Human Thymic Epithelial Cells Inhibit IL-15- and IL-2-Driven Differentiation of NK Cells from the Early Human Thymic Progenitors

Antibodies

The following Abs were used in this study: anti-CD1a (OKT-6), anti-CD2 (35.1), anti-CD3 (OKT-3), anti-CD4 (OKT-4), and anti-CD8 (51.1) (American Type Culture Collection, Manassas, VA); anti-CD1a-CyChrome (CyChr; HI149), anti-CD56-PE (B159), anti-CD3-CyChr (UTCHT1), anti-CD34-APC (581), and anti-CD38-FITC (HIT2) (PharMingen, San Diego, CA); anti-NKR-P1A (DX12, HP3G10, 191B8), anti-p58.1 (HP3E4, EB6), anti-p58.2 (CHL, GL183), anti-CD94 (HP3B1, XA185, HP3D9) (Leukocyte Typing VI); and goat anti-IgG (H + L)-FITC (human serum adsorbed) (Life Technologies, Gaithersburg, MD). All Abs were titrated to obtain optimal concentrations for flow cytometric analyses.

Human TEC cultures

Normal human thymi were obtained as discard tissues through the Department of Cardiovascular Surgery and Pathology, Loyola University Medical Center from otherwise healthy children <1 year of age undergoing corrective cardiovascular procedures for congenital heart diseases. TEC cultures were initiated by an explant technique and propagated in supplemented medium as previously described (31). Contaminating thymic fibroblasts were removed by treatment with 0.02% EDTA, followed by anti-fibroblast mAb (1B10) treatment, and complement-mediated cell lysis (32). TEC cultures were monitored for contaminating macrophages and fibroblasts using Mo-1 and TE-7 mAb, respectively (American Type Culture Collection). TEC cultures used in the experiments were 85–93% positive for keratin as determined by reactivity to anti-keratin mAb NMF-116 (Dako, Carpinteria, CA).

Thymocyte cultures

Human thymocytes were gently teased from thymus tissue, isolated by Ficoll-Hypaque centrifugation, and slowly frozen in RPMI 1640 (Mediatech, Herndon, VA) containing 20% FCS (Life Technologies), 7.5% DMSO (Sigma, St. Louis), and 10 μg/ml gentamicin (Life Technologies). Frozen thymocytes were thawed quickly and incubated with RPMI 1640 containing 30% FCS, 100 μg/ml DNase I (Sigma), and 10 μg/ml gentamicin for 1 h at 37°C. Dead cells were removed by Ficoll-Hypaque centrifugation. In some experiments, thymocytes were used immediately after isolation from thymus tissue without freezing. Thymocytes were cultured in RPMI 1640 supplemented with 10% FCS (Life Technologies), 10% human pooled AB serum (Sigma), and 10 μg/ml gentamicin. For coculturing with TEC, thymocytes were placed in insert chambers (0.2 μm; Nunc Nalge, Naperville, IL) which were then placed in six-well cluster dishes previously plated with TEC. Thymocytes at 3.0 × 10⁶ cells were cultured with 5 × 10⁵ TEC in a total volume of 3.0 ml/well. The following cytokines were used: human rIL-2 (100 U/ml; Cetus, Emeryville, CA), IL-7 (10 ng/ml), SCF (10 ng/ml; BioSource International, Camarillo, CA), and IL-15 (10 ng/ml; R&D Systems, Minneapolis, MN). Human TGF-β3 was used at 5 ng/ml. The cells were cultured with IL-2 or IL-15 alone, or with combinations of IL-2, IL-7, and SCF or IL-15, IL-7, and SCF. Thymocytes were removed, washed, adjusted to equal cell density, and cultured with newly plated TEC every 5–7 days. Thymocytes were cultured up to 30 days and were harvested at various time points and assayed for cell surface markers by flow cytometry as described elsewhere (33).

