Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Cell Type Preference of a Novel Human Derived Cell-Permeable Peptide dNP2 and TAT in Murine Splenic Immune Cells

  • Sangho Lim ,

    Contributed equally to this work with: Sangho Lim, Jung-ah Lee

    Affiliations Department of Life Science, College of Natural Sciences, Hanyang University, Seoul, 133–791, Korea, Research Institute for Natural Sciences, Hanyang University, Seoul, 133–791, Korea

    ⨯
  • Jung-ah Lee ,

    Contributed equally to this work with: Sangho Lim, Jung-ah Lee

    Affiliations Department of Life Science, College of Natural Sciences, Hanyang University, Seoul, 133–791, Korea, Research Institute for Natural Sciences, Hanyang University, Seoul, 133–791, Korea

    ⨯
  • Ja-Hyun Koo,

    Affiliations Department of Life Science, College of Natural Sciences, Hanyang University, Seoul, 133–791, Korea, Research Institute for Natural Sciences, Hanyang University, Seoul, 133–791, Korea

    ⨯
  • Tae Gun Kang,

    Affiliation Department of Biochemistry, College of Life Science and Biotechnology, Yonsei University, Seoul, 120–749, Korea

    ⨯
  • Sang-Jun Ha,

    Affiliation Department of Biochemistry, College of Life Science and Biotechnology, Yonsei University, Seoul, 120–749, Korea

    ⨯
  • Je-Min Choi

    jeminchoi@hanyang.ac.kr

    Affiliations Department of Life Science, College of Natural Sciences, Hanyang University, Seoul, 133–791, Korea, Research Institute for Natural Sciences, Hanyang University, Seoul, 133–791, Korea

    ⨯

Abstract

Cell-permeable peptides (CPPs) have been widely studied as an attractive drug delivery system to deliver therapeutic macromolecules such as DNA, RNA, and protein into cells. However, its clinical application is still limited and controversial due to the lack of a complete understanding of delivery efficiency in target cells. Previously we identified and characterized the novel and superior CPP, named dNP2, and here we comparatively analyzed intracellular delivery efficiency of dNP2 and TAT in various immune cells of mouse spleen to demonstrate their cell type preference. dNP2- or TAT-conjugated fluorescent proteins were most efficiently taken up by phagocytic cells such as dendritic cells and macrophages while little protein uptake was seen by lymphocytes including T cells, B cells, and NK cells. Interestingly CD8+ lymphoid dendritic cells and CD62LloCD44hi memory like T cell subsets showed significantly better uptake efficiency in vitro and in vivo relative to other dendritic cells or T cells, respectively. In addition, activated macrophages, T cells, and B cells took up the proteins more efficiently relative to when in the resting state. Importantly, only dNP2, not TAT, shows significant intracellular protein delivery efficiency in vivo. Collectively, this study provides important information regarding heterogeneous intracellular delivery efficiency of CPPs such as dNP2 and TAT with cell type preference in the spleen needed for its application in phagocytic cells or activated immune cells.

Citation: Lim S, Lee J-a, Koo J-H, Kang TG, Ha S-J, Choi J-M (2016) Cell Type Preference of a Novel Human Derived Cell-Permeable Peptide dNP2 and TAT in Murine Splenic Immune Cells. PLoS ONE 11(5): e0155689. https://doi.org/10.1371/journal.pone.0155689

Editor: Yeonseok Chung, Seoul National University College of Pharmacy, REPUBLIC OF KOREA

Received: January 30, 2016; Accepted: May 3, 2016; Published: May 17, 2016

Copyright: © 2016 Lim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information file.

Funding: This work was supported by grants from the Korea Health Technology R&D project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (HI14C0234 to J.M.C., HI14C2680 to S.J.H.) (URL: http://www.khidi.or.kr/eps). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Cell permeable peptides (CPPs), either identified from various native proteins or designed as synthetic peptides, are being utilized to deliver therapeutic macromolecules into cells in vitro and in vivo [1, 2]. Since the TAT protein from HIV was found to localize to the nucleus and cytoplasm of cultured cells without the need for transfection reagents [3], the intracellular delivery of proteins by cell permeable peptides has been intensively studied. Arginine or lysine rich cationic peptides such as TAT [4], Antp [5], VP22 [6] and R9 [7] have been used over the last few decades to deliver various macromolecular cargos including siRNA [8, 9], oligonucleotides [10], peptides [11], and transcription factors [12, 13] to alter cellular behavior and modulate disease pathogenesis as a macromolecular therapeutics. Cationic CPPs bind to negatively charged cell membrane molecules such as proteoglycans on glycoproteins or negatively charged membrane lipid molecules and trigger endocytosis by the cells to uptake the CPP-cargo complex [14, 15]. Recently, human protein derived CPPs have been identified such as VectoCell [16], Lactoferin [17], Hph-1 [18], Sim-2 [11], LPIN [19], 2IL-1a [20] to minimize possible immunogenicity and toxicity in vivo. However, it has been demonstrated that almost all cells are permeable to CPPs without specificity, which remains one of the major challenges to the use of CPPs as an advantageous drug delivery tool [2]. Although a significant number of clinical trials have been performed using TAT, there are as of yet no available FDA approved CPP-drugs for any human disease. Here, we hypothesized that the understanding of heterogeneous delivery efficiency by a CPP would increase the success of CPP drug development.

Recently, we identified the novel human-derived CPP named dNP2 that was efficiently delivered into primary immune cells including T cells and was even localized to brain tissue through the blood-brain barrier (BBB) in mice [21]. dNP2 contains amphiphilic amino acids with basic sequences and hydrophobic amino acids with helix-loop-helix structure as prediction that seem to enhance cellular uptake by endocytosis. To better use this attractive CPP for valuable drug development, the preferential targeting of dNP2 to specific cells should be investigated. Therefore, the present study was designed to determine the cell type preference of dNP2 and TAT in various immune cells of mouse spleen. Enhanced green fluorescent proteins (EGFPs) fused to CPPs were used to treat splenocytes in vitro and flow cytometric analyses defined the specific cell types targeted by each CPP protein. In addition, intravenous administration of dNP2-EGFP and TAT-EGFP proteins in mice confirmed the in vivo relevance of heterogeneous delivery efficiency among the immune cells. Here we found that phagocytic cells such as DCs and macrophages uptake CPP-proteins more efficiently than lymphocytes. In addition, activated T cells, B cells, and macrophages are better targeted by CPP then these cells in their resting state. In addition, in vivo relevance suggests that there is obvious cell type preference for CPPs with heterogeneous delivery efficiency among various cell population, which should be considered for therapeutic drug design.

