In contrast to lymphocyte homing, little is known about molecular cues controlling the motility of lymphocytes within lymphoid organs. Applying intravital two-photon microscopy, we demonstrate that chemokine receptor CCR7 signaling enhances the intranodal motility of CD4+ T cells. Compared to wild-type (WT) cells, the average velocity and mean motility coefficient of adoptively transferred CCR7-deficient CD4+ T lymphocytes in T cell areas of WT recipients were reduced by 33 and 55%, respectively. Both parameters were comparably reduced for WT T lymphocytes migrating in T cell areas of plt/plt mice lacking CCR7 ligands. Importantly, systemic application of the CCR7 ligand CCL21 was sufficient to rescue the motility of WT T lymphocytes inside T cell areas of plt/plt recipients. Comparing the movement behavior of T cells in subcapsular areas that are devoid of detectable amounts of CCR7 ligands even in WT mice, we failed to reveal any differences between WT and plt/plt recipients. Furthermore, in both WT and plt/plt recipients, highly motile T cells rapidly accumulated in the subcapsular region after subcutaneous injection of the CCR7 ligand CCL19. Collectively, these data identify CCR7 and its ligands as important chemokinetic factors stimulating the basal motility of CD4+ T cells inside lymph nodes in vivo.
Naive T cells constantly recirculate through LNs in search of cognate antigen (1). Upon encountering antigen in the context of peptide–MHC complexes on the surface of antigen-presenting cells, naive T cells become activated, change their movement behavior, and start to undergo rapid cell division that leads to clonal expansion of T cells specific for the antigen recognized (2–4). In an antigen-inexperienced host, the frequency of naive T cells specific for any single given antigen is extremely low, typically 1 in 105–106 T cells (5). Therefore, to allow for an effective induction of a primary immune response, it is of utmost importance that naive T cells are able to scan large areas within LNs to encounter those DCs that actually present their cognate antigen.
Several two-photon microscopy studies analyzing intranodal T cell migration in situ or after LN explantation (3, 6–8) have established a model of naive T cells migrating in a mode of random walk while scanning the T cell area. However, it remains largely unknown which factors actually determine the intranodal motility of naive T cells. Different chemokines, most prominently CCL19 and CCL21, are constitutively present in the T cell area with corresponding chemokine receptors expressed on naive T lymphocytes (9, 10). Therefore, it seems possible that chemokines act as modulators of the basal motility level of naive T cells. The key functions of CCR7 for the homing of B and T lymphocytes (10, 11) and DCs (12) to LNs, the positioning of activated B and T cells within LNs (13, 14), and the chemotaxis of activated B cells (15) and thymocytes (16) are well established. Furthermore, recent studies have suggested a chemokinetic effect of CCR7 ligands on naive T cells in vitro (17, 18); however, it is still unclear whether CCR7 signaling affects the intranodal motility of naive T lymphocytes in vivo.
In this study, we provide a detailed analysis of the role of CCR7 signaling for the motility of CD4+ T cells inside the popliteal LNs (pLNs) of living animals by intravital two-photon laser- scanning microscopy. We identify CCR7 and its ligands as important chemokinetic factors stimulating the basal motility of naive CD4+ T cells inside the T cell area, potentially contributing to the effective scanning of resident DCs for cognate antigen.
Results And Discussion
To directly address whether CCR7 signaling affects the intranodal motility of naive T lymphocytes, we used intravital two-photon laser-scanning microscopy of the pLNs of anesthetized mice as described previously (3), imaging the movement of CD4+ T cells in different LN regions. Between 24 and 48 h after adoptive i.v. transfer of TAMRA-labeled WT CD4+ T cells, WT recipients were anesthetized and the pLN was micro-surgically prepared for intravital imaging. Only very few transferred T cells were present in the subcapsular region of the LN, displaying a considerably lower average velocity (6.4 μm/min; Fig. 1 C, leftmost panel) than described for T cells in the paracortical T cell zone (10–15 μm/min) (4, 6, 15, 19, 20). This is reminiscent of earlier reports by Huang et al. (21) suggesting a correlation between T cell motility and distance to the LN surface. As shown in Fig. 1 A, CCR7 ligands are abundantly present in the paracortical T cell area but completely absent in the subcapsular region of WT LNs, fostering the idea that absence of CCR7 ligands might cause a lower overall T cell motility. To test this hypothesis, we artificially altered the regular distribution of CCR7 ligands within the pLNs by s.c. injection of 1 μg CCL19 together with FITC-labeled dextran of 150 kD as a tracer into the right hind footpad. Approximately 5 min after injection, the first TAMRA-labeled WT T cells were visibly migrating inside the subcapsular sinus (SCS), and within the next 15 min, this sinus was filled with highly motile TAMRA-labeled T cells (Fig. 1 B and Video S1, which is available at http://www.jem.org/cgi/content/full/jem.20061706/DC1). In addition to the adoptively transferred T cells, numerous nonlabeled host cells, visible as black “shadows” inside the green field of fluorescence elicited by the FITC-dextran, migrated inside the SCS. We could not identify the exact “route of entry” of these cells in our imaging setup; however, we can exclude a recruitment into the SCS via afferent lymphatics (Video S2). We therefore speculate that under the chemotactic influence of high CCL19 concentrations inside the SCS, WT cells are possibly recruited into the SCS via the medullary sinus by retrograde migration. When, on the other hand, CCR7−/− CD4+ T cells had been transferred into WT recipients, the s.c. injection of CCL19 failed to induce mobilization of TAMRA-labeled CCR7−/− cells into the SCS, whereas unlabeled endogenous WT cells were amply present in this area (Fig. 1 B and Video S3).