Culture of CD34⁺-enriched TN thymocytes

CD34⁺-enriched TN thymocytes were isolated from freshly obtained thymocytes (not frozen) using the negative selection StemSep kit (Stem Cell Technology, Vancouver, Canada). Between 5 × 10⁸ and 2 × 10⁹ thymocytes were used for each enrichment separation. Freshly obtained human thymocytes were first incubated with anti-CD1a, anti-CD3 mAbs. The cells were then treated with goat anti-mouse Ab coupled to magnetic colloidal at a ratio of 80 particles/cell (BioMag; Perceptive BioSystems, Framingham, MA) for 30 min at 4°C. The CD1a⁺ and CD3⁺ cells were then removed using a strong magnetic plate. This step was repeated twice or was repeated until the cell suspension was clear of magnetic colloidal particles. The unbound cellular fraction containing predominantly the CD1a⁻CD3⁻ cells was then collected and subjected to a second round of negative selection using the StemSep system. In the second step, the cells were treated with a mAb mixture containing Abs to CD2, CD3, CD14, CD16, CD19, CD24, CD56, CD66, and glycophorin A (StemSep) for 30 min on ice. A total of 2 μl of the Ab mixture was used for each 1 × 10⁶ cells as suggested by the manufacturer. After incubation with magnetic colloid (60 μl/1.0 ml of cell suspension) for 30 min, the cells were loaded into a magnetic column and washed with 8 ml of medium according to the manufacturer’s protocol. The flow-through fraction containing lineage-negative cells enriched for CD34⁺ cells was collected. Purity of the isolated cell fraction was determined by flow cytometric analysis using CD34-APC, CD38-FITC, CD3-PerCP, CD56-PE, and CD1a-PE. The isolated cells (1 × 10⁵ cells/well) were cultured in 96 U-bottom well culture dishes in RPMI 1640 plus 10% FCS and 10% human AB serum. The following combination of cytokines were included: IL-2 (100 U/ml) + IL-7 (10 ng/ml) + SCF (10 ng/ml) and IL-15 (10 ng/ml) + IL-7 + SCF. In the coculture system, TEC were plated onto transwells at 1.5 × 10⁴cell/well. Flow cytometric analysis was performed at day 8 or 12. We also cultured CD34⁺-enriched TN thymocytes with cytokines in the presence of two molecular mass fractions (8,000–30,000 Da, >30,000 Da) isolated from TEC culture supernatants.

Fractionation of TEC culture supernatants

Culture supernatants from confluent cultures of human TEC were collected every 3 days. The pooled supernatants were first clarified by centrifugation and then filtered through a 0.2-μm filter. A total of 160 ml of culture supernatants was concentrated using Centricon Plus-80 centrifugal filter devices (Millipore, Bedford, MA). The supernatants were first filtered through a 30,000-Da cutoff and the filtrates were then filtered through a 8,000-Da cutoff filter. The molecular mass fractions were designated as >30,000 and 8,000–30,000 Da, respectively. The fractions were tested at various final concentrations.

NK cytotoxicity assay

Measurement of NK cell lytic activity for K562 and Daudi cells was performed as previously described (34). Briefly, 1 × 10⁷ K562 and Daudi cells were labeled with 100 μCi of [⁵¹Cr]sodium chromate for 1 h at 37°C. Labeled target tumor cells were cultured with effector cells at various E:T ratios for 4 h. The supernatants were removed using a Skatron harvesting press (Skatron, Sterling, VA) and associated radioactivity determined with a 4/600 ME PLUS automated gamma counter. Maximum release of radioactivity was obtained by adding 0.05% Nonidet P-40 (Sigma). Lytic activity was expressed as percent cytotoxicity as calculated by the formula: percent cytotoxicity = [(experiment dpm) − (minimum dpm)]/[(maximum dpm) − (minimum DPM)] × 100. All experimental means were calculated from triplicate values.

