| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Laboratory of Immunobiology for Research and Application to Diagnosis, Center for Transfusion and Tissue Bank, Badalona, Spain; and
Cytometry Unit, Hospital Universitari "Germans Trias i Pujol," Department of Cell Biology, Physiology, and Immunology, Universitat Autònoma de Barcelona, Barcelona, Spain
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, lymphotoxin 
, C-C chemokine ligand (CCL) 21, CXC chemokine ligand (CXCL) 12, CXCL13, and CCL22 were increased in thyroids from autoimmune patients, whereas CXCL12, CXCL13, and CCL22 levels were significantly higher in autoimmune glands with ectopic secondary lymphoid follicles than in those without follicles. Interestingly, thyroid epithelium produced CXCL12 in response to proinflammatory cytokines providing a possible clue for the understanding of how tissue stress may lead to ectopic follicle formation. The finding of a correlation between chemokines and thyroid autoantibodies further suggests that intrathyroidal germinal centers play a significant role in the autoimmune response. Unexpectedly, the percentage of circulating CXCR4+ T cells and CCR7+ B and T cells (but not of CXCR5) was significantly reduced in PBMCs of patients with autoimmune thyroid disease when they were compared with their intrathyroidal lymphocytes. This systemic effect of active intrathyroidal lymphoid tissue emerges as a possible new marker of thyroid autoimmune disease activity. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hashimoto’s disease (HT) and Graves-Basedow thyrotoxicosis (GD) are two clinical forms of autoimmune thyroid disease (AITD) (18) in which lymphoid infiltration commonly evolves into secondary LFs (19). Despite longtime suspicion (20), only recently it has been shown that these structures are fully functional, yet their role in pathogenesis is unclear (19). To sustain their activity, intrathyroidal LFs must be included in the migration circuits of B and T lymphocytes. However, migration by itself is not sufficient, because massive infiltration can occur in the thyroid without LF neogenesis. Different types of organizing factors must be present in the tissue. It is currently accepted that lymphotoxin (LT) , LT, C-C chemokine ligand (CCL) 21 (secondary lymphoid tissue chemokine), CXC chemokine ligand (CXCL) 12 (stromal cell-derived factor-1), CXCL13 (B cell-attracting chemokine-1), and CCL22 (macrophage-derived chemokine) are probably the most important mediators for GC formation (21). CCL21 selectively attracts naive lymphocytes and DCs via CCR7 (CD197) (22), and DCs in turn produce CCL22 that induces segregation of lymphocytes into T and B zones (23). CXCL12 has homeostatic chemotactic activity via CXCR4 (CD184) on most T and B cells subpopulations and contributes decisively to the development of GCs and maintenance of LFs (24). CXCL13 organizes the distribution of B and T cells within the GCs via CXCR5 (25).
In previous studies, CCL3 (macrophage-inhibitory protein-1), CCL4 (macrophage-inhibitory protein-1), CCL5 (26), CCL21, CXCL12, and CXCL13 (19) have been detected in AITD glands and elevated levels of macrophage chemoattractant protein-1 in the patient’s sera (27), whereas CXCR3 and CCR5 (28) were associated with GD. In contrast, in experimental autoimmune thyroiditis (29), several chemokines have been involved in the disease. There are no systematic studies on the expression of the chemokines responsible for the GC organization and their corresponding receptors in AITD.