To quantitatively characterize this CCR7-mediated cell motility in the SCS, we imaged different parts of the subcapsular region of the pLN at depths of ∼20–60 μm in WT as well as in plt/plt (plt) recipients, with the latter completely lacking intranodal CCR7 ligands. Although the few WT CD4+ T cells present under physiological conditions in the subcapsular area of the pLN exhibited equally low average velocities in both recipients (WT, 6.4 μm/min; plt, 5.4 μm/min; Fig. 1 C), we observed a dramatic increase in the motility of WT donor T cells in the SCS region in the presence of exogenous CCL19. Importantly, this holds true for WT as well as for plt recipients, as the median of the average cell velocities equally increased by a factor of 2.6 (from 6.4 to 16.6 μm/min in WT, and from 5.4 to 14.0 μm/min in plt recipients; Fig. 1 C). These results demonstrate that the basic migration behavior of adoptively transferred WT T cells is comparable in superficial LN areas of WT and plt recipients and seems to be influenced by the presence or absence of CCR7 ligands in a given region of the organ.
To further elucidate the importance of CCR7 signaling for the intranodal motility of CD4+ T cells, we subsequently analyzed the LN compartment in which CCR7 ligands are naturally present (Fig. 1 A). We therefore imaged adoptively transferred WT or CCR7−/− CD4+ T cells in the paracortical T cell areas of WT and plt recipients at imaging depths of 120–200 μm below the LN capsule. At first glance, the overall mode of movement seemed to be comparable in the presence or absence of CCR7 signaling (Fig. 2 A). However, the velocity was obviously reduced for CCR7−/− T cells in WT recipients and for WT T cells migrating in the T cell zone of plt recipients (Videos S4–S6, available at http://www.jem.org/cgi/content/full/jem.20061706/DC1). The quantitative analysis by automated cell tracking confirmed these observations, revealing a significant decrease of the median of the average cell velocities of CCR7−/− T cells (WT, 14.6 μm/min vs. CCR7−/−, 9.7 μm/min; Fig. 2 B). Of interest, WT CD4+ T cells in the paracortical T cell areas of plt recipients showed a comparably reduced motility (median of average cell velocities, 9.4 μm/min; Fig. 2 B). This indicates that CCR7 signaling indeed affects the basal velocity level of T lymphocytes inside the T cell area of noninflamed LNs.
Analogous to Miller et al. (19) we calculated the mean motility coefficient (MC), representing the propensity of a cell to move away from its original starting position. This general motility parameter is largely reduced once CCR7 or its ligands are missing (WT T cells in WT recipient, 61.0 ± 5.8 μm2/min; CCR7−/− T cells in WT recipient, 27.7 ± 3.3 μm2/min; WT T cells in plt recipient, 30.7 ± 3.5 μm2/min; for comparison, WT B cells in WT recipient, 11.1 ± 2.5 μm2/min; Fig. 2 C). These findings strongly indicate that the reduced intranodal velocity of T lymphocytes in the absence of CCR7 signaling is associated with a reduced ability to effectively sweep large parts of the T cell area that could lead to T cells contacting fewer DCs per time in their search for cognate antigen.