NK-CFU colony-forming assay

NK colony assays were performed according to a method previously described by Ogawa and colleagues (35). Briefly, postnatal CD34⁺ TN thymocytes were cultured at 50 × 10³ and 100 × 10³cells in 35-mm vented lid suspension culture dishes (Nagel Nunc). The cells were suspended in 1.0 ml of Iscove’s medium (Mediatech) containing 1.2% methylcellulose (Stem Cell Technology), 10% FCS, 10% human AB serum, 10 U/ml penicillin-streptomycin, and 5 × 10⁻⁵ M 2-ME. The cells were cultured with a combination of IL-2 (100 U/ml), IL-7 (10 ng/ml), and SCF (10 ng/ml) or IL-15 (10 ng/ml), IL-7, and SCF. The cells were cultured for 20 days at 37°C in a humidified atmosphere with 5% CO₂/95% air. The NK-CFU were identified by their large, diffuse morphology as previously described and confirmed by Ogawa and colleagues (35). The colonies were counted and the average number of colonies per duplicate well was determined. To confirm NK phenotype, single colonies were plugged and stained with Wright stain (New Comer Supply, Middleton, WI) to detect cytoplasmic granules.

Flow cytometric analysis

Cells were harvested and washed twice in PBS/2% BSA. Two hundred thousand cells (2 × 10⁵) were stained and analyzed by flow cytometry using a FACScalibur (Becton Dickinson, San Jose, CA). For three-color analysis, cells were stained with each mAb for 30 min on ice and washed three times with PBS/2% BSA. If nonconjugated mAb were used, cells were first stained with these mAb followed by washing and incubation with goat anti-mouse Ig (H + L)-FITC (human serum adsorbed). Data analyses were performed with CellQuest (Becton Dickinson) or WinMDI version 2.8 (Scripps Research Institute, San Diego, CA). Appropriate isotype control Abs were used to set markers where >98% of the cells are negative.

RT-PCR

Cells were lysed in 4 M guanidine isothiocyanate and total RNA was isolated by centrifugation through a 5.7 M cesium chloride gradient as previously described (36). Alternatively, mRNA was purified using the QuickPrep Micro mRNA purification kit (Pharmacia, Piscataway, NJ). Synthesis of first-strand cDNA was performed with 3 μg of total RNA or 100 ng of mRNA using a first-strand synthesis kit (Pharmacia) according to the manufacturer’s instructions. The PCR amplification reactions contain 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.2 mM MgCl₂, 0.125 mM dNTP, 0.25 μM primers, and 0.025 U/μl Taq (Life Technologies). For IL-7, two rounds of PCR amplification were performed: the first-round PCR was performed with 2 μl of cDNA samples for 25 cycles, and the second-round PCR was performed with 3 μl of the first-round PCR products for 30 cycles. IL-15 and GAPDH were performed with one round of 35 cycles of amplification. Each PCR cycle includes 15 s at 94°C, 20 s at 55°C, and 40 s at 72°C; the last cycle was followed by a final 7-min extension step at 72°C. PCR were performed with a GeneAmp PCR 2400 instrument (Perkin-Elmer/Cetus, Norwalk, CT). PCR products were separated in 2% agarose gel and visualized with ethidium bromide staining. Gel images were captured using a Kodak digital camera with one-dimensional image analysis software (version 2.1; Kodak, Rochester, NY). Primer sequences for human IL-7 and IL-15 were as follows: IL-7: forward, 5′-atg gcaaac aat atg aga gtg-3′; reverse, 5′-cca aat ttt att cca aca agt-3′; IL-15: forward, 5′-ttt cga aac cac att tga gaa-3′; reverse, 5′-cct cca gtt cct cac att ctt tg-3′.

Most ex vivo human thymocytes from postnatal thymus are committed to the T cell lineage and only a fraction of thymocytes is NK phenotype

