The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

The Journal of Immunology, 2007, 179, 7535-7543
Copyright © 2007 by The American Association of Immunologists, Inc.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bekiaris, V.
Right arrow Articles by Lane, P. J. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bekiaris, V.
Right arrow Articles by Lane, P. J. L.

Role of CD30 in B/T Segregation in the Spleen1

Vasileios Bekiaris2,*, David Withers2,*, Stephanie H. Glanville2,*, Fiona M. McConnell2,*, Sonia M. Parnell3,*, Mi-Yeon Kim3,*, Fabrina M. C. Gaspal3,*, Eric Jenkinson3,*, Clive Sweet3,{dagger}, Graham Anderson3,* and Peter J. L. Lane2,4,*

* Medical Research Council Centre for Immune Regulation, Birmingham Medical School; and {dagger} School of Biosciences, Birmingham, United Kingdom


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
In this report, we identify an important function for CD30 signals in the effective segregation of B and T lymphocytes in the murine spleen, additional to the recognized requirement for lymphotoxin signals. We show that CD30 signals are not required for transcription or protein expression of homeostatic chemokines, but CD30-deficient mice display impaired B/T segregation. This defect correlates with defective expression as detected by Abs of the transmembrane mucin-type protein podoplanin on T zone stroma, although expression at other sites is normal. Defective segregation is not intrinsic to CD30-deficient lymphocytes which segregate normally following transfer into RAG-deficient mice and significantly up-regulate the expression of both CCL21 and podoplanin on T zone stroma of RAG-deficient mice. During development, induction of expression of the CD30 ligand by lymphoid tissue inducer cells and podoplanin by T zone stroma are temporally linked, and the spatial association of these cells suggests that lymphoid tissue inducer cells are capable of providing the CD30 signals. Finally, we show that the appearance of podoplanin on T zone stroma in development is associated with B/T segregation of splenic white pulp areas. Our studies indicate that homeostatic chemokine expression by itself is not sufficient for B/T segregation and our data point to a significant role for podoplanin or molecules associated with podoplanin expressing stroma in the effective segregation of lymphocytes.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
In the white pulp structures of secondary lymphoid tissues, B and T lymphocytes are organized by homeostatic chemokines. There are at least three identified components that regulate this process: specialized stromal cell subsets, the chemokines they secrete, and an intercellular conduit system. VCAM-1+ stromal cells produce either T zone or B zone chemokines: in the T zone, CCL19 and the genetically closely linked gene, CCL21; in the B follicle, CXCL13. In spleen (1) and lymph node (2, 3), the chemokine-expressing stromal cells ensheathe a conduit network of collagen fibers that are permeable to low molecular weight molecules, including chemokines.

The conduit system performs two functions. It provides a support structure for the fibroreticular cells that ensheathe it and also transports the chemokines they secrete to the luminal surface of blood vessels in the spleen or high endothelial venules in lymph nodes. The chemokine gradients thus established guide the recruitment of lymphocytes and APCs. In addition to VCAM-1 expression, the T zone stroma that expresses CCL19 and CCL21 (4) coexpresses podoplanin (gp38) (5), a mucin-type transmembrane protein. Analysis of cells transfected with podoplanin has implicated this protein in cell adhesion and tube formation, functions clearly relevant to the formation of the conduit system (6). Podoplanin also appears on a subset of thymic medullary epithelial cells (5) and lymphatic endothelium, as well as in locations outside the conventional immune system, such as the glomeruli of the kidney, where it is expressed on podocytes (7). Podoplanin expression in all of these tissues is associated with expression of the T zone chemokine CCL21 (8, 9).

The secretion of chemokines and expression of podoplanin by stroma depends on signals delivered through their lymphotoxin β receptor (LTβR)5 (10). There is evidence for B lymphocyte provision of these signals to stromal cells with regard to the expression of CXCL13, the B cell chemoattractant (10), as well as of the T zone chemokines (11). However, there is also clear evidence that cells other than mature B and T cells play pivotal roles. We have highlighted the role of a cell with a similar phenotype (12, 13) and genotype (14) to the embryonic CD4+ lymphoid tissue inducer cell (LTI) population linked with the development of secondary lymphoid structures in embryonic life (15). We have shown that following transfer into adult LT{alpha}–/– mice, both embryonic and adult LTI populations but not dendritic cell (DC) subsets or splenocytes induce a significant degree of B/T segregation, with up-regulation of both VCAM-1 and the T zone chemokine CCL21, but not the B zone chemokine CXCL13 (14). Significantly, injection of LTI into adult LT{alpha}–/– mice did not up-regulate podoplanin, consistent with previous reports showing a developmental requirement for lymphotoxin {alpha} (LT{alpha}) signals in the neonatal spleen (16). These studies showed that podoplanin expression was lymphocyte dependent, but that once podoplanin expression was established, normal B/T segregation was much more difficult to disrupt by blockade of LT{alpha} signals.

In addition to their expression of LTβR ligands, mature LTI express high levels of the TNF ligands, OX40-ligand (L) (CD252) and CD30L (CD153) which we have linked to CD4 T cell memory survival (12, 17). Unlike the expression of LTβR ligands, the expression of both OX40L and CD30L is missing on embryonic populations, but both molecules can be induced in vivo, following adoptive transfer of juvenile populations into an adult environment (14), as well as in vitro (14, 18), Although absent at birth, OX40L and CD30L are both expressed by 7 days, but expression is not normal until mice are 3 wk old. In this manuscript, we show that in addition to LTβR signals, CD30 signals also contribute to the process of B/T segregation. The expression of podoplanin on T zone stroma mirrors the up-regulation of CD30L on LTI populations, and CD30-deficient (CD30–/–) mice lack expression of podoplanin on T zone stroma. This defect in T zone stroma is functional, as CD30–/– mice demonstrate impaired B/T segregation, but CXCL13, CCL19, and CCL21 mRNA expression occurs independently of CD30 signaling. Impairment of B/T segregation is associated with the failure to exclude B cells from CCL21-expressing regions of white pulp areas of CD30–/– mice. This defect is not intrinsic to CD30–/– lymphocytes, which segregate comparably to wild-type (WT) splenocytes following transfer into RAG–/– recipient mice. Significantly, this normal B/T segregation in RAG–/– mice is associated with the appearance of podoplanin expression.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Mice

All experiments were performed in accordance with U.K. laws and with the approval of the University of Birmingham Ethics Committee. WT mice were of the C57BL/6 strain and were bred in our animal facility. CD30–/– mice on the C57BL/6 background were generated as previously described (19). Adult mice were ages between 6 and 10 wk. For experiments with neonatal mice, the day of birth was considered day 0.

