Human Thymocytes Express CCR-3 and Are Activated by Eotaxin

Human thymus, removed as a consequence of cardiac surgery (average age, 6 months; age range, 1 week to 15 months; n = 10) was gently teased apart in cold Hanks’ balanced salt solution (HBSS), under sterile conditions. Total thymocytes were separated after Ficoll density gradient centrifugation and multiple HBSS washes at 4°C. Minor contaminating populations of monocytes were removed after Dynal bead separation of antibody-labeled (CD14) cells according to standard protocols. Populations of whole thymocytes were then stored on ice before use or stained for fluorescence-activated cell sorting (FACS) of individual subpopulations.

Human thymus was frozen in liquid nitrogen and 7-μm sections were cut and fixed in acetone (5 minutes). Nonspecific antibody binding was prevented by preblocking with 10% human (type AB) and goat serum (Pel-Freez Clinical Systems, Brown Deer, WI) and subsequently with avidin/biotin Blocking Kit (Vector Laboratories, Burlingame, CA). Fixed sections were incubated with mouse monoclonal antibody to human eotaxin (1 hour; 10 μg/mL; R&D Systems, Minneapolis, MN) or an isotype-matched control. Slides were subsequently rinsed in TBS/0.1% bovine serum albumin (BSA) and incubated with biotinylated goat antimouse Ig (2.5 μg/mL; 1 hour). After a second wash, sections were incubated with alkaline phosphatase/avidin-biotin complex for 30 minutes. Sections were incubated with Vector Red alkaline phosphatase substrate for 10 minutes (room temperature), and antibody staining was visualized by microscopy.

Thymocytes were labeled with anti-CD4-fluorescein isothiocyanate (FITC) and anti-CD8-phycoerythrin (PE) (Becton Dickinson, Mountain View, CA). After gating on appropriate populations, single-positive (SP; CD4⁺ or CD8⁺) populations, CD4⁺CD8⁺ double-positive (DP), and CD4⁻CD8⁻ double-negative (DN) populations were sorted on a FACStar Plus (Becton Dickinson) cell sorter using standard methods. Sorted populations were used immediately for bioassay or frozen at −80°C before RNA preparation.

Chemotaxis of thymocytes was performed as previously described22 using the 48-well modified Boyden chamber from NeuroProbe (Cabin John, MD). Cells that had migrated onto the underside of an 8-μm polyvinyl-pyrrolidone free (PVPF) polycarbonate membrane after 1 hour of incubation at 37°C were quantified under high power microscopy (400×), and results are expressed as the mean ± SEM of 5 high power fields.

Whole thymocyte populations were labeled with Indo-1AM (3 μmol/L final concentration; Molecular Probes, Eugene, OR) for 45 minutes at room temperature. After labeling, cells were resuspended in 1 mL HBSS containing 1% BSA. Measurement of calcium flux was performed according to previously published methods23 using a PTI fluorimeter (South Brunswick, NJ).

Unfractionated thymocytes (10⁷ cells) were washed once in phosphate-buffered saline (PBS), and the pellets were used for DNA-free RNA isolation using the SNAP Total RNA Isolation Kit (Invitrogen, San Diego, CA) according to the manufacturer’s protocol. First-strand cDNA synthesis for use in PCR reactions was generated using the cDNA Cycle Kit (Invitrogen). Parallel reactions were performed without AMV Reverse Transcriptase enzyme to control for genomic DNA contamination in the subsequent PCR reactions. PCR analysis was then performed using primers specific for human CCR-3 (sense, 5′-ATGACAACCTCACTAGATACAGTTG; antisense, 5′-CTAAAACACAATAGAGAGTTCCGG)12 for 40 cycles of 94°C for 1 minute, 55°C for 2 minutes, and 72°C for 3 minutes. Similar reactions were run with human CCR-3 cDNA (positive control) and water in place of first-strand cDNA (negative control). Amplified product (1.1 kb) was resolved on a 1% agarose gel containing ethidium bromide. Subsequently, the gel was denatured in Hybridization denaturing solution (1.5 mol/L NaCl, 0.5 mol/L NaOH; 5-Prime–3-Prime, Boulder, CO) for 30 minutes at room temperature. The PCR products were then transferred in 20× SSC (3 mol/L NaCl, 0.3 mol/L Na citrate, pH 7.0; Boehringer Mannheim, Indianapolis, IN) to Nytran nitrocellulose membranes (Schleicher & Schuell, Keene, NH) overnight. The transferred products were cross-linked to the membrane (UV Stratalinker; Stratagene, San Diego, CA), and the membrane was neutralized for 30 minutes in Hybridization neutralization solution (1 mol/L Tris-HCl, pH 7.4, 1.5 mol/L NaCl; 5-Prime–3-Prime). Prehybridization was performed at 42°C in Church buffer (7% sodium dodecyl sulfate [SDS], 0.5 mol/L NaHPO₄, 0.5 mmol/L EDTA, pH 8), and the membrane was probed with 10⁶ cpm/mL³²P-labeled oligomer corresponding to an internal sequence of the CCR-3 gene: 5′-CGTTATGGCCATCTGCTACACAGG-3′. The blots were washed 3 times in 0.2% SSC/0.1% SDS (42°C) and then exposed on Biomax radiography film (Eastman Kodak, Rochester, NY).