We report here that autoimmune thyroid glands hyperexpress LT, IFN-
, CXCL12, CXCL13, and CCL22. Those containing secondary lymphoid follicles overexpress IFN-, CXCL12, CXCL13, and CCL22 and express LT and CCL21, and the corresponding receptors CXCR4, CXCR5, LTR, and CCR7. We found that thyroid epithelial cells produce CXCL12 in vivo and in vitro; this may explain the relatively high frequency of LFs in thyroid glands. In patients with intrathyroidal LFs, which are rich in CXCR4+ T lymphocytes, we detected a concomitant reduction in circulating CXCR4+ T cells and also in CCR7+ T and B lymphocytes. This constitutes an unexpected effect on the periphery of intrathyroidal lymphoid neogenesis and suggests the existence of a general disturbance in lymphocyte recirculation circuit as part of AITD.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thyroid tissue was obtained from 34 patients (27 of them had been included in a previous study (19)): 4 HT, 12 GD with LFs, 11 GD without LFs, 1 multinodular goiter (MNG) with thyroiditis, 4 MNGs and 2 multiorgan donor glands. MNG and donor glands were used as nonautoimmune thyroid controls. Clinical diagnosis was made based on usual thyroid tests and confirmed by histopathology. Thyroid peroxidase (TPO) and thyroglobulin Abs were measured by ELISA (Immunowell, San Diego, CA), and anti-thyroid-stimulating hormone receptor Abs were measured by RIA (Brahms Diagnostica, Berlin, Germany) (Table I). A human thymus, two palatine tonsils (PT), and two lymph nodes (one of them from the HT385 patient) were used as reference lymphoid tissue. Because the palatine tonsils were not inflamed at the time of surgery, we considered their values basal. Cervical lymph nodes containing highly active lymphoid follicles were considered positive controls. For molecular biology techniques and intrathyroidal lymphoid follicle staining, 4-µm cryosections from thyroids (3 of 4 HTs, 16 of 23 GDs, 4 of 5 MNGs, and 2 of 2 normal donors) and tissue controls were used.
|
Thyroid monolayers were maintained in RPMI containing 10% FCS, and the percentage of FCS was decreased to 2% 2 h before exposure to the cytokines recombinant human (rh) IL-1 200 IU/ml (DAKO, Glostrup, Denmark), rhIFN- 500 IU/ml, or rhTNF- 500 IU/ml (supplied by G. R. Adolph, Boehringer Institute, Vienna, Austria). The human thyroid cell line HT93 was also used. It was generated by transfection with a SV40 early region construct and originally produced thyroglobulin (31). It has maintained epithelial features, and we have already used it in studies of HLA regulation.
PBMCs from 12 patients (obtained at the same time as the thyroid resection procedure) and 12 healthy donors were separated using a Ficoll (Sigma-Aldrich, St. Louis, MO) discontinuous density gradient.
Immunofluorescent histochemistry and FACS
Sources of Abs were: goat Abs to CXCL12, CCL21 and CXCL13, and murine mAbs to CXCR4 and CXCR5 from R&D Systems (Minneapolis, MN); mAb to cytokeratin 8–18 from Novocastra Laboratories (Newcastle upon Tyne, U.K.); mAb to CD83 from T. F. Tedder (Duke University Medical Center, Durham, NC); mAbs to human CD19, CD20, CD3, and CD5 from R. Vilella (Hospital Clínic Barcelona, Barcelona, Spain), murine mAb to CD21L from P. Garrone (Laboratory for Immunological Research, Schering-Plough, Paris, France); mAb to CD4 and CD8 from M. Bofill (Fundcaó IRSI-Caixa, London, U.K.); mAbs to CD3-PerCP, CCR7, CD45RA-FITC, and CD45RO-PE from BD PharMingen (San Diego, CA), and mAb to CD19-PECy5 from DAKO; and fluorochrome-labeled secondary Abs from Southern Biotechnology (Birmingham, AL). For thyroperoxidase detection, a high titer anti-TPO-positive human serum was used.
Cryosections were fixed in 4% paraformaldehyde (CXCL12 and CXCL13), or cold acetone (CCL21). Independent observers blindly examined slides. Images were acquired with a high resolution video camera and analyzed using Openlab software (Improvision, Coventry, U.K.).
Three-color flow cytometry was performed using a FACScan (BD Biosciences, San Jose, CA) to assess chemokine receptors expression in ITLs and PBMCs. Data were acquired and analyzed with the CellQuest software (BD Biosciences). Intrathyroidal T and B cells were sorted in a FACSVantage cell sorter (BD Biosciences); their purity ranged from 94 to 99%.
CXCL12 ELISA
CXCL12 levels were measured in supernatants from HT93 cells and from thyroidal cultures by an in-house sandwich ELISA (T. Gallart, Hospital Clínic Barcelona, Barcelona, Spain). Plates containing 96 wells were coated with Abs to human CXCL12 (R&D Systems). Soluble CXCL12 was detected with a biotinylated Ab to human CXCL12, followed by streptavidin-peroxidase incubation. The ELISA detection range was 0.39–200 ng/ml (human CXCL12; PeproTech, Rocky Hill, NJ).
