This work was approved by the Institutional Review Board of Seoul National University Hospital (# 0905-014-280). Peripheral blood was obtained from healthy volunteers who were taking no immunosuppressive drugs and had no diseases that could potentially affect the immune system such as autoimmunity, infections, and malignancies (13). Skin specimens (5 mm diameter) were obtained from a patient who had moderate atopic dermatitis with chronic lesional and non-lesional skin. Written informed consent was obtained from all subjects according to the Declaration of Helsinki.

Peripheral blood mononuclear cells (PBMCs) in heparinized peripheral blood were isolated using a Ficoll-Histopaque gradient (1.077 g/mL; GE Healthcare Biosciences, Piscataway, NJ, USA). The following anti-human Abs were used for flow cytometry staining: allophycocyanin-anti-CD8, phycoerythrin-anti-CCR7 (all from BD Biosciences, San Jose, CA, USA), and eFluor^® 450NC-anti-IL-7Rα (CD127) (eBioscience, San Diego, CA, USA).

Fluorescence-activated cell sorter (FACS) staining was performed as described (13). Staining intracellular molecules was performed after permeabilization using a Cytofix/Cytoperm kit (BD Biosciences). The stained cells were analyzed on a BD LSRII^® (BD Biosciences) with FACSDiva software, and the data were analyzed using FlowJo^® software (TreeStar, Ashland, OR, USA).

CD8⁺ T cells were enriched from PBMCs using the Human CD8⁺ T Cell Isolation kit (Miltenyi Biotec) and then stained with Abs and sorted into naïve (CCR7⁺CD45RA⁺), IL-7Rα^high, and IL-7Rα^low EM (CCR7⁻CD45RA^+/−) CD8⁺ T cells using a BD FACSAria^® (BD Biosciences).

Sorted CD8⁺ T cell subsets were cultured in RPMI 1640 media containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (i.e., RPMI 1640 complete media; all reagents from Life Technologies, Carlsbad, CA, USA). To reverse CD8⁺ T cells by IL-2 stimulation, freshly sorted CD8⁺ T cell subsets were incubated for 5 days with anti-CD3 (HIT3a, 10 µg/mL, plate-bound)/CD28 (28.2, 5 µg/mL, soluble) Abs (both from eBioscience) in the presence of recombinant human IL-2 (20 IU/mL, R&D Systems, Minneapolis, MN, USA) and then stimulated for 10 days in the presence of IL-2 alone with media changes every 3 days with IL-2 replenishment (16) (Sim et al., manuscript submitted).

To stimulate CD8⁺ T-cell subsets with cytokines, cells were stimulated for 3 days with anti-CD3/CD28 Abs in the presence or absence of IL-2 (20 IU/mL), IL-15 (5 ng/mL; PeproTech, Rocky Hill, NJ, USA), or IL-4 (5 ng/mL; PeproTech).

For cytokine measurement, sorted CD8⁺ T cell subsets were stimulated for 24 h with anti-CD3/CD28 Abs. Culture supernatants were analyzed for cytokines using a MILLIPLEX^® MAP kit (EMD Millipore, Billerica, MA, USA) according to the manufacturer’s instructions.

For measuring cell proliferation, sorted CD8⁺ T cell subsets were labeled with 0.5 µg/mL carboxyfluorescein diacetate (CFSE, Life Technologies) and incubated for 6 days with anti-CD3/CD28 Abs. Cell proliferation was analyzed using a flow cytometer.

Cells were incubated with 50 µM 1-ethyl-2-benzimidazolinone (1-EBIO) or with 5 µM (1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole, TRAM-34) to activate or inhibit the KCa3.1 channel, respectively, or with margatoxin (5 nM) (all from Sigma-Aldrich, St. Louis, MO, USA) to inhibit the Kv1.3 channel during TCR stimulation with anti-CD3/CD28 Abs.

Electrophysiological measurements were performed using the conventional whole-cell recording mode at room temperature. Patch pipettes with a free-tip resistance of 5–7 MΩ were connected to the head stage of an Axopatch 200B patch-clamp amplifier (Molecular Devices, Sunnyvale, CA, USA). pCLAMP software v9.0 and Digidata-1332A (Molecular Devices) were used to acquire data and apply command pulses. Cells were transferred into a bath (~0.15 mL) mounted on the stage of an inverted microscope (IX50; Olympus, Tokyo, Japan) and perfused with HEPES-buffered normal Tyrode (NT) solution at 5 mL/min. The NT external solution was composed of 145 mM NaCl, 5 mM glucose, 3.6 mM KCl, 1.3 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES (adjusted to pH 7.4 with NaOH). For whole-cell patch clamp analysis of Kv1.3 and KCa3.1 channels, the pipette solution contained 140 mM KCl, 5 mM NaCl, 2 mM MgATP, 5 mM EGTA, 4.37 mM CaCl₂, 0.5 mM MgCl₂, and 10 mM HEPES (pH 7.2) with free Ca²⁺ concentration fixed at 1 µM. Phosphatidylinositol 3-kinase (PI3K) inhibitor (10 µM LY294002, Sigma-Aldrich) was included in the intracellular pipette solution in several experiments as indicated.

