotin in DMF to the peptide-resin. The coupling reaction was complete after 1 h at room temperature. The peptides were finally cleaved from the resin with liquid HF at 0°C for 30 min. Deprotection of the side chains, cleavage of the peptides, and purification on HPLC have been described in detail (14). The purity of the peptides was >99% as checked by HPLC analysis on Nucleosil 120–3 C₁₈ column (0.4 cm x 10 cm). The molecular mass of each synthetic peptide was determined on a Plasma Desorption Mass Spectrometer (Bioion 20, Applied Biosystems), and the calculated values were obtained in each case.

Preparation of ¹²⁵I-Tyr⁹, N¹³-biotinyl-transportan (¹²⁵I-biotinyl-transportan)
The iodination of N¹³-biotinyl-transportan (referred to as biotinyl-transportan) was carried out by the chloramine-T method (15) with small modifications. The reaction time was prolonged to 2 min, and an eightfold excess of biotinyl-transportan over Na¹²⁵I was used. The iodinated biotinyl-transportan was separated from free label by gel filtration on Sephadex G-25.

Cell cultures
Bowes' melanoma cells were cultivated in minimal essential medium (MEM), supplemented with 10% fetal calf serum, 2 mM glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. Rin m5F cells were grown in RPMI-1640 with 10% fetal calf serum and antibiotics; all other cell lines were grown in Dulbecco's modified Eagle medium. For binding experiments, the cells were washed with phosphate-buffered saline (PBS), scraped off, and harvested by centrifugation.

Delivery of ¹²⁵I-biotinyl-transportan into Bowes' melanoma cells
Detached and washed Bowes' cells were suspended in ice-cold phosphate-buffered Krebs-Ringer (KR) solution (10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, 137 mM NaCl, 2.68 mM KCl, 1.8 mM CaCl₂, mM MgCl₂, 1 g/l glucose, 2 g/l BSA, pH 7.4) and aliquoted. The concentration of ¹²⁵I-biotinyl-transportan was adjusted (5 to 500 nM) and the suspensions were realiquoted to about 100 000 cells per assay tube. The cells were incubated for various periods at 37°C in a shaking water bath. The cell-bound/internalized radioactivity was separated from the free label by centrifuging the cells through a solution of 40% dinonylphthalate in dibutylphthalate (v/v) for 15 s at 6000 g. The bottoms of the tubes were cut off and the radioactivity in the solutions and precipitates was counted. The loosely bound fraction of ¹²⁵I-biotinyl-transportan (not internalized or inserted into membranes) was removed by incubating the cells in ice-cold 0.2 M acetic acid, 0.5 M NaCl (pH 2.5) for 5 min before centrifuging.

Detection of biotinylated peptides in Bowes' cells
The cells used to determine the intracellular localization of the biotin labeled peptides were grown on round glass coverslips in 24-well plates to about 50% confluence. Serum containing medium was replaced by a serum-free one, and peptides were added directly into the medium. Biotinyl-transportan was added to concentrations ranging from 1 to 10 µM and biotinyl-mastoparan to 10–50 µM. The cells were incubated with peptides from 1 min to 4 h at 0°, 4°, or 37°C. The coverslips with cells were washed twice with warm PBS and fixed with 4% paraformaldehyde and 1% glutaric aldehyde in PBS. Fixed cells were permeabilized with cold methanol for 15 min at -20°C, and sites for nonspecific binding were blocked with 5% BSA in PBS. Samples incubated for less than 30 min were not fixed with aldehydes; instead, the cells were rinsed briefly with cold PBS, instantly fixed, and permeabilized with cold methanol. The biotinylated peptides were visualized by incubating the treated cells with streptavidin-FITC (1:100 dilution) or streptavidin-Texas red (1:50) for 1 h at room temperature. The cell nuclei were stained with Hoechst 33258 (0.5 µg/ml), and the nucleoli with acridine orange (0.3 µg/ml) in PBS. After rinsing with PBS, the preparates were mounted in a mixture of PBS and glycerol (1:3) containing 0.1% p-phenylenediamine. The stainings were examined and photographed by Vanox AH2-NAS microphotographic system (Olympos).

Confocal microscopy
The cells were prepared as usual. After rinsing with PBS, the preparates were mounted in a mixture of PBS and glycerol (1:3) containing 0.1% p-phenylenediamine. The stainings were examined and photographed under a Bio-Rad MRC-600 laser scanning confocal imaging system.

Treatment of Bowes' melanoma cells with phenylarsine oxide and hyperosmolar sucrose solution.
To block the endocytosis, Bowes' cells were treated with 80 µM phenylarsine oxide for 5 min at 37°C in serum-free medium. The cells were washed twice with PBS and incubated with biotinyl-transportan in MEM and processed as usual. The hyperosmolar solution used for avoiding endocytosis contained 0.45 M sucrose in MEM. The peptides were added directly into this medium.

Staining of intracellular structures in Bowes' melanoma cells
The detection of structural proteins in Bowes' cells was achieved by incubating fixed and permeabilized cells with monoclonal antibodies against ß-tubulin diluted 1:200 (TUB 2.1 Sigma), actin 1:200 (AC-40 Sigma), or pan-cytokeratin 1:400 (C-11 Sigma) in BSA containing PBS for 1 h, followed by incubation with anti-mouse antibody coupled to fluorescein (diluted 1:100) for 30 min. Membranes were stained with fluorescein-labeled concavalin A (2 µg/ml) for 30 min (16). The stainings were carried through before the detection of the localization of transportan with Texas red-conjugated streptavidin. All stainings were performed at room temperature.