RT-PCR and quantitative real time PCR
RNA was obtained using a modified Chomczynski method (32); genomic DNA was removed and then was retrotranscribed with oligo(dT)15 (Pharmacia Biotech, Peapack, NJ) and SuperScript-II (Pharmacia Biotech). cDNA duplicates were amplified by semiquantitative PCR using DynaZyme II (Finnzymes OY, Helsinki, Finland) and specific primers for LT, LT, LTR, and IFN- (conditions listed in Table II). Amplimers were hybridized with [-32P]ATP oligoprobes and quantified in a Bio-Rad Molecular Imager-FX (Bio-Rad, Hercules, CA) by using Quantity-One. Results were normalized by GAPDH expression.
|
|
The transmigration assay was conducted on Transwell plates (5-µm pore size; Corning, Corning, NY), by triplicate or duplicate, depending on cell availability: 3 x 104–105 sorted CD3+ and CD19+ cells were plated in the upper chamber; and the lower chamber was filled with medium containing 10–1000 ng/ml human CXCL12 (R&D Systems). Supernatants from thyrocytes cultured with IL-1, IFN-, or TNF- (at 200, 500, and 500 IU/ml, respectively) taken at peak CXCL12 protein (quantified by the previous ELISA) or control medium were also used to measure spontaneous migration. In some experiments, cells were preincubated with blocking anti-CXCR4 (12G5, 20 µg/ml). Migrated cells were counted using Flow-Count Fluorospheres (Coulter, Miami, FL).
Statistical analysis
Pearson’s correlation coefficient, together with parametric (Student t test) and nonparametric (Mann-Whitney) tests, were applied to normal and nonnormal distributed data, respectively, using the SPSS software package (Microsoft, Seattle, WA). A level of significance of 5% was used in all the statistical evaluations.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IFN-, CXCL12, CXCL13, and CCL22, but not LT or LTR, are overexpressed in autoimmune thyroid glands
The levels of chemokine transcription were analyzed in relation to the diagnosis and presence of intrathyroidal GCs. LT, IFN-, CXCL12, and CCL22 mRNA levels were higher in AITD glands than in nonautoimmune glands (p = 0.022, p = 0.024, p = 0.030, p = 0.010, respectively). Elevated CXCL13 was a constant and exclusive feature of AITD glands (p < 0.01) whereas CCL21 showed a nonstatistically significant association with disease (p = 0.09). Presence of GC was associated to higher levels of CXCL12, CXCL13, CCL22, and IFN- (n = 20; p = 0.022, p < 0.001, p = 0.007, and p = 0.024, respectively). Lymphocytic infiltration was also associated with CXCL13 and CCL22 levels (CXCL13, r = 0.972, p < 0.0009; CCL22, r = 0.983, p < 0.0009). Finally, expression of LT and/or of LTR did not bear a clear relation to diagnosis (p = 0.6 and p = 0.54) (Fig. 1).
The immunohistochemistry tests showed CCL21+ cells around the GCs in up to 30% of capillaries with features of high endothelial venules (CLA+; Fig. 2). CXCL12 was found in thyroid epithelium (TPO+ and cytokeratin+; Fig. 2) and also in clusters of nonidentified cells amid the lymphocytic infiltration. By contrast, CXCL13+ cells were distributed in the mantle zone of LFs. Although we were unable to clearly identify CD83+ mature DCs or follicular DCs as CXCL13+ producers (due to nonappropriate reagents), the pattern of staining points to the former as the CXCL13 source (Fig. 2).
