Enhanced Cellular Entry and Efficacy of Tat Conjugates by Rational Design of the Auxiliary Segment

Abstract

Conjugation with a cell penetrating peptide such as Tat presents an effective approach to improve the intracellular accumulation of molecules with low membrane permeability. The same strategy, unfortunately, leads to a reduced cellular entry of molecules that can cross cell membrane effectively. We report here that covalent linkage of an additional hydrophobic unit that mimics a hydrophobic domain near the Tat sequence to the resulting Tat conjugates can further improve the cellular uptake of the conjugate into cancer cells regardless of the membrane permeability of the unconjugated molecule. Both fluorescent imaging and flow cytometry measurements confirmed the effect of palmitoylation on the increased internalization of the Tat conjugates with either 5-carboxyfluorescein (5-FAM), a non-membrane penetrating dye, or doxorubicin, an anticancer cancer drug that can readily diffuse across cell membranes. In the case of the Tat-doxorubicin conjugate, palmitoylation improves the conjugate’s anticancer activity in both drug sensitive and resistant cervical cancer cell lines. We further demonstrate that modification of a Tat-5-FAM conjugate with a hydrophobic quencher could not only efficiently quench the fluorescence outside of cancer cell but also facilitate its entry into MCF-7 breast cancer cells. These results highlight the importance of rational molecular design of using peptide conjugation chemistry in cancer therapeutics and diagnostics.

Introduction

Therapeutic or diagnostic agents with intracellular targets must gain access into cells to produce the desired biological effects. The importance of cellular internalization is highlighted by the observation that ~50% of known drug targets¹ and almost all oligonucleotides’ targets are located in the cells². In the particular case of cancer treatments, almost all the targets (DNA, enzymes, cytoskeletons, hormone receptors, etc.) of approved chemotherapeutics and all the potential targets for cancer gene therapy are found within the cells.³ However, a large portion of therapeutic agents do not possess the ability to effectively cross cell membranes, or have difficulty reaching a therapeutic dose within cells due to various drug resistance mechanisms. For instance, peptides, proteins and oligonucleotides are known to have low membrane permeability, and as a result they are considered to be of limited therapeutic value unless this issue can be addressed.⁴ Some small molecular drug candidates, for a number of reasons, also lack the ability to attain sufficient intracellular accumulation.⁵ There is also an increasing interest in monitoring disease-related dynamic changes in intracellular signaling, regulation and metabolism networks, for the altered expression levels of intracellular biomacromolecules in cancer cells relative to normal cells could be targets for potential disease prevention, diagnosis and treatment.^6,7 Therefore, it is important and necessary to design strategies to improve the intracellular accumulation of molecules to reach the desired therapeutic levels.

Several versatile methods have been developed to increase the intracellular accumulation of cargos, including electroporation,⁸ microfluidics,⁹ nanocarrier encapsulation^10–14 and conjugation,^15–17 and cell penetrating peptides (CPP) modification.⁴ Among these approaches, CPP modification shows the broadest application with its ability to aid the intracellular delivery of cargos ranging from small molecules to nanoparticles both in vitro and in vivo.¹⁸ The HIV-derived Tat^48–60 peptide (GRK₂R₂QR₃P₂Q from Tat protein)¹⁹ is one of the most investigated CPPs, and has been conjugated to many different types of cargos (e.g. nanoparticles) to enhance their intracellular accumulation since its discovery.^20,21 However, we recently found that the intracellular accumulation of doxorubicin, an anticancer drug with reasonably good cell membrane permeability, in drug-sensitive cells was actually reduced in a significant way upon conjugate to Tat peptide, although advantages were observed in drug-resistant cells.²² Similar observations have been reported when comparing cellular uptake of doxorubicin and doxorubicin-conjugates on other drug-sensitive cell lines.²³