Methods

Cell lines and cell culture

EL4 (mouse lymphoma cell line) and Jurkat (human lymphoma cell line) cells were purchased from the American Type Culture Collection (ATCC) and cultured using Roswell Park Memorial Institute (RPMI) 1640 media with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin antibiotics. All cells were maintained in a 5% CO2 incubator at 37stitute (RPMI) 1640 media with 10% fetal bovine seThermo Scientific Hyclone.

Purification of recombinant proteins

Escherichia coli BL21 (DE3) Star pLysS bacterial cells were transformed with CPP-conjugated fluorescent proteins containing pRSET-b vectors and incubated in 50 ml Lysogeny broth (LB) containing ampicillin for 12 h at 37l and shaking at 200 rpm. The cultures were transferred to 500 ml of LB media without antibiotics and incubated under the same conditions for another 1–2 h until the optical density at 600 nm (OD600) reached 0.4–0.6. The protein production was induced using 0.2 mM isopropyl-beta-D-thiogalactopyranoside (IPTG) for 12 h at 20°C with shaking at 150 rpm. The cells were then harvested by centrifugation at 4arv 5,000 rpm for 15 min and sonicated using the Vibra-cell VCX-130 ultrasonic processor. The supernatants were filtered through a 0.45 μM cellulose nitrate syringe filter. The 6-His tag proteins were purified through Ni-NTA affinity chromatography, and then desalted and buffer exchanged to PBS containing 10% glycerol using a PD-10 size exclusion column.

Mice

Wild type C57BL/6 mice were purchased from Orient Bio Inc. and bred. All mice were housed and maintained in a specific pathogen-free animal facility at Hanyang University. All animal experiments were approved by the Animal Experimentation Ethics Committee of Hanyang University (Permit Number: 2015–0004, 2015–0006) and were performed in strict accordance with the guidelines of Institutional Animal Care and Use Committees of Hanyang University. And all animal research protocols followed the guidelines of Korean Association for Laboratory Animal Sciences (KALAS). All surgery was performed after CO2 asphyxiation euthanasia protocol and all efforts were made to minimize suffering.

Splenocyte isolation

Spleens were removed from 6 to 10-week-old mice and chopped with scissors. The chopped tissue was incubated with 2 mg/ml collagenase D at room temperature for 20 min and red blood cells were depleted by incubating with RBC lysis buffer for 2 min. After centrifugation, the cells were resuspended in 10% FBS and 1% penicillin/streptomycin containing RPMI media and counted.

Flow cytometric analysis of CPP-proteins delivery efficiency of in vitro

To analyze the delivery efficiency of CPP-conjugated proteins or CPP-peptide, 5 μM of each molecule were added to the cells and incubated for 1 h. We used 2.5X105 cells/well for Jurkat cells and EL4 cells compared with splenocytes and human PBMCs, and 8X106 cells/well for analyzing various splenocyte cell types. Following incubation, cells were washed twice with PBS and trypsinized with 0.25% (for cell line) or 0.05% (for primary cells) trypsin-EDTA at 37°C for 10 min. Cells were washed with PBS and stained with fluorochrome-conjugated cell surface marker specific antibodies and the intracellular delivery efficiency was analyzed using flow cytometry.

Surface staining for flow cytometric analysis

After incubation with CPP-conjugated molecules, the cells were resuspended in 100 μl antibody mixture diluted in PBS and incubated at 4°C for 15 min. To discriminate each cell subset from the mixed cell population, different antibody combinations were used. For cell type analysis, anti-mouse CD4-PerCP-Cy5.5 (BioLegend, Clone:RM4-5), anti-mouse CD19-PE-Cy7 (BioLegend, Clone:6D5), anti-mouse NK1.1-APC (BioLegend, Clone:PK136), anti-mouse CD11b-PE-Cy7 (eBioscience, Clone:H57-597), anti-mouse CD11c-APC (eBioscience, Clone:N418), anti-mouse I-A/I-E-APC-Cy7 (BioLegend, Clone:M5/114.15.2), and anti-mouse F4/80-PerCP-Cy5.5 (BioLegend, Clone:BM8) were used. Also, anti-mouse CD16/32 (BioLegend, Clone:93) was used for blocking the Fc receptor. For activated T cell analysis, anti-mouse CD69-FITC (BioLegend, Clone:H1.2F3), anti-mouse CD4-APC (eBioscience, Clone:GK1.5), and anti-mouse CD8-PerCP-Cy5.5 (eBioscience, Clone:53–6.7) were used. For activated B cell analysis, anti-mouse CD86-PE-Cy5 (BioLegend, Clone:GL-1), anti-mouse IgG2a, kappa-isotype (BioLegend, Clone:RTK2758), and anti-mouse CD19-PE (BioLegend, Clone:6D5) were used.

Immune cell activation

To activate splenic T cells, anti-mouse CD3 antibody (2 μg/ml, BD Pharmingen) was coated onto 96-well plates (BD Falcon) at 3796-well plateellmousesplenocytes were seeded at a density of 2.5X105 cells/well and incubated overnight at 37 r in a 5% CO2 incubator. To activate B cells, DCs, or macrophages, isolated splenocytes were cultured in 12-well plates at 4X106 cells/well with 5 μg/ml (for B cells) or 100 ng/ml (for DCs and macrophages) of Escherichia coli 055:B5 (Sigma-Aldrich) in each well and incubated overnight at 37°C in 5% CO2 incubator.

Flow cytometric analysis of CPP-protein delivery efficiency in vivo

5 mg of EGFP, CPP-EGFP, or PBS was intravenously injected into 6-week-old C57BL/6 female mice. After 1 h, the mice were sacrificed and spleens were removed for isolating splenocytes from a single cell suspension. The cells were stained with fluorochrome-conjugated surface marker specific antibodies and analyzed via flow cytometry.

Statistical analysis

The data were analyzed using one-way ANOVA, two-way ANOVA with multiple comparison tests or Student’s t-test. p values <0.05 were considered significant. Statistical analysis was performed using Prism 6 (Graphad Software, Inc.).