Immunostaining for confocal microscopy

The procedure for preparing and staining mouse spleen tissue for confocal microscopy has been described before (12). B cells were detected by staining for IgM either with a donkey anti-mouse IgM-Rhodamine Red Ab (Jackson ImmunoResearch Laboratories) or a goat anti-mouse IgM-aminomethyl coumarin acetate Ab (Jackson ImmunoResearch Laboratories). T cells were detected by staining for CD3 with either a biotinylated or a purified hamster anti-CD3{epsilon} Ab (BD Biosciences). Biotinylated anti-CD3{epsilon} was detected with streptavidin-Alexa Fluor 647 (Molecular Probes and Invitrogen Life Technologies.) and the purified anti-CD3{epsilon} was detected with a goat anti-hamster-Cy5 Ab (Jackson ImmunoResearch Laboratories). CCL21 and CXCL13 were detected using the following four-step amplification staining method: goat anti-mouse CCL21 or CXCL13 (R&D Systems), followed by a donkey anti-goat-FITC (Jackson ImmunoResearch Laboratories), a third step of rabbit anti-FITC (BioSource International), and a fourth step of goat anti-rabbit-FITC (Southern Biotechnology Associates). VCAM-1 was detected using a three-step amplification staining: rat anti-mouse VCAM-1-FITC (BD Biosciences), followed by rabbit anti-FITC (BioSource International), then goat anti-rabbit-FITC (Southern Biotechnology Associates). Podoplanin (gp38) staining was achieved using a hamster anti-mouse gp38 (provided by A. Farr, University of Washington, Seattle, WA) Ab, which was detected by a second step of goat anti-hamster-biotin (Vector Laboratories) and a third step of streptavidin-Alexa Fluor 647 (Molecular Probes). CD11c+ DCs were stained using a purified hamster anti-mouse CD11c Ab (clone N418, grown in-house), which was detected with goat anti-hamster-Cy5 (Jackson ImmunoResearch Laboratories). CD4+ cells were stained with a purified rat anti-mouse CD4 Ab (BD Biosciences) and detected with goat anti-rat TRITC (Southern Biotechnology Associates). ER-TR7 (a gift from W. von Ewjik; Erasmus University, Rotterdam, The Netherlands) was detected with an anti-rat Ab followed by goat ant-rat TRITC. CD45.1+ cells were detected using a CD45.1-FITC (BD Biosciences) Ab followed by the FITC amplification as described for VCAM-1.

Image acquisition and storage for confocal microscopy

Confocal images were acquired using a Zeiss LSM510 laser scanning confocal head with a Zeiss Axio Imager Z1 microscope. FITC-labeled Abs were excited with a 488-nm argon laser, TRITC, or Rhodamine Red-labeled Abs with a 561-nm helium laser, Cy5 or Alex Fluor 647-labeled Abs with a 633-nm helium laser, and aminomethyl coumarin acetate-labeled Abs with a 405 diode laser. Data were stored in four different channels, one for each color, which were scanned separately. All detectors were set so as to exclude the possibility of more than one fluorochrome appearing in each channel. Confocal micrographs were stored as digital arrays of 2048 x 2048 pixels with an 8-bit sensitivity; detectors were routinely set so that intensities in each channel spanned the 0–255 scale optimally.

Analysis of confocal images

All recorded images were analyzed using the Zeiss LSM510 (laser scanning microscope) software. The software permits delineation of specified areas with automatic calculation of their size (in square micrometers). This feature was used routinely to define and quantify areas of interest (see Figs. 1 and 2). The LSM510 software also tabulates pixel frequency for each intensity. We used this facility to quantify the presence of CCL21 and CXCL13. Assuming that the number of pixels positive for the chemokine is proportional to the amount present, we defined total presence of the chemokine as the number of pixels with intensities for CCL21 or CXCL13 higher than 25 (avoiding electronic and autofluorescent noise) and lower than 225 (avoiding "spillover" from off-scale intensities).


Figure 1
View larger version (103K):
[in this window]
[in a new window]

  		FIGURE 1. B/T segregation and CCL21 localization in WT and CD30–/– mice. A, Spleen sections from WT and CD30–/– mice were stained for IgM (green), CD3 (red), and CCL21 (white) and analyzed by confocal microscopy; each confocal micrograph was taken with a x10 objective (scale bar, 100 µm). The T zones were defined either as the B cell-free area using IgM staining as a guide (yellow line) or as the T cell-rich area using CD3 staining as a guide (turquoise line); the extent of the white pulp was defined as the limit of the IgM staining (white line); the same staining was repeated three times for six mice per strain; and a minimum of 10 white pulps were photographed and analyzed for each mouse. B, WT or CD30–/– spleen cells were transferred into RAG–/– recipients that were sacrificed 4 wk later; spleen sections from the RAG–/– recipients were stained as in A and analyzed by confocal microscopy (scale bar, 50 µm). C, Spleen sections from WT and CD30–/– mice were stained for IgM (green) and CXCL13 (white) and analyzed by confocal microscopy; each confocal micrograph was taken with a x10 objective (scale bar, 100 µm).