The interaction between Tat conjugates and targeting cells surface is apparently crucial for the cellular entry. It is widely accepted that the internalization of Tat (or Tat-cargos) is initialized through electrostatic interactions with cell membrane components such as heparan sulfate proteoglycans,²⁴ and involves cytoskeleton rearrangement during internalization.²⁵ However, our recent work on Tat conjugates containing different chemical moieties confirms that factors other than electrostatic interactions must also contribute to the cell penetrating efficiency.^22,26,27 Indeed, there have been several reports in the literature on the effect of amphiphilic or hydrophobic peptide sequences and lipid modification to (PRR)^3,28 octaarginine (R8),^29,30 peptide analogues,³¹ peptide amphiphiles^32,33 and other CPPs^34,35 on their intracellular delivery efficiency of poor cell penetrating molecules. A recent finding suggested the hydrophobic domain of the Tat protein (a36–47) could complex with lipid post its encounter with lipid bilayer,³⁶ further highlighting the importance of hydrophobic interaction during cellular entry. In this manuscript, we report that palmitoylation to Tat conjugates facilitates efficient intracellular accumulation of both poor and good cell penetrating molecules, and eliminates the difference in intracellular accumulation and efficacy of doxorubicin due to cancer cell heterogeneity. More importantly, we further demonstrate the extension of the palmitoylation strategy to the design of molecular beacon for intracellular enzyme detection by replacing the palmitoyl tail with a hydrophobic quencher.

Materials and Methods

Materials

All Fmoc amino acids were purchased from Advanced Automated Peptide Protein Technologies (AAPPTEC, Louisville, USA) and Rink Amide MBHA and Wang resins were purchased from Novabiochem (San Diego, CA). 5-carboxyfluorescein (5-FAM) was purchased from AnaSpec, Inc. (Fremont, CA). Vinblastine and doxorubicin (Dox) were purchased from Avachem Scientific LLC (San Antonio, TX). Palmitic acid was obtained from Sigmal-Aldrich (St. Louis, MO). 3-Maleimido-propionic acid was purchased from Bachem (Torrance, CA). All other reagents were of analytical grade and used as received without further purification. C₈-Tat, NTF, TF, TFB and NTD were synthesized and characterized as we reported previously.^22,26

Cell culture

MCF-7 breast cancer cells were kindly provided by Dr. Wirtz (Department of Chemical and Biomolecular Engineering, Johns Hopkins University). KB-3-1 and KB-V1 cervical cancer cells were gifted by Dr. Gottesman (Center for Cancer Research, National Cancer Institute). All the cells were grown in DMEM (Invitrogen) containing 10% fetal bovine serum (FBS, Invitrogen) and 1% of antibiotics (Invitrogen) at 37 °C in a humidified incubator (OASIS 6300, CARON, OH) with a 5% CO₂ atmosphere. For KB-V1 cells, 1 µg/mL of vinblastine was added to maintain its multidrug resistance.

Tat conjugates synthesis and characterization

All peptide conjugates were synthesized using standard 9-fluorenylmethoxycarbonyl (Fmoc) solid phase synthesis techniques. For C₁₆NTF and palmitic acid-modified cleavable Tat-doxorubicin conjugate (C₁₆NTD), Fmoc-CGRK₂R₂QR₃P₂Q-Rink, Fmoc-K(Mtt)CGRK₂R₂QR₃P₂Q-Rink and maleimide-modified enzyme degradable peptide-drug conjugates (Mal-GFLG-Dox) were synthesized according to our previous report.²² For C₁₆NTF, Mtt deprotections were carried out using a mixture of TFA/TIS/DCM with a ratio of 3:5:92 for 5 min, repeating twice, and then the Lys ε-amine was reacted with C₁₆/HBTU/DIEA at a ratio of 4:4:6 relative to the peptide, shaking overnight at room temperature. 5-FAM was manually coupled at the peptide N-terminus (after Fmoc removal) with 5-FAM/HBTU/DIEA at a ratio of 4:4:6 relative to the peptide, shaking overnight at room temperature. For C₁₆NTD, the N-terminal of the peptide was reacted with C₁₆/HBTU/DIEA at a ratio of 4:4:6 relative to the peptide, shaking overnight at room temperature. Mal-GFLG-Dox was then reacted with Ac-CGRK₂R₂QR₃P₂Q-NH₂ to obtain C₁₆NTD according to the procedure described in our previous work.²² The two conjugates were purified by preparative RP-HPLC using a Varian ProStar Model 325 HPLC (Agilent Technologies, Santa Clara, CA) equipped with a fraction collector. Separations were performed using a Varian PLRP-S column (100 Å, 10 µm, 150 × 25 mm) monitoring at 480 nm. Collected fractions were analyzed by ESI-MS (LDQ Deca ion-trap mass spectrometer, Thermo Finnigan, USA) and those containing the target molecules only were combined and lyophilized (FreeZone −105 °C, Labconco, Kansas City, MO), and then stored at −30 °C. The purity of two conjugates was determined by analytical HPLC using the same method we described before.²²