Results

dNP2- and TAT-proteins are delivered more efficiently to phagocytic cells than lymphocytes

To determine cell type heterogeneity of dNP2-protein delivery efficiency in various immune cells of mouse spleen, we treated dNP2-EGFP or TAT-EGFP to splenocytes and intracellular delivery efficiency was analyzed by flow cytometry by the staining of cell type specific markers (Fig 1A–1D). The relative mean fluorescent intensity (Relative MFI) represents fold increase that each MFI is normalized to PBS treated sample (Fig 1B–1D). Statistical significance among the all groups were comparatively analyzed by one-way ANOVA test with Sidak’s multiple comparison test (S1 Fig). Both dNP2-EGFP and TAT-EGFP proteins were delivered into CD11cloCD11bhiF4/80+ macrophages and CD11c+MHCII+ DCs, which are phagocytic and antigen presenting cells, with significantly higher efficiency than into lymphocytes. In the case of CD4 T cells, B cells, and NK cells, only dNP2-EGFP showed significant protein delivery efficiency comparing with PBS control while TAT-EGFP did not at 5 μM. This pattern is almost identical at various incubating time points including 10 min, 30 min, 1 h, 2 h or 6 h (S2 Fig) and also consistent in the FACS-sorted each cell population (S3 Fig). Interestingly, even EGFP protein without CPPs was significantly taken up by macrophages as well as DCs although their relative MFI value is much lower than that of dNP2- or TAT-EGFP treated group (Fig 1B). These results collectively suggest that phagocytic cells such as macrophages and DCs uptake CPP-protein more efficiently than lymphocytes, and CPPs such as dNP2 and TAT could significantly deliver a cargo protein into those cells to modulate innate immunity including antigen presentation, inflammatory response, etc.

thumbnail
Download:
Fig 1. Protein delivery efficiency of dNP2 and TAT in various immune cells.

(A) 5 μM of EGFP, TAT-EGFP, dNP2-EGFP, or PBS were treated to splenocytes from 6-week-old C57BL/6 mice and protein delivery efficiencies were analyzed by flow cytometry after cell staining with cell type specific markers for CD4+ CD4 T cells, CD19+ B cells, NK1.1+ NK cells, CD11clowCD11bhighF4/80+ macrophages, and MHCII+CD11chigh DCs. (B-D) The experiments were independently repeated three times and are presented as the relative mean fluorescence intensity (relative MFI). The values were normalized with the MFI of PBS treated samples. Bar graphs indicate the mean ± s.d. One-way ANOVA was used for statistical analysis and * indicates p<0.05, ** indicates p<0.01, *** indicates p<0.001, **** indicates p<0.0001.

https://doi.org/10.1371/journal.pone.0155689.g001

CD8+ dendritic cells are the major cell type to uptake dNP2- and TAT-proteins

Splenic DCs can be divided into subsets including lymphoid CD8+ DCs, myeloid CD8- DCs, and plasmacytoid DCs (pDCs) based on different biological characteristics [22–24]. To determine CPP-protein delivery efficiency in splenic DC subsets, splenocytes were treated with red fluorescent dTomato protein fused with dNP2 or TAT (dNP2-dTomato or TAT-dTomato). Intracellular red fluorescence intensity and each DC subset marker were analyzed by flow cytometry. First, we determined the DC subsets as Siglec H+ pDC, Siglec H-CD11c+MHCII+CD8+ lymphoid CD8+ DC, and Siglec H-CD11c+MHCII+CD8- myeloid CD8- DC (Fig 2A). Both TAT-dTomato and dNP2-dTomato proteins were delivered into the CD8+ DCs with significantly higher efficiency relative to CD8- DCs. In addition, pDCs did not uptake CPP-proteins suggesting that the major cellular subset to uptake CPP-proteins is CD8+ lymphoid DCs (Fig 2B, 2C and 2D). These results suggest that among the DC subsets, CPP-protein uptake efficiency is quite variable even in DC subsets and that lymphoid DCs have higher delivery efficiency than myeloid DCs and pDCs indicating that immunization with CPP-conjugated antigen cargos would be efficient application using CPPs such as dNP2 and TAT. mDCs also showed significantly higher delivery efficiency than pDCs (Fig 2C and 2D) suggesting that targeting mDCs by CPP-mediated cargo delivery would be good approaches such as anti-inflammatory functions.

thumbnail
Download:
Fig 2. Protein delivery efficiency of dNP2 and TAT among dendritic cell subsets.

5 μM dTomato, TAT-dTomato, dNP2-dTomato, or PBS were treated to splenocytes from 6-week-old C57BL/6 mice and the cells were analyzed by flow cytometry. (A) The cells were subdivided into Siglec H-CD11c+MHCII+CD8+ lymphoid CD8+ DCs, Siglec H-CD11c+MHCII+CD8- myeloid CD8- DCs, and Siglec H+ pDCs. (B) The protein delivery efficiencies into each subset of DCs were analyzed by flow cytometry. (C-D) The experiments were independently repeated three times and are presented as the relative mean fluorescence intensity (relative MFI). The values were normalized with the MFI of PBS treated samples. Bar graphs indicate the mean ± s.d. and * indicates p<0.01 and *** indicates p<0.0001 from one-way ANOVA with Tukey’s multiple comparison test (n = 3, (C) F = 59.46, P<0.0001 and (D) F = 134.5, P<0.0001).

https://doi.org/10.1371/journal.pone.0155689.g002

Activated lymphocytes uptake dNP2- and TAT-proteins more efficiently than resting state lymphocytes

Because activated lymphocytes are rapidly proliferating and their membrane proteins are dynamically regulated, we hypothesized that CPP-proteins would be delivered into activated or memory-like lymphocytes with higher efficiency relative to resting state lymphocytes. To examine this, we treated dNP2-dTomato or TAT-dTomato to mouse splenocytes. Then, we at first subdivided CD4 T cells into CD4+CD62LhiCD44lo naïve CD4 T cells, CD4+CD62LloCD44hi memory-like CD4 T cells, and CD4+Foxp3+ regulatory T cells by flow cytometry and determined heterogeneous delivery efficiency by CPP-proteins (Fig 3A). Both dNP2-dTomato and TAT-dTomato protein, but not dTomato without CPPs, delivery efficiencies were significantly higher in CD62LloCD44hi memory like activated CD4 T cells than in naïve CD4 T cells and Foxp3+ regulatory T cells, though the efficiency of dNP2 is much higher than TAT (Fig 3B and 3C). Also the delivery efficiency was significantly higher in CD4+Foxp3+CD44high cells than that of CD4+Foxp3+CD44low cells (Fig 3D and 3E). This result suggests that among the peripheral T cell subsets, CPP-protein delivery efficiency is heterogeneous that memory-like T cells or effector T cells are preferred by CPP-protein delivery than naïve T cells.

thumbnail
Download:
Fig 3. Protein delivery efficiency of dNP2 and TAT on naïve, regulatory, or memory-like CD4 T cells.