Nonsaturability of transportan internalization
For assaying the possible saturability of the internalization of biotinyl-transportan, first a 10-fold excess of the competitor peptide was added into the cell medium. Galparan, galanin, and mastoparan were used at 0.1 mM concentration; biotinyl-transportan to a concentration of 10 µM was subsequently added. The internalization of biotinyl-transportan was followed by immunohistochemical staining as described above.

Trypsin resistance of internalized transportan
The assay for protection of internalized biotin-transportan against degradation by trypsin was performed analogously to Derossi et al. (17). About 1 million Bowes' cells were gently shaken for 1 h at 37°C with 50 µM biotin transportan in 0.5 ml PBS. The cells were washed three times with PBS, incubated in 0.5 ml of 0.25% trypsin solution for 30 min at 37°C, and washed again. The cell material was suspended in 0.1 ml of Schägger sample buffer and sonicated for 15 s. For comparison, 0.5 ml of 50 µM biotinyl-transportan was incubated with 0.25% trypsin prior to incubation with the cells.

Noninternalized peptide as negative control
N²⁵ Gly-Gly-biotinyl-galanin (1–29) was used as negative control. Cells were incubated with 50 µM control peptide in 0.5 ml PBS and then treated as previously described (see above).

Polyacrylamide gel electrophoresis, electroblotting, and immunodetection
Peptides at the concentrations indicated or extracted from cells were separated by using Tris-tricine-SDS (sodium dodecyl sulfate) polyacrylamide gel (18). The peptides and cell extracts were incubated for 30 min at 40°C in 4% (w/v) SDS 12% (w/v) glycerol, 50 mM Tris, 2% (v/v) 2-mercaptoethanol, 0.01% Serva blue G, adjusted to pH 6.8 with HCl. The separation was performed in urea containing 16.5% small pore gel, overlaid by a 10% spacer and a 4% stacking gel. Gels were fixed in 1% glutaric aldehyde for 1 h and visualized by Coomassie brilliant blue or silver staining (19).

The material was electrotransferred from unfixed gels to ProBlott membranes (Applied Biosystems, Sweden) in 10 mM CAPS 10% methanol pH 11.0 (20). The peptides were cross-linked on the blot with 1% glutaric aldehyde and stained with 0.1% Coomassie brilliant blue in 10% acetic acid and 40% methanol for 10 min. Destaining was performed in the same dye-free solution.

Immunodetection was carried out on the same membranes after removing all the stain by washing blots with several changes of methanol. The sites for nonspecific binding were blocked with 5% nonfat dry milk powder solution in 20 mM Tris, 150 mM NaCl, and 0.1% Tween 20, pH 7.5 (TBST). The blots were incubated with streptavidin-alkaline phosphatase for 1 h in 1% blocking solution in TBST, and the biotin-containing peptides were visualized with NBT/BCIP.

Modeling of internalization of ¹²⁵I-biotinyl-transportan
In the simulation of uptake and release of transportan by Bowes' cells, we considered degradation of the peptide in the cellular interior according to:

where A and B are ¹²⁵I-biotinyl-transportan outside and inside the cells, C and D are radioactive peptide fragment (or fragments) inside and outside the cells, and k's are the first-order rate constants for the corresponding processes.

To calculate the amount of the radioactive peptides in the cells (B + C) and to determine the corresponding rate constants, a numerical treatment of the experimental data was applied and a modified regression computer program of Stojan (see ref 21) was used with boundary conditions at t = 0: A = A₀, B = 0, C = 0, and D = 0. The curves for all five initial concentrations (A₀) of ¹²⁵I-biotinyl-transportan were fitted simultaneously to the obtained experimental points using proportional weighting. Consequently, values obtained for the rate constants are equally valid for all curves.

Binding experiments
Cells were lysed by suspending them in hypotonic ice-cold 5 mM HEPES buffer containing 2.5 mM MgCl₂, 0.5 mM EDTA (pH 7.3), and subsequently incubated for 10 min on ice. The resulting microsomal membrane fraction was collected by centrifugation at 10,000 g for 10 min at 4°C and weighed. The membranes were resuspended in HEPES-buffered Krebs-Ringer (HKR) solution, supplemented with 0.05% (w/v) bovine serum albumin and 0.1% (w/v) bacitracin, and homogenized with a glass-Teflon homogenizer. The equilibrium binding experiments were performed in a final volume of 400 µl HKR containing 50 pM [¹²⁵I]-galanin (NEN Research), Bowes' cell membranes, and displacer (22). Samples were incubated for 30 min at 37°C in a shaking water bath. The incubation was terminated by the addition of 2 x 10 ml ice-cold HKR, followed by a rapid filtration over Whatman GF/C glass fiber filters precoated for 2–3 h in 0.3% (v/v) polyethylenimine solution. The binding of ¹²⁵I-biotinyl-transportan was assayed in the presence of peptidase inhibitors (1 µM leupeptin, 1 µM pepstatin, 0.2 µM aprotinin, and 0.1 mM PMSF). The radioactivity retained on the filters was determined in a Packard gamma counter. The specific binding was determined as the part of the total binding that could be displaced with a large excess of unlabeled galanin (1 µM). The specific binding amounted to at least 90% of the total binding. Since there was no detectable difference in the binding affinities of either porcine or human galanin to the receptors, porcine galanin was used as displacer throughout the binding studies. Membrane protein concentration was determined according to Peterson (23). All tubes and tips used in binding experiments were coated with Sigmacote (Sigma).