Expression of CCR-3 on human thymocytes was analyzed using specific antibody 7B1124 (NIH AIDS Reagent Program, Rockville, MD). Highly purified thymocytes were stained according to standard methods with 7B11 for 30 minutes (1/500 dilution) and then with secondary FITC-coupled antibody (goat antimouse IgG2a; Pharmingen). Cells were washed once and stained with a cocktail of CD4-PE and CD8-PE-Cychrome (Becton Dickinson) for three-color analyses using a Becton Dickinson FACScan.

Equilibrium binding of ¹²⁵I-labeled eotaxin in the presence of various concentrations of unlabeled human chemokines was performed as previously described.25 ¹²⁵I-labeled eotaxin was obtained from DuPont NEN (Boston, MA). Recombinant chemokines were obtained from PeproTech (Rocky Hill, NJ) or R&D Systems (Minneapolis, MN). Briefly, 2.5 × 10⁶ highly purified thymocytes were suspended in a 200 μL reaction volume containing 0.1 nmol/L¹²⁵I-labeled eotaxin and various concentrations of unlabeled chemokines in the following buffer (25 mmol/L HEPES, 1 mmol/L CaCl₂, 5 mmol/L MgCl₂, and 0.5% BSA, adjusted to pH 7.4) for 90 minutes at 22°C. All assays were performed with 6 identical reactions per condition. The reactions were then filtered onto PEI-treated Whatman GF/C filters using a Packard cell harvester (Meriden, CT). Filters were washed with the following buffer (25 mmol/L HEPES, 1 mmol/L CaCl₂, 5 mmol/L MgCl₂, and 0.5 mol/L NaCl, adjusted to pH 7.4). Radioactivity retained on the filters was quantified using a Packard Top Count (Downers Grove, IL). The data obtained were analyzed using WaveMetrics IGOR software (Lake Oswego, OR) and/or the LIGAND program to determine the displacement binding characteristics, dissociation constants (kd), and average number of binding sites per cell.

Isolated human thymocytes (10⁸/well of a 6-well Costar culture plate) were placed in culture for 48 hours in the presence of increasing concentrations of recombinant human eotaxin (PeproTech, Rocky Hill, NJ), hrIL-2, hrIL-7 (DNAX Research Institute, CA), or phytohemagglutinin (PHA; Wellcome, Chapel Hill, NC). Cell-free supernatants were then collected and frozen at −80°C before assay. Commercially available ELISA kits for the chemokines IL-8, MIP-1α, MIP-1β, and RANTES and the cytokines IL-4, IL-5, IL-10, and γ-interferon (γ-IFN; R&D Systems) were used according to the manufacturer’s protocols.

Immunohistochemical analyses of eotaxin, using an antihuman eotaxin antibody showed a distinct pattern of expression in human thymus. Sections showed a punctate staining in large cells with dendritic morphology, primarily in the medulla (Fig1). A significant proportion of eotaxin-positive cells also expressed CD11b in double-staining experiments (not shown). Additionally, there was staining of eotaxin in most Hassle’s corpuscles examined.

To further establish the significance of eotaxin expression in the thymus, equilibrium binding experiments assessing the interaction of¹²⁵I-eotaxin with highly purified thymocytes were performed. These binding analyses indicated that a receptor for eotaxin was expressed on thymocytes (Fig 2A). Eotaxin displayed a binding affinity of 2 nmol/L, and Scatchard analysis indicated the expression of an average of 1,100 binding sites per cell over the entire population. Heterologous chemokine competition for ¹²⁵I-eotaxin binding to thymocytes was consistent with thymocyte expression of CCR-3, the receptor known to bind eotaxin,12-14 because eotaxin tracer was competed from thymocytes by MCP-2, MCP-3, and MCP–4 that are also bound by CCR-3 (Fig 2B). Additionally, IL-8 and Gro-α were used as negative controls and in both cases failed to compete for eotaxin binding. Although this specific, displaceable binding is represented in Fig 2, in both homologous and heterologous competition analyses we noted a considerable level of nonspecific binding of the¹²⁵I-eotaxin to thymocytes. The nature of such binding is unresolved and under investigation; however, we cannot rule out a role for cell surface glycosaminoglycans and or other cell surface moieties.

As a further test of specificity, we analyzed unfractionated thymocytes for the presence of CCR-3 message by RT-PCR. Figure 2c shows the PCR data, with the presence of a specific 1.1-kb message in the fractions containing reverse transcriptase, but not in the −RT or water controls. To further confirm the specificity of this message, we demonstrated that the 1.1-kb mRNA bound a ³²P-labeled oligonucleotide corresponding to an internal sequence in the CCR-3 gene (Fig 2D).

Using a specific anti–CCR-3 antibody (7B11), CCR-3 expression was shown on human thymocytes (Fig 3B) when compared with an isotype-matched negative control. Further analyses, gating on individual subsets of thymocytes, showed differences in expression intensity; CD8⁺ SP cells stained with greatest intensity with little if any staining on CD4⁺/CD8⁺ DP or CD4⁻/CD8⁻ DN cells. CD4⁺ SP staining was more ambiguous, with expression observed in 3 of 4 thymocyte samples. However, even in the positive samples, there was a high intensity of staining only on a very small percentage of cells.

We next investigated the biological activity of eotaxin on human thymocytes, preliminarily analyzing standard readouts for chemokine-induced cell activation, including chemotaxis and calcium mobilization. Up to 10% of the unfractionated thymocytes were found to migrate in response to eotaxin in a dose-dependent manner (Fig 4A). Significant migration was obtained with 0.01 nmol/L eotaxin and maximal eotaxin-induced migration was observed at 10 nmol/L, in contrast to the positive control lymphotactin, which induced maximal migration at 1 nmol/L. Additionally, chemotaxis in response to either chemokine was abolished by preincubation of thymocytes with 100 ng/mL pertussis toxin (PTX; not shown), suggesting coupling of these functional chemokine receptors to a Gαi G-protein. Further analysis of the effects of eotaxin on thymocytes showed that this chemokine was capable of inducing a rapid transient mobilization of Ca²⁺, typical of chemokine-induced Ca²⁺ flux profiles in leukocytes (Fig4B), but contrasting that induced by anti-CD3 monoclonal antibody, which was more prolonged (Fig 4B, inset).

On investigation of fractionated thymocyte subpopulations, it was observed that the DP (CD4⁺ CD8⁺) and SP CD4⁺ and CD8⁺ populations migrated in response to eotaxin, whereas the DN (CD4⁻CD8⁻) cells failed to migrate (Fig 5). In the three responder populations, maximal migration was also observed at 10 nmol/L eotaxin; however, significant numbers of thymocytes had migrated in response to 0.01 nmol/L eotaxin, indicating a potent action of eotaxin in this assay. In addition, it was interesting to note that, in accordance with the higher levels of CCR-3 on CD8⁺ thymocytes than on CD4⁺ thymocytes, a considerably greater number of CD8⁺ SP cells than CD4⁺ SP cells migrated in response to eotaxin. In a similar manner to unfractionated cells, migration was completely inhibited by PTX pretreatment (not shown).

To further assess the biological activities of eotaxin, thymocytes were cultured in the presence of increasing concentrations of eotaxin to determine whether there were any proliferative or phenotypic changes. Over a period of 48 hours, eotaxin induced little if any change in CD25, CD3, CD4, and CD8 expression, as measured by FACS analysis (not shown). In addition, there was no change in total cell numbers or³H-thymidine incorporation, in contrast to IL-7, IL-2, or PHA, which induced thymocyte proliferation as expected (not shown). However, interestingly, eotaxin induced significant increases in chemokine release from thymocytes at doses of 1 to 100 nmol/L (Table 1). Specifically, a mean of 1.4 ng of MIP-1β (range, 1.17 to 1.84 ng) was released from 10⁶thymocytes stimulated with 100 nmol/L eotaxin, whereas a mean of 100 pg (range, 82 to 131 pg) was released from PBS-stimulated thymocytes. Likewise, 205 pg (range, 65 to 371 pg) and 229 pg (range, 106 to 345 pg) of RANTES and IL-8, respectively, were secreted from eotaxin-stimulated thymocytes. In contrast, only 35 pg (range, 16 to 55 pg) and 130 pg (range, 45 to 198 pg) of these two chemokines were detected in PBS-stimulated thymocytes.