5 μM dTomato, TAT-dTomato, dNP2-dTomato or PBS were treated to splenocytes from Foxp3-GFP transgenic 6-week-old C57BL/6 mice and the cells were analyzed by flow cytometry. (A) The cells were subdivided into CD4+Foxp3+ regulatory T cells, CD4+CD62LhighCD44low naïve CD4 T cells, and CD4+CD62LlowCD44high memory-like CD4 T cells by flow cytometry after cell surface marker staining. Also regulatory T cells were subdivided into Foxp3+CD44low and Foxp3+CD44high cells. (B) The protein delivery efficiencies into each subset of CD4 T cells were analyzed by flow cytometry. (C) The experiments were independently repeated three times and are presented as the relative mean fluorescence intensity (relative MFI). The values were normalized with the MFI of PBS treated samples. Bar graphs indicate the mean ± s.d. and within groups (dTomato, TAT-dTomato and dNP2-dTomato) the differences of means of memory like CD4 T cells and regulatory CD4 T cells were compared with means of naïve CD4 T cells by two-way ANOVA with Dunnett’s multiple comparison test, * indicates p<0.05 and **** indicates p<0.0001 (n = 3). (D) The protein delivery efficiencies into each subset of regulatory T cells were analyzed by flow cytometry. (E) The values were normalized with the MFI of PBS treated samples. Bar graphs indicate the mean ± s.d. and the data were analyzed by student’s t-test, *indicates p<0.05 (n = 3).

https://doi.org/10.1371/journal.pone.0155689.g003

Next, we stimulated isolated mouse splenocytes with plate bound anti-CD3 and anti-CD28 antibodies to determine whether activated CD4 and CD8 T cells would uptake CPP-proteins more efficiently relative to unstimulated cells (Fig 4A). Both dNP2-EGFP and TAT-EGFP proteins were delivered into activated CD4 and CD8 T cells with significantly higher efficiency than unstimulated cells (Fig 4C, 4E and 4F). In addition, we stimulated splenocytes with lipopolysaccharide (LPS) to examine whether activated B cells would uptake CPP-proteins (Fig 4B). From the results, LPS stimulated B cells uptake dNP2-EGFP and TAT-EGFP proteins more efficiently than resting B cells (Fig 4D and 4G). These results collectively suggest that CPPs such as dNP2 and TAT could deliver a cargo protein into activated immune cells with better efficiency than their resting state that it would be valuable to apply on modulating abnormally activated T or B cell responses in allergic or autoimmune diseases.

thumbnail
Download:
Fig 4. Protein delivery efficiency of dNP2 and TAT on stimulated or unstimulated T or B cells.

Splenocytes from 6-week-old C57BL/6 mice were stimulated with (A) 2 μg/ml anti-mouse CD3 antibody for 12 h or (B) 5 μg/ml LPS for 24 h and additionally incubated in the presence of 5 μM EGFP, TAT-EGFP, dNP2-EGFP, or PBS for 30 min. (C) The protein delivery efficiency into stimulated or unstimulated CD4 or CD8 T cells was analyzed by flow cytometry. (D) The protein delivery efficiency into stimulated or unstimulated B cells was analyzed by flow cytometry. (E-G) The experiments were independently repeated three times and are presented as the mean fluorescence intensity (MFI). Bar graphs indicate the mean ± s.d. and * indicates p<0.05 and ** indicates p<0.01 from Student’s t-tests (n = 5).

https://doi.org/10.1371/journal.pone.0155689.g004

Activated macrophages, but not activated dendritic cells, uptake dNP2- and TAT-proteins more efficiently than resting state

We next examined CPP-protein delivery efficiency in activated macrophages and DCs compared to their resting state. Isolated mouse splenocytes were incubated with or without LPS stimulation overnight, and then cells were treated with dNP2-EGFP or TAT-EGFP protein for 1 h. The intracellular fluorescence of CD11cloCD11bhiF4/80+ macrophages and CD11c+MHCII+ DCs was analyzed by flow cytometry. dNP2-EGFP treated LPS-stimulated macrophages showed significantly higher intracellular fluorescence intensity than unstimulated macrophages and TAT-EGFP treated macrophages also showed a similar delivery pattern even though the delivery efficiency of dNP2 is much higher than TAT (Fig 5A, 5C and 5E). However, CPP-protein delivery efficiency in LPS-stimulated DCs was significantly lower than in resting DCs (Fig 5B, 5D and 5F) possibly due to maturated DCs have less antigen uptake ability then immature state [25]. Interestingly EGFP without CPPs also shows consistently reduced protein uptake efficiency by DCs while no difference was observed in macrophages (Fig 5E and 5F). These results suggest that CPP-protein delivery efficiency is more significant in activated macrophages and immature DCs that it could be valuable to apply it on modulating inflammatory responses mediated by macrophages or on initiating immune responses mediated by DCs.

thumbnail
Download:
Fig 5. Protein delivery efficiency of dNP2 and TAT on unstimulated or LPS-stimulated macrophages or dendritic cells.