Plotting the mean displacement of all tracked cells for any given experimental condition over the square root of time yields in all cases a pattern that is in accordance with random walk movements (Fig. 2 D). Therefore, lack of CCR7 signaling obviously does not alter the general mode of T cell movement. The meandering index (total displacement/path length of a cell track) allows for a more detailed analysis of the straightness of T lymphocyte movement, with a value of 1 representing a completely linear cell track. As displayed in Fig. 2 E, there was only a minor reduction of the mean meandering index in the absence of CCR7 signaling (WT T cells in WT recipient, 0.61 ± 0.01; CCR7−/− T cells in WT recipient, 0.56 ± 0.02; WT T cells in plt recipient, 0.55 ± 0.02; for comparison, WT B cells in WT recipient, 0.50 ± 0.03). Analysis of the turning angle distribution confirms this result, as CD4+ T cells in the absence of CCR7 signaling turn at slightly larger turning angles, indicating sharper turns and therefore a less linear movement (not depicted).
Analyzing lymphocytes isolated from pooled LNs and spleens of WT and CCR7−/− animals, we found comparable percentages of naive (CD62L+ CD44low) cells within the CD4+ T cell population (WT, 76.1%; CCR7−/−, 70.7%; Fig. 3 A). However, transfer of MACS-purified bulk CD4+ T cells might nonetheless have introduced a selectional bias in the motility analysis because nonnaive T cell subsets, potentially exhibiting a different intrinsic motility, could have been preferentially recruited to LNs in the absence of CCR7 function. We therefore repeated these experiments with WT and CCR7−/− CD4+ T cells that were additionally sorted for low expression of CD44 before adoptive transfer. In both cases, the total percentage of CD4+ CD44low T cells was equally increased from ∼19 to ∼84% by FACS sorting (not depicted). Imaging the motility of the transferred naive CD4+ CD44low T lymphocytes in the T cell area of WT recipients, we obtained results comparable to the experiments using bulk CD4+ T cells. The median of the average cell velocities was slightly increased for both WT and CCR7−/− cells (WT, 15.1 μm/min vs. CCR7−/−, 11.4 μm/min; Fig. 3 B). The same holds true for the MC (WT T cells in WT recipient, 71.8 ± 11.2 μm2/min; CCR7−/− T cells in WT recipient, 34.1 ± 3.3 μm2/min; Fig. 3 C), whereas the mean meandering indices were almost unchanged (WT T cells in WT recipient, 0.61 ± 0.01; CCR7−/− T cells in WT recipient, 0.56 ± 0.01; Fig. 3 D). After completion of intravital microscopy, the recipient mice were analyzed by FACS for the subset composition of TAMRA-labeled cells that had homed into peripheral LNs (and were therefore subject to the imaging analysis). As depicted in Fig. 3 (E and F), TAMRA+ cells found in peripheral LNs were mostly naive (CD62L+ CD44low) CD4+ T cells after transfer of WT (96.6% of transferred CD4+ T cells) as well as of CCR7−/− (93.3% of transferred CD4+ T cells) cells, ruling out an influence of biased homing of different T cell subsets on the motility analysis performed in these experiments (for a direct comparison of CD44 expression levels see Fig. S1, available at http://www.jem.org/cgi/content/full/jem.20061706/DC1). Collectively, these results demonstrate that CCR7 and its ligands positively influence the intranodal cell migration speed of naive CD4+ T lymphocytes, whereas the directionality of T cell movement behavior is largely unaffected by lack of CCR7 function.
Finally, we aimed to directly visualize the alteration of intranodal T cell motility by manipulation of CCR7 signaling during imaging (Fig. 4). 24 h after adoptive transfer of CD4+ WT T cells into plt recipients, mice were i.v. injected with 200 μg anti-CD62L mAb to block any further homing of lymphocytes to LNs (3). Imaging the movement behavior of T lymphocytes in the paracortical T cell zone 2 h later, we observed a significant reduction of T cell velocity and MC compared with the movement of WT T cells in WT recipients (Fig. 4, A and B; see also Fig. 2). To restore the presence of CCR7 ligands within the LNs of the plt recipients, we subsequently i.v. injected 100 μg CCL21. During the next 2.5 h of imaging, the motility of WT T cells in the plt T cell area increased substantially. Within this time period, velocity and MC reached approximately levels observed imaging WT T cells in WT recipients (Fig. 4, A and B; see also Fig. 2), whereas the meandering index was only moderately increased (Fig. 4 C). This reversal of the motility reduction observed in the absence of CCR7 ligands (T cell zone of untreated plt recipients) by systemic application of CCL21 strongly argues for a direct positive influence of CCR7 signaling on the intranodal motility of T lymphocytes.