Current–voltage (I–V) curves reflecting the functional expression of Kv1.3 and KCa3.1 channels were obtained using ramp-like pulse protocols. Using a whole-cell configuration, the total membrane conductance was initially obtained by applying a depolarizing ramp pulse (from −120 to 60 mV, held at −60 mV). Subsequently, in the same cell, a hyperpolarizing ramp pulse from 60 to −120 mV was applied with −10 mV of depolarized holding voltage to induce inactivation of Kv1.3. The remaining non-inactivating conductance recorded with 1 µM of free Ca²⁺ concentration was regarded as the KCa3.1 current based on the characteristic weak inwardly rectifying I/V curve and the activation by KCa3.1 activator (50 µM 1-EBIO, Figures 1A,B). Digital subtraction of the KCa3.1 I/V curve from the initial total I/V curve provided the Kv1.3 current, demonstrating the characteristic outwardly rectifying voltage-dependent property (Figure 1A, lower panel).

For migration assays (17), a flat polyurethane acrylate (PUA) surface on glass coverslips was coated with recombinant human intercellular adhesion molecule 1 (ICAM-1) Fc chimera (ICAM-1, R&D Systems) and/or recombinant human stromal cell-derived factor (SDF)-1α (CXCL12, PeproTech) by incubating ICAM-1 (10 µg/mL) in the absence or presence of SDF-1α (2 µg/mL) in PBS for 3 h at 37°C after a brief (~60 s) air plasma treatment (100 W; Femto Science, Hwaseong, South Korea). ICAM-1 was used because it binds to integrin LFA-1 on the T cell membrane and causes T cell attachment, polarization, and random migration (18).

For migration assays, CD8⁺ T cell suspensions in RPMI complete media were added on flat PUA surfaces in the Chamlide magnetic chamber (Live Cell Instrument, Seoul, South Korea) and maintained in a cell culture incubator at 37°C for 1 h before starting the migration experiments. Ultra-low-melt agarose (USB, Cleveland, OH, USA) at a final concentration of 1% was added to RPMI media to minimize convection during live cell imaging. Unless otherwise noted, we used freshly sorted cells without TCR stimulation in cell migration experiments.

For KCa3.1 channel inhibition, cells were treated with 5 µM TRAM-34 for 2 h before starting the migration experiments.

A modified Zeiss Axio Observer Z1 epifluorescence microscope (Jena, Germany) with a 40× (numerical aperture = 1.3; Plan-Neofluar) objective lens and a Roper Scientific CoolSnap HQ CCD camera (Photometrics, Tucson, AZ, USA) were used for imaging (17). The CD8⁺ T cell-seeded surfaces were mounted on the microscope stage equipped with a Chamlide TC incubator system (Live Cell Instrument) and maintained at 37°C and 5% CO₂ for live cell imaging. Time-lapse microscopy was initiated with differential interference contrast images recorded at 15 s intervals for 10 min. The trajectory of the CD8⁺ T cells was analyzed using the manual tracking plugin of ImageJ (http://imagej.nih.gov/ij/, National Institute of Health, Bethesda, MD, USA), and quantitative analysis was performed with a custom built program on Matlab (MathWorks, Natick, MA, USA).

For transendothelial migration analysis (19), human umbilical vein endothelial cells (HUVECs, Lonza, Basel, Switzerland) were grown on 8-µm microporous membranes in transwell chambers (Life Technologies) under EGM™-2 (Lonza) plus FBS 5% for 48 h. CD8⁺ T cell subsets (1.5 × 10⁵ cells/well) were added above the HUVEC monolayers on the filters. Then, the transwells were placed into a 24-well plate containing 100 ng/mL SDF-1α as a chemoattractant. After 20 h of incubation, the number of cells that had migrated into the lower chambers or underneath the upper transwells was determined by counting the number of cells. The results are expressed as the number of cells that migrated across the filter.