Circular Dichroism (CD) measurement

The CD spectra of TF, TFB, NTF, C₁₆NTF, NTD and C₁₆NTD (50 µM in Dulbecco’s Phosphate-Buffered Saline, DPBS) were recorded on a J-710 spectropolarimeter (JASCO, Easton, MD) from 195 nm to 260 nm, and the signal was converted from ellipticity (mdegs) to mean molar ellipticity per residue (deg·cm²·dmol⁻¹·residue⁻¹).

Cellular uptake

MCF-7, KB-3-1 or KB-V1 cells were seeded onto a 24-well plate at 1 × 10⁵ cells/well, and incubated overnight. The media were replaced with fresh medium containing free fluorophore (5-FAM), free drug (Dox), or conjugates (TF, TFB, NTF, C₁₆NTF, NTD and C₁₆NTD) at a concentration of 5 µM, and were incubated with cells for 2 h. If necessary, caprylic acid-modified Tat (C₈-Tat) of various concentrations was added to evaluate the effect of free Tat peptide on the internalization of tested conjugates. The retention experiment was performed according to our previous protocol.²² Briefly, the cells were incubated with 5 µM Dox, NTD or C₁₆NTD for 2 h followed by additional 2 h incubation in drug-free medium. For quantitative evaluation, the cells were washed once with fresh medium, trypsinized, washed twice with DPBS, and the fluorescence intensity of the cells was determined using a FACScalibur flow cytometer (BD Biosciences, San Jose, CA) using the FL1 (530/30) and FL3 (670LP) channels for 5-FAM and doxorubicin, respectively. For qualitative evaluation, the cells were washed 3 times with phenol red free medium (containing 10% FBS), and imaged using an epifluorescence microscope (JENCO, Portland, OR).

Colocalization

KB-3-1 or KB-V1 cells were seeded onto a collagen-pretreated 8-well glass-bottom plate at 4 × 10⁴ cells/well and incubated overnight. The cells were then incubated with fresh medium containing 5 µM Dox, NTD or C₁₆NTD for 2 h. Thirty minutes before imaging, Hoechst 33342 (10 µg/mL, Invitrogen) and Lysotracker green (100 nM, Invitrogen) were added, and the cells were then washed twice with phenol red-free medium (Corning, Tewksbury, MA) and imaged using a Zeiss 510 confocal laser scanning fluorescent microscope (Frankfurt, Germany) in phenol red-free medium (containing 10% FBS). Blue filter (BP 420–480 nm), green filter (BP 505–550 nm), and red filter (604–690 nm spectrum of the Meta detector) were used to recorded the fluorescence from Hoechst 33342, Lysotracker green and doxorubicin, respectively.

Cytotoxicity

KB-3-1 or KB-V1 cells were seeded onto a 96-well plate at 5 × 10³ cells/well, and incubated overnight. The cells were first treated with fresh medium containing 5 or 15 µM Dox, NTD, C₁₆NTD or C₈-Tat control molecule for 2 h, and further incubated with drug free medium for 46 h. Equal molar C₈-Tat was added into some wells to evaluate the effect of C₈-Tat. The cell viability was determined using the SRB method according to the manufacturer’s protocol (TOX-6, Sigma, USA).