Measurement of GTPase activity
Measurement of GTPase activity was performed radiometrically according to Cassel and Selinger (24), with the modifications suggested by McKenzie (25). To the diluted membranes was added an ice-cold reaction cocktail containing ATP (1 mM), 5'-adenylylimido-diphosphate (1 mM), ouabain (1 mM), phosphocreatine (10 mM), creatine phosphokinase (2.5 units/ml), dithiothreitol (4 mM), MgCl₂ (5 mM), NaCl (100 mM), and trace amounts of [-³²P]GTP to 50,000–100,000 cpm in an aliquot of the cocktail (with the addition of GTP to reach a total GTP concentration of 0.01–2 µM). The incubation medium was TE buffer (10 mM Tris-HCl, 0.1 mM EDTA), pH 7.5. Background low-affinity hydrolysis of [-³²P]GTP was assessed by incubating parallel tubes supplemented with 100 µM GTP. Blank values were determined by replacing the Bowes' cell membrane solution with assay buffer. Transportan was added to the assay tubes in concentrations ranging from 0.1 µM to 0.1 mM. The GTPase reaction was started by transferral of the assay tubes to a 30°C water bath in which they were subsequently incubated for 12 min. GTP was removed by a 5% suspension of activated charcoal in 20 mM H₃PO₄. The amount of radioactive phosphate produced was determined in an LKB 1214 Rackbeta or a Packard 3255 liquid scintillation counter. The basal GTPase activity of the Bowes' cellular membrane preparation was 22 pmol/(min/mg protein).

Partition of ¹²⁵I-biotinyl-transportan in an octanol–water system
¹²⁵I-Biotinyl-transportan was used to determine the partition coefficient of transportan in a water/octanol system. To water solutions ranging from 1.5 to 150 nM of ¹²⁵I-biotinyl-transportan was added an equal volume of n-octanol. After vigorous vortexing of the solutions, the phases were separated by centrifugation for 5 min at 10,000 g. All samples were prepared in triplicate, and three aliquots from each phase were counted.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Synthesis of ¹²⁵I-biotinyl-transportan
The biotinylated analogs of galparan and mastoparan were synthesized ( Table 1) in order to detect intracellular localization of the peptides. Biotinylated transportan was iodinated by the chloramine-T method, enabling us to follow the kinetics and yield of binding/internalization. The reaction time for iodination was prolonged (compared to the original method) in order to increase labeling efficacy. Incorporation of iodine, however, corresponded to only 5.1% of the total recovered radioactivity. The resulting specific activity was 12.6 Ci/mmol. The low yield can be explained by sterical hindrances (26) or oligomerization of the peptide in solution, which cause an additional decrease in reaction speed.

Delivery of ¹²⁵I-biotinyl-transportan into Bowes' melanoma cells
The time course of the uptake of ¹²⁵I-biotinyl-transportan by Bowes' melanoma cells is shown in Fig. 1. The internalization/binding process seems to be fast and efficient, since 58–80% of the maximal binding is achieved in the approximate interval of 3 min between adding ¹²⁵I-biotinyl-transportan to the suspension and the centrifugation used to separate free and bound label, during which time all solutions are kept on ice. The rate of uptake is comparable to insertion of the hydrophobic membrane probe TMA-DPH (1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene) into cell membranes. TMA-DPH can already be found in the outer leaflet of the cell membrane after 10 s at 37°C (27). The high speed of uptake is probably due to the positive charge in the structure, which quickly associates the probe with the negativly charged phospholipids of the cell membranes.

View larger version (17K): [in this window] [in a new window]   Figure 1. Uptake of 125I-biotinyl-transportan by Bowes' melanoma cells. The cells were incubated at 37°C with 125I-biotinyl-transportan in various concentrations: empty triangles = 500 nM, filled boxes = 167 nM, empty boxes = 50 nM, filled circles = 17 nM, and empty circles = 5 nM in phosphate-buffered Krebs-Ringer solution. Cell-bound peptide was separated from free peptide by centrifugation through `oil.' Points are experimental; curves were fitted with the aid of the modified regression program by J. Stojan.

Incubation of the cell suspension at 37°C with labeled transportan induces additional uptake for 15–25 min. This binding is strong, if not irreversible, since only a small fraction of the ¹²⁵I-labeled peptide is released from cells in acidic medium (data not shown). The fraction of the labeled peptide bound to the cells did not persist on the maximal level, but rather decreased over the period studied, accompanied by an increase of radioactivity in the cell supernatant. The decrease in bound radioactivity can be explained by active transport of the peptide from the cells or by release of radioactive degradation products.

¹²⁵I-Biotinyl-transportan is internalized at all concentrations under study (5–500 nM), and the time course of uptake is similar. The maximal uptake of labeled transportan depends slightly on the peptide concentration in the solution. The relative uptake is higher at lower concentrations: 16.3% of the total amount of peptide at 5 nM and 9.2% at 500 nM. Hence, the concentration of ¹²⁵I-biotinyl-transportan bound to or inside the cells is at least twofold higher than the concentration of free ligand, since the cells occupy about 2–5% of the suspension volume (visual estimation after centrifugation).