The velocities of WT T cells reported in this study are in good agreement with data of existing two-photon microscopy studies of in situ and explanted LNs. Zinselmeyer et al. (20) also reported 15 μm/min for CD4+ T cells in explanted LNs, and Okada et al. (15) found ∼14 μm/min for helper T cells in a similar setup. Earlier studies by Miller et al. (4, 6, 19) resulted in somewhat lower T cell velocities of 10–12 μm/ min. Importantly, Miller et al. performed the cell tracking on z-projections of the imaging volumes, thereby obtaining lateral velocities for movements in the xy-plane only. This should result in velocity values that are by a factor of square root (3/2) = 1.225 lower compared with three-dimensional cell tracking, as the z-compound of the movement vector is not contributing to the velocity calculation (assuming that a T lymphocyte migrating in the T cell area of a LN by random walk does not exhibit any directional bias). The velocity of WT B cells (7.1 μm/min) is also comparable to previously published data (15, 19), further confirming the validity of the experimental setup used.
Previous in vitro studies suggested that the higher motility of T cells migrating in a three-dimensional collagen gel in the presence of DCs could be due to soluble factors secreted by DCs (22), and that CCL19 might have a positive effect on the scanning behavior of naive T cells (17). During the submission process of this article, Stachowiak et al. (18) reported that CCL19 and CCL21 induce long-lived chemokinesis in CD4+ T cells migrating on different adhesion molecules in vitro, further supporting the in vivo analysis of the present study.
Using intravital microscopy, we provide for the first time direct in vivo evidence for a chemokinetic function of CCR7 and its ligands stimulating the intranodal basal motility of naive T lymphocytes. Interestingly, a recent study by Bajenoff et al. (23) provides evidence that the fibroblastic reticular cell network essentially defines and supports the apparent random walk movement of naive T cells within the LN paracortex. It therefore seems possible that CCR7 ligands immobilized on fibroblastic reticular cells or extracellular matrix surfaces, rather than soluble chemokines, are mediating the observed CCR7-dependent stimulation of the intranodal T cell velocity. Regardless of the actual state of the chemokines involved, the enhanced motility allows for a faster, more widespread movement of T cells in the T cell area, thereby potentially enhancing the likelihood of encounters with resident DCs presenting rare cognate antigens. In light of these findings, it seems possible that a reduced scanning performance, in addition to the known homing defects, could contribute to the delayed and in some cases impaired immune responses observed in CCR7−/− animals. It will be interesting to see in subsequent studies how other parameters, such as recruitment of inflammatory cells or ongoing immune responses, influence intranodal T cell motility.
Materials And Methods
C57BL/6, C57BL/6 CCR7−/–, and C57BL/6 plt/plt were bred at the animal facility of Hannover Medical School under specific pathogen-free conditions. Animal experiments have been approved by the institutional review board and local authorities.
Antibodies and reagents.
Anti-CD3 (clone 17A2), anti-CD4 (clone RMCD4), anti-CD62L (clone MEL-14), and anti-B220 (clone TIB 146) antibodies were provided by E. Kremmer (GSF, Munich, Germany) and used in the following conjugations: anti–CD3-Cy5, anti–CD4-PacificOrange, anti–CD4-Cy5, and anti–B220-Cy5. These antibodies were also used in this study: anti–CD44-APC, anti–CD62L-FITC, StreptAvidin–PE-Cy7 (BD Biosciences), goat anti-mCCL19, goat anti-mCCL21 (R&D Systems), donkey anti–goat IgG-peroxidase (Jackson ImmunoResearch Laboratories). 150 kD of FITC-labeled dextran was obtained from Sigma-Aldrich, and recombinant mCCL19 was from R&D Systems.
Acetone-fixed 8-μm cryosections of pLNs were rehydrated, preincubated with 10% mouse serum, and stained with a cocktail of antibodies in 2.5% serum. Nuclei were stained with DAPI. Images were acquired using a Zeiss Axiovert 200 M microscope. Detection of anti-mCCL19/anti-mCCL21 antibody binding was enhanced using the TSA-Cy3 Tyramid Signal Amplification system (PerkinElmer).