 

Figure 2
View larger version (22K):
[in this window]
[in a new window]

  		FIGURE 2. Quantification of B/T segregation and chemokine production. A, Spleen sections from WT ({circ}) and CD30–/– ({triangleup}) mice were stained and analyzed by confocal microscopy as shown in Fig. 1. i, Total white pulp area defined by the white line in Fig. 1. ii, T zone (TZ) area defined by CD3 staining (turquoise line in Fig. 1). iii, T zone area defined by IgM staining (yellow line in Fig. 1). iv, B/T interface area defined as the difference between the areas defined in ii and iii. v, The area defined in iv corrected for the CD3-defined T cell area. Bi, Splenic mRNA expression for CCL19, CCL21, and CXCL13 corrected to the housekeeping gene β-actin (WT ({square}), CD30–/– ({blacksquare})); results are the mean of two to three mice; error bar, SD. Results are representative of two separate experiments. ii, Quantitation from confocal micrographs of protein expression (positive pixels) of CCL21 and CXCL13 in total white pulp areas defined by IgM staining in Fig. 1. iii, The proportion of CCL21 present in the B cell follicle defined by subtracting the CD3-defined T zone from the total white pulp area. iv, The proportion of CXCL13 present in the B cell follicle defined by subtracting the IgM-defined T zone from the total white pulp area. A, i–v, and B, ii–iv, Each symbol represents a single confocal micrograph that was assigned areas as shown in Fig. 1. For this figure, 18 micrographs per mouse strain from six mice were analyzed. WT and CD30–/– values were compared using nonparametric Mann-Whitney U statistics; p values for individual comparisons are shown.

 
Preparation of cDNA from mouse spleen

Total RNA from frozen 8-µm-thick spleen sections was prepared using the RNeasy micro kit (Qiagen) according to the manufacturer’s instructions. Total RNA was immediately reverse-transcribed into cDNA using oligo(dT) (Amersham Pharmacia Biotech) and Moloney Murine Leukemia Virus reverse transcriptase (Invitrogen Life Technologies). The reaction was conducted in a MJ Research thermal cycler (Global Medical Instrumentation).

Quantitative real-time PCR

Splenic cDNA was used to determine mRNA expression of CCL19, CCL21, CXCL13, podoplanin, and LTβ. All real-time PCR were performed in a 7900HT Fast Real-Time PCR System (Applied Biosystems) in 96-well clear plates (Applied Biosystems). Primers and probes were designed using the TaqMan primer and probe design software and were ordered and synthesized by Eurogentec. All primers and probes were used at limiting concentrations optimized for real-time PCR. Quantification of gene expression was based on the expression of the control housekeeping gene β-actin for each sample. The copy number of the gene of interest relative to the control gene was calculated by setting a threshold within the logarithmic phase of each PCR and determining the cycle number at which the signal reached that threshold (Ct). The Ct for the control gene was subtracted from the Ct for the gene of interest and the relative amount of that gene was calculated as 2–{Delta}Ct. Primer and probe sequences (5'->3') were: β-actin, forward: CGT-GAA-AAG-ATG-ACC-CAG-ATC-A, reverse: TGG-TAC-GAC-CAG-AGG-CAT-ACA-G; probe: VIC-TCA-ACA-CCC-CAG-CCA-TGT-ACG-TAG-CC-TAMRA; CCL19, forward: CCT-TCC-GCT-ACC-TTC-TTA-ATG-AAG, reverse: ACA-GAG-CTG-ATA-GCC-CCT-TAG-TGT; probe: 6-FAM-TGC-AGG-GTG-CCT-GC-TAMRA; CCL21, forward: TCC-CGG-CAA-TCC-TGT-TCT-C, reverse: TTC-TGC-ACC-CAG-CCT-TCC-T; probe: 6-FAM-CCC-CGG-AAG-CAC-TCT-AAG-CCT-GAG-CTA-T-BHQ-1; CXCL13, forward: ACA-TCA-TAG-ATC-GGA-TTC-AAG-TTA-CG, reverse: TCT-TGG-TCC-AGA-TCA-CAA-CTT-CAG; and probe: 6-FAM-CCT-GGG-AAT-GGC-TGC-CCC-AAA-TAMRA. Podoplanin, LTβ, and TNF-{alpha} were used as ready-made TaqMan gene expression assays (podoplanin: Mm00494716_m1, LTβ: Mm00434774_g1, and TNF-{alpha}: Mm00434774_m1).

Podoplanin RT-PCR

The following primers (5'->3') were used to detect 98% of the published podoplanin coding sequence (NM010329; National Center for Biotechnology Institute): forward: GCC-AGT-GTT-GTT-CTG-GGT-TT and reverse: GCT-CTT-TAG-GGC-GAG-AAC-CT. The PCR consisted of a 5-min 94°C denaturation step followed by 35 cycles of 30 s at 94°C, 40 s at 58°C, and 60 s at 72°C. The primers were ordered and synthesized by Sigma-Genosys and were designed using Primer3 (http://frodo.wi.mit.edu/cgibin/ primer3/primer3_www.cgi).

Cell transfers

Splenocytes were prepared from WT or CD30–/– mouse spleens by cutting them into small pieces and crushing them through gauze, followed by RBC lysis. After counting, ~107 WT or CD30–/– splenocytes were transferred per RAG–/– mouse i.p. Two weeks later, each mouse received 108 WT or CD30–/– splenocytes i.v. and 2 wk later all mice were sacrificed and their spleens were analyzed by confocal microscopy.

Flow cytometry for surface LTβR ligands

Splenocytes from WT, CD30–/–, or LT{alpha}–/– mice were prepared as above and cultured overnight with 10 ng/ml recombinant murine IL-7 (R&D Systems) before staining for surface LTβ ligands using LTβR-Ig fusion protein as described previously (20). Samples were acquired using a BD FACScan flow cytometer.

Isolation of white pulp cells

The capsule of the spleen was teased apart so that individual white pulps were visible under the dissecting microscope. White pulps were then picked up using fine forceps and cells were prepared by treatment with collagenase D (Roche) and DNase I (Sigma-Aldrich) followed by trypsin (Sigma-Aldrich). Cells were then FACS stained for VCAM-1 and CD45 and sorted on a Mo-Flo high-speed cell sorter (DakoCytomation).