|
Because production of CXCL12 by thyrocytes has pathogenic implications, we tested the effects of stimuli (IL1, IFN-, or TNF-) presumably acting in vivo on primary thyroid cell cultures. These cytokines clearly induced a raise of CXCL12 secretion (peak at 6 h) over baseline levels more marked for IL-1, which produced a 2-fold increase: 7.56 ± 0.75 vs 3.72 ± 1.4 ng/ml (Fig. 3A). To rule out the possibility that CXCL12 was derived from contaminating macrophages or fibroblasts, it was also measured in the supernatants of HT93 cells, and similar results were obtained (Fig. 3A). Semiquantitative RT-PCR and hybridization detected CXCL12 mRNA in both tissues, confirming ELISA results (Fig. 3A). Interestingly, these supernatants from cytokine-stimulated thyrocytes showed a chemotactic activity that was proportional to CXCL12 levels, and chemotaxis was blocked with anti-CXCR4 Abs (Fig. 3B).
|
Thyroid autoantibody levels correlate with CXCL13 and CCL22
In a previous study, we reported the association between thyroid autoantibody titers and the presence of intrathyroidal GCs. Here, we describe a correlation between thyroid autoantibody titers and chemokine levels: anti-thyroglobulin Abs correlated with CXCL13 (r = 0.951, p < 0.0001) and CCL22 (r = 0.906, p < 0.0001) levels; anti-TPO Abs correlated with CXCL13 (r = 0.940, p < 0.0001) and CCL22 (r = 0.910, p < 0.0001) levels. When GD were analyzed vs nonautoimmune glands, anti-thyroid-stimulating hormone receptor Abs correlated with IFN- (n = 14, r = 0.726, p = 0.025) and CCL21, CXCL13 and CCL22 levels (n = 18; r = 0.829, p < 0.0001; r = 0.831, p < 0.0001; and r = 0.877, p < 0.0001, respectively).
Infiltrating leukocytes express the corresponding chemokine receptors
In ITLs, chemokine receptors were expressed in a higher proportion of B cells than T cells following the hierarchy CXCR4>CXCR5>CCR7 (Table III and Fig. 4A); a higher density was reached in B cells than in T cells (MFI of CXCR4 109.9 vs 38.2; CXCR5 140 vs 49.7; and CCR7 27.8 vs 20.7). Preliminary experiments excluded a significant effect of the trypsin-collagenase solution on their membrane expression on ITLs but confirmed an overall 30% reduction in CD45 expression; therefore, the relative (but not absolute) distribution of chemokine receptor among the CD45RA and CD45RO subsets could be analyzed. Taking the predominance of CD45RO over CD45RA cells into account (ratio, 4:3), the distribution of CXCR4 (ratio, 3:1) indicates its preferential expression in the CD45RO+ subset. The same applies to CXCR5 and CCR7 (Table III). The immunohistochemistry studies confirmed the flow cytometry results: CXCR4+CD3+ or CXCR4+CD19+ cells were mainly located outside the LFs (Fig. 4A), CXCR5+CD3+ or CXCR5+CD19+ lymphocytes were mainly distributed in the mantle zone and within the GC, whereas only some CD4+ T cells were occasionally CXCR5+ in areas of diffuse infiltration (Fig. 4A). In contrast, CCR7 was little expressed by few T cells and B cells, possibly because those cells lose their receptor on arrival to the GC. In addition, mature DCs (CD83+) outside the lymphoid follicles were brightly stained (Fig. 4A).
|
|
Given that several factors (extracellular matrix, cell activation, soluble cytokines, and others) can determine molecular functions, we assessed the functionality of CXCR4 molecules on the surface of intrathyroidal B and T lymphocytes by chemotaxis assays with sorted CD3+ and CD19+ ITLs. Both sorted intrathyroidal T and B lymphocytes dose dependently migrated in a CXCL12 gradient, thereby confirming the functionality of CXCR4; maximal migration index was obtained at 100 ng/ml (33) for T cells and at 1000 ng/ml for B cells (Fig. 4B).