Because the T/NK bipotential precursors were found in the human thymus (8, 10, 37), we determined whether the thymus affects NK development in vivo. Using the CD1a as a marker for T lineage commitment, we determined that >80% of total thymocytes express CD1a (Fig. 1⇓B), while only a small fraction of thymocytes are positive for the NK cell marker CD56 (Fig. 1⇓, C–F; Table I⇓). Further phenotypic analysis showed that the CD56⁺ thymocyte population is heterogeneous. We determined that 89–90% of the CD56⁺ thymocytes also express CD1a and CD3 and only 10–11% are negative for T cell-specific markers CD1a and CD3 (Fig. 1⇓, C–F). Thus, in postnatal human thymus, only 0.2% (n = 7) of total thymocytes are the classical NK cells that are CD1a⁻CD3⁻CD56⁺ (Table I⇓). Thus, the development of T/NK precursors is skewed toward the T lineage in postnatal human thymus. The poor development of intrathymic NK cells is not caused by the absence of cytokines, because we showed by RT-PCR that the human TEC do express IL-15 mRNA, the potent cytokine required for NK differentiation (Fig. 2⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Phenotypes of CD56-positive ex vivo human thymocytes. Three-color flow cytometric analysis was used to determine the phenotypes of CD56-positive ex vivo human thymocytes. Thymocytes were stained with anti-CD3-FITC, anti-CD56-PE, and anti-CD1a-CyChr and analyzed by flow cytometry. A, Forward light scatter and side light scatter profile; R1 depicted the analyzed thymocyte population; B, CD1a-CyChr, CD3-FITC; C, CD56-PE, CD3-FITC; D, CD56-PE, CD1a-CyChr gated on thymocytes (R1) and CD56⁺ (R3); E, CD56-PE, CD1a-CyChr; F, CD56-PE, CD3-FITC gated on thymocytes (R1) and CD56⁺ (R2). One representative result of six experiments performed with thymocytes from six different human pediatric thymi.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Expression of IL-15 and IL-7 mRNA by human TEC. IL-15 and IL-7 mRNA were detected by RT-PCR. GAPDH was used to monitor the quality of the isolated mRNA and cDNA synthesis. Data were from five different TEC cultures generated from five different human pediatric thymi.

Do TEC regulate NK cell development?

Because T/NK progenitors are present in the human thymus, but only a fraction of thymocytes are CD1a⁻CD3⁻CD56⁺ NK cells, we wanted to determine whether TEC affect NK cell development. Total thymocytes were cultured with a combination of IL-15 (10 ng/ml), IL-7 (10 ng/ml), and SCF (10 ng/ml) and were analyzed for expression of CD56 and CD3 by two-color flow cytometric analysis. The combination of the above cytokines has been shown to optimize the induction of NK differentiation from thymic precursors in vitro (8). Increased percentages of CD3⁻CD56⁺ cells were observed with increasing culture periods (Fig. 3⇓). In addition to the increases in the percentage of CD3⁻CD56⁺ cells, we also observed increases in the percentage of CD3⁺CD56⁺ cells and a concomitant decrease in the percentage of CD3⁺CD56⁻ cells; the percentage of CD3⁻CD56⁻ cells remained unchanged (Fig. 3⇓). In contrast, TEC, when cultured with thymocytes in transwells, inhibited the in vitro development of CD3⁻CD56⁺ (Fig. 3⇓). The decreases in the percentage of CD3⁻CD56⁺ cells were associated with increases in the percentage of CD3⁺CD56⁻ T cells and CD3⁻CD56⁻ cells (Fig. 3⇓). TEC had no effect on the development of CD3⁺CD56⁺ cells (Fig. 3⇓). Similar results were obtained when thymocytes were cultured with IL-2, IL-7, and SCF in the presence of TEC (data not shown). To confirm that the inhibition of CD3⁻CD56⁺ cell development by TEC was not due to selective TEC-induced outgrowth of other subsets, we determined the absolute number of CD3⁻CD56⁺ cells. Fig. 4⇓ shows results from six independent experiments in which cells were treated with either IL-2, IL-7, and SCF (left panels) or IL-15, IL-7, and SCF (right panels) in the presence or absence of TEC. We demonstrated that the number of NK cells increased with time when thymocytes were cultured with mixtures of cytokines in the absence of TEC. In contrast, a dramatic reduction of NK cell numbers was observed when TEC were present in thymocyte cultures. We observed a 6- to12-fold reduction in the number of CD3⁻CD56⁺ cells in thymocyte cultures treated with IL-2, IL-7 and SCF, and a 3- to 25-fold reduction in cultures treated with IL-15, IL-7, and SCF (Fig. 4⇓). We also determined that the absolute number of cells in the CD3⁺CD56⁻ and CD3⁺CD56⁺ subsets were not affected by TEC; a moderate increase in the absolute number of CD3⁻CD56⁻ cells however were observed (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Human TEC inhibit the development of CD3⁻CD56⁺ NK cells from total human thymocytes. Thymocytes were cultured with IL-15 (10 ng/ml), IL-7 (10 ng/ml), and SCF (10 ng/ml) in the presence or absence of human TEC as described in Materials and Methods. Thymocytes were harvested at days 5, 12, and 19, and 2.0 × 10⁵ cells were analyzed for the expression of CD3 and CD56 by flow cytometric analysis. At days 5 and 12, cells were adjusted to an equal cell number and cultured until the next indicated time points. Results from one representative experiment of six experiments performed with thymocytes isolated from six pediatric thymi.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Human TEC inhibit the development of CD3⁻CD56⁺ NK cells in thymocyte cultures. Thymocytes were cultured with IL-2 (100 U/ml), IL-7, and SCF (left panels, experiments 1–3) or IL-15, IL-7, and SCF (right panels, experiments 4–6) in the presence or absence of TEC as described. At a specific time, thymocytes were harvested and analyzed by flow cytometric analysis as described in the legend to Fig. 3⇑. The absolute number of NK cells was calculated from the percentage of positive cells and total cell count at the end of each culture period. Thymocytes were replated at an equal number at the beginning of each culture periods as described in Materials and Methods. ▪, Without TEC; ▧, with TEC.