Splenocytes from 6-week-old C57BL/6 mice were stimulated with 5 μg/ml LPS for 24 h and additionally incubated in the presence of 5 μM EGFP, TAT-EGFP, dNP2-EGFP, or PBS for 30 min. (A-D) The protein delivery efficiency into stimulated or unstimulated macrophages or DCs was analyzed by flow cytometry. (E-F) The experiments were independently repeated three times and are presented as the relative mean fluorescence intensity (relative MFI). The values were normalized with the MFI of PBS treated samples. Bar graphs indicate the mean ± s.d. and *** indicates p<0.001 from Student’s t-tests (n = 3).

https://doi.org/10.1371/journal.pone.0155689.g005

dNP2- but not TAT-proteins shows consistent cell type preference in various immune cells of mouse spleen in vivo

Finally, we investigated the applicability of in vitro results in vivo by intravenous administration of dNP2-EGFP or TAT-EGFP protein in mice. One hour after the intravenous injection of 5 mg dNP2-EGFP or TAT-EGFP into mice, splenocytes were isolated and various immune cell types were stained with cell type specific marker antibodies (Fig 6A). As consistent with the in vitro results, F4/80highCD11bint macrophage cells (blue) are the most major cells to uptake proteins. In addition, delivery efficiency of dNP2-EGFP shows higher delivery efficiency on Ly6C+CD11bhigh monocytes than Ly6G+CD11bhigh neutrophils (Fig 6B and 6C). Among the DC subsets, CD8+ Siglec H- lymphoid DCs (blue) showed higher delivery efficiency than CD8- Siglec H- myeloid DCs and Siglec H+ pDCs (Fig 6D and 6E). CD62LloCD44hi memory-like activated cells (red) in both CD4 and CD8 T cells preferentially transduced by dNP2-EGFP protein relative to CD62LhiCD44lo naïve CD4 and CD8 T cells (Fig 6F and 6G). However, importantly, TAT-EGFP delivery efficiency was not significant in any cell type population suggesting its limitation for in vivo application. These results suggest that dNP2 efficiently deliver a cargo protein into various immune cells in vivo with heterogeneous delivery efficiency and this should be considered for its application design to modulate cellular responses in vivo.

thumbnail
Download:
Fig 6. Protein delivery efficiency of dNP2 and TAT in various immune cells in vivo.

(A) Experimental scheme is indicated. EGFP, TAT-EGFP, dNP2-EGFP, or PBS (5 mg) were injected to 6-week-old C57BL/6 mice and after 1 h the splenocytes were isolated and analyzed by flow cytometry after cell surface marker staining. (B-C) Protein delivery efficiency into Ly6C+CD11bhigh monocytes (red), F4/80highCD11bint macrophage cells (blue), and Ly6G+CD11bhigh neutrophils (orange) was analyzed by flow cytometry. (D-E) Protein delivery efficiency into CD8-Siglec H- myeloid DCs (red), CD8+Siglec H- lymphoid DCs (blue), and Siglec H+ pDC (orange) was analyzed by flow cytometry. (F-G) Protein delivery efficiency into CD4+CD62LhighCD44low, CD4+CD62LlowCD44high, CD8+CD62LhighCD44low, or CD8+CD62LlowCD44high cells was analyzed by flow cytometry. Bar graphs indicate the mean ± s.d. and within groups (EGFP, TAT-EGFP, dNP2-EGFP) the differences of means of each cell types were compared with each other by two-way ANOVA with Tukey’s multiple comparison test, * indicates p<0.05, ** indicates p<0.01, *** indicates p<0.001, **** indicates p<0.0001 (n = 3).

https://doi.org/10.1371/journal.pone.0155689.g006

Discussion

In the present study, we characterized and demonstrated cell type preference of the recently identified human derived BBB-permeable peptide dNP2 [21] and the widely used HIV-derived CPP, TAT [4, 26] for the uptake of proteins in various immune cells. Both dNP2 and TAT conjugated fluorescent proteins were more preferentially delivered into phagocytic cells such as macrophages and DCs relative to lymphocytes. In addition, we confirmed that activated immune cells including macrophages, T cells, and B cells could uptake proteins with better efficiency relative to their resting state, collectively demonstrating heterogeneous intracellular delivery efficiency of dNP2 and TAT in vitro and in vivo should be considered for better application of these CPPs in disease modulation.

CPPs are generally regarded as a non-specific macromolecular delivery tool, which has been considered a major disadvantage of CPP usage for therapeutic purposes [27–29]. However, previous studies reported that the arginine-based CPP r8 peptide, without any targeting sequences, spontaneously accumulated in xenografted tumors of tumor-bearing nude mice [30]. Even though the R8 peptide was delivered into normal tissues, it showed a preference for tumor cells. In addition, in the early days of CPPs, Dowdy et al. reported an odd delivery pattern in spleen tissue in vivo where TAT-β-galactosidase was only detected in red pulp while it was clearly not detected in white pulp in mice [4]. These studies suggested that CPPs could be delivered with different efficiencies and specificities depending on target cell type. These previous results suggest the possibility of cell type preference in various types of cells in vivo.

In the present study, we revealed that both dNP2 and TAT have cell type preference in various immune cells. Although the proportion of DCs and macrophages are just 1–2% of total spleen cells, their CPP-protein uptake efficiency is much dominant compare to the lymphocytes including T cells or B cells. This suggests that the phagocytic action or cell’s own characteristics are important in effective CPP-protein delivery. High efficiency in phagocytic cells including DCs and macrophages also suggest that the application of CPPs such as dNP2 as an immunization tool to delivery of antigens for vaccine development could be one of the best approaches. Especially, lymphoid DCs were shown to uptake proteins much more efficiently than myeloid DCs and plasmacytoid DCs. Previously, it has been reported that lymphoid DCs have much higher T cell stimulation capacity than myeloid DCs [31]. While, myeloid DCs have been known as having higher migratory capacity compared with lymphoid DCs [32–34]. In addition, only lymphoid DC can cross-prime cytotoxic T cells in vivo [35]. Thus, lymphoid DCs which is preferred by CPP-cargo delivery is appropriate target for DC vaccination using cell-penetrating peptide [36]. Similar to the preference and high delivery efficiency of TAT to lymphoid DCs in this study, previous studies showed that TAT-conjugated protein antigen could be a useful DC vaccination methodology for boosting antigen specific CD8 T cell immunity [37, 38]. For example, treatment of the TAT conjugated Leishmania antigen TAT-LACK to DCs successfully induced Leishmania specific CD8 T cells in vitro while LACK alone could not [39]. Also, numerous studies reported that DC vaccination is one of the most attractive approaches for cancer immune therapy using specific cancer antigens [40–43]. TAT conjugated breast cancer specific antigen generated cancer specific CD8 T cells and infiltrated into tumor tissues in vivo [44]. Our study previously showed that dNP2 has higher delivery efficiency relative to TAT suggesting that dNP2 could be more valuable to deliver antigens for the generation of pathogen or cancer specific immunity.