Marked peripheral reduction in CCR7+ and CD3+CXCR4+ lymphocytes of AITD-GC+ patients
The comparison between ITLs and PBMCs showed that CXCR4+ lymphocytes were more abundant in the former (63.6 ± 12.9% vs 31.6 ± 12.9%; Fig. 5). This was due to their low proportion among circulating T cells (59.6 ± 15.0% vs 21.9 ± 15.1%; p < 0.0001); no such differences were found for B cells. This reduction of circulating CXCR4+ T cells was particularly marked in the CD45RO subset (32.6 ± 10.7% vs 5.1 ± 8.6% p < 0.001; Table III). No significant differences in the proportion of CXCR5+ cells were found when the main lymphocyte subsets were compared. The distribution of CCR7+ cells was the opposite to that of CXCR4; CCR7+ lymphocytes were significantly less abundant among ITLs than among PBMCs from the same patient (4.7 ± 2.8% vs 32.9 ± 15.1%; p < 0.0001), and this trend was found in both T cells and B cells (2.5 ± 1.4% vs 39.3 ± 13.6%, p = 0.018; and 11.4 ± 7.1 vs 39.8 ± 17.0 p = 0.028, respectively; Fig. 5).
|
Regarding CXCR5, no significant differences were found between PBMCs from AITDs and healthy donors (23.1 ± 2.7% vs 19.1 ± 4.2%; p = 0.391; Fig. 5). The same applies to distribution analysis conducted among T and B cells and to the CD45RO and CD45RA T cell subsets.
The percentage of CCR7+ cells was lower in AITD patients than in controls (35.3 ± 12.8% vs 68.5 ± 7.0%; p = 0.0021; Fig. 5). This was attributable to both the T cell (39.3 ± 13.6% vs 67.3 ± 7.9%; p = 0.013) and the B cell compartments (40.3 ± 17.0% vs 82.6 ± 3.8%; p = 0.0005; Fig. 5). Among T cells, the lowest proportion of CCR7+ cells was found in the memory CD45RO+ subset (8.9 ± 5.0% vs 31.7 ± 13.0%; p = 0.006; Table III).
To simplify the representation of results concerning the percentage of cells expressing CD45RA and CD45RO, the results were recalculated, and the graph was plotted taking the percentage of positive cells as 100% (Fig. 6).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. The low number of CCR7+-infiltrating lymphocytes which cannot be explained by the lack of high endothelial venules expressing CCL21, the initial event in LFs formation and prime chemoattractor of naive T cells (35), because we detected CCL21 in intrathyroidal high endothelial venules to an extent that is similar to those in other autoimmune (36, 37) or infectious diseases (38). The most plausible explanation is that once in the tissue, CCR7 is down-regulated; this would indicate indirectly that the lymphocyte turnover in intrathyroidal LFs is relatively low at this stage. By contrast, CXCR4 and CXCR5 receptors remain expressed by both T and B cells, probably because of their role in maintaining the distribution of the different cell types in distinct areas of the intrathyroidal LFs.
2. The lack of correlation of LT and LTR expression, factors known to play a fundamental role in the initiation of LFs, with the presence of ectopic LFs. This difference, as the former, may be due to the late stage of maturation of the intrathyroidal LFs in our material. In fact, by the time patients undergo surgery, the disease has been active for years, and therefore early events are not necessarily present.
3. Thyrocytes are the main source of CXCL12 in AITD. This is important but not totally unexpected because CXCL12 is an ancestral chemokine produced by a variety of cell types (24). Thyrocyte production of biologically active CXCL12 and infiltrating lymphocyte responsiveness to it are conclusively demonstrated by a combination of RT-PCR, immunofluorescent histochemistry, and chemotaxis techniques. Thyrocytes can also produce CCL2 (macrophage chemoattractant protein-1) (39), CXCL19 (monokine induced by IFN-), and CXCL10 (IFN--inducible protein-10) (40, 41), and this may explain why LFs arise so often in the thyroid gland even in MNG, a clinical entity of uncertain etiology. It is conceivable that the initial event leading to the formation of intrathyroidal LFs is local nonspecific stress, e.g., iodine overload as in nonobese diabetic mouse thyroiditis (42). Iodine may induce thyrocyte necrosis, which would stimulate resident macrophages to produce IL-1 and TNF, which may in turn induce CXCL12 synthesis by adjacent thyrocytes.
The finding of a correlation between the levels of CCL21, CCL22, and CXCL13 and thyroid autoantibodies is well in agreement with our previous report of a correlation between the presence of intrathyroidal LFs and thyroid Ab titers (19). The capability of the thyroid tissue to express such a diversity of chemokines may contribute decisively to the local differentiation of autoreactive B cells and to the pathogenesis of AITDs. This concept is also in line with the geographical view of immune response proposed by Zinkernagel et al. (4); i.e., the normal regional recirculation and distribution of lymphocytes would be crucial to prevent autoimmunity, and therefore its disruption, as here reported, may be a decisive event in the development of autoimmunity.