TEC also inhibits NK cell development from the CD34⁺-enriched CD3⁻CD4⁻CD8⁻ TN thymocytes

To differentiate between mature NK expansion and differentiation of NK progenitor cells, we next determined whether TEC inhibit NK development from the isolated CD34⁺ TN progenitors. As previously shown by others (9, 10), we determined that the human CD34⁺-enriched TN thymocytes can differentiate into functional NK cells in our culture system. The ex vivo-enriched progenitor population was negative for CD3 and CD56 (Fig. 5⇓A). The CD34⁺-enriched TN cells were positive for CD38 and CD7, but were negative for the early T cell marker CD1a (data not shown), characteristic of the previously described T/NK progenitors (9, 10). After 14 days in culture with IL-2, IL-7, and SCF, 80% of the cells differentiated into CD1a⁻ CD3⁻CD56⁺ NK cells (Fig. 5⇓, B and C). However, we did not detect CD3⁺, TCRαβ⁺, or TCRγδ⁺ T cells (data not shown). Further phenotypic analysis showed that >60% of CD1a⁻ CD3⁻CD56⁺ cells express CD94 and the NK cell marker NKRP1A (CD161); however, the cells do not express the killer inhibitory receptor receptors p58.1 (HP3E4, EB6) and p58.2 (CHL, GL 183) (Fig. 5⇓, D–I). Functionally, the thymic-derived CD3⁻CD56⁺ NK cells display potent cytotoxicity for both K562 and Daudi target cells (Fig. 6⇓). Thus, we could generate functionally mature NK cells in vitro from the human CD34⁺-enriched TN thymocytes. Similar results were obtained with IL-15, IL-7, and SCF (data not shown). This in vitro culture system thus provides a functional model where regulation of NK differentiation can be investigated. We next examined whether TEC regulate NK differentiation from the CD34⁺-enriched TN thymocytes. As observed with total thymocyte cultures, TEC also inhibited the differentiation of CD3⁻CD56⁺ cells from the early thymic progenitor cells; while 85% of CD3⁻CD56⁺ cells were detected in the absence of TEC, only 42% of CD56⁺ cells were detected in the presence of TEC (Fig. 7⇓, top three panels). In a separate series of experiments, a 8,000- to 30,000-Da fraction derived from TES inhibits NK differentiation from thymic progenitors (Fig. 7⇓, bottom four panels). Although we noticed that the inhibition was not as completed as in the case of total thymocytes (Figs. 3⇑ and 4⇑), the consistent results from several independent experiments strongly indicate that TEC do regulate NK cell development. IL-7 and SCF have been shown to synergize with IL-2 or IL-15 in the induction of NK cell differentiation (26, 30). However, IL-15 or IL-2 themselves can induce NK differentiation from thymic and bone marrow precursors. Thus, we determined whether TEC directly inhibit IL-15- or IL-2-driven NK cell differentiation. We showed that TEC or the 8,000- to 30,000-Da fraction are potent inhibitors of IL-15- or IL-2- driven NK cell development from the CD34⁺ TN progenitors; the >30,000 Da showed no inhibitory effect (Fig. 8⇓). Furthermore, the development of both the CD3⁻CD56⁺ and CD3⁻CD56^Hi cells was dramatically reduced by TEC or by the TEC-derived 8,000- to 30,000-Da fraction (Fig. 8⇓). Since TGF-β has been shown to inhibit NK development from murine bone marrow (BM) (38) and we have shown that human TEC produce active TGF-β3 (39), we tested whether TGF-β3 affects NK development from thymocytes. Fig. 9⇓ shows that TGF-β3 has no effect on cytokine-driven NK development.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Phenotypic characterization of human CD34⁺ TN-derived NK cells in vitro. Isolated CD34⁺ TN cells were analyzed for the expression of CD3 and CD56 at day 0 (before culture, A) and at day 14 for CD3, CD1a, CD56, NKRP1A (CD161), CD94, p58.1 (HP3E4, EB6), and p58.2 (CHL, GL183) (B–I) by two-color flow cytometric analysis.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Cytotoxic activity of the CD34⁺ TN-derived CD3⁻CD56⁺ NK cells as determined by ⁵¹Cr release assay. Both K562- and Daudi-labeled cells were used as targets at various E:T ratios.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