Macrophages are another potent target cell type for modulating the immune system because of their high delivery efficiency with both dNP2 and TAT. Macrophages are one of the most abundant cells existing in every tissue including lymphoid, lung, liver, bone, and brain tissues with phenotypic diversity [45, 46]. For this reason, the preference of dNP2-protein delivery to macrophages could be a useful tool for modulating macrophage functions regarding inflammation, disease pathogenesis, or bone remodeling.

Both dNP2 and TAT conjugated protein were taken up by memory like CD4 T cells or activated lymphocytes including CD4 and CD8 T cells or B cells with higher efficiency relative to these cells in their resting state. This preference would have the advantage of modulating activated lymphocytes and memory T and B cells by delivering signaling modulatory molecules. Previously we have attempted to deliver cytoplasmic domain of CTLA-4 protein which inhibits T cell receptor signaling pathway in activated T cells modulated abnormal immune responses in allergic or autoimmune disease models [18, 21, 47]. In addition, CPP-conjugated Foxp3 protein could modulated OVA-induced allergic T cell responses in vivo without affecting the other T cells response to KLH antigen which was still intact [12]. Other previous report showed that CPP-based antisense-oligomer delivery successfully modulated T cell response by altering gene expression [48]. These results suggest CPPs such as dNP2 and TAT could be applied to modulate activated lymphocyte reaction by passive targeting or preference moiety.

In consistency of relative preference on activated T cells, regulatory T cells which is highly expressing CD44, that could be regarded as activated/effector regulatory T cell [49], showed significantly higher CPP-protein uptake efficiency than resting regulatory T cells which have lower level of CD44. Regulatory T cell subsets were reported as CD45RA+Foxp3low resting regulatory T cells, CD45RA-Foxp3high activated regulatory T cells and CD45RA-Foxp3low nonsuppressive T cells [50]. In the case of carcinoma patients, the activated regulatory T cells are increased in the peripheral circulation and antitumor immunity is suppressed [51]. The CPP-drugs could be preferably delivered into activated regulatory T cells for cancer therapy which possibly be a novel approach for cancer-immunotherapy.

Collectively, we demonstrate cell type preference of CPP delivery efficiency using dNP2 and TAT in various immune cells in vitro and in vivo. First, CPP delivery efficiency is the highest into phagocytic cells, especially lymphoid DCs. Second, CPP delivery efficiency is much higher into the activated immune cells than non-activated immune cells. Last, in vivo protein delivery efficiency of TAT is as little as negative control suggesting significant advantage of dNP2 usage in vivo. These findings suggest that heterogeneity of CPP-protein delivery efficiency would be dependent on cells own character such as phagocytic activity, etc. Cell type preference analysis in non-immune cells also should be considered and it would provide important additional information for the accurate application of CPPs in appropriate target cells.

Supporting Information

S1 Fig. One-way ANOVA with Sidak’s multiple comparison test results of Fig 1B–1D.

https://doi.org/10.1371/journal.pone.0155689.s001

(TIF)

S2 Fig. Time dependent protein delivery efficiency of CPP-EGFPs in various immune cells.

5 μM of EGFP, TAT-EGFP, dNP2-EGFP or PBS were treated to splenocytes from 6-week-old C57BL/6 mice for various reaction times including 10 min, 30 min, 1 h, 2 h or 6 h and protein delivery efficiencies were analyzed by flow cytometry after cell staining with cell type specific markers (CD4+ CD4 T cells, CD19+ B cells, NK1.1+ NK cells, CD11clowCD11bhighF4/80+ macrophages, and MHCII+CD11chigh DCs). The values were normalized with the mean fluorescence intensity (MFI) of PBS treated samples (relative MFI). The graphs indicate the mean ± s.d. In the case of TAT-EGFP and dNP2-EGFP, macrophages and DCs showed significantly higher delivery efficiency than lymphocytes within every time points. We performed statistical analysis by two-way ANOVA with Tukey’s multiple comparisons test and *** indicates p<0.001 and **** indicates p<0.0001.

https://doi.org/10.1371/journal.pone.0155689.s002

(TIF)

S3 Fig. Protein delivery efficiency of CPP-EGFPs into FACS sorted various immune cells.

5 μM of EGFP, TAT-EGFP, dNP2-EGFP or PBS were treated to FACS sorted CD4 T cells (CD4+), CD8 T cells (CD8+), B cells (CD19+), Dendritic cells (MHCII+CD11chigh) or Macrophages (CD11clowCD11bhighF4/80+). The delivery efficiencies were analyzed by flow cytometry. In the bar graphs, the values were normalized with MFI of PBS treated samples (relative MFI).

https://doi.org/10.1371/journal.pone.0155689.s003

(TIF)

Acknowledgments

This work was supported by grants the Korea Health Technology R&D project through the Korea Health Industry Development Institute (KHIDI) funded by the Ministry of Health & Welfare, Republic of Korea (HI14C0234 to J.M.C., HI14C2680 to S.J.H.) URL; http://www.khidi.or.kr/eps. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author Contributions

Conceived and designed the experiments: JMC JL. Performed the experiments: SL JL JHK. Analyzed the data: SL JL JHK TGK SJH JMC. Contributed reagents/materials/analysis tools: TGK SJH. Wrote the paper: SL JMC.