Statistical analysis

All statistical analyses were performed with the nonparametric Mann-Whitney U test using StatView 5.0 (p < 0.05 is considered significant).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Impaired B/T segregation in CD30–/– mice

Although CD30–/– mice were initially described to have normally organized tissues (19), we noticed that the T cell areas of CD30–/– mice were smaller and less well demarcated than those of WT mice (compare immunostaining for T cells (CD3) and B cells (IgM) in Fig. 1A). Overall, the white pulp areas were comparable between WT and CD30–/– mice. To quantify the impaired B/T segregation in CD30–/– mice, T zone areas, defined by IgM (yellow line) and CD3 (turquoise line) staining, were outlined in WT and CD30–/– mice. Total white pulp areas were defined by IgM staining (white line). The size of each area was calculated as described in Materials and Methods.

The size of the total B and T white pulp areas was not statistically different between WT and CD30–/– mice (p = 0.7; Fig. 2Ai). However, when we compared T zones, defined either by CD3 (p = 0.0002; Fig. 2Aii) or IgM (p < 0.0001; Fig. 2Aiii), T zone areas in CD30–/– mice were significantly reduced. To quantify the degree of B/T mixing (impaired B/T segregation), the difference in area between the IgM and CD3-defined T zone was plotted (Fig. 2Aiv), demonstrating significantly larger differences in the CD30–/– mice (p = 0.0024), particularly when corrected for the CD3-defined T zone size (p < 0.0001; Fig. 2Av). One possible explanation for these findings was intrinsic failure of CD30–/– lymphocytes to respond to homeostatic chemokines. To investigate this, we transferred WT or CD30–/– lymphocytes into RAG–/– mice and demonstrated that their B and T cell segregation was comparable (Fig. 1B). This experiment demonstrates that there is no intrinsic defect in the response to chemokines and shows, in addition, that the failure to segregate cannot be explained by the reduction in T cell numbers in CD30–/– mice (data not shown) that is due to their impaired survival (17).

The cluster of TNFRs on human chromosome 1 and mouse chromosome 4 containing CD30 also contains TNFR2 in close proximity. We excluded by mRNA analysis the possibility that TNFR2 had also been disrupted by the CD30 gene-targeting process (data not shown).

Sustained expression of homeostatic chemokines in CD30–/– mice

The above data suggested that there was impaired B/T segregation and that this was dependent on the absence of signals through CD30. B and T cells are segregated by the influence of T cell (CCL19 and CCL21) and B cell (CXCL13) chemokines expressed in stromal cell subpopulations in B follicles and the T zone. The production of these chemokines, particularly CXCL13 and CCL21, but to a lesser extent CCL19 (10), depends on LT{alpha} and LTβ signals (10). To test whether the expression of these genes depended also on signals through the related TNFR CD30, we compared mRNA expression for these homeostatic chemokines in WT and CD30–/– mice (Fig. 2Bi). Quantitative determination of levels of chemokine mRNA (after correction for total mRNA levels) showed them to be greater in CD30–/– than in WT mice (p = 0.05), implying that defective transcription of the chemokine genes could not account for the impaired B/T segregation.

Although we were unable to identify expression of CCL19 protein on tissue sections using polyclonal antisera (possibly related to the reduced mRNA expression compared with CCL21, see Fig. 2Bi), expression of CCL21 (Fig. 1A) and CXCL13 (Fig. 1C) proteins was readily detected both in WT and CD30–/– tissues. The specificity of expression was confirmed by the lack of staining in LT{alpha}–/– mice, which have grossly impaired expression of both CXCL13 and CCL21 (data not shown). To quantify chemokine protein expression, we calculated the numbers of CXCL13+ and CCL21+ pixels (see Materials and Methods) in the white pulp areas (defined by the boundary of the IgM+ cells) in WT and CD30–/– mice and found no significant difference (Fig. 2Bii) even when pixel intensity was taken into account, suggesting that neither transcription nor translation of these proteins was reduced in CD30–/– mice. Similarly, when the extent of CCL21 staining relative to total white pulp area was determined, there was found to be no difference between WT and CD30–/– mice (data not shown). Because T zones in the CD30–/– mice were smaller and B follicles larger than in their WT counterparts (Figs. 1A and 2A), this resulted in the encroachment of CCL21 into B cell rich white pulp in CD30–/– mice, whereas the boundary of CCL21 staining in WT mice demarcated the CD3+IgM– T zone (compare WT and CD30–/– in Fig. 1A).

To describe the modified relationship between CCL21 expression and lymphocyte distribution quantitatively, we expressed the number of CCL21+ pixels in B cell areas as a percentage of the total CCL21+ pixels in the white pulp areas in both WT and CD30–/– mice. Whether B cell areas were defined by CD3 staining (p < 0.0001; Fig. 2Biii) or by IgM (p < 0.0001; data not shown), there was significantly greater expression of CCL21 in the B cell-rich follicles of CD30–/– mice, i.e., CCL21 was significantly less well limited to areas enriched with T cells (T zones). In contrast to CCL21, where the extent of staining was independent of B and T cell areas, the extent of CXCL13 staining remained correlated with the limit of B follicles and was therefore greater in CD30–/– B follicles (Fig. 2Biv).

These data indicate that production of CCL21 is not impaired in CD30–/– mice. Therefore, CCL21 protein production by itself is not sufficient to segregate B and T cells.

Intact conduit system, but no expression of podoplanin on T zone stroma in CD30–/– mice

Our data suggested that CCL21 was unable to segregate lymphocytes and that this defect was not intrinsic to the lymphocytes themselves. Another possibility was that CCL21 was presented inappropriately in CD30–/– spleen. Homeostatic chemokines are secreted proteins but they are bound to glycoproteins in a heparin-sensitive process (21). We therefore looked for defects in the T zone stroma of CD30–/– mice. Normal T zone stroma coexpresses VCAM-1 and podoplanin (4, 5). In normal WT mice, the podoplanin+VCAM-1+ T zone stroma ensheathes the collagen fibers of the conduit system, and this appears as tube-like structures when sectioned longitudinally and as holes when sectioned transversely (Fig. 3A). What these images clearly demonstrate is that VCAM-1 and podoplanin are expressed on stromal cells on both their inner surfaces, where they surround the conduit, and their outer surfaces, where they interact with lymphocytes and accessory cells.