6–8-wk-old donor mice (C57BL/6 or C57BL/6 CCR7−/–) were killed by CO2 inhalation, and single cell suspensions were prepared from spleens and LNs (mesenteric, inguinal, brachial, and axillary). After erylysis, cells were labeled with 10 μM 5-(and-6-)-TAMRA SE or 5 μM CFSE (Invitrogen) for 15 min at 37°C. Untouched CD4+ T lymphocytes were isolated using a MACS CD4+ T cell isolation kit together with an AutoMACS (Miltenyi Biotec). Alternatively, cells were stained with the biotin antibody cocktail followed by StreptAvidin–PE-Cy7 and anti–CD44-APC, and cell sorting for PE-Cy7−/CD44low cells was performed on a FACSAria (BD Biosciences). B220+ B cells were purified using anti-B220 MACS microbeads. The purity of cell preparations after MACS was always >90%. 6–8-wk-old WT or plt recipients received 1–1.25 × 107 WT CD4+ T or B220+ B cells or 1.5–2.5 × 107 CCR7−/− CD4+ T cells by i.v. injection. Dye labeling (CFSE or TAMRA) was swapped between experiments. In most experiments, only TAMRA-labeled cells were used.
Between 24 and 48 h after adoptive transfer, mice were anaesthetized by an initial i.p. injection of 100 mg/kg ketamine and 1 mg/kg medetomidine. In some experiments, mice received a single i.v. injection of 200 μg anti-CD62L mAb (clone MEL-14) 2 h before imaging commenced to prevent further homing of lymphocytes to LNs. The animal was placed on a custom-built preparation stage fixating the right hind leg. The right pLN was micro-surgically prepared free of overlying connective and adipose tissue, submerged in normal saline and covered with a glass coverslip. Special care was taken to spare blood vessels and afferent lymphatic vessels during microsurgery. A thermocouple was placed next to the LN to monitor local temperature, which was maintained at 36.5 ± 1°C. Two-photon laser scanning microscopy was performed with an upright Leica DM LFSA microscope equipped with a 20× 0.95 NA water immersion objective (Olympus) and a MaiTai Ti:sapphire-pulsed laser (Spectra-Physics). For two-photon excitation, the MaiTai laser was tuned to 858 nm. Green (CFSE or FITC) and red (TAMRA) fluorescence emission was detected with nondescanned detectors fitted with 535/50 and 610/75 band path filters, respectively. To generate time-lapse series, stacks of 10–11 x-y sections with 4–4.5-μm spacing were acquired every 15 or 20 s with 1.5–2.5× electronic zoom providing imaging volumes 40–50 μm in depth and 215–360 μm in width. Exogenous chemokine (1 μg of recombinant mCCL19 in 30 μl of normal saline containing 5 μg of FITC-labeled dextran of 150 kD as tracer) was injected s.c. into the right hind footpad.
Imaris (Bitplane) was used for four-dimensional image analysis and automated tracking of cells. The accuracy of the automated tracking was manually controlled, and only tracks with durations of >60 s were included in the analysis. Average cell velocity and meandering index were calculated using Imaris, and turning angle distribution, mean displacement, and MC were calculated using SciLab after exporting the x, y, and z coordinates of spot positions. Statistical analysis was performed with GraphPad Prism 4. Results are displayed as individual data points plus median or mean ± SEM summarizing collective data from all experiments performed. All significant values were determined using the unpaired two-tailed t test.
Online supplemental material.
Video S1 shows the accumulation of highly motile transferred TAMRA-labeled WT CD4+ T cells as well as of unlabeled endogenous WT T cells in the SCS of the pLNs after s.c. injection of exogenous CCL19 into the hind footpad. As exemplified by Video S2, there is no evidence for immigration of T cells into the SCS via the afferent lymphyatics after injection of exogenous CCL19. Video S3 shows the absence of TAMRA-labeled CCR7−/− CD4+ T cells inside the SCS in the same experimental setting, whereas unlabeled endogenous cells of the WT recipient exhibit a strong migratory response. Videos S4–S6 illustrate the motility of transferred TAMRA-labeled T cells within the T cell zone of the pLNs (Video S4, WT T cells in WT recipient; Video S5, CCR7−/− T cells in WT recipient; Video S6, WT T cells in plt recipient). As depicted in Fig. S1, no enrichment of nonnaive CD44high lymphocytes was observed after adoptive transfer of CD4+ T cells in the absence of CCR7 signaling.
We are especially grateful to Stefanie Willenzon for excellent technical assistance and Oliver Pabst and Günter Bernhardt for discussion and critical comments on the manuscript.
This study has been supported by DFG grants SFB566-A14 and SFB587-B3 to R. Förster.
The authors have no conflicting financial interests.