References

  1. 1. Heitz F, Morris MC, Divita G. Twenty years of cell-penetrating peptides: from molecular mechanisms to therapeutics. British journal of pharmacology. 2009;157(2):195–206. pmid:19309362; PubMed Central PMCID: PMC2697800.
  2. 2. van den Berg A, Dowdy SF. Protein transduction domain delivery of therapeutic macromolecules. Current opinion in biotechnology. 2011;22(6):888–93. pmid:21489777.
  3. 3. Frankel AD, Pabo CO. Cellular uptake of the tat protein from human immunodeficiency virus. Cell. 1988;55(6):1189–93. pmid:2849510.
  4. 4. Schwarze SR, Ho A, Vocero-Akbani A, Dowdy SF. In vivo protein transduction: delivery of a biologically active protein into the mouse. Science. 1999;285(5433):1569–72. pmid:10477521.
  5. 5. Derossi D, Joliot AH, Chassaing G, Prochiantz A. The third helix of the Antennapedia homeodomain translocates through biological membranes. The Journal of biological chemistry. 1994;269(14):10444–50. pmid:8144628.
  6. 6. Elliott G, O'Hare P. Intercellular trafficking and protein delivery by a herpesvirus structural protein. Cell. 1997;88(2):223–33. pmid:9008163.
  7. 7. Rothbard JB, Garlington S, Lin Q, Kirschberg T, Kreider E, McGrane PL, et al. Conjugation of arginine oligomers to cyclosporin A facilitates topical delivery and inhibition of inflammation. Nature medicine. 2000;6(11):1253–7. pmid:11062537.
  8. 8. Uchida T, Kanazawa T, Kawai M, Takashima Y, Okada H. Therapeutic effects on atopic dermatitis by anti-RelA short interfering RNA combined with functional peptides Tat and AT1002. The Journal of pharmacology and experimental therapeutics. 2011;338(2):443–50. pmid:21531792.
  9. 9. Moschos SA, Williams AE, Lindsay MA. Cell-penetrating-peptide-mediated siRNA lung delivery. Biochemical Society transactions. 2007;35(Pt 4):807–10. pmid:17635153.
  10. 10. Freire JM, Veiga AS, Rego de Figueiredo I, de la Torre BG, Santos NC, Andreu D, et al. Nucleic acid delivery by cell penetrating peptides derived from dengue virus capsid protein: design and mechanism of action. The FEBS journal. 2014;281(1):191–215. pmid:24286593.
  11. 11. Choi JM, Sohn JH, Park TY, Park JW, Lee SK. Cell permeable NFAT inhibitory peptide Sim-2-VIVIT inhibits T-cell activation and alleviates allergic airway inflammation and hyper-responsiveness. Immunology letters. 2012;143(2):170–6. pmid:22342853.
  12. 12. Choi JM, Shin JH, Sohn MH, Harding MJ, Park JH, Tobiasova Z, et al. Cell-permeable Foxp3 protein alleviates autoimmune disease associated with inflammatory bowel disease and allergic airway inflammation. Proceedings of the National Academy of Sciences of the United States of America. 2010;107(43):18575–80. pmid:20937878; PubMed Central PMCID: PMC2972952.
  13. 13. Lee J, Sayed N, Hunter A, Au KF, Wong WH, Mocarski ES, et al. Activation of innate immunity is required for efficient nuclear reprogramming. Cell. 2012;151(3):547–58. pmid:23101625; PubMed Central PMCID: PMC3506423.
  14. 14. Gump JM, June RK, Dowdy SF. Revised role of glycosaminoglycans in TAT protein transduction domain-mediated cellular transduction. The Journal of biological chemistry. 2010;285(2):1500–7. pmid:19858185; PubMed Central PMCID: PMC2801275.
  15. 15. Ziegler A. Thermodynamic studies and binding mechanisms of cell-penetrating peptides with lipids and glycosaminoglycans. Advanced drug delivery reviews. 2008;60(4–5):580–97. pmid:18045730.
  16. 16. De Coupade C, Fittipaldi A, Chagnas V, Michel M, Carlier S, Tasciotti E, et al. Novel human-derived cell-penetrating peptides for specific subcellular delivery of therapeutic biomolecules. The Biochemical journal. 2005;390(Pt 2):407–18. pmid:15859953; PubMed Central PMCID: PMC1198920.
  17. 17. Duchardt F, Ruttekolk IR, Verdurmen WP, Lortat-Jacob H, Burck J, Hufnagel H, et al. A cell-penetrating peptide derived from human lactoferrin with conformation-dependent uptake efficiency. The Journal of biological chemistry. 2009;284(52):36099–108. pmid:19858187; PubMed Central PMCID: PMC2794725.
  18. 18. Choi JM, Ahn MH, Chae WJ, Jung YG, Park JC, Song HM, et al. Intranasal delivery of the cytoplasmic domain of CTLA-4 using a novel protein transduction domain prevents allergic inflammation. Nature medicine. 2006;12(5):574–9. pmid:16604087.
  19. 19. Lim S, Kim WJ, Kim YH, Choi JM. Identification of a novel cell-penetrating peptide from human phosphatidate phosphatase LPIN3. Mol Cells. 2012;34(6):577–82. pmid:23263658; PubMed Central PMCID: PMCPMC3887828.
  20. 20. Koo JH, Yoon H, Kim WJ, Lim S, Park HJ, Choi JM. Cell membrane penetrating function of the nuclear localization sequence in human cytokine IL-1alpha. Molecular biology reports. 2014;41(12):8117–26. pmid:25205122.
  21. 21. Lim S, Kim WJ, Kim YH, Lee S, Koo JH, Lee JA, et al. dNP2 is a blood-brain barrier-permeable peptide enabling ctCTLA-4 protein delivery to ameliorate experimental autoimmune encephalomyelitis. Nature communications. 2015;6:8244. pmid:26372309; PubMed Central PMCID: PMC4579786.
  22. 22. Shortman K, Liu YJ. Mouse and human dendritic cell subtypes. Nature reviews Immunology. 2002;2(3):151–61. pmid:11913066.
  23. 23. Merad M, Sathe P, Helft J, Miller J, Mortha A. The dendritic cell lineage: ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting. Annual review of immunology. 2013;31:563–604. pmid:23516985; PubMed Central PMCID: PMC3853342.
  24. 24. Mildner A, Jung S. Development and function of dendritic cell subsets. Immunity. 2014;40(5):642–56. pmid:24837101.
  25. 25. Granucci F, Ferrero E, Foti M, Aggujaro D, Vettoretto K, Ricciardi-Castagnoli P. Early events in dendritic cell maturation induced by LPS. Microbes and infection / Institut Pasteur. 1999;1(13):1079–84. pmid:10572310.
  26. 26. Vives E, Brodin P, Lebleu B. A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. The Journal of biological chemistry. 1997;272(25):16010–7. pmid:9188504.