Human skin specimens were snap frozen in liquid nitrogen and stored at −80°C. OCT compound (Tissue-Tek^®, Sakura Finetek USA, Torrance, CA, USA) was used in embedding tissues for frozen section. Tissue sections (7 µm) were fixed in 4% paraformaldehyde, blocked with a blocking buffer (5% goat serum and 5% BSA in PBS) for 30 min at room temperature, and stained with purified anti-human perforin Ab (Mabtech, Nacka Strand, Sweden), Alexa647-conjugated anti-human CD8 Ab (BD Biosciences), biotin-conjugated anti-human IL-7Rα Ab (13-1278, eBioscience), Streptavidin-Cy3 (Life Technologies), and Alexa488-conjugated anti-mouse IgG Ab (Life Technologies) at 4°C overnight. After mounting with DAPI-containing media (AR-6501-01, ImmunoBioScience, Mukilteo, WA, USA), the stained tissue sections were analyzed using a confocal microscopy system (A1, Nikon, Tokyo, Japan) and NIS-Elements viewer (Nikon).

For calculating the number of cells in CD8⁺ T cell subsets, IL-7Rα⁺ and perforin⁺ CD8⁺ T cells were counted for each tissue section, with at least four fields including the epidermis and dermis. Cells were enumerated in the tissue fields of the epidermis and dermis by manual counting.

All data are expressed as mean ± SEM. Data were compared using one-way analysis of variance followed by Tukey’s post hoc test, two-tailed Student’s t-test, Wilcoxon matched pairs test, and Mann–Whitney U test. p-Values <0.05 were considered to indicate significance. All statistical analyses were performed using GraphPad Prism 6.01 (GraphPad Software, La Jolla, CA, USA).

We conducted whole-cell patch clamp analysis and dissected the membrane conductance for KCa3.1 and Kv1.3 based on the specific inhibitor and biophysical properties (see Subsection “Electrophysiology” in Section “Materials and Methods”) in IL-7Rα^high and IL-7Rα^low EM CD8⁺ T cell subsets, derived from PBMCs of healthy individuals, after TCR stimulation with anti-CD3/CD28 Abs for 24 h derived from PBMCs of healthy individuals. In contrast to the significant activity of KCa3.1 in IL-7Rα^high EM CD8⁺ T cells (mean ± SEM, 13.66 ± 5.55 pA/pF), the corresponding current was almost absent in IL-7Rα^low EM CD8⁺ T cells (2.17 ± 1.26 pA/pF). The Kv1.3 current was observed in both cell groups, and the amplitudes appeared to be higher in IL-7Rα^high than IL-7Rα^low EM CD8⁺ T cells (329.55 ± 30.61 versus 237.74 ± 33.81 pA/pF), although the difference was not statistically significant (Figure 1A). To further confirm the large difference in KCa3.1 activity, we compared the effects of KCa3.1 activator (1-EBIO, 50 µM). In contrast to the consistent increase of KCa3.1 current in IL-7Rα^high EM CD8⁺ T cells (control; 13.52 ± 4.46 pA/pF, 1-EBIO; 34.08 ± 10.41 pA/pF), no significant increase was observed in the IL-7Rα^low EM CD8⁺ T cells in the presence of 1-EBIO (Figure 1B). These results suggest that the lack of KCa3.1 current was due to the insignificant expression of functional KCa3.1 in the cell membrane. Meanwhile, the expression of Kv1.3 and KCa3.1 channels at the transcriptional level were comparable between IL-7Rα^high and IL-7Rα^low EM CD8⁺ T cells (Figure S2A in Supplementary Material). Thus, it is presumed that the posttranslational regulation of potassium channel proteins could lead to the functional difference of ionic currents (20, 21). Phosphatidylinositol 3-kinase-dependent signaling positively regulates KCa3.1 activity (22). However, treatment with PI3K inhibitor LY294002 did not affect the conductance of KCa3.1 in IL-7Rα^high EM CD8⁺ T cells (0 min; 11.35 ± 2.17 pA/pF, 5 min; 11.09 ± 2.15 pA/pF, 10 min; 7.49 ± 2.24 pA/pF) (Figure 1C). We also compared the changes in [Ca²⁺]_i upon TCR stimulation between the two groups. We found that the relatively slow increase in [Ca²⁺]_i reflecting Ca²⁺ influx was reduced in IL-7Rα^low EM CD8⁺ T cells (Figure S1 in Supplementary Material), suggesting that the potassium channel activities may be involved in the relatively low Ca²⁺ signal.

Clonal anergy of T cells is reversed by IL-2 stimulation (i.e., IL-2 reversal) (16). In fact, IL-2-reversed IL-7Rα^low EM CD8⁺ T cells developed into functionally competent cells in terms of cell proliferation and IL-2 production (Sim et al., manuscript submitted). Interestingly, IL-2 reversal rescued the KCa3.1 current in the IL-7Rα^low EM CD8⁺ T cells (16.41 ± 2.80 pA/pF). In addition, the Kv1.3 current became slightly larger in the IL-2-reversed IL-7Rα^low EM CD8⁺ T cells than in the IL-2-reversed IL-7Rα^high EM CD8⁺ T cells (162.46 ± 13.39 versus 106.32 ± 16.09 pA/pF) (Figure 1D).

We investigated the role of potassium channels in T cell proliferation and cytokine production using specific inhibitors for KCa3.1 (TRAM-34) and Kv1.3 (margatoxin). Margatoxin treatment significantly, although not completely, suppressed the proliferation of IL-7Rα^high EM CD8⁺ T cells (Figure 2A) and completely inhibited the secretion of IL-2 and TNF-α, but not IFN-γ in IL-7Rα^high EM CD8⁺ T cells upon TCR stimulation with anti-CD3/CD28 Abs (Figure 2B). TRAM-34 treatment had no effect on cytokine production or cell proliferation in IL-7Rα^high EM CD8⁺ T cells (Figures 2A,B).

The importance of KCa3.1 in cell migration has been highlighted by its expression patterns in migrating cells including immune cells (23). In migrating human T cells, KCa3.1 is localized in the uropod rear portion (24). Additionally, adenosine suppresses the motility of fully activated T cells with phytohemagglutinin or anti-CD3/CD28 Abs for 72−96 h via inhibition of KCa3.1 channel activity (25).

We examined the migration of IL-7Rα^low EM CD8⁺ T cells that had negligible levels of KCa3.1 activity. The motility of CD8⁺ T cells was determined on surfaces coated with ICAM-1 in the presence or absence of chemokine SDF-1α, which enhances motility of lymphocytes by inducing actin polymerization (26). After isolating EM CD8⁺ T cell subsets without activation, we assessed their migration by performing time-lapse microscopy and observed that IL-7Rα^high EM CD8⁺ T cells showed markedly higher mean velocity (V_mean) than IL-7Rα^low EM CD8⁺ T cells (Figure 3A). Although SDF-1α modestly increased motility in both IL-7Rα^high and IL-7Rα^low EM CD8⁺ T cells, IL-7Rα^high EM CD8⁺ T cells showed substantially higher motility than IL-7Rα^low EM CD8⁺ T cells (Figure 3A). In line with these results, the TRAM-34 KCa3.1 inhibitor significantly inhibited the motility of IL-7Rα^high EM CD8⁺ T cells on ICAM-1 surfaces, even in the presence of SDF-1α (Figure 3B). Interestingly, the margatoxin Kv1.3 inhibitor did not affect T cell motility on the ICAM-1 surface; surprisingly, it increased motility in the presence of additional SDF-1α (Figure 3B).

Consistent with the recovery of KCa3.1 activity by IL-2 reversal (Figure 1D), the IL-2-reversed IL-7Rα^low EM CD8⁺ T cells showed similar levels of migration activity to those of IL-7Rα^high EM CD8⁺ T cells (Figure 3C). These data suggest that KCa3.1, but not Kv1.3, regulates the motility of IL-7Rα^low EM CD8⁺ T cells in the absence of TCR stimulation.

We next examined whether the KCa3.1 channel activity of EM CD8⁺ T cells affected transendothelial migration. We utilized transwell chambers to measure T cell migration across a HUVEC monolayer, mimicking the in vivo microenvironment for T cell migration into local tissues. EM CD8⁺ T cell subsets were allowed to migrate through a HUVEC monolayer in the presence or absence of SDF-1α in the lower chamber. The transendothelial migration was significantly higher in IL-7Rα^high EM CD8⁺ T cells than in IL-7Rα^low EM CD8⁺ T cells. The low level of transendothelial migration of IL-7Rα^low EM CD8⁺ T cells was enhanced by adding SDF-1α (Figure 4A). Furthermore, IL-2 reversal restored the decreased transendothelial migration of IL-7Rα^low EM CD8⁺ T cells to a level surpassing even that of IL-7Rα^high EM CD8⁺ T cells (Figure 4B). However, in contrast to a previous result (Figure 3B), the KCa3.1 inhibitor failed to suppress transendothelial migration of IL-7Rα^high EM CD8⁺ T cells and IL-2-reversed EM CD8⁺ T cell subsets (data not shown). We postulate that an unknown mechanism, other than the KCa3.1 channel, might facilitate the complex process of transendothelial migration.