Localization of transportan in Bowes' melanoma cells
When Bowes' melanoma cells were incubated with biotinyl-transportan at 37°C, the peptide could be detected throughout the cell interior by indirect immunofluorescence ( Fig. 2A). Transportan was accumulated markedly in the plasma membrane and nuclear membrane, but all other intracellular membrane structures were heavily stained as well. The rate of internalization of transportan, followed by indirect immunofluorescence, was similar to the uptake of ¹²⁵I-labeled biotinyl-transportan (preceding text). The initial process of internalization was very fast: the cells were intensely stained after 1 min incubation at 37°C with 10 µM biotinyl-transportan. After the first 5 min, the peptide was localized mostly in the plasma membrane and cytosolic membranous structures (endosomes, endoplasmatic reticulum, Golgi). Staining of the nuclear membrane and nuclei was slight, but clearly visible. After 15–30 min, biotinyl-transportan was preferentially concentrated in the nuclear membrane and the nuclei ( Fig. 2B). The peptide was not spread evenly throughout the nucleus, but had a distinct localization. The fact that nuclear staining with acridine orange coincided with transportan staining suggests that the nucleoli were the intranuclear regions where transportan was concentrating. Transportan showed a cell-penetrating ability even at 0°C ( Fig. 2C). When Bowes' cells were incubated with a 10 µM solution of biotinyl-transportan for 30 min at 0°C, the peptide was detectable in all membranous structures, but mostly in the plasma membrane. All cells were stained, but not uniformly; some cells (about 20%) showed more efficient uptake after 30 min. Incubation of the cells with biotinyl-transportan for 1 to 2 h increased staining of the nuclear membrane. Either the peptide was not penetrating into the nuclei at 4 or 0°C ( Fig. 3A) or its concentration inside the nuclei remained below the detection limit of this method.

View larger version (361K): [in this window] [in a new window]   Figure 2. Internalization of biotinyl-transportan by Bowes' melanoma cells. The cells were incubated with 10 µM biotinyl-transportan at 37°C for 10 min (A), 30 min (B), or at 0°C for 30 min (C). The cells were fixed and incubated with avidin-TRITC (A, C) or streptavidin-FITC (B) for staining of transportan.

View larger version (155K): [in this window] [in a new window]   Figure 3. Confocal microphotographs showing internalization of biotinyl-transportan. Bowes' cells were incubated with 10 µM biotinyl-transportan at 0°C (A–C) or at 37°C (D–F) for 60 min. Transportan was visualized by streptavidin-Texas red (B, E) and membranes were stained with concavalin A-FITC (A, D); the combined image is shown in panels C, F.

The cell-penetrating property of biotinyl-transportan was not restricted to the Bowes' melanoma cell line, but is a general feature of this peptide, as demonstrated in other frequently used mammalian cell lines that originate from different tissues: HeLa, HEK 293, SAOS-1, CaSki, U937, COS-7, Jurkat, and Rin m5F. Biotinyl-transportan readily entered all cell lines studied. Localization of biotinyl-transportan is similar in all the cell lines mentioned above (data not shown).

To better characterize the cytosolic structures where transportan was localized, monoclonal antibodies against structural proteins were used. The intracellular localization of biotinyl-transportan did not coincide with the staining of actin, ß-tubulin, or cytokeratines (data not shown). We conclude that transportan was not bound to the cytoskeleton nor was it concentrated there after internalization.

Concavalin A conjugated with fluorochromes has been used to stain cellular membranous glycoproteins with a high mannose content. It is known to efficiently label the plasma and nuclear membrane, but other membranous structures are stained as well (16). The interior of the nucleus is not immunoreactive to this lectin as judged by confocal fluorescence microscopy ( Fig. 3E). After incubation with the peptide at 4°C for 30 min, the distribution of biotinyl-transportan in Bowes' melanoma cells is almost identical to concavalin A-FITC staining ( Fig. 3A). Plasma and the nuclear membranes were stained intensely; cytosol was labeled to a lesser extent, and the nuclear interior was stained very faintly, if at all. The combined image shows a nearly overlapping distribution of biotinyl-transportan and concavalin A-FITC in Bowes' melanoma cells ( Fig. 3C).

The distribution of biotinyl-transportan in Bowes' cells after incubation at 37°C ( Fig. 3D) was not limited to membranous structures ( Fig. 3E). The intranuclear concentration of biotinyl-transportan seemed to be relatively high, since the nuclei stained intensely red on the combined image ( Fig. 3F). The intranuclear localization of transportan coincided exactly with DNA staining by Hoechst 33258 (not shown). Fluorescence confocal microscopy data confirm distribution of transportan throughout the cell, including the nuclei.

Nonsaturability of the internalization of transportan
The uptake of biotinyl-transportan was not saturable in the concentration range we used (5–50 µM). Internalization was not blocked by pre- or coincubation with a 10-fold excess of galparan, mastoparan, or galanin. Biotinyl-transportan is judged to enter all Bowes' melanoma cells because they were heavily stained in the indirect immunofluorescence assay. When the cells were treated with a mixture of 10 µM biotinyl-transportan and 100 µM galparan for a shorter period (5–15 min) at 37°C, the staining was more uniform than with an incubation with only biotinyl-transportan. This can be explained by a requirement for multimerization of biotinyl-transportan and/or galparan for efficient internalization. The uptake of ¹²⁵I-biotinyl-transportan in a 50 nM solution was not blocked by 100 µM galparan or 100 µM mastoparan (not shown).

Transportan is not internalized by receptor-mediated endocytosis
Endocytosis can be inhibited by a hyperosmolar solution of sucrose, which blocks the formation of clathrin-coated pits (28), or by phenylarsine oxide treatment, which cross-links the thiol groups of membrane surface proteins (29). Incubation of Bowes' melanoma cells in 0.5 M sucrose containing medium caused subtle changes in the cell volume; the cells were more rounded, but internalization of biotinyl-transportan was not blocked. Pretreatment of Bowes' cells with phenylarsine oxide caused a fraction of the cells to detach from the surface; nevertheless, the cells that remained on the coverslips took up biotinyl-transportan efficiently. Transportan was not internalized by a receptor-mediated mechanism, since the internalization of galanin receptors in Bowes' cell in responce to galanin is abolished in phenylarsine oxid-treated cells or in hyperosmolar solution (M. Pooga, unpublished observation). Moreover, endocytosis can be excluded at temperatures under 18°C; transportan, however, is internalized under all these conditions.