  27. 27. Lee SH, Castagner B, Leroux JC. Is there a future for cell-penetrating peptides in oligonucleotide delivery? European journal of pharmaceutics and biopharmaceutics: official journal of Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik eV. 2013;85(1):5–11. pmid:23958313.
  28. 28. Vives E, Schmidt J, Pelegrin A. Cell-penetrating and cell-targeting peptides in drug delivery. Biochimica et biophysica acta. 2008;1786(2):126–38. pmid:18440319.
  29. 29. Regberg J, Srimanee A, Langel U. Applications of cell-penetrating peptides for tumor targeting and future cancer therapies. Pharmaceuticals. 2012;5(9):991–1007. pmid:24280701; PubMed Central PMCID: PMC3816645.
  30. 30. Nakase I, Konishi Y, Ueda M, Saji H, Futaki S. Accumulation of arginine-rich cell-penetrating peptides in tumors and the potential for anticancer drug delivery in vivo. Journal of controlled release: official journal of the Controlled Release Society. 2012;159(2):181–8. pmid:22285548.
  31. 31. Martin P, del Hoyo GM, Anjuere F, Ruiz SR, Arias CF, Marin AR, et al. Concept of lymphoid versus myeloid dendritic cell lineages revisited: both CD8alpha(-) and CD8alpha(+) dendritic cells are generated from CD4(low) lymphoid-committed precursors. Blood. 2000;96(7):2511–9. pmid:11001905.
  32. 32. Leenen PJ, Radosevic K, Voerman JS, Salomon B, van Rooijen N, Klatzmann D, et al. Heterogeneity of mouse spleen dendritic cells: in vivo phagocytic activity, expression of macrophage markers, and subpopulation turnover. J Immunol. 1998;160(5):2166–73. pmid:9498754.
  33. 33. Ruedl C, Bachmann MF. CTL priming by CD8(+) and CD8(-) dendritic cells in vivo. Eur J Immunol. 1999;29(11):3762–7. pmid:10556833.
  34. 34. Steinman RM, Inaba K. Myeloid dendritic cells. J Leukoc Biol. 1999;66(2):205–8. pmid:10449155.
  35. 35. den Haan JM, Lehar SM, Bevan MJ. CD8(+) but not CD8(-) dendritic cells cross-prime cytotoxic T cells in vivo. J Exp Med. 2000;192(12):1685–96. pmid:11120766; PubMed Central PMCID: PMCPMC2213493.
  36. 36. Lim S, Koo JH, Choi JM. Use of Cell-Penetrating Peptides in Dendritic Cell-Based Vaccination. Immune Netw. 2016;16(1):33–43. pmid:26937230; PubMed Central PMCID: PMCPMC4770098.
  37. 37. Sun Y, Hu YH. Cell-penetrating peptide-mediated subunit vaccine generates a potent immune response and protection against Streptococcus iniae in Japanese flounder (Paralichthys olivaceus). Veterinary immunology and immunopathology. 2015;167(3–4):96–103. pmid:26232860.
  38. 38. Sakuma S, Suita M, Inoue S, Marui Y, Nishida K, Masaoka Y, et al. Cell-penetrating peptide-linked polymers as carriers for mucosal vaccine delivery. Molecular pharmaceutics. 2012;9(10):2933–41. pmid:22953762.
  39. 39. Kronenberg K, Brosch S, Butsch F, Tada Y, Shibagaki N, Udey MC, et al. Vaccination with TAT-antigen fusion protein induces protective, CD8(+) T cell-mediated immunity against Leishmania major. The Journal of investigative dermatology. 2010;130(11):2602–10. pmid:20574442.
  40. 40. Cho HI, Kim EK, Park SY, Lee SK, Hong YK, Kim TG. Enhanced induction of anti-tumor immunity in human and mouse by dendritic cells pulsed with recombinant TAT fused human survivin protein. Cancer letters. 2007;258(2):189–98. pmid:17931768.
  41. 41. Shibagaki N, Udey MC. Dendritic cells transduced with TAT protein transduction domain-containing tyrosinase-related protein 2 vaccinate against murine melanoma. European journal of immunology. 2003;33(4):850–60. pmid:12672050.
  42. 42. Kim SG, Park MY, Kim CH, Sohn HJ, Kim HS, Park JS, et al. Modification of CEA with both CRT and TAT PTD induces potent anti-tumor immune responses in RNA-pulsed DC vaccination. Vaccine. 2008;26(50):6433–40. pmid:18812201.
  43. 43. Mitsui H, Inozume T, Kitamura R, Shibagaki N, Shimada S. Polyarginine-mediated protein delivery to dendritic cells presents antigen more efficiently onto MHC class I and class II and elicits superior antitumor immunity. The Journal of investigative dermatology. 2006;126(8):1804–12. pmid:16645583.
  44. 44. Viehl CT, Becker-Hapak M, Lewis JS, Tanaka Y, Liyanage UK, Linehan DC, et al. A tat fusion protein-based tumor vaccine for breast cancer. Annals of surgical oncology. 2005;12(7):517–25. pmid:15889213.
  45. 45. Gordon S, Pluddemann A, Martinez Estrada F. Macrophage heterogeneity in tissues: phenotypic diversity and functions. Immunological reviews. 2014;262(1):36–55. pmid:25319326; PubMed Central PMCID: PMC4231239.
  46. 46. Epelman S, Lavine KJ, Randolph GJ. Origin and functions of tissue macrophages. Immunity. 2014;41(1):21–35. pmid:25035951; PubMed Central PMCID: PMC4470379.
  47. 47. Choi JM, Kim SH, Shin JH, Gibson T, Yoon BS, Lee DH, et al. Transduction of the cytoplasmic domain of CTLA-4 inhibits TcR-specific activation signals and prevents collagen-induced arthritis. Proceedings of the National Academy of Sciences of the United States of America. 2008;105(50):19875–80. pmid:19066215; PubMed Central PMCID: PMC2604944.
  48. 48. Marshall NB, Oda SK, London CA, Moulton HM, Iversen PL, Kerkvliet NI, et al. Arginine-rich cell-penetrating peptides facilitate delivery of antisense oligomers into murine leukocytes and alter pre-mRNA splicing. Journal of immunological methods. 2007;325(1–2):114–26. pmid:17673254.
  49. 49. Lim HW, Broxmeyer HE, Kim CH. Regulation of trafficking receptor expression in human forkhead box P3+ regulatory T cells. J Immunol. 2006;177(2):840–51. pmid:16818738.
  50. 50. Miyara M, Yoshioka Y, Kitoh A, Shima T, Wing K, Niwa A, et al. Functional delineation and differentiation dynamics of human CD4+ T cells expressing the FoxP3 transcription factor. Immunity. 2009;30(6):899–911. pmid:19464196.
  51. 51. Sun W, Li WJ, Wu CY, Zhong H, Wen WP. CD45RA-Foxp3high but not CD45RA+Foxp3low suppressive T regulatory cells increased in the peripheral circulation of patients with head and neck squamous cell carcinoma and correlated with tumor progression. J Exp Clin Cancer Res. 2014;33:35. pmid:24761979; PubMed Central PMCID: PMCPMC4022051.