According to our data published previously (13, 27), IL-2 and IL-15 are both capable of reviving the impaired proliferative function of IL-7Rα^low EM CD8⁺ T cells (Sim et al., manuscript submitted). Thus, we analyzed the KCa3.1 activity and motility in T cells stimulated for 3 days with anti-CD3/CD28 Abs in the presence of IL-2, IL-15, or IL-4 (i.e., short-term cytokine stimulation). Short-term cytokine stimulation with IL-2 (IL-7Rα^high versus IL-7Rα^low, 49.26 ± 8.28 versus 144.51 ± 26.15 pA/pF) or IL-15 (27.37 ± 5.56 versus 54.27 ± 20.50 pA/pF), but not IL-4 (24.76 ± 11.83 versus 22.63 ± 7.22 pA/pF), increased the KCa3.1 activity considerably in IL-7Rα^low EM CD8⁺ T cells (Figure 5A). However, the cytokine stimulation had no effect on Kv1.3 activity (Figure 5B). But, as with the above results (Figure S2A in Supplementary Material), expression levels of Kv1.3 and KCa3.1 transcript did not reflect both of the potassium channel activities in IL-7Rα^high and IL-7Rα^low EM CD8⁺ T cells even under the cytokine-stimulated condition (Figure S2B in Supplementary Material). Likely for the same reason, the expression of functional potassium channel seems to be regulated by posttranslational regulation. Although neither IL-2 nor IL-15 affected the motility of IL-7Rα^low EM CD8⁺ T cells on the ICAM-1 surface, both cytokines increased the motility markedly in the presence of SDF-1α (Figure 5C). This result suggests that the functionally revived IL-7Rα^low EM CD8⁺ T cells by short-term stimulation with IL-2 or IL-15 regained motility in the presence of SDF-1α due to increased KCa3.1 activity. Our findings suggest that T cells activated with TCR stimulation and inflammatory cytokines could revive cell motility via increasing KCa3.1 expression, eventually leading to the accumulation of innately motility-impaired IL-7Rα^low EM CD8⁺ T cells in the inflammatory site.

Next, we hypothesized that IL-7Rα^high EM CD8⁺ T cells, which have significant KCa3.1 activity, are more frequently present in tissue than IL-7Rα^low EM CD8⁺ T cells are because of their excellent in vitro migration activity. Thus, we analyzed the tissue infiltration of EM CD8⁺ T cell subsets in non-lesional (healthy, HC) and lesional (dermatitis) skin. Our previous finding showed that IL-7Rα^low EM CD8⁺ T cells are generally found to be largely perforin^high compared to IL-7Rα^high EM CD8⁺ T cells (13). Moreover, IL-15 stimulation increases intracellular perforin expression (28), suggesting that CD8⁺ T cells could increase perforin expression in inflammatory conditions. Thus, we performed immunofluorescence staining on skin-infiltrating IL-7Rα^high and IL-7Rα^low EM CD8⁺ T cells using an Ab combination for IL-7Rα and CD8, or perforin and CD8, respectively.

In the non-lesional HC tissues, the number and frequency of IL-7Rα⁺ CD8⁺ T cells reflecting IL-7Rα^high EM CD8⁺ T cells were significantly higher than the number and frequency of perforin⁺ CD8⁺ T cells that correspond to IL-7Rα^low EM CD8⁺ T cells (Figures 6A,B). Meanwhile, in the lesional (dermatitis) tissue, both IL-7Rα⁺ and perforin⁺ CD8⁺ T cells were increased (Figures 6A,B). However, the ratio of IL-7Rα⁺ CD8⁺ T cells to perforin⁺ CD8⁺ T cells in non-lesional HC tissue was higher than that in the lesional tissue (Figure 6B), indicating that IL-7Rα^high EM CD8⁺ T cells comprised the majority of the population in non-inflammatory sterile tissue.

These results suggest that IL-7Rα^high EM CD8⁺ T cells may preferentially reside in non-inflammatory sterile tissue and that increased perforin expression in lesional tissue may be secondary to either the acquisition of perforin molecules by IL-7Rα⁺ CD8⁺ T cells or the enhanced migration of perforin⁺ CD8⁺ T cells by inflammatory stimuli such as cytokines, as supported by previous data (Figure 5C).