Figure 3
View larger version (90K):
[in this window]
[in a new window]

  		FIGURE 3. Podoplanin and VCAM-1 expression on T zone stroma. A, Spleen sections from WT and CD30–/– mice were stained for IgM (gray), VCAM-1 (green), and podoplanin (red) and analyzed by confocal microscopy. Upper panel, Representative confocal micrographs that were taken with a x10 objective (scale bar, 100 µm); bottom panel, confocal micrographs that were taken with a x63 objective (scale bar, 20 µm) and correspond to the boxes in the top panel micrographs. This staining was performed for at least 10 different CD30–/– mice with similar results. B, Spleen sections from WT and CD30–/– mice were stained for IgM (green) and ER-TR7 (red) and analyzed by confocal microscopy; confocal micrographs were taken with a x10 objective (scale bar, 100 µm); C, Splenic mRNA expression for podoplanin corrected to the housekeeping gene β-actin (WT ({circ}) and CD30–/– ({triangleup}). D, Each symbol is an individual mouse; the inset shows RT-PCR for podoplanin using primers covering ~98% of its coding sequence. +ve, Positive control;–ve, negative control. D, Inguinal lymph node (iLN) or thymus sections from WT and CD30–/– mice were stained for IgM (green; inguinal lymph node only) and podoplanin (red) and analyzed by confocal microscopy; each confocal micrograph was taken with a x10 objective (scale bar, 100 µm).

 
To characterize splenic T zone stroma more carefully, we stained for VCAM-1 (Fig. 3A) and ER-TR7 (Fig. 3B). Although the T cell area was smaller in CD30–/– mice, both of these Ags were expressed. In contrast, podoplanin was almost completely absent (Fig. 3A). Despite the lack of podoplanin expression as detected by two different mAbs, podoplanin mRNA was detected in CD30–/– spleens (Fig. 3C), a situation similar to LT{alpha}-deficient mice which also fail to stain for podoplanin with Abs (22). This is in keeping with the idea that both LT{alpha}1β2 and CD30 signals regulate podoplanin posttranscriptionally. Importantly, the regulation of podoplanin by CD30 was spleen specific since its expression was normal in both lymph nodes and thymus of CD30–/– mice (Fig. 3D).

The sustained expression of homeostatic chemokines, at both mRNA and protein (as assessed by Ab staining) levels, indicates that the effect of CD30 signals is not to regulate LTβR ligands, especially because there are normal levels of LTβ mRNA and normal surface expression of LTβR ligands in CD30–/– spleens (Fig. 4, A and B). Moreover, the levels of TNF-{alpha} mRNA were normal in CD30–/– mice (Fig. 4A). Expression of podoplanin, however, was independent of CD30 signals to lymphocytes, because transfer of CD30–/– lymphocytes into RAG–/– mice was sufficient to restore podoplanin expression (Fig. 4C), whereas WT lymphocytes were unable to restore podoplanin expression in CD30–/– hosts despite being able to home to the white pulps (Fig. 4D).


Figure 4
View larger version (80K):
[in this window]
[in a new window]

  		FIGURE 4. LTβR ligands, TNF-{alpha}, and lymphocyte function in relation to CD30. A, Splenic mRNA expression for LTβ or TNF-{alpha} (red) corrected to the housekeeping gene β-actin (WT ({circ}) and CD30–/– ({triangleup})); each symbol is an individual mouse. B, WT, CD30–/–, or LT{alpha}–/– spleen cells were stained for surface LTβR ligands using LTβR-Ig fusion protein after overnight culture with rIL-7 and were analyzed by flow cytometry. Histograms represent cells in the lymphocyte gate. C, WT or CD30–/– spleen cells were transferred into RAG–/– recipients that were sacrificed 4 wk later. Spleen sections from the RAG–/– recipients were stained for IgM (green) and podoplanin (red) and analyzed by confocal microscopy; spleen sections from RAG–/– mice that had not received cells were also stained for CD4 (green) and podoplanin (red); (scale bars, 100 µm). D, CD45.1+ WT spleen cells were transferred into CD45.1– WT or CD30–/– recipients that were sacrificed 3 days later and their spleens were stained for IgM (green), podoplanin (red), and CD45.1 (blue) and analyzed by confocal microscopy (scale bar, 100 µm).

 
A normal relationship between CD30L-expressing CD4+ LTI and T zone stromal cells in CD30–/– mice

Our results show that in the absence of CD30, impaired B/T segregation is associated with lack of podoplanin expression on T zone stroma in the presence of CCL21 production. Previous studies have also linked LT{alpha}-dependent podoplanin expression to normal B/T segregation (16). Adult CD4+ LTI associate closely with VCAM-1+ T zone stroma, express high levels of LT{alpha} and LTβ, are functionally capable of up-regulating the expression of CCL21 protein but not podoplanin in T zone stroma, and can restore a degree of B/T segregation following transfer into LT{alpha}–/– mice (20). Their constitutive expression of levels of CD30L (12) qualifies them as a possible source of CD30 signals required for T zone stromal podoplanin expression.

One possibility was that LTI failed to associate with the conduit in CD30–/– mice. Associations between adult CD4+ LTI and T zone stromal cells were identified by costaining simultaneously for CD4, CD11c, and CD3 along with VCAM-1 (IgM staining was also performed for orientation purposes; data not shown.) Fig. 5A shows clear CD4+ LTI cell association with VCAM-1+ stromal cells in CD30–/– mice and, furthermore, emphasizes the microenvironment formed among the conduit, VCAM-1+ stroma, CD4+ LTI, and CD3+ T cells and CD11c+ DCs.