Results and Discussion

Conjugates design and characterization

In this work, Tat peptide was chosen as the model CPP because of its wide application in drug delivery and relative clear mechanism of cellular entry.^20,25 For easy molecular tracking and broad representativeness consideration, a poor cell penetrating fluorescent dye (5-fluorescein, 5-FAM) and a good cell penetrating fluorescent dye and also potent anticancer drug (doxorubicin, Dox) were chosen to conjugate to the Tat peptide. When the release of free drug is necessary, an enzymatically cleavable tetrapeptide linker, GFLG, was introduced between the drug and Tat peptide. This linker has been reported to be specifically responsive to a lysosomal enzyme, cathepsin B (CatB).^37–40 Palmitic acid was chosen for Tat modification because of its less crystalline nature at body temperature, which was shown to be important for activity of modified peptides.⁴¹ Additionally, palmitoylation has been widely used to construct peptide amphiphiles to promote their self-assembling feature in aqueous environments and to achieve new bioactivities.^42–45

Figure 1 listed all the studied molecules in this paper. Details of the synthesis and purification can be found in the previous reports^22,26 and also in the supporting information. The purities of the two new conjugates (C₁₆NTF and C₁₆NTD) were all above 98% according to analytical HPLC analysis (Figure S1a). The m/z of C₁₆NTF was observed to be 2442.025 Da (Figure S1b), according to the MALDI-TOF mass spectrum, in agreement with the expected exact mass of the conjugate (2241.440 Da calculated from C₁₁₃H₁₈₄N₃₈O₂₃). However, due to the chemical instability of our synthesized Tat-doxorubicin conjugates under the MALDI-TOF operation conditions,²² high resolution ESI was used instead to identify their molecular masses. The m/z of C₁₆NTD was observed to be 3127.7119 Da, according to ESI mass spectrum, in agreement with the expected exact mass of the conjugate (3126.703 Da calculated from C₁₄₂H₂₂₇N₄₃O₃₅S). Since the hydrophilicity and hydrophobicity balance value (represented by logarithm of partition coefficient, logP) can be used to predict a molecule’s ability to diffuse across cell membranes,⁴⁶ we have calculated the partition coefficients of all the studied molecules using Marvinsketch (ver. 5.12.3, ChemAxon, MA) (Table S1). Due to the hydrophilic and highly positively charged nature of Tat peptide, all the Tat conjugates were calculated to have a logP (clogP) value lower than −20 in biological relevant pH. These values are out of the typical range (−0.4–5.6) desired for molecules to effectively cross cell membranes, suggesting that free diffusion is unlikely the dominant mechanism for these Tat conjugates to get into cells.

Figure 1.

Chemical structures of the synthesized Tat conjugates (top to bottom): non-cleavable Tat-5-fluorescein conjugate (NTF), palmitic acid modified non-cleavable Tat-5-fluorescein conjugate (C₁₆NTF), cleavable Tat-5-fluorescein conjugate (TF), black hole quencher 1 modified cleavable Tat-5-fluorescein conjugate (TFB), cleavable Tat-doxorubicin conjugate (NTD) and palmitic acid-modified cleavable Tat-doxorubicin conjugate (C₁₆NTD).

Palmitoylation significantly improves the cellular uptake of Tat-5-FAM conjugates

To verify whether hydrophobic modification could improve cellular uptake of Tat conjugates, the cellular uptake of NTF and C₁₆NTF was compared. The only difference between these two conjugates is the introduction of palmitic acid in C₁₆NTF. Both the 5-FAM and the palmitic acid were conjugated to Tat peptide directly through an amide bond to achieve stable conjugation in order to exclude the possibility of premature degradation. As shown in Figure 2, C₁₆NTF treated cells exhibited a great improvement in fluorescence intensity which was about 40 and 6 times of the intensity of 5-FAM and NTF treated cells, respectively. This is a remarkable increase considering that palmitic acid only makes up ~10% of the whole conjugate by mass. Hydrophobic modification was reported to affect the secondary structure of conjugated peptide,^47–49 which is important to the cell penetrating efficiency.⁵⁰ To evaluate whether the observed difference in cellular uptake of NTF and C₁₆NTF was linked to a change in secondary structure, we recorded the circular dichroism spectrum of each conjugate (Figure S3). These results reveal that there is no significant difference in the conformation of the two conjugates. The slight increase in the intensity for C₁₆NTF could perhaps be a result of enhanced intermolecular interactions caused by fatty acid acylation.⁵¹

Figure 2.