The question of the formation of intrathyroidal LFs is part of a more general question of the reasons for the development of TLTs in many conditions. Secondary lymphoid organs are designed to maximize the chances of a scarce Ag meeting a rare lymphocyte bearing the appropriate receptor. When the immune system fails to create an efficient immune response against a highly localized Ag (e.g., a microorganism), it may be advantageous to move a critical anatomical element of the immune response, the lymphoid follicle, to the site of the infection. The stimulus that triggers lymphoid neogenesis may be the continuous production of cytokines and other inflammatory mediators by the stromal cells (43). TLTs lack the intricate circulation channels that distribute scarce incoming Ag through the lymph node. They are probably not needed in TLT because their LFs are responding to local Ags that are both abundant and readily accessible (and this may be the main difference between TLTs and secondary lymphoid tissue). Lymphoid neogenesis in autoimmunity may be an undesirable consequence of this mechanism and may arise as results of chronic local stress (either mechanical stress in the joints or chemical stress in the thyroid). The formation of extranodal LFs does not inevitably lead to autoimmunity, because there is tolerance to local self-Ags, but breaking tolerance may be easier within these extranodal LFs because of the huge excess of Ag and the abundance of costimulatory signals (such as intrathyroidal chemokines) that may activate ever present low affinity T cells and B cells. Unrestrained receptor revision occurring within these LFs may eventually lead to the formation of high affinity autoreactive B cells. Intrathyroidal LFs are highly activated (19), and this is not surprising because there is plenty of Ag and inflammatory mediators to keep the reaction going.
Finally, one of the most striking results from this work is the marked decrease in the proportion of circulating CXCR4+ T cells and CCR7+ lymphocytes, and the less marked but still significant reduction in CXCR5+ B cells in AITD patients with ectopic LF. These phenotypic changes may constitute good markers of the activity of the cellular immune response to the thyroid Ags and may be useful tests for the clinical assessment of the patients. Ongoing work at our laboratory will determine whether this pattern of chemokine receptor expression in PBLs of AITD patients is a feature of the group whose glands contain LFs and also whether its is present in patients suffering other autoimmune diseases characterized by the formation of TLT.
Autoimmune diseases are the end result of a multistep process in which genetic and environmental factors interact for a long time. In humans, it is very difficult to define the main checkpoints, but organization of TLT may be an important step in the consolidation of an autoimmune response. The data presented in this article define a broad and complex network of chemoattractant molecules determining the migration of lymphocytes to the thyroid and the formation of intrathyroidal lymphoid tissue, a process that, given its correlation with thyroid autoantibody levels and its repercussion in the phenotype of circulating lymphocytes, we believe constitutes a decisive checkpoint in the development of AITD.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. R. Pujol-Borrell, Laboratory of Immunobiology for Research and Application to Diagnosis-Immunology Unit, Hospital Universitari "Germans Trias i Pujol," Ctra. Canyet s/n, 08916 Badalona (Barcelona), Spain. E-mail address: ricardo.pujol{at}ub.es
3 Abbreviations used in this paper: DC, dendritic cell; LF, lymphoid follicle; CCL, C-C chemokine ligand; CXCL, CXC chemokine ligand; LT, lymphotoxin; GC, germinal center; HT, Hashimoto’s disease; GD, Graves-Basedow thyrotoxicosis; TLT, tertiary lymphoid tissue; AITD, autoimmune thyroid disease; MNG, multinodular goiter; TPO, thyroid peroxidase; PT, palatine tonsils; ITLs, intrathyroidal lymphocytes; rh, recombinant human.