TEC and TEC-derived soluble factor(s) inhibit the development of CD3⁻CD56⁺ NK cells from the CD34⁺ TN thymocytes. Isolated CD34⁺ TN thymocytes (1 × 10⁵ cells/well) were cultured with IL-2, IL-7, and SCF in the presence of TEC (1.5 × 10⁴) or two fractions of TES. The 80-fold concentrated fractions of 8,000–30,000 Da and >30,000 Da were tested at a 1/5 dilution (16× concentrated). Flow cytometric analysis was performed after 12 days in cultures. Top and bottom panels, Two independent experiments performed with two preparations of CD34⁺ TN thymocytes obtained from different pediatric thymi.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Similar experiments as described in the legend to Fig. 7⇑; however, CD34⁺ TN thymocytes were cultured with either IL-2 (100 U/ml) or IL-15 (10 ng/ml). TEC-derived fractions were from a different preparation (120-fold concentrated) and tested at a 1/5 dilution (24× concentrated). Rectangles depict R3 and R4 regions corresponding to CD3⁻CD56⁺ and CD3⁻CD56^high populations, respectively. Ig-CyChr and Ig-PE were used as isotype control Abs. Flow cytometric analysis was performed at day 12.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

Effects of TGF-β3 on NK cell development from human thymocytes. Thymocytes (2 × 10⁵ cells/200 μl/well; 96-well plate) were cultured with IL-7 and SCF (10 ng/ml) for 3 days, and the cells were washed and replated at the same cell density in the presence of IL-2 (100 U/ml) or IL-15 (10 ng/ml) alone or IL-2 or IL-15 plus IL-7 and SCF. TGF-β3 was used at a 5 ng/ml final concentration. Thymocytes were harvested after 12 additional days in culture and analyzed for CD3 and CD56 expression by two-color flow cytometry.

We also performed NK colony assays to determine the effect of TEC on NK cell development at the single precursor cell level. As shown in Table II⇓, TES served as a potent inhibitor of NK-CFU when CD34⁺ TN thymocytes were cultured with IL-2, IL-7, and SCF or IL-15, IL-7, and SCF; however, TES did not affect GM-CFU formation. Thus, TEC specifically inhibit the development of NK-CFU from thymic CD34⁺ TN progenitors.