Cellular uptake of 5-FAM, NTF and C₁₆NTF by MCF-7 cells. Epifluorescent images of MCF-7 cells (post-wash with DPBS) after 2 h incubation with 5 µM free 5-FAM (a), NTF (b), and C₁₆NTF (c). Quantitative comparison of endocytosed free 5-FAM, NTF, and C₁₆NTF by MCF-7 cells after 2 h incubation with 5 µM molecules and trypsin treatment (removes non-endocytosed conjugates) using flow cytometry at FL-1 channel (d). Quantitative comparison of endocytosed C₁₆NTF (e) and NTF (f) by MCF-7 cells after 2 h incubation with 5 µM molecules in the presence or absence of C₈-Tat (5, 10, 20 and 50 µM). Both the qualitative and quantitative results show greatly improved cellular uptake after lipidation of the Tat conjugates. The cellular uptake of the conjugate cannot be inhibited by even 10 times Tat peptide.

Since the cellular entry of Tat peptide is generally regarded to start with its interaction with negatively charged cell membrane components (such as heparan sulfate and phospholipids),^24,25 it is very likely that the palmitic acid can further enhance this interaction and the subsequent adsorption-mediated endocytosis (AME) pathway,⁵² by inserting into the hydrophobic domain of cell membrane. In an effort to verify the AME pathway that in general cannot be directly inhibited by the presence of other absorption molecules, we performed experiments to investigate the cellular uptake of C₁₆NTF in the presence of C₈-Tat of various concentrations. As shown in Figure 2e, the amount of C₈-Tat appears to have no influence on the cellular entry of C₁₆NTF. Even in the presence of a 10-fold excess of C₈-Tat relative to C₁₆NTF, the inhibition effect was still not observed. Similar results were also found in the cellular uptake experiments with the NTF molecule (Figure 2f). The slight difference in fluorescence intensity of cells incubated with different amount of C₈-Tat is within the margin of experiment errors due to the measured low fluorescence intensity as a result of NTF’s poor cellular entry. These experiments suggest that AME is the most plausible mechanism for NTF and C₁₆NTF internalization.

Efficient cancer cell killing by palmitoylation of Tat-Dox conjugate regardless of P-glycoprotein expression level

Co-existence of cancer cells with different phenotypes, termed as heterogeneity, is a big challenge in cancer chemotherapy whose outcome is maximized by killing cancer cells as much as possible.^48,49 In the case of doxorubicin, a widely used chemotherapeutic for cancer therapy, its application is greatly hampered by the development of multi-drug resistance.⁵³ Conjugation of Dox to the CPPs is a facile way to increase the intracellular accumulation of Dox in multi-drug resistant cells, but unfortunately decreases accumulation of doxorubicin in sensitive cells.^22,23 Therefore, one type of doxorubicin derivative that can kill both sensitive and resistant cancer cells is desired for successful cancer chemotherapy. Based upon the above results from 5-FAM conjugates, C₁₆NTD was therefore synthesized by introducing a palmitic acid at the N-terminal of the NTD conjugate (Figure 1), with the aim of increasing drug accumulation in multi-drug resistant cells while retaining its efficacy in sensitive cells.

The cellular uptake of NTD and C₁₆NTD by drug sensitive KB-3-1 cells and drug resistant KB-V1 cells was first compared using flow cytometry to confirm the effect of hydrophobic modification on cellular uptake. Consistent with our previous results,²² after 2 h incubation on drug resistant cells, the intracellular accumulation of free doxorubicin (Geo mean = 11.7) was greatly inhibited by the expressed P-glycoprotein, which effluxes Dox during its diffusion across the cell membrane (Figure 3a).⁵⁴ The accumulation of the NTD (Geo mean = 18.2) is not affected by P-glycoprotein,²² and showed improvement against free doxorubicin (Figure 3a). The palmitoylation showed an exciting improvement in intracellular accumulation of Tat-Dox conjugate, evidenced by 5-time increasing when comparing C₁₆NTD (Geo mean = 110) with NTD. It is even more exciting to find that C₁₆NTD (Geo mean = 113) showed a comparable intracellular accumulation to free doxorubicin observed in KB-3-1 cells (Geo mean = 152), while that of NTD (Geo mean = 22.3) was much lower (Figure 3b). The difference between free doxorubicin and C₁₆NTD is possibly due to their difference in internalization pathway, as free diffusion (Dox) is more efficient than endocytosis (Tat conjugates) in most cases.²³ The secondary structures of NTD and C₁₆NTD (50 µM in DPBS) were similar according to the CD spectra (Figure S4), consistent with the results from the 5-FAM conjugates.