Cell-internalized biotinyl-transportan is protected against degradation by trypsin
In solution, biotinyl-transportan molecules are organized as an ensemble of different multimers, as judged by polyacrylamide electrophoresis. The prevailing form is dimer, but monomers and higher oligomers are also present ( Fig. 4, lane 1) to a relatively high extent. This could explain the difficulties in ¹²⁵I labeling of biotinyl-transportan. However, the SDS-induced multimerization of peptide in sample buffer, proposed for the Antennapedia (43–58) peptide by Derossi et al. (17), cannot be excluded. Biotin-transportan is efficiently taken up by Bowes' cells and is detectable in cell lysates by Coomassie brilliant blue staining ( Fig. 4, A2) and by the biotinyl moiety ( Fig. 4, B2). The dimeric form of biotinyl-transportan is dominating, but monomers as well as larger oligomers are present. Internalized biotinyl-transportan is protected against degradation by trypsin ( Fig. 4, lane 2), whereas preincubation of an equal amount of the peptide with trypsin under identical conditions, before incubation with cells, completely abolishes any uptake ( Fig. 4, lane 3). Trypsin solution (0.25%) completely digests 50 µM biotinyl-transportan in 30 min at 37°C. No peptide was detected when this solution was analyzed by gel electrophoresis; a silver-satined or identical blot was developed with streptavidin-alkaline phosphatase (not shown). Under identical conditions, peptides that are internalized by endocytosis cannot be detected in cells. We used N²⁵ Gly-Gly-biotinyl-galanin (1–29) and were not able to visualize the peptide in cell lysates by gel electrophoresis or immunoblot (not shown).

View larger version (84K): [in this window] [in a new window]   Figure 4. SDS-PAGE gel and blot demonstrating that internalized peptides are protected against degradation by trypsin. Bowes cells were incubated with 50 µM biotin transportan in 0.5 ml PBS for 1 h at 37°C and analyzed by electrophoresis and electroblotting. The peptides were visualized by Coomassie brilliant blue (A) and alkaline phosphatase staining (B). Lane 1 shows biotinyl transportan directly applied to the gel. Lane 2 shows Bowes cells incubated with biotin-transportan and treated with trypsin for 30 min at 37°C. Lane 3 shows Bowes cells incubated with trypsin-treated biotin-transportan. M: molecular weight marker.

Modeling of the internalization of ¹²⁵I-biotinyl-transportan
We calculated the characteristic rate constants from scheme 1 in order to describe the uptake and release of biotinyl-transportan and its degradation products by Bowes' melanoma cells. The rate constants are as follows: k₁ = 0.019 min^-1, k₂ = 0.15 min^-1, k₃ = 0.058 min^-1, k₄ = 0.27 min^-1 and k₅ = 0.0039 min^-1. The ratio between constants for uptake and release and the ratio of volumes outside and inside the cells are in accordance with a maximal accumulation of ¹²⁵I-biotinyl-transportan in the cells. The ratio of the constants for uptake and release of the degradation products corresponds to the ratio of intracellular (2–5 µl) and extracellular (100 µl) volumes. This suggests that the degradation products do not accumulate in the membranes and that transportan could be degraded in the cells.

Biotinyl-transportan and biotinyl-mastoparan as ligands to human galanin receptors
The effects of the biotin-containing analogs of transportan and mastoparan on the binding of ¹²⁵I-galanin to galanin receptors in Bowes' melanoma cellular membranes are presented in Table 1. The association of the respective biotinylated peptides with human galanin receptors is fairly close to the results reported with rat galanin receptors (1). The K_D values for galanin, biotinyl-transportan, and biotinyl-mastoparan in Bowes' cell membranes are 0.53 nM, 17.4 nM, and 10 µM, respectively. The Hill coefficients are close to unity for all three peptides. We have characterized the binding of ¹²⁵I-biotinyl-transportan to Bowes' cell membranes by using nonlabeled biotinyl-transportan and biotinyl-mastoparan as displacers. The K_D values—31.6 nM for biotinyl-transportan and 3.6 µM for biotinyl-mastoparan—are similar to their respective binding characteristics at galanin receptors when measured with ¹²⁵I-galanin. The K_D value for biotinyl-transportan is 1.8-fold higher, and for biotinyl-mastoparan 2.9-fold lower, with ¹²⁵I-transportan as ligand. This could reflect binding of biotinyl-transportan to the galanin receptors in Bowes' cell membranes. When ¹²⁵I-biotinyl-transportan was incubated with Bowes' cell membranes in a buffer that contained only bacitracin as peptidase inhibitor, no binding to membranes was detected at low transportan concentrations, probably due to degradation of the labeled ligand. We assume that mastoparan influences the binding of ¹²⁵I-galanin to galanin receptors by some mechanism that involves G-protein receptor interaction. The mechanisms of this phenomenon, however, are not clear.

Inhibition of GTPase activity by transportan
The effect of biotinyl-transportan on the GTPase activity of Bowes' melanoma cell membranes is shown in Fig. 5. Unlike mastoparan, which increases GTPase activity (4), biotinyl transportan inhibits basal GTPase activity in Bowes' cell membranes. Fitting the data with a sigmoidal dose-response curve results in the following parameters: maximal inhibition of the basal activity 21.7 ± 8.3%, EC₅₀ 21.1 ± 1.2 µM, Hill coefficient 2.2 ± 0.6. The inhibitory effect of biotinyl-transportan on the activity of GTPases in membrane preparation is probably caused by a direct interaction of the peptide with the enzyme or by changes in the structure/properties of the membranes, since transportan is located predominantly in the membranous structures of the cell.