Figure 5
View larger version (57K):
[in this window]
[in a new window]

  		FIGURE 5. LTI cells in CD30–/– mice and expression of podoplanin, CD30, and CD30L during development. A, Spleen sections from CD30–/– mice were stained for CD4 (red), VCAM-1 (green), CD3, and CD11c (blue). This figure is two adjacent micrographs taken with a x63 objective (scale bar, 20 µm) and encompassing a longitudinal section of a single T zone. Boxes 1–4 show the areas enlarged (the 2-µm scale bar applying to all) and presented with the plots of the stain intensities registered along the line of the red arrows. These identify cell profiles by color and confirm the associations of the various cell types; green = VCAM-1+ stromal cells, red = CD4+ LTI cells, and red and blue = CD4+CD3/CD11c+ T cells or DCs. B, Spleen sections from day 0, day 4 WT, and day 7 WT or day 7 CD30–/– mice were stained for IgM (green) and podoplanin (red) and analyzed by confocal microscopy. Micrographs were taken with a x10 objective (scale bar, 100 µm). C, Splenic mRNA expression for CD30 and CD30L corrected to the housekeeping gene β-actin; results are from one or the mean of two to five mice; error bar, SD. D, mRNA expression for CD30 corrected to the housekeeping gene β-actin from VCAM-1+CD45– ({blacksquare}), VCAM-1–CD45– (Figure 5), and VCMA-1–CD45+ ({square}) spleen cells of day 4- and day 8-old mice.

 
A close correlation between CD30, CD30 ligand expression, and splenic stromal podoplanin expression

The close association between the stroma and CD4+ LTI demonstrates that the latter are ideally situated to provide the CD30 signals essential for the expression of podoplanin on VCAM-1+ T zone stroma, which in turn requires LTβR signals in the first week of life but not thereafter (16). Whereas LTβR ligands are expressed at comparable levels on neonatal and adult LTI, CD30L is lacking in the neonatal population and is only found toward the end of the first week of life in LTI (18). To test whether the expression of podoplanin on developing spleen mirrored CD30L expression on LTI, we examined podoplanin expression at birth and at intervals thereafter (Fig. 5B). Podoplanin was missing on day 0, it was barely detectable on day 4, but was present on day 7 (Fig. 5B), correlating with the up-regulation of CD30L on LTI (18). Analysis of relative mRNA expression for CD30 and CD30L from splenic tissue showed that expression kinetics mirrored that of podoplanin (Fig. 5C and cf with 5B). Because of the expression kinetics for podoplanin and CD30 and CD30L, we isolated stromal (VCMA-1+ or VCAM-1–) or hemopoietic cells from day 4- and day 8-old white pulps and measured mRNA levels of CD30 (Fig. 5D). We found that at day 4, CD30 was expressed by all three cell populations (Fig. 5D). By day 8, VCAM-1+ stromal cells expressed the highest levels of CD30, which correlates with the appearance of podoplanin in the developing spleen (Fig. 5D).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
LT{alpha} signals are required for the normal segregation of B and T cells in the spleen. This is because LT{alpha} signals to stroma regulate expression of homeostatic chemokines, particularly the T zone chemokine CCL21 and the B cell chemokine CXCL13 (10). In addition, LT{alpha} signals in the neonatal period (16) regulated the expression of the transmembrane mucin-type glycoprotein podoplanin on T zone stroma (4, 5). This study showed that blockade of LT{alpha} signals neonatally not only prevented splenic podoplanin expression, but that normal splenic B/T segregation failed to develop even when LT{alpha} signals were restored. In this study, we report further evidence implicating podoplanin in B/T segregation: we found that CD30-deficient mice lacked splenic podoplanin expression and had grossly impaired B/T segregation, despite levels of homeostatic chemokines that were at least as good as WT. The defective segregation was not intrinsic to lymphocytes since CD30–/– splenocytes segregated comparably to WT lymphocytes following transfer into RAG-deficient mice.

The association of defective podoplanin expression with impaired B/T segregation in this and previous studies (16) suggests that either this protein is implicated in the process of B/T segregation or that podoplanin deficiency is a marker for a more generalized defect in stromal glycoprotein expression. Previous studies have suggested that CCL21 might bind to the stromal-associated protein ER-TR7, (1) but we found that expression of this Ag appeared normal in CD30–/– mice. Either way this study emphasizes that the expression of homeostatic chemokines by themselves is not sufficient to segregate lymphocytes and that the matrix associated with the T zone stroma is important.

The defect in podoplanin Ag expression in CD30–/– mice was specific to spleen because expression in lymph node and thymus was normal. Furthermore, splenic podoplanin mRNA was detected, a situation similar to that reported for LT{alpha}–/– mice (22). This suggested that the lack of splenic podoplanin expression might be consequent on posttranslational modification rather than transcription and translation of protein. Podoplanin is highly glycosylated and, in humans, podoplanin and CCL21 proteins colocalize in vivo (23). Additionally, the binding of recombinant human podoplanin to CCL21 is a high-affinity interaction (Kd ~ 100 nM), which can be blocked by heparin and depends on the glycosylation level of podoplanin (23). These data are consistent with the idea that a key function of the podoplanin expressed on T zone stroma is to trap the CCL21 locally. We suggest that this ensures a steep CCL21 gradient between B and T cell areas, lost in CD30–/– mice. An alternative explanation for our findings for poor B/T segregation is that there is overlap in expression between the stromal cells that produce CXCL13 and CCL21, with consequent broadening of the B/T interface in CD30–/– mice. This is currently under investigation.

Expression of CCL21 and podoplanin is lymphocyte dependent, because RAG–/– mice express neither, and both also depend on LT{alpha} signals (16). However, we have recently reported that lymphocyte expression of LT{alpha} is not essential for the induction of podoplanin, which we found to be up-regulated when LT{alpha}–/– lymphocytes were transferred into RAG–/– mice (24). This finding suggested that a cell type other than a mature T or B cell might provide the LT{alpha} signal. For a number of years, we have investigated the function of a LT{alpha}-expressing CD4+ accessory cell that is genetically and phenotypically similar to the LTI responsible in embryonic life for lymph node development (15). Recently, we reported that in the embryonic spleen, LTI were found clustered with VCAM-1-expressing stromal cells around blood vessels and that the expression of VCAM-1 depended on LT{alpha} signals from LTI (24).

However, this was not sufficient for expression of CCL21, which also required lymphocytes, although LT{alpha} expression on lymphocytes was not essential (24). Direct evidence that LTI could activate stroma with LT{alpha}–/– lymphocytes was provided by the transfer of LTI into adult LT{alpha}–/– mice. LTI transfer was associated with up-regulation of both VCAM-1 and CCL21 on splenic stroma, but expression of podoplanin remained absent (20), consistent with a specific requirement for neonatal LT{alpha} signals for the induction of podoplanin (16).