Figure 3.

Cellular uptake of doxorubicin (Dox), NTD and C₁₆NTD. Intracellular accumulation of free doxorubicin, NTD or C₁₆NTD by drug-resistant cervical cancer cells KB-V1 (a) and drug-sensitive KB-3-1 cells (b) after 2 h incubation with 5 µM molecules and trypsin treatment (removes the non-endocytosed conjugates) determined by flow cytometry at FL-3 channel. Quantitative comparison of cellular uptake of Dox, NTD or C₁₆NTD (5 µM) in the absence or presence of equal molar C₈-Tat (5 µM) (c). Studies on the retention ability of endocytosed Dox or C₁₆NTD (2 h and 5 µM) in KB-3-1 cells or KB-V1 cells after additional 2 h incubation in drug-free medium (d). * p < 0.05 compared with cells after 2 h incubation with doxorubicin. ** p < 0.01 compared with cells after 2 h incubation with doxorubicin.

In order to investigate the role that the Tat peptide plays in the cellular uptake of doxorubicin conjugates, we conducted a series of experiments on drug resistant KB-V1 cells that are known to have a high expression level of p-glycoprotein. These experiments involved the incubation of the cells with each individual molecule (Dox, NTD, or C₁₆NTD) in the absence and presence of equal molar of C₈-Tat. Figure 3c shows the flow cytometry measurement results, revealing that the intracellular accumulation of all the three studied molecules is almost identical between cells with or without C₈-Tat treatment. This suggests soluble Tat peptide has little impact on the p-glycoprotein’s drug efflux capacity, highlighting the importance of the covalent linkage of Dox to the Tat to achieve enhanced cellular uptake.

To explore further how this covalent linkage contributes to the cellular uptake, we evaluated the retention ability of drug conjugates in both drug sensitive and resistant cell lines. These cells were incubated for 2 h with 5 µM either Dox or C₁₆NTD, followed by additional 2 h incubation in drug-free medium. We found that ~20% of the internalized free doxorubicin was effluxed out of the drug sensitive KB-3-1 cells, compared to the ~80% clearance by the drug-resistant KB-V1 cells (Figure 3d). In sharp contrast, the clearance of endocytosed C16NTD from the KB-3-1 cells was almost negligible, and only ~20% loss was recorded on KB-V1 cells. These results were consistent with our previous report on the retention of NTD,²² again showing conjugation with Tat peptide presents an effective strategy to prolong the drug retention within cells. These results also demonstrate that hydrophobic modification could be used as a versatile means to enhance the cellular uptake of hydrophilic cargo (such as 5-FAM) or P-glycoprotein substrates (such as doxorubicin) in both drug sensitive and resistant cell lines.

Once increased intracellular accumulation was observed, we asked whether this could lead to better antitumor activity as the drug must still be released from the conjugate for it to exert an effect. Since doxorubicin is usually given as a bolus injection with a half-life of several hours,⁵⁵ we determined the short-term (2 h incubation) cytotoxicity of the drugs only. Consistent with the cellular uptake results, potent anticancer activity of C₁₆NTD was observed on both cell lines regardless of P-glycoprotein level (Figure 4), while NTD showed only little activity. The mixture of Dox with equal molar Tat peptide did not display any advantageous efficacy over free Dox in killing KB-V1 cells (Figure 4b), again suggesting that soluble Tat plays little role in overcoming drug resistance. Free Dox only showed toxicity toward drug-sensitive cells after diffusion across the cell membrane,⁵⁶ but not on drug-resistant cells because of P-glycoprotein overexpression.⁵⁴ The Tat peptide showed no cytotoxicity to both the drug-sensitive and -resistant cells (Figure 4a and S5), indicating the peptide itself was biocompatible. These results imply that C₁₆NTD could be a potential treatment for heterogeneous tumors containing both drug-sensitive and -resistant cells.^57,58

Figure 4.