View larger version (15K): [in this window] [in a new window]   Figure 5. The effect of biotinyl-transportan on the GTPase activity of Bowes' melanoma cell membranes.

Partition of ¹²⁵I-biotinyl-transportan between water and n-octanol
The concentration of ¹²⁵I-biotinyl-transportan in water was 15.1 ± 1.1 times higher than in octanol. The relative distribution of ¹²⁵I-biotinyl-transportan between water and n-octanol did not depend on the concentration of the peptide in solution. The finding that the concentration of biotinyl-transportan in water was higher than in n-octanol does not contradict the idea that transportan is a membrane-specific peptide. The partition coefficient for ¹²⁵I-biotinyl-transportan is sixfold higher than for the ¹²⁵I^- ion from Na¹²⁵I, showing that the peptide may enter membranes. Interpretation of these results should be made with caution, because the partition between octanol/water in the model system does not describe the details of a peptide/phopholipids interaction.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
We have previously characterized some physiological effects of the peptide galparan, consisting of the smallest fragment of the neuropeptide galanin with high (submicromolar) affinity, galanin (1–13), and mastoparan. Transportan ([Lys¹³]galparan, or galparan substituted with lysin in position 13) was designed to study the cellular uptake and distribution of this type of peptide. The uptake of transportan is rapid and efficient in all cell lines tested. In Bowes' melanoma cells, nearly half of the maximal labeling is obtained in the course of mixing ¹²⁵I-biotinyl-transportan with the cells for about 3 min at 0°C. Immunochemical labeling experiments confirm this result. Exposure of Bowes' melanoma cells to biotinyl-transportan for 1 min at 37°C is sufficient to detect the peptide in the plasma membranes and cell interior. Rapid binding of biotinyl-transportan to cells may be induced by the positive charges in the molecule, concentrated in the carboxy-terminal part of the peptide (Lys^13,17,24,25) (27).

Trypsin treatment of transportan-loaded cells, followed by separation of the cell lysate on an SDS-polyacrylamide gel electrophoresis (PAGE), shows that the peptide is protected from degradation and can be found intact in the cellular interior. Transportan pretreated with trypsin cannot be found in cell lysates under the same conditions. This is another indication that transportan truly is internalized and not adsorbed to the membranes.

Internalization of biotinyl-transportan can follow the mechanism proposed for the peptide corresponding to the third -helix of the Drosophila homeoprotein Antennapedia (30). In the Antennapedia fragment, the presence of two (17), or at least one, tryptophan residue (30) is shown to be necessary for internalization. For the Antennapedia-derived peptide, formation of an inverted micelle is proposed as a mechanism for cell entry (30). Transportan contains Trp² and Tyr⁹, which can induce the formation of inverted micelles.

The maximal intracellular concentration of ¹²⁵I-biotinyl-transportan is reached in 15–25 min at 37°C. The very fast uptake is followed by a slow decrease in the amount of cell-bound radioactivity. In parallel, the fraction of radioactivity outside the cells starts to increase. We cannot observe by immunochemical staining any decrease in label in the cells after 4 h. The decrease in the fraction of radioactivity bound to the cells can be caused by cell lysis or pore formation by transportan, followed by outflow of the internalized radioactivity. We cannot observe any lysis of cells in the immunofluorescence experiment. This could be because the immunofluorescence method is not sensitive enough to detect subtle differences in fluorescence intensity and cell shape. We interpret the decrease in the fraction of radioactivity bound to cells as an outflow of peptide fragments after degradation of the peptide because ¹²⁵I-biotinyl-transportan is degraded when incubated with membranes. It has been demonstrated that the peptide bond between amino acids Gly¹² and Pro¹³ in galanin is cleaved by membrane-bound proteases in spinal cord material (31). Cleavage of the 12th peptide bond between Gly¹² and Lys¹³ leaves the ¹²⁵I-label on Tyr⁹ and the biotin label on Lys¹³ on different peptide fragments, which could explain our results.

Analogous to the 'parent' compound galparan, biotinylated transportan is recognized by galanin receptors in Bowes' cell membranes with relatively high affinity. Galparan displaces ¹²⁵I-galanin from galanin receptors in rat brain cortical membranes (1), and biotinyl-transportan shows comparable binding to galanin receptors in Bowes' melanoma cellular membranes. The K_D value for binding of biotinyl-transportan to the receptors in Bowes' cell membranes is 2.7-fold higher than that for binding of galparan to the receptors in rat cortical membranes. This slight difference may be caused by Pro¹³ to Lys-biotinyl modification or by subtle differences in receptor structures and surroundings that exist in different tissues and species.

Unlike mastoparan, transportan inhibits total basal GTPase activity, possibly by preventing docking of G-proteins to the receptors. Characterization of the mechanisms behind this behavior is in progress (M. Zorko et al., unpublished results).

We suggest that at concentrations of 1 µM and higher, transportan binds to cellular membranes in the form of micelles or oligomers. This interpretation is supported by the following observations. 1) Labeling of transportan with Na¹²⁵I was inefficient, even when the reaction time was prolonged. 2) The degradation of ¹²⁵I-biotinyl-transportan is inhibited by transportan and galparan, but not by galanin or mastoparan. 3) Immunohistochemical staining shows that biotinyl-transportan is located in membranous structures. 4) Transportan moves predominantly as a dimer or a tetramer on SDS-PAGE.

We suggest here that transportan is internalized by a receptor-independent mechanism, since it has been observed inside every cell line we have studied. The cellular penetration is not blocked by an excess of either unlabeled transportan, galparan, galanin, or mastoparan. N²⁵ Gly-Gly-biotinyl-galanin (1–29) is undetectable in the cell under the same conditions where transportan is clearly detectable. The receptor internalization probably is too slow to generate a detectable intracellular concentration. Receptor-mediated endocytosis can generally be excluded at temperatures below 18°C; in hyperosmolar solutions or by cross-linking the proteins on the cell surface, the application of these methods has not prevented the internalization of biotinyl-transportan (see Results).

We suggest that transportan first enters the outer leaflet of the plasma membrane and then spreads throughout cellular membranous structures. The intracellular distribution of transportan coincides with the staining of high mannose-containing membrane proteins by FITC-conjugated concavalin A. We have not yet defined the localization of transportan in a more detailed manner (e.g., using antibodies against specific proteins in membranous cell compartments). In longer incubations (starting with 1 h), transportan concentrates in the nuclear membrane and nuclei, where it localizes predominantly in the nucleoli. The mechanisms and forces causing transportan to enter the cell and concentrate in the nuclei are not yet understood, but can be analogous to those proposed for Drosophila's Antennapedia-derived peptide. The peptide corresponding to the third -helix of Antennapedia is spread evenly throughout the cell nucleus in primary neurons in culture (32). Transportan concentrates predominantly in the nucleoli in Bowes' cells, but we suggest that the forces that convey these different peptides into nuclei can be similar, if not identical.

In conclusion, transportan is penetrating into every cell type examined in a rapid and efficient way. In the cell interior, it localizes mostly in membranous structures and is conveyed into the nuclei, where it concentrates in the nucleoli. The internalization of transportan is receptor/protein independent and seems to be a general feature of this peptide. In moderate concentrations it does not affect the growth of cells in the culture. Transportan is degraded or expelled from the cells when the cells are loaded with transportan and grown for several days. Together, these features make transportan a potentially useful peptide vector for the introduction of different hydrophilic macromolecules into the cell interior in order to regulate finely tuned cellular processes.

	ACKNOWLEDGMENTS

 
We thank Zhi-Qing Xu and Prof. Thomas Hökfelt from Karolinska Institute, Stockholm, Sweden, for help with confocal microphotographs, Anne Mardimäe from Estonian Biocentre, Peeter Kukk from Tartu University, Tartu, Estonia, for help in preparing figures, and Jure Stojan University of Ljubljana, Slovenia, for his help with the curve fitting and modeling. This work is supported by the Swedish Research Council for Engineering Sciences (T.F.R.), the Royal Swedish Academy (stipend to M.P.), and the Wenner-Gren's Foundation (stipend to M.Z.).

	FOOTNOTES

 
¹ Correspondence: Department of Neurochemistry and Neurotoxicology, Arrhenius Laboratories, Stockholm University, S-106 91 Stockholm, Sweden. E-mail: ulo{at}neurochem.su.se

² Abbreviatins: MBHA, p-methylbenzylhydrylamine; ¹²⁵I-biotinyl-transportan, ¹²⁵I-Tyr⁹, N¹³-biotinyl-transportan; biotinyl-transportan, N¹³-biotinyl-transportan; MEM, minimal essential medium; KR, phosphate-buffered Krebs-Ringer; HKR, HEPES-buffered Krebs-Ringer; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate; TBST, 20 mM Tris, 150 mM NaCl, and 0.1% Tween 20, pH 7.5; TMA-DPH, (1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene; PAGE, polyacrylamide electrophoresis gel.