Received for publication December 20, 2002. Accepted for publication March 27, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chemokine, TCA4, with activity on T-lymphocytes and mesangial cells. J. Immunol. 159:5671.[Abstract]
cells in type I diabetes. Mol. Biol. Med. 3:159.[Medline]
This article has been cited by other articles:
|
|
 |
|
  E. E. McCandless, L. Piccio, B. M. Woerner, R. E. Schmidt, J. B. Rubin, A. H. Cross, and R. S. Klein Pathological Expression of CXCL12 at the Blood-Brain Barrier Correlates with Severity of Multiple Sclerosis Am. J. Pathol., March 1, 2008; 172(3): 799 - 808. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Manzo, S. Bugatti, R. Caporali, R. Prevo, D. G. Jackson, M. Uguccioni, C. D. Buckley, C. Montecucco, and C. Pitzalis CCL21 Expression Pattern of Human Secondary Lymphoid Organ Stroma Is Conserved in Inflammatory Lesions with Lymphoid Neogenesis Am. J. Pathol., November 1, 2007; 171(5): 1549 - 1562. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. C. Furtado, T. Marinkovic, A. P. Martin, A. Garin, B. Hoch, W. Hubner, B. K. Chen, E. Genden, M. Skobe, and S. A. Lira Lymphotoxin beta receptor signaling is required for inflammatory lymphangiogenesis in the thyroid PNAS, March 20, 2007; 104(12): 5026 - 5031. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G.-Z. Liu, L.-B. Fang, P. Hjelmstrom, and X.-G. Gao Increased CD8+ central memory T cells in patients with multiple sclerosis Multiple Sclerosis, March 1, 2007; 13(2): 149 - 155. [Abstract] [PDF]
|
|
|
|
 |
|
 D. Viemann, M. Goebeler, S. Schmid, U. Nordhues, K. Klimmek, C. Sorg, and J. Roth TNF induces distinct gene expression programs in microvascular and macrovascular human endothelial cells J. Leukoc. Biol., July 1, 2006; 80(1): 174 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Aust, K Krohn, N G Morgenthaler, S Schroder, A Schutz, J Edelmann, and E Brylla Graves' disease and Hashimoto's thyroiditis in monozygotic twins: case study as well as transcriptomic and immunohistological analysis of thyroid tissues Eur. J. Endocrinol., January 1, 2006; 154(1): 13 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Chen, Y. Wei, A. Alter, G. C. Sharp, and H. Braley-Mullen Chemokine expression during development of fibrosis versus resolution in a murine model of granulomatous experimental autoimmune thyroiditis J. Leukoc. Biol., September 1, 2005; 78(3): 716 - 724. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Burman, O. Haworth, D. L. Hardie, E. N. Amft, C. Siewert, D. G. Jackson, M. Salmon, and C. D. Buckley A Chemokine-Dependent Stromal Induction Mechanism for Aberrant Lymphocyte Accumulation and Compromised Lymphatic Return in Rheumatoid Arthritis J. Immunol., February 1, 2005; 174(3): 1693 - 1700. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. Martin, E. C. Coronel, G.-i. Sano, S.-C. Chen, G. Vassileva, C. Canasto-Chibuque, J. D. Sedgwick, P. S. Frenette, M. Lipp, G. C. Furtado, et al. A Novel Model for Lymphocytic Infiltration of the Thyroid Gland Generated by Transgenic Expression of the CC Chemokine CCL21 J. Immunol., October 15, 2004; 173(8): 4791 - 4798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Carragher, R. Johal, A. Button, A. White, A. Eliopoulos, E. Jenkinson, G. Anderson, and J. Caamano A Stroma-Derived Defect in NF-{kappa}B2-/- Mice Causes Impaired Lymph Node Development and Lymphocyte Recruitment J. Immunol., August 15, 2004; 173(4): 2271 - 2279. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-S. Li, N. Kanamoto, Y. Hataya, K. Moriyama, H. Hiratani, K. Nakao, and T. Akamizu Transgenic Mice Producing Major Histocompatibility Complex Class II Molecules on Thyroid Cells Do Not Develop Apparent Autoimmune Thyroid Diseases Endocrinology, May 1, 2004; 145(5): 2524 - 2530. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||
) are linked by a line; 
, PBMCs from healthy donors; horizontal bars, mean of 8–10 independent experiments. *, p < 0.05; **, p < 0.01; ns, nonsignificant. As an additional reference, the mean ± SD of CXCR4-positive cells in lymph nodes from healthy donors is given with the symbol I–.