Cell viability of KB-3-1 and KB-V1 cells. Cell viability of drug-sensitive KB-3-1 (a) and drug-resistant KB-V1 (b) cells after 2 h treatment with 5 or 15 µM of free doxorubicin, NTD or C₁₆NTD in the presence or absence of equal molar C₈-Tat and a further 46 h incubation in drug-free medium. Caprylic acid-modified Tat (C₈-Tat) was used as control to determine the toxicity of the Tat peptide. Data is presented as mean ± s.d. (n = 3). * p < 0.01 compared with NTD; ^# p < 0.01 compared with free doxorubicin.

The GFLG linker incorporated into our design is specifically sensitive to the lysosomal enzyme, CatB,⁵⁹ and thus doxorubicin is expected to be released within lysosomes upon their encounter with CatB.²² The liberated free doxorubicin would gradually diffuse out of the lysosomes.⁴⁰ We therefore used confocal microscopy to monitor the subcellular localization of the endocytosed conjugates. After 2 h incubation, C₁₆NTD showed a higher intracellular accumulation than NTD, some of which colocalized with the lysosomes (yellow dots) and some located on the cell membrane or peri-membrane area in both drug-sensitive and -resistant cells (Figure 5, S6 and S7). In sharp contrast, Dox showed high intracellular accumulation predominantly in the nuclei of sensitive cells due to its ability to readily penetrate the cell membrane and its affinity toward DNA.⁶⁰ No discernible Dox was observed in drug-resistant cells owing to drug efflux (Figure 5 and S7).⁵⁴ Since there is a strong electrostatic interaction between the positively charged Tat peptide and negatively charged cell membrane components,²⁴ adsorptive mediated endocytosis could be the most probable pathway for internalization as observed for a Tat-insulin conjugate⁶¹ and other peptides.^52,62 The amount of macromolecules adsorbed on the cell membrane will likely determine the extent of intracellular accumulation.⁶³ In the presence of salts and negatively charged proteins in the cell culture medium, the electrostatic attraction will be significantly screened, and the introduction of a hydrophobic block could enhance the interaction through membrane insertion. Since both the adsorption and insertion are not cell membrane location specific and we also do not expect any self-assembled nanostructures at the studied concentrations, it is likely the bounded molecules somehow concentrated to several areas before internalization.⁶³ This perhaps explains why red fluorescent dots were observed rather than continuous red staining of the cell membrane. Most of the endocytosed NTD and C₁₆NTD were found to be colocalized with lysosomes in both KB-3-1 and KB-V1 cell (Figure 5, S6 and S7), ensuring the release of free doxorubicin from the conjugates after endocytosis.

Figure 5.

Subcellular drug localization. Subcellular colocalization of Dox, NTD or C₁₆NTD (red) in sensitive KB-3-1 and resistant KB-V1 cells with lysosome (green, Lysotracker Green) and nucleus (blue, Hoechst 33342) after 2 h drug (15 µM) incubation. Scale bar: 20 µm. The enlarged version of Figure 5 was provided in the supporting information as Figure S8 and S9.

Improved imaging contrast by BHQ1 modification

Intracellular enzymatic activity detection (e.g. the activity of CatB) in live cells or organisms is critical for cancer diagnosis and staging.^7,64–66 Due to their intracellular localization, efficient intracellular accumulation is critical to achieve high contrast between intracellular signal and extracellular background noise. Since the hydrophobic modification significantly enhanced the cellular uptake of the Tat conjugate, we applied this strategy in our design of cancer imaging beacons for CatB detection.²⁶ 5-FAM was used as a reporting fluorophore, and was conjugated to Tat peptide through GFLG linker. The black hole quencher 1 (BHQ1) was chosen because of its near-perfect quenching of fluorescence from 5-FAM and the hydrophobicity required for enhanced intracellular accumulation. The successful quenching of 5-FAM fluorescence by BHQ1 in TFB was evidenced by epifluorescence images of the samples (Figures S8), which ensures low background noise during CatB activity detection.

We first compared the fluorescence intensity of the cells treated by TF or TFB (two conjugates with similar secondary structure²⁶) after the removal of conjugate-containing media using epifluorescence microscope (qualitative) and flow cytometry (quantitative). The fluorescence intensity of the TFB-treated cells was approximately 8 times higher than TF, although both conjugates showed increased cellular uptake compared with 5-FAM (about 20 times for TF and 160 times for TFB) (Figure 6). It is noteworthy that the intracellular TFB may not be fully activated at this time point, and therefore the potential of hydrophobic modification might have been underestimated. The GLFG linker, because of its hydrophobicity, also seems to contribute to cellular uptake when comparing the relative fluorescence intensity of TF and NTF to 5-FAM, similar to the observed effect of the FFLIPKG sequence on octaarginine (R8).³⁰ To mimic the conditions in potential in vivo application, under which images would be taken in the presence of non-endocytosed beacon, the intracellular fluorescence from TFB was determined without replacing the TFB-containing media previously. A significant contrast between intra- and extra-cellular fluorescence was clearly observed for TFB (greater than 50% compared to control) but not for TF due to signal saturation (Figure S9). These internalization results, consistent with those from the comparison of NTF and C₁₆NTF, clearly show that hydrophobic blocks other than fatty acids can also improve the intracellular accumulation of a Tat conjugate. In this specific case, the BHQ1 acts as dual-functional component, first as a hydrophobic block to enhance cellular uptake and second as a quencher to 5-FAM. Taken in tandem, these two effects contribute to improved imaging contrast, a property that is critical for cancer imaging based on intracellular biomacromolecule detection.

Figure 6.

Cellular uptake of 5-FAM, TF and TFB by MCF-7 cells. Epifluorescent images of MCF-7 cells (post-wash with DPBS) after 0.5 and 2 h incubation with 5 µM free 5-FAM, TF, and TFB (a). Quantitative comparison of endocytosed free 5-FAM, TF, and TFB by MCF-7 cells after 2 h incubation with 5 µM molecules and trypsin treatment (removes non-endocytosed conjugates) using flow cytometry at FL-1 channel (b).

Summary

In conclusion, our results collectively suggest the advantages of hydrophobic modification of the Tat conjugates (Figure 7). First, the hydrophobic modified Tat conjugates can efficiently deliver cargos with different cell membrane permeability into cell. The cellular uptake of the hydrophobic modified Tat conjugates was comparable to the free form of drug with good cell permeability and was greatly improved compared to the drug with poor cell permeability. Second, the hydrophobic modified Tat conjugates can deliver substrate of P-glycoprotein regardless of its expression level in cancer cells. Doxorubicin after conjugation with Tat peptide with a palmitoyl tail can achieve considerable intracellular accumulation and effective anticancer activity in both drug-sensitive and -resistant cells. Third, the hydrophobic unit can be functional molecules other than fatty acids, which could bring additional features to the conjugates. Hydrophobic quencher was used in our study instead of palmitic acid, which substitution did not deteriorate the intracellular accumulation and in the contrary endow the conjugate an additional function as a beacon that could be effectively used as diagnostic reagent for intracellular enzyme activity in living cells. The potential of using other functional hydrophobic molecules such as hydrophobic drugs or dyes in drug-Tat conjugates could obtain synergetic and theranostic properties.

Figure 7.

Schematic illustration of the mechanism of improved intracellular accumulation of drug through Tat and hydrophobic unit modification. The accumulation of a free drug in the cells is depending on its logarithm of partition coefficient P (logP) and whether it is a substrate of efflux transporter. Conjugation of the drug to Tat peptide can improve the intracellular accumulation of drug with poor cell permeability and help the drug to bypass P-gp efflux. Further hydrophobic modification with inert fatty acid or other functional block such as quencher could improve the cellular uptake of drug further and at the same time bring in new properties to the conjugate as a beacon.