Received for publication June 5, 1997. Accepted for publication October 13, 1997.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Langel, Ü., Pooga, M., Kairane, C., Zilmer, M., and Bartfai, T. (1996) A galanin-mastoparan chimeric peptide activates the Na⁺,K⁺-ATPase and reverses its inhibition by ouabain. Regul. Peptides 62, 47–52[Medline]
2. Land, T., Langel Ü., Löw, M., Berthold M., Undén A., and Bartfai T. (1991) Linear and cyclic N-terminal galanin fragments and analogs as ligands at the hypothalamic galanin receptor. Int. J. Peptide Protein Res. 38, 267–272[Medline]
3. Juréus, A., and Langel, Ü. (1996) Galanin and galanin antagonists. Acta Chim. Slovenica 43, 51–60
4. Higashijima, T., Uzu, S., Nakajima, T., and Ross, E. M. (1988) Mastoparan, a peptide toxin from wasp venom, mimics receptors by activating GTP-binding regulatory proteins (G proteins). J. Biol. Chem. 263, 6491–6494[Abstract/Free Full Text]
5. Tanimura, A., Matsumoto, Y., and Tojyo, Y. (1991) Mastoparan increases membrane permeability in rat parotid cells independently of action on G-proteins. Biochem. Biophys. Res. Commun. 177, 802–808[Medline]
6. Danilenko, M., Worland, P., Carlson, B., Sausville, E. A., and Sharoni, Y. (1993) Selective effects of mastoparan analogs: separation of G-protein-directed and membrane-perturbing activities. Biochem. Biophys. Res. Commun. 196. 1296–1302[Medline]
7. Matsuzaki, K., Yoneyama, S., Murase, O., and Miyajima, K. (1996) Transbilayer transport of ions and lipids coupled with mastoparan X translocation. Biochemistry 35, 8450–8456[Medline]
8. Weidman, P. J., and Winter, W. M. (1994) The G protein-activating peptide, mastoparan, and the synthetic NH₂-terminal ARF peptide, ARFpl3, inhibit in vitro golgi transport by irreversibly damaging membranes. J. Cell Biol. 127, 1815–1827[Abstract/Free Full Text]
9. Pfeiffer, D. R., Gudz, T. I., Novgorodov, S. A., and Erdahl, W. L. (1995) The peptide mastoparan is a potent facilitator of the mitochondrial permeability transition. J. Biol. Chem. 270, 4923–4932[Abstract/Free Full Text]
10. Ahrén, B., Arkhammar, P., Berggren, P.-O., and Nilsson, T. (1986) Galanin inhibits glucose-stimulated insulin release by a mechanism involving hyperpolarization and lowering of cytoplasmic free Ca²⁺ concentration. Biochem. Biophys. Res. Commun. 140, 1059–1063[Medline]
11. Östensson, C.-G., Zaitsev, S., Berggren, P.-O., Efendic, S., Langel, Ü., and Bartfai, T. (1997) Galparan, a powerful insulin releasing chimeric peptide acting at a novel site. Endocrinology 138, 3308–3313[Abstract/Free Full Text]
12. Consolo, S., Baldi, G., Nannini, L., Uboldi, M. C., Pooga, M., Langel, Ü., and Bartfai, T. (1997) Galparan induces in vivo acetylcholine release in the frontal cortex. Brain Res. 756, 174–178[Medline]
13. Tam, J. P., Heath, W. F., and Merrifield, R. B. (1985) Mechanisms for the removal of benzylprotecting groups in synthetic peptides by trifluoromethanesulfonic acid-trifluoroacetic acid-dimethylsulfid. J. Am. Chem. Soc. 108, 5242–5251
14. Langel, Ü., Land, T., and Bartfai, T. (1992) Design of chimeric peptide ligands to galanin receptors and substance P receptors. Int. J. Pept. Protein Res. 39, 516–522[Medline]
15. Kask, K., Berthold, M., Bourne, J., Andell, S., Langel, Ü., and Bartfai, T. (1995) Binding and agonist/antagonist actions of M35, galanin(1–13)-bradykinin(2–9)amide chimeric peptide, in Rin m 5F insulinoma cells. Regul. Peptides 59, 341–348[Medline]
16. Söderqvist, H., and Hallberg, E. (1994) The large C-terminal region of the integral pore membrane protein, POM121, is facing the nuclear pore complex. Eur. J. Cell Biol. 64, 186–191[Medline]
17. Derossi, D., Joliot, A. H., Chassaing, G., and Prochiantz, A. (1994) The third helix of the Antennapedia homeodomain translocates through biological membranes. J. Biol. Chem. 269, 10444–10450[Abstract/Free Full Text]
18. Schägger, H., and von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379[Medline]
19. Merril, C. R., Goldman, D., Sedman, S., and Ebert, M. H. (1981) Ultrasensitive stain for proteins in polyacrylamide gels shows regional variation in cerebrospinal fluid proteins. Science 211, 1437–1438[Abstract/Free Full Text]
20. Matsudaira, P. (1987) Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membranes. J. Biol. Chem. 262, 10035–10038[Abstract/Free Full Text]
21. Stojan, J. (1997) Analysis of progress curves in an acetylcholine-esterase reaction: a numerical integration treatment. J. Chem. Inf. Comput. Sci. In Press
22. Kask, K., Berthold, M., Kahl, U., Nordvall, G., and Bartfai, T. (1996) Delineation of the peptide binding site of the human galanin receptor EMBO J. 15, 236–244
23. Peterson, G. L. (1977) A simplification of the protein assay Mmthod of Lowry et al. Which is more generally applicable? Anal. Biochem. 83, 346–356[Medline]
24. Cassel, D., and Selinger, Z. (1976) Catecholamine stimulated GTPase activity in turkey erythrocyte membranes. Biochim. Biophys. Acta 452. 538–551[Medline]
25. McKenzie, F. R. (1992) Basic tecniques to study G-protein function. In Signal Transduction. A Practical Approach (Milligan, G., ed) pp. 31–56, Oxford University Press, New York
26. Land, T., Langel, Ü., Fisone, G., Bedecs, K., and Bartfai, T. (1991) Assay for galanin receptor. Methods Neurosci. 5, 225–234
27. Illinger, D., and Kuhry, J. G. (1994) The kinetic aspects of intracellular fluorescence labeling with TMA-DPH support the maturation model for endocytosis in L929 cells. J. Cell Biol. 125, 783–794[Abstract/Free Full Text]
28. Heuser, J. E., and Anderson, R. G. (1989) Hypertonic media inhibit receptor-mediated endocytosis by blocking clathrin-coated pit formation. J. Cell Biol. 108, 389–400[Abstract/Free Full Text]
29. Frost, S. C., and Lane, M. D. (1985) Evidence for the involvement of vicinal sulfhydryl groups in insulin-activated hexose transport by 3T3-L1 adipocytes. J. Biol. Chem. 260, 2646–2652[Abstract/Free Full Text]
30. 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[Abstract/Free Full Text]
31. Bedecs, K., Langel, Ü., and Bartfai, T. (1995) Metabolism of galanin and galanin(1–16) in isolated cerebrospinal fluid and spinal cord from rat. Neuropeptides 29, 137–143[Medline]
32. Prochiantz, A., and Theodore, L. (1995) Nuclear/growth factors. Bioessays 17, 39–44[Medline]