We have demonstrated that eotaxin protein is expressed in human thymus and that, upon binding its specific receptor CCR-3, it elicits chemotaxis, Ca²⁺ mobilization, and cytokine release by thymocytes. These results provide information on a potentially novel biological function for eotaxin, beyond the activation of eosinophils and Th2 lymphocytes. The functional correlate of this restricted expression of eotaxin in the medullary region and Hassle’s corpuscles of the thymus remains to be elucidated. Although the eotaxin knock-out mouse does not show compromized T-cell development, a systematic analysis of the migratory patterns of thymic precursors and thymocytes has yet to be undertaken. T-cell development in the thymus requires the regulated migration of selected progenitors from the cortex to the medulla, ultimately resulting in the emigration of selected T cells into the periphery. Interestingly, CD8⁺ cells expressed the highest levels of CCR-3 receptor, and a greater number of cells in this subpopulation migrated in response to eotaxin, suggesting a preferential activation role for eotaxin in CD8⁺thymocytes. Although far from unequivocal, the data presented here suggest that eotaxin may be involved in the regulation of cell trafficking in the thymus, because (1) its expression is highly localized to the medulla and (2) thymocytes at different stages of maturation respond differently to eotaxin stimulation. In contrast, Zhang et al26 did not detect any lymphoid cell staining when using the 7B11 antibody directed against CCR-3. The reasons for this discrepancy remain unclear, although it is relevant to note that the antibody is specific for human CCR-3; thus, species differences as well as methodological and preparative differences in obtaining thymocytes may account for any contrasts between the two studies.

Eotaxin induced a greater than 10-fold increase in the chemokines RANTES and MIP-1β, suggesting a prominent role for these two chemokines in thymic physiology. However, the significance of such a robust chemokine release in the thymus is unclear. Of the chemokines measured, only MIP-1β has thus far been shown to have significant effects on thymocytes,27 although both RANTES and MIP-1β can bind to the same receptor (CCR-5). Interestingly though, both RANTES and IL-8 have been shown to play significant migration and activation roles in lymphoid cells.28-30 The induction of these chemokines may therefore represent an autocrine feedback mechanism in the regulation of thymocyte migration within and from the thymic environment. However, the validation of such an hypothesis does await further experimentation. It is relevant to point out that recent studies have demonstrated the induction of IFN-γ and tumor necrosis factor-α release after T-cell activation by various chemokines.31 Although we have not directly measured other cytokines, it is interesting to speculate that this precedent for chemokine-induced chemokine release may be of functional significance in lymphocyte populations.

Although we have not begun to investigate the maturation stages of the lymphocyte populations expressing CCR-3 or their Th1/Th2 polarization, our findings of CD8⁺ lymphocyte expression of CCR-3 supports the recent findings of CD8⁺ T-cell clone (Th2) expression.32 Whereas CD4⁺ cells were less responsive to eotaxin in a chemotaxis assay than CD8⁺cells, the CD4⁺ SP population still showed fairly robust migratory potential to this chemokine. Because the levels of detectable CCR-3 on CD4⁺ SP cells were very low, the reason for this biological effect of eotaxin is as yet unclear. However, it may support the phenomenon of spare receptors or suggest that eotaxin can act through an alternative receptor, such as CXCR-333; alternatively, there may be indirect activation capacities through the release of other chemoattractant factors. The secretion of MIP-1β, RANTES, and IL-8 may indeed fulfill such a role.

Although CCR-3 has been shown to function as a coreceptor for HIV-1,19,20 the role of this receptor in HIV-1 infection in the embryo and neonate remains questionable. There are significantly fewer CCR-3 than CXCR-4 and CCR-5 receptors in the thymus (Res et al, manuscript submitted; and Zhang et al26); however, its presence complements the repertoire of primary HIV coreceptors at this site of T-cell development and thus may provide a mechanism for efficient HIV entry.

These observations detail the expression of both eotaxin and its specific receptor in a tightly regulated T-cell environment. Whereas the biological importance of this action is, as yet, not fully understood, these data suggest that eotaxin may play an important role in thymic physiology.