On the basis of these observations, we propose a model showing how LTI and lymphocytes program T zone stromal development through LT{alpha} and CD30 signals (Fig. 6). In the first phase, LTI colonize the embryonic spleen and activate embryonic T zone stroma (up-regulating VCAM-1 but not CCL21 or podoplanin) by LT{alpha} signaling (Fig. 6A). The entry of lymphocytes to the spleen around birth (Fig. 6B) is associated with the up-regulation of CCL21 but not with podoplanin expression. Our evidence suggests that lymphocytes are required but that lymphocyte LT{alpha} itself is not essential, although under normal circumstances it is likely that both lymphocytes and LTI provide LT{alpha} signals. Finally, up-regulation of CD30 ligand on LTI around day 7 (18) is correlated with the appearance of stromal podoplanin (Fig. 6C) and the development of a mature well-segregated white pulp. The CD30-expressing cellular target is not yet identified; one possibility is that CD30 is expressed on stroma.


Figure 6
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 6. Model of integrated roles of LTI cells, stroma, and lymphocytes in white pulp development. A, LTI cells interact with stroma and induce stromal VCAM-1 expression in a LT{alpha}-dependent fashion (24 ); CCL21 and podoplanin are not induced. B, Lymphocyte entry is associated with up-regulation of CCL21, but not of podoplanin; lymphocytes, but not the LT{alpha} they express, are essential for CCL21 expression (24 ). C, Around neonatal day 7, LTI up-regulation of CD30L coincides with the up-regulation of podoplanin and normal B/T segregation. Lymphocytes are required for podoplanin expression, but do not need to express either CD30 or LT{alpha} (24 ).

 
In summary, we have demonstrated a role for CD30 signals in establishing B/T segregation, not by affecting homeostatic chemokine expression, but by regulating the normal development of T zone stroma. Although lack of podoplanin expression may just be a marker for a more generalized defect of T stroma, others have shown podoplanin to bind CCL21 specifically (23) and this could account for the impaired B/T segregation in CD30–/– mice. Our studies highlight the point that production of chemokines alone is insufficient for effective B/T segregation in the spleen, which depends additionally on mechanisms for localizing CCL21 appropriately.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by a Wellcome Programme Grant to P.L. and G.A. Back

2 V.B., D.W., F.M.M., S.H.G., and P.J.L.L. designed and performed the research, collected and analyzed the data, and wrote the article. Back

3 S.M.P., M.-Y.K., F.M.C.G., E.J., C.S., and G.A. contributed to designing and performing the research. Back

4 Address correspondence and reprint requests to Dr. Peter J. L. Lane, Medical Research Council Centre for Immune Regulation, Institute for Biomedical Research, Birmingham Medical School, Birmingham B15 2TT, U.K. E-mail address: p.j.l.lane{at}bham.ac.uk Back

5 Abbreviations used in this paper: LTβR, lymphotoxin β receptor; LTI, lymphoid tissue inducer; LT, lymphotoxin; WT, wild type; DC, dendritic cell; TRITC, tetramethylrhodamine isothiocyanate; Ct, cycle threshold. Back

Received for publication June 11, 2007. Accepted for publication September 26, 2007.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Nolte, M. A., J. A. Belien, I. Schadee-Eestermans, W. Jansen, W. W. Unger, N. van Rooijen, G. Kraal, R. E. Mebius. 2003. A conduit system distributes chemokines and small blood-borne molecules through the splenic white pulp. J. Exp. Med. 198: 505-512. [Abstract/Free Full Text]
  2. Anderson, A. O., S. Shaw. 1993. T cell adhesion to endothelium: the FRC conduit system and other anatomic and molecular features which facilitate the adhesion cascade in lymph node. Semin. Immunol. 5: 271-282. [Medline]
  3. Sixt, M., N. Kanazawa, M. Selg, T. Samson, G. Roos, D. P. Reinhardt, R. Pabst, M. B. Lutz, L. Sorokin. 2005. The conduit system transports soluble antigens from the afferent lymph to resident dendritic cells in the T cell area of the lymph node. Immunity 22: 19-29. [Medline]
  4. Luther, S. A., H. L. Tang, P. L. Hyman, A. G. Farr, J. G. Cyster. 2000. Coexpression of the chemokines ELC and SLC by T zone stromal cells and deletion of the ELC gene in the plt/plt mouse. Proc. Natl. Acad. Sci. USA 97: 12694-12699. [Abstract/Free Full Text]
  5. Farr, A. G., M. L. Berry, A. Kim, A. J. Nelson, M. P. Welch, A. Aruffo. 1992. Characterization and cloning of a novel glycoprotein expressed by stromal cells in T-dependent areas of peripheral lymphoid tissues. J. Exp. Med. 176: 1477-1482. [Abstract/Free Full Text]
  6. Schacht, V., M. I. Ramirez, Y. K. Hong, S. Hirakawa, D. Feng, N. Harvey, M. Williams, A. M. Dvorak, H. F. Dvorak, G. Oliver, M. Detmar. 2003. T1{alpha}/podoplanin deficiency disrupts normal lymphatic vasculature formation and causes lymphedema. EMBO J. 22: 3546-3556. [Medline]
  7. Matsui, K., S. Breiteneder-Geleff, D. Kerjaschki. 1998. Epitope-specific antibodies to the 43-kD glomerular membrane protein podoplanin cause proteinuria and rapid flattening of podocytes. J. Am. Soc. Nephrol. 9: 2013-2026. [Abstract]
  8. Banas, B., M. Wornle, T. Berger, P. J. Nelson, C. D. Cohen, M. Kretzler, J. Pfirstinger, M. Mack, M. Lipp, H. J. Grone, D. Schlondorff. 2002. Roles of SLC/CCL21 and CCR7 in human kidney for mesangial proliferation, migration, apoptosis, and tissue homeostasis. J. Immunol. 168: 4301-4307. [Abstract/Free Full Text]
  9. Kriehuber, E., S. Breiteneder-Geleff, M. Groeger, A. Soleiman, S. F. Schoppmann, G. Stingl, D. Kerjaschki, D. Maurer. 2001. Isolation and characterization of dermal lymphatic and blood endothelial cells reveal stable and functionally specialized cell lineages. J. Exp. Med. 194: 797-808. [Abstract/Free Full Text]
  10. Ngo, V. N., H. Korner, M. D. Gunn, K. N. Schmidt, D. S. Riminton, M. D. Cooper, J. L. Browning, J. D. Sedgwick, J. G. Cyster. 1999. Lymphotoxin {alpha}/β and tumor necrosis factor are required for stromal cell expression of homing chemokines in B and T cell areas of the spleen. J. Exp. Med. 189: 403-412. [Abstract/Free Full Text]
  11. Ansel, K. M., L. J. McHeyzer-Williams, V. N. Ngo, M. G. McHeyzer-Williams, J. G. Cyster. 1999. In vivo-activated CD4 T cells upregulate CXC chemokine receptor 5 and reprogram their response to lymphoid chemokines. J. Exp. Med. 190: 1123-1134. [Abstract/Free Full Text]
  12. Kim, M. Y., F. M. Gaspal, H. E. Wiggett, F. M. McConnell, A. Gulbranson-Judge, C. Raykundalia, L. S. Walker, M. D. Goodall, P. J. Lane. 2003. CD4+CD3– accessory cells costimulate primed CD4 T cells through OX40 and CD30 at sites where T cells collaborate with B cells. Immunity 18: 643-654. [Medline]
  13. Lane, P. J. L., M. C. Gaspal, M.-Y. Kim. 2005. Two sides of a cellular coin: CD4+CD3– cells orchestrate memory antibody responses and lymph node organisation. Nat. Rev. Immunol. 5: 655-660. [Medline]
  14. Kim, M. Y., K. M. Toellner, A. White, F. M. McConnell, F. M. Gaspal, S. M. Parnell, E. Jenkinson, G. Anderson, P. J. Lane. 2006. Neonatal and adult CD4+CD3– cells share similar gene expression profile, and neonatal cells up-regulate OX40 ligand in response to TL1A (TNFSF15). J. Immunol. 177: 3074-3081. [Abstract/Free Full Text]
  15. Mebius, R. E.. 2003. Organogenesis of lymphoid tissues. Nat. Rev. Immunol. 3: 292-303. [Medline]
  16. Ngo, V. N., R. J. Cornall, J. G. Cyster. 2001. Splenic T zone development is B cell dependent. J. Exp. Med. 194: 1649-1660. [Abstract/Free Full Text]
  17. Gaspal, F. M., M. Y. Kim, F. M. McConnell, C. Raykundalia, V. Bekiaris, P. J. Lane. 2005. Mice deficient in OX40 and CD30 signals lack memory antibody responses because of deficient CD4 T cell memory. J. Immunol. 174: 3891-3896. [Abstract/Free Full Text]
  18. Kim, M.-Y., G. Anderson, I.-L. Martensson, L. Erlandsson, W. Arlt, A. White, P. J. L. Lane. 2005. OX40-ligand and CD30-ligand are expressed on adult but not neonatal CD4+CD3– inducer cells: evidence that IL7 signals regulate CD30-ligand but not OX40-ligand expression. J. Immunol. 174: 6686-6691. [Abstract/Free Full Text]
  19. Amakawa, R., A. Hakem, T. M. Kundig, T. Matsuyama, J. J. Simard, E. Timms, A. Wakeham, H. W. Mittruecker, H. Griesser, H. Takimoto, et al 1996. Impaired negative selection of T cells in Hodgkin’s disease antigen CD30-deficient mice. Cell 84: 551-562. [Medline]
  20. Kim, M. Y., F. M. McConnell, F. M. Gaspal, A. White, S. H. Glanville, V. Bekiaris, L. S. Walker, J. Caamano, E. Jenkinson, G. Anderson, P. J. Lane. 2007. Function of CD4+CD3– cells in relation to B- and T-zone stroma in spleen. Blood 109: 1602-1610. [Abstract/Free Full Text]
  21. Rot, A., U. H. von Andrian. 2004. Chemokines in innate and adaptive host defense: basic chemokinese grammar for immune cells. Annu. Rev. Immunol. 22: 891-928. [Medline]
  22. Benedict, C. A., C. De Trez, K. Schneider, S. Ha, G. Patterson, C. F. Ware. 2006. Specific remodeling of splenic architecture by cytomegalovirus. PLoS Pathog. 2: e16[Medline]
  23. Kerjaschki, D., H. M. Regele, I. Moosberger, K. Nagy-Bojarski, B. Watschinger, A. Soleiman, P. Birner, S. Krieger, A. Hovorka, G. Silberhumer, et al 2004. Lymphatic neoangiogenesis in human kidney transplants is associated with immunologically active lymphocytic infiltrates. J. Am. Soc. Nephrol. 15: 603-612. [Abstract/Free Full Text]
  24. Withers, D. R., M.-Y. Kim, V. Bekiaris, S. W. Rossi, W. E. Jenkinson, F. Gaspal, F. McConnell, G. Anderson, and P. J. L. Lane. 2007. The role of lymphoid tissue inducers in splenic white pulp development. Eur. J. Immunol. In press.



This article has been cited by other articles:


Home page
 J. Immunol.Home page
V. Bekiaris, F. Gaspal, M.-Y. Kim, D. R. Withers, F. M. McConnell, G. Anderson, and P. J. L. Lane
CD30 Is Required for CCL21 Expression and CD4 T Cell Recruitment in the Absence of Lymphotoxin Signals
J. Immunol., April 15, 2009; 182(8): 4771 - 4775.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
V. Bekiaris, O. Timoshenko, T. Z. Hou, K. Toellner, S. Shakib, F. Gaspal, F. M. McConnell, S. M. Parnell, D. Withers, C. D. Buckley, et al.
Ly49H+ NK Cells Migrate to and Protect Splenic White Pulp Stroma from Murine Cytomegalovirus Infection
J. Immunol., May 15, 2008; 180(10): 6768 - 6776.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bekiaris, V.
Right arrow Articles by Lane, P. J. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bekiaris, V.
Right arrow Articles by Lane, P. J. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS