Published online before print May 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Tumor Progression and Tumor Defense, German Cancer Research Center, Heidelberg, Germany; and
Department of Applied Genetics, University of Karlsruhe, Germany
Correspondence: Margot Zöller, Department of Tumor Progression and Immune Defense, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. E-mail: m.zoeller{at}dkfz.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: rodent • autoimmunity • adhesion molecules • apoptosis • signal transduction
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-knockout mice [8
9
10
11
], which spontaneously develop colitis on drug-induced models, such as trinitrobenzene sulfonic acid (TNBS) and dextran sodium sulfate, and on the transfer of a selected T cell population into severe combined immunodeficiency mice to induce colitis [12
, 13
]. These models provided convincing evidence for T helper cell type 1 (TH-1) cytokines as an essential component in IBD [3 , 14 , 15 ]. Yet, apparently IL-12 plays a pivotal role only during the early period of IBD [16 ]. It also has been shown that IL-2-/- mice with a targeted deletion of CD28 still develop colitis [17 ]. Both findings suggest that additional regulatory mechanisms are involved in the persisting activation of T cells in IBD. It is known that IBD is associated with high levels of CD40 and CD40L expression [18 ], but mainly, it has been suggested that regulatory elements may be altered [19 ]. Besides a suggested control of DC activation via CD4+CD25+ regulatory T cells [20 ], it was described that CD4+CD25+ T cells signal via cytotoxic T-lymphocyte antigen (CTLA)-4 and that constitutive expression of CTLA-4 is mainly restricted to these regulatory T cells [21 ]. It has also been noted that lamina propria lymphocytes (LPL) from inflamed tissue do not respond to down-regulatory signals such as IL-4 [22 ]. Recently, it also has been reported that LPL of patients with IBD display a marked reduction in apoptosis [23 ], and also that despite normal levels of CD95 expression, T cells are more resistant to CD95-induced apoptosis, which may be a result of elevated levels of Bcl-2 [24 ].
We described that CD44v7 appears to be directly involved in the chronification of gut-associated inflammatory processes, as an inflammation but no persisting colitis could be induced in CD44v7-/- mice. Initial experimental evidence pointed toward elevated levels of activation-induced cell death (AICD) in CD44v7-/-mice [25 ]. Furthermore, whereas mice with a targeted deletion of IL-10 spontaneously develop colitis, mice with a targeted deletion of IL-10 and CD44v7 do not develop colitis [25 ]. Finally, the development of a chronic colitis could be prevented by the application of a CD44v7-specific monoclonal antibody (mAb) [26 ].
Despite the involvement of CD44v7 in IBD and delayed-type hypersensitivity reactions, not much is known about its mode of action [25 26 27 ]. CD44 comprises a set of transmembrane glycoproteins, whose members differ by glycosylation [28 ] and by insertion of up to 10 variant exons between exon 5 and exon 6 of the CD44 standard isoform (CD44s) [29 ], which is expressed rather ubiquitously. The multitude of so-called variant isoforms (CD44v) is expressed in a tissue-related and developmentally restricted manner [30 ]. Considering functions of CD44 on hematopoietic cells, CD44s has been described as a lymphocyte homing receptor [30 , 31 ]. In addition, the molecule is involved in lymphocyte maturation [32 , 33 ], traffic [34 , 35 ], and activation [36 37 38 ]. CD44s functions as a costimulatory molecule in T cell activation as a result of its constitutive association with lck and its apposition toward the TCR upon stimulation and formation of the immunological synapse [39 ]. It is well known that formation of the immunological synapse is accompanied by redistribution of the actin cytoskeleton [40 41 42 ], and cross-linking CD44, too, leads to actin bundle formation in a RAC-dependent manner [43 ].
Already in 1991, Haynes et al. [44 ] suggested that the multitude of CD44 isoforms mediates a multitude of functions, a prediction that experimentally has been well supported. Nevertheless, with the exception of the proteoglycan isoforms CD44v3 and CD44v10 [45 , 46 ], knowledge of ligands and/or signal transduction initiated selectively via CD44v isoforms is still scarce. It has been suggested that CD44 associates with the cytoskeletal linker proteins ankyrin and the ezrin, radixin, moesin (ERM) family including merlin [47 48 49 50 51 52 ] and that the association with the ERM family may depend on the expression of CD44v isoforms [48 ]. It should also be mentioned that anchoring lipid rafts to the cytoskeleton through Cbp-EBP50-ERM has a negative effect on immune-synapse formation [53 , 54 ] and that the association with ERM is responsible for the exclusion of CD43 from the immunological synapse [55 , 56 ].
We investigated here functional activity of CD44v7. Based on the observations that CD44v7 apparently has a major impact on the development of chronic colitis and that resistance of CD44v7-/- mice toward TNBS-induced colitis was accompanied by a significantly increased number of apoptotic cells in the lamina propria, we speculated that CD44v7 interferes with the expression of pro-apoptotic genes or supports expression of anti-apoptotic genes. We tested the hypotheses by comparing responsiveness of lymphocytes deficient in or overexpressing CD44v7. We analyzed the expression level of the anti-apoptotic proteins Bcl-2, Bcl-Xl, and A1 and the phosphorylation status of the pro-apoptotic protein BAD, which exists in a phosphorylated (pBAD), inactive state and a dephosphorylated, active state [57 , 58 ]. We also show that T cell activation is not influenced by CD44v7 expression and provide evidence that CD44v7 is excluded from the immunological synapse, possibly via its preferential association with ezrin. Instead, CD44v7 expression correlates with Bcl-2 and Bcl-Xl expression and pBAD, which supports survival of activated T cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50%) but significantly exceed expression levels of endogenous CD44v4-v7. Mice were bred at the central animal facilities of the German Cancer Research Center (Heidelberg) and were kept under specific pathogen-free conditions. They were fed with sterilized TAP food and water ad libitum. Mice were used at the age of 8 weeks. One day before treatment with TNBS, mice were transferred from specific pathogen-free to conventional conditions. They received an intrarectal injection of 200 µl TNBS (2.5% w/v) in 50% ethanol [25
, 26
].
Antibodies
The following mAb were used: anti-CD44s [IM-7, recombinant immunoglobulin G (rIgG)2b; IM-7 does not recognize the CD44v4-v7 TG; American Type Culture Collection (ATCC), Manassas, VA], anti-Thy1.2 [YTS154.7.7.10, rIgG2b; European Collection of Cell Cultures (ECACC), Wiltshire, UK], anti-CD4 (YTA3.2.1, rIgG2b; ECACC), anti-CD8 (YTS169.4.1, rIgG2b; ECACC), anti-interferon-
(IFN-; R4-6A2, anti-rIgG; ATCC), anti-Mac-1 (YBM6.1.1, rIgG2a; ECACC), anti-µ (131.12, rIgG2b; ATCC), anti-CD3 (145-2C11, hamster IgG; ATCC), and anti-rat CD44v6 [A2.6, mouse IgG1]. The A2.6 mAb [60
], a subclone of which has been named ASML1.1 [59
], has been obtained after vaccination of mice with a metastasizing rat tumor line (BSp73ASML), which expresses CD44v4-v7 at an extraordinary high level. The mAb has been shown to exclusively recognize the v6 exon product [61
]. Culture supernatants were purified by passage over Protein G-Sepharose, and purified mAb were used in vitro at a concentration of 10 µg/ml. Anti-CD25, -CD28, -CD40, -CD69, -CD80, -CD86, -CD95, -CD95L, -CD152, and -CD154; dye-labeled [fluorescein isothiocyanate (FITC) or phycoerythrin] secondary antibody and Streptavidin; anti-ankyrin, -ezrin, and -moesin; and Phalloidin–FITC were obtained commercially (PharMingen, Hamburg, Germany). Anti-Bcl-2 (clone 124, mIgG1) and anti-pBAD (ser 112, sheep IgG) were obtained from Upstate Biotechnology (Lake Placid, NY); antiactin (polyclonal goat antiserum), anti-A1 (polyclonal goat antiserum), anti-Bcl-Xl (clone HS, mIgG1), and anti-BAD (polyclonal goat antiserum) were from Santa Cruz Biotechnology (Santa Cruz, CA).
For fluorescein-activated cell sorter (FACS) analysis, 3 x 105 cells were stained according to routine procedures. Expression of CD152 was evaluated in permeabilized cells. The percentage of apoptotic cells was evaluated by double-staining with annexin V–FITC and propidium iodine (PI). Fluorescence was determined using a FACSstar (Becton Dickinson, Heidelberg).
Preparation of lymphoid cells
Mesenteric lymph node cells (mLNC), spleen, and PP were removed under sterile conditions; peritoneal exudate cells (PEC) were collected by flushing the peritoneal cavity with 3 ml phosphate-buffered saline (PBS) containing 100 U heparin. Tissues were pressed through sterile gauze, and cell suspensions were washed several times with PBS. Seeding cells in flat-bottomed, microtiter plates enriched monocytes (M
). After 2 h of incubation at 37°C, nonadherent cells were removed by vigorous washing. T cells were enriched by depletion of plastic adherent cells (2x 1 h) and of B cells by panning on anti-mIg-coated plates (90 min at 4°C), collecting the nonadherent fraction.
In vitro stimulation and proliferation
Lymphocytes (2x105/well) were stimulated in vitro by seeding on anti-CD3 (10 µg/ml)-coated plates or on plates coated with 1 µg/ml anti-CD3 plus 10 µg/ml anti-CD44s or anti-CD44v6 or by culturing in the presence of 2 µg/ml phytohemagglutinin (PHA). For antigenic stimulation, LNC or T cells were cultured in the presence of trinitrophenyl-ovalbumin (TNP-OVA)-pulsed peritoneal exudate M, which served as antigen-presenting cells (APC). After 48–72 h of culture, cells were collected for additional analyses. Where indicated, 3H-thymidine (10 µCi/ml) was added during the last 8 h of culture. Cells were harvested, and thymidine incorporation was determined in a ß-counter.
Cells lysis and immunoprecipitation (IP)
Cells (107) were washed twice in TNE buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 1 mM Na2VO4, 10 mM NaF) and were lysed in the same buffer containing 1% 3-[(3-cholamidopropyl) dimethylammonio]-1-propane sulfonate (CHAPS) for IP or 1% Nonident P-40 + 0.1% sodium dodecyl sulfate (SDS) + 0.25% sodium deoxycholate for Bcl-2 protein extraction. In all conditions, 1 mM phenylmethylsulfonyl fluoride (PMSF) and a protease inhibitor cocktail (Boehringer Mannheim, Germany) were added to the lysis buffer. After 30 min at 4°C, lysates were centrifuged at 13,000 g for 10 min at 4°C. Lysates were analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting (WB) or were used for IP. For IP, lysates were precleared by the addition of 1/10 vol Protein G-Sepharose for 1 h at 4°C. Precleared lysates were incubated for 1 h at 4°C with 5 µg anti-CD44 followed by an additional 1 h incubation with Protein G-Sepharose. Immune complexes were washed four times with lysis buffer. Immunoprecipitated proteins were analyzed by SDS-PAGE, followed by WB.
WB
Samples were resolved on 10% or 4–20% gradient (for Bcl-2 proteins) SDS-PAGE under nonreducing conditions, and the proteins were transferred to Immobilon P (Millipore, Bedford, MA) at 30 V for 16 h at 4°C. Membranes were blocked [5% bovine serum albumin (BSA)], and immunoblotting was performed with the indicated antibodies, followed by the appropriate secondary horseradish peroxidase (HRP)-conjugated antibody. Blots were developed with the enhanced chemiluminescence detection system (Amersham, Little Chalfont, UK). Where indicated, antibody stripping was performed according to the manufacturer’s recommendation. Densitometric analysis was performed with NIH Image 1.60 software.
Purification of glycolipid-enriched membrane (GEM) fractions and subcellular fractionation
Cells (107) were seeded into antibody-coated Petri dishes and incubated for different times at 37°C. Stimulation was terminated by transferring the dishes to ice and immediate lysis of the cells for 30 min in 1 ml ice-cold TNE buffer containing 0.5% Triton X-100 and a protease inhibitor cocktail (Boehringer Mannheim). The lysate was adjusted to 40% (w/v) sucrose in TNE buffer. After transfer of the lysate to the centrifuge tube, 2 ml 30% (w/v) and 1 ml 5% (w/v) sucrose in TNE were overlaid. Samples were centrifuged for 16–18 h at 200,000 g at 4°C. Gradient fractions (0.4 ml) were collected from the top. Fractions were analyzed by dot blot using biotinylated cholera toxin B subunit (Sigma Chemical Co., St. Louis, MO) as a GM1 marker. GEM fractions (fractions 2–4) and soluble fractions (fractions 10–12) were pooled and analyzed on a 10% SDS-PAGE followed by WB with anti-CD44 mAb.
Statistics
Significance of differences was evaluated by the Student’s t-test or by multifactorial ANOVA using Statistical Analysis Software version 8.1.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
View this table: [in a new window] |
Table 1. Expression Level of CD44v7 in Comparison with panCD44
|
|
View this table: [in a new window] |
Table 2. Induction of Apoptosis in CD44v7-Competent and -Deficient Lymphocytes
|
CD44v7-/- mLNC were more susceptible to apoptosis than CD44v7+/+ mLNC. This was irrespective of whether the cells were only stimulated in vitro or were restimulated. Expression of CD44v7 on APC did not influence the susceptibility of CD44v7+/+ and CD44v7-/- mLNC to apoptosis. Overexpression of CD44v7 in mLNC of CD44v4-v7 TG mice protected cells from apoptosis, although the effect became apparent only after restimulation (Fig. 1 ).
 View larger version (19K): [in a new window] |
Figure 1. Protection from cell death is mediated preferentially by CD44v7-expressing T cells. Adherent PEC from untreated and TNBS-treated CD44v7-/- and CD44v7+/+ mice as well as from NTG and CD44v4-v7 TG mice were loaded with TNP-OVA. After 3 h of incubation, the adherent PEC were washed to remove free TNP-OVA, and mLNC from CD44v7+/+ versus CD44v7-/- 129SVEV mice as well as from CD44v4-v7 TG and NTG BALB/c mice, respectively, were added. Cells were incubated for 48 h. The population of nonadherent lymphocytes was collected and stained with annexin V-FITC plus PI. Mean values of triplicate cultures are presented. *Significant differences (P < 0.01) between the percentage of annexin-only-positive plus annexin/PI double-positive cells in mLNC of CD44v7+/+ versus CD44v7-/- and CD44v4-v7 TG versus NTG mice. The experiment was repeated three times revealing similar results. Mo, Monocytes.
|
|
View this table: [in a new window] |
Table 3. Influence of CD44v7 Expression on Proliferative Activity
|
CD44v7 is engaged in the activation of anti-apoptotic signaling molecules
CD95 expression did not differ in unstimulated or stimulated (in vivo and/or in vitro) mLNC and PPL of CD44v7+/+, CD44v7-/-, NTG, and TG mice (data not shown). However, mLNC and PPL of CD44v7-/- mice contained a higher percentage of CD95L+ cells. Intensity of expression was increased at a statistically significant level in PPL only (Fig. 2A
). The percentage of CD95L+ cells did not differ significantly between unstimulated lymphocytes from TG versus NTG mice. However, the percentage of CD95L-expressing cells increased during stimulation in NTG but not in TG mice (Fig. 2B)
. After co-culture with TNP-OVA-pulsed APC, expression of CD95L was slightly increased in mLNC of CD44v7-/- mice and decreased, although not at a statistically significant level, in mLNC from CD44v4-v7 TG mice. This was independent of whether the APC were CD44v7-competent or -deficient (Fig. 2C)
.
 View larger version (34K): [in a new window] |
Figure 2. Activation of apoptosis-inducing molecules and expression of CD44v7. mLNC and PPL from CD44v7+/+ and CD44v7-/- 129SVEV mice (A) and from CD44v4-v7 TG and NTG mice (B) were freshly harvested from untreated mice or from mice 8 days after intrarectal TNBS treatment. Where indicated, mLNC were stimulated in vitro with PHA or with anti-CD3 (10 µg/ml) for 48 h or with antigen-pulsed APC (C) as described in Figure 1
. Thereafter, expression of CD95L was evaluated by flow cytometry. Mean values of three independently performed experiments of the percentage of CD95L expressing mLNC and PPL (A–C) and of the mean intensity of CD95L expression (A and B) are presented. *, Significant differences (P < 0.01) in the percentage of CD95L+ lymphocytes of CD44v7+/+ versus CD44v7-/- and CD44v4-v7 TG versus NTG mice;  , significant differences (P < 0.01) in the mean intensity of expression ( , differences are not significant).
|
pBAD and the expression level of anti-apoptotic proteins Bcl-2, Bcl-Xl, and A1 were evaluated after stimulation of CD44v7-competent and -deficient LNC on pulsed APC (Fig. 3 ). mLNC were lysed, and proteins were separated by gel electrophoresis and were transferred for blotting. pBAD and the expression of Bcl-2 and Bcl-Xl were strongly reduced in mLNC from CD44v7-/- mice, irrespective of whether mLNC were co-cultured with CD44v7-competent or -deficient APC. We did not detect differences in the expression of A1. Differences in pBAD and Bcl-2 expression were less pronounced in mLNC from CD44v4-v7 TG versus NTG controls. Yet, expression of Bcl-Xl was up-regulated on mLNC of TG mice. As revealed by densitometry of three independently performed experiments (Table 4 ), differences in Bcl-2, Bcl-Xl, and pBAD expression in CD44v7+/+ versus CD44v7-/- mLNC and differences in Bcl-Xl expression between NTG and TG mLNC were highly significant. No differences in Bcl-2, Bcl-Xl, and pBAD expression were seen in unstimulated mLNC.
 View larger version (52K): [in a new window] |
Figure 3. Regulation of Bcl-2 family proteins and CD44v7 expression. (A) Freshly harvested mLNC from CD44v7+/+ and CD44v7-/- 129SVEV mice as well as from CD44v4-v7 TG and NTG BALB/c mice and (B) mLNC co-cultured with antigen-pulsed APC as described in Figure 1
were lysed and after electrophoresis, blotted with anti-Bcl-Xl, anti-Bcl-2, anti-pBAD, anti-BAD, and anti-A1. Blotting with anti-actin is included as control.
|
|
View this table: [in a new window] |
Table 4. Expression of Anti-apoptotic Genes in Activated mLNC of CD44v7-/- and CD44v4-v7 TG Mice
|
CD44v7 is involved in maintenance of the activated state
To confirm that CD44v7 is involved in regulation rather than in induction of a T cell response, expression of accessory molecules on T cells and of costimulatory molecules on APC was evaluated in mLNC after in vivo and in vitro stimulation (Fig. 4
). No differences in expression of CD44s, lymphocyte function-associated antigen-1 (LFA-1), and intercellular adhesion molecule-1 (ICAM-1) were seen in lymphocytes from TG versus NTG and CD44v7-/- versus CD44v7+/+ mice (data not shown). Deletion of the CD44v7 exon product was accompanied by a more pronounced up-regulation of CD69 and CD25 on mLNC after in vitro stimulation or restimulation as compared with CD44v7+/+ lymphocytes. A higher percentage of unstimulated mLNC from CD44v4-v7 TG as compared withNTG mice expressed CD69 and CD25. However, in TG mice, the percentage of CD69+ and CD25+ mLNC hardly increased after stimulation (Fig. 4A) .
 View larger version (35K): [in a new window] |
Figure 4. Expression of accessory molecules and their ligands on mLNC in relation to CD44v7 expression. mLNC from CD44v7-/-, CD44v7+/+, CD44v4-v7 TG, and NTG mice were stimulated in vivo by TNBS and/or in vitro by anti-CD3. Expression of the activation markers CD69 and CD25 (A), CD40 and CD40L (B), and CD152 (C) was evaluated by flow cytometry. Expression of these markers on freshly harvested mLNC of untreated mice is shown for comparison. The experiment was repeated three times and revealed comparable results. *, Significant differences (P < 0.01) between mLNC of CD44v7+/+ versus CD44v7-/- and CD44v4-v7 TG versus NTG mice.
|
CD40 expression is also involved in the up-regulation of CD80 and CD86, which can influence CD28 and CD152 (CTLA-4) expression. CD80 and CD86 expression did not differ between CD44v7+/+ and CD44v7-/- or between NTG and TG mice. Furthermore, no significant differences were seen in CD28 expression. This accounted for freshly harvested mLNC as well as for cells harvested after TNBS treatment (data not shown). Expression of CD152 was slightly increased in mLNC from CD44v7-/- as compared with CD44v7+/+ mice and was slightly reduced in mLNC of TG as compared with NTG mice (Fig. 4C) . An analysis of costimulatory and accessory molecules on PPL of CD44v7-competent, CD44v7-deficient, and CD44v4-v7-TG mice revealed similar features as described for mLNC (data not shown).
These features provided mainly two insights. Except for the influence of constitutively expressed CD44v7 (TG) on CD40 expression, CD44v7 did not influence expression of costimulatory molecules on APC. Yet, CD44v7 (over)expression was accompanied by reduced up-regulation of the activation markers CD25 and CD69 and a distinct reduction in CD152+ T cells, which comprise regulatory T cells. These features are compatible with CD44v7, acting during regulation rather than during induction of response.
CD44v7 and CD44s are recovered from different membrane microdomains and associate with different cytoskeletal-linker proteins
We and others [39
, 43
, 64
] demonstrated that CD44s is constitutively associated with lck and that upon cross-linking CD44s together with the CD3 complex, CD44s is driven into the immunological synapse, where it supports T cell activation or apoptosis depending on the maturation state of the T cell [39
]. The opposing features described for CD44v7 could be a result of CD44s and CD44v7 being located in different membrane microdomains and associating with distinct linker proteins. To test this hypothesis, we first looked for the presence of CD44s and CD44v7 in light (Triton X-100-insoluble) and dense (Triton X-100-soluble) membrane fractions, which were separated by sucrose gradient centrifugation.
mLNC were stimulated in vitro by cross-linking CD3 or CD44 or in vivo by TNBS treatment. CD44v7-/- and CD44v7+/+ mLNC (Fig. 5A and Table 5 ) and CD44v4-v7 TG and NTG mLNC (Fig. 5B and Table 5 ) were lysed, and the lysates were separated by sucrose density-gradient centrifugation. Proteins in light and dense fractions of the sucrose gradient were separated by gel electrophoresis and were blotted with IM-7 (all preparations) and with anti-rCD44v6 (only mLNC from TG mice). More CD44 was raft-associated (light fraction) in unstimulated and TNBS-stimulated CD44v7-/- mLNC than in CD44v7+/+ mLNC. After primary, in vitro stimulation by cross-linking CD3 or CD44, CD44 was enriched in the light fraction. This accounted for mLNC of CD44v7+/+ and CD44v7-/- mice. After in vitro restimulation with anti-CD3 of mLNC of TNBS-treated mice, enrichment of CD44 in the light (raft) fraction was mostly seen with mLNC of CD44v7-deficient mice. Furthermore, as opposed to mLNC from NTG mice, CD44 was not enriched in the light fraction of TG mLNC after cross-linking CD3 or CD44. Cross-linking CD44v4-v7 via anti-rat CD44v6 did not support translocation of CD44 or of the CD44v4-v7 TG into the light raft fraction. The statistical evaluation of densitometric data from three independently performed experiments revealed that in the absence of any stimulation and after in vivo TNBS treatment, significantly more CD44 was recovered in the light fraction of CD44v7-/- than of CD44v7+/+ mice. By anti-CD3 or IM-7 cross-linking, significantly less CD44 was recruited into the light fraction of TG mLNC than of NTG mLNC. Also, significantly less CD44v4-v7 than CD44 was recovered from the light fraction of TG mLNC (Table 5) .
 View larger version (49K): [in a new window] |
Figure 5. CD44v7 is excluded from the immunological synapse. Unstimulated and in vivo (TNBS)-stimulated mLNC from CD44v7+/+ versus CD44v7-/- 129SVEV mice (A) and unstimulated mLNC from CD44v4-v7 TG and NTG BALB/c mice (B) were cross-linked with anti-CD3, anti-CD44, or anti-rat CD44v6 (only mLNC of TG mice). LNC were lysed and subjected to sucrose gradient centrifugation. Light (Triton X-100-insoluble) and dense (Triton X-100-soluble) fractions were collected and pooled, and proteins were separated by gel electrophoresis and blotted with anti-mouse CD44 (IM-7) and/or anti-rat CD44v6 (A2.6).
|
|
View this table: [in a new window] |
Table 5. CD3 and CD44 Cross-Linking Induces Enrichment Only of CD44s in the Light Fraction of the Immunological Synapse
|
 View larger version (35K): [in a new window] |
Figure 6. Co-localization of CD44v7 with moesin and ezrin. (A) mLNC from CD44v7-/- and CD44v7+/+ and (B) CD44v4-v7 TG and NTG mice were stimulated in vitro by anti-CD3, anti-CD44, or anti-CD44v6 (only TG mLNC) cross-linking. Cells were lysed and immunoprecipitated with anti-mouse CD44 (IM-7) and/or anti-rat CD44v6 (A2.6). Precipitated proteins were separated by gel electrophoresis and were blotted with anti-ezrin, anti-moesin, anti-CD44, or anit-CD44v6 and the respective HRP-labeled secondary antibodies.
|
|
View this table: [in a new window] |
Table 6. CD44v7 Preferentially Associates with Ezrin
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
IBD is a TH-1-mediated autoimmune disease [3
, 4
, 14
, 15
], and we have shown before that anti-CD44v7 treatment supports the secretion of IL-10, which leads to down-regulation of IFN- secretion [26
]. These features were a result of a direct activation of APC by cross-linking via the CD44v7-specific antibody [26
]. However, such a functional principle does not provide an explanation for the resistance of CD44v7-/- mice toward induction of colitis [25
]. Furthermore, there is evidence that IL-12 is pivotal only at the starting period of IBD [16
]. The fact that CD44v7-/-/IL-10-/- mice do not develop chronic colitis [25
] also argues against an overshooting TH-1 reaction as the exclusive principle of persisting inflammation. Hence, additional activities of CD44v7 should be important for maintenance of the inflammatory state. It is known that LPL are rather resistant toward apoptosis [23
, 68
], and significantly more apoptotic cells are seen in the inflamed gut of CD44v7-/- as compared with CD44v7+/+ mice, which could be confirmed by annexin V and PI staining of in vitro, restimulated mLNC from TNBS-treated CD44v7-/- and CD44v4-v7 TG mice. The finding pointed toward CD44v7, blocking activation of the apoptotic cascade or sustaining high-level expression of anti-apoptotic molecules.
To differentiate between these possibilities, expression of pro- and anti-apoptotic molecules was evaluated. CD95 expression was unaltered, and CD95L expression was not significantly reduced in lymphocytes of CD44v7-competent as compared with CD44v7-deficient mice. Yet, Bcl-Xl expression was clearly elevated in lymphocytes of CD44v4-v7 TG mice and was further increased after repeated stimulation (data not shown), whereas in control mice, Bcl-Xl and pBAD but not BAD expression was lowered. Bcl-2, Bcl-Xl, and pBAD were hardly detectable in stimulated LNC of CD44v7-/- mice. Differences in expression of anti-apoptotic molecules were most pronounced after antigenic stimulation, i.e., were less striking after stimulation by PHA or cross-linking CD3 (data not shown). The features argue against a dominating influence of CD44v7 on the CD95/CD95L pathway. Instead, CD44v7 obviously supports expression of anti-apoptotic molecules.
The analysis of lymphocyte activation markers corroborated that CD44v7 may well be important for prevention of AICD but is unlikely to play a major role in lymphocyte activation. Neither the constitutive overexpression nor the absence of CD44v7 had any bearing on lymphocyte activation markers such as LFA-1, CD44s, and CD28 and did not trigger CD80 and CD86 expression. Yet, expression of some costimulatory and accessory molecules apparently was not fully independent of CD44v7 expression. First, CD69 and CD25 expression was elevated in unstimulated mLNC of CD44v4-v7 TG mice but hardly increased during activation. The feature could well be connected to the accelerated response of LNC of CD44v4-v7 TG mice toward a nominal antigen [59 ] but is not of relevance in chronic inflammation (R. M., unpublished). Second, CD40 and CD40L expression was elevated in mLNC of CD44v4-v7 TG mice, whereas CD40L expression was low in lymphocytes of CD44v7-/- mice. High CD40/CD40L expression levels have also been described in patients with IBD [18 ]. A CD40/CD40L interaction is supposed to trigger early up-regulation of a not yet defined CD44 isoform, which in turn influences CD40/CD40L expression [62 ]. The CD44 isoform may well be CD44v7, which is rapidly up-regulated by CD40L cross-linking [25 ]. Third, expression of CTLA-4 was low in TG mice but high in CD44v7-/- mice, and triple-fluorescence staining for CD4, CD25, and cytoplasmic CTLA-4 revealed a higher percentage of CD4+CD25+CTLA-4+ mLNC in CD44v7-/- as compared with CD44v7+/+ mice (M. Z., unpublished finding). It has recently been described that CD4+CD25+CTLA+ regulatory T cells are important in the control of IBD [20 , 21 ].
Taken together, expression of CD44v7 appears to be linked to CD40L expression and possibly has a negative impact on regulatory T cells. More essential, the analysis of activation markers provided no evidence for an involvement of CD44v7 in lymphocyte activation.
The minor impact on CD95L expression and the failure to support expression of activation markers such as CD25 and CD69 expression are in clear distinction to CD44s, which functions as an accessory molecule in T cell activation and AICD. Cross-linking CD44s supports phosphorylation of lck, fyn, and additional kinases [39 ], known to become activated during the initiation of a T cell response [69 , 70 ]. Importantly, the accessory function of CD44s depends on its recruitment in the lipid-rich microdomain of the immunological synapse [43 ]. Thus, we speculated that the failure to support T cell activation and the stabilization of anti-apoptotic gene expression may be a result of CD44v7 locating in different membrane compartments and associating with different adaptor molecules.
It has been described that upon TCR engagement, membrane molecules become resorted [41 , 42 , 54 55 56 , 71 , 72 ]. Although the TCR/CD3 complex, CD4/CD8, and additional accessory molecules are recovered from the central immunological synapse, other molecules such as integrins form an inner rim, and e.g., CD43 being excluded from the synapse is recovered in an outer rim [73 , 74 ]. Sorting can be guided by the cytoskeletal-linker proteins ezrin and moesin [54 55 56 ]. Therefore, it became of interest whether CD44v7, too, may become excluded from the synapse by associating with ezrin and/or moesin. Ezrin and moesin were present in IM-7 precipitates of CD44v7-competent and -deficient mice. Yet, a densitometric analysis revealed that only a minor portion of CD44s was associated with ezrin, a fivefold higher amount of ezrin coprecipitated with CD44v4-v7, and more moesin coprecipitated with CD44 in CD44v7+/+ compared with CD44v7-/- cells. It is interesting that although CD44 was the first transmembrane protein reported to associate with the ERM family [48 ], it was also described that only a minor part of CD44 associates with ERM proteins during lymphocyte activation [54 ]. There are at least two possible explanations for the preferential association of ERM proteins with CD44v. First, CD44s, which is rapidly recruited into the immunological synapse, supports the phosphorylation of ERM proteins as a result of its constitutive association with lck and fyn. The activated ERM proteins migrate out of the synapse and reassociate with membrane proteins as well as polymerized actin [75 , 76 ]. As CD44 variant isoform expression becomes up-regulated only during the process of lymphocyte activation, the newly appearing CD44v molecules may more likely associate with the active form of ezrin/moesin than CD44s, which is already engaged in the TCR/CD3 signaling complex in the inner synapse. The strong association of ezrin with constitutively expressed CD44v4-v7 in TG mice strengthens this hypothesis. Second, CD44v isoforms are associated with additional transmembrane proteins, which are excluded from the immunological synapse. A CD44 variant isoform-selective association with other membrane molecules could be demonstrated for CD44v-expressing tumor cells [77 ]. It remains to be explored whether this also accounts for CD44v7 expressed on lymphocytes. Irrespective of the mechanism underlying the preferential association of CD44v with ERM proteins, this association exludes CD44v7 from the immunological synapse and thus provides a convincing explanation for CD44v7 not functioning as an accessory molecule in T cell activation. Could the association with ezrin also support prevention of apoptosis?
Activated ERM proteins are known to associate with signal-transducing molecules such as regulators of Rho-family GTPases and phosphatidylinositol-3 kinase (PI-3K) [78 79 80 ]. Upon association, the latter translocates to the membrane and phosphorylates phosphatidylinositol-4,5-bisphosphate, which leads to activation of AKT [81 ], which is known to transduce antiapoptotic signals [57 , 58 , 80 , 81 ]. It was recently described that ICAM-2 induces tyrosine phosphorylation of ezrin and PI-3K translocation, AKT activation, and phosphorylation of AKT targets, thus preventing B cell death [82 ]. Our findings suggest that CD44v7 may support T cell survival via a similar signaling pathway.
We briefly want to discuss whether the presented data are in line with our hypothesis that CD44v7 sustains chronic IBDs by preventing down-regulation of anti-apoptotic genes and as a consequence, AICD, and whether this can be a physiologically relevant mechanism. With respect to the potential, physiological function of CD44v7, it is important to remember that CD44v7 expression on T cells is mainly induced via CD40L engagement [25 ] or by (repeated) antigenic stimulation but hardly by mitogens or by cross-linking the TCR/CD3 complex. Inflammatory reactions start with the recruitment of elements of nonadaptive-immune defense, followed by the infiltration of CD4+ cells, the first wave of infiltrating T cells mostly being nonantigen-specific. As outlined above, CD44v7 expression on T cells is apparently restricted to TH cells, which become activated by directly contacting APC. Taking this expression profile, it becomes unlikely that CD44v7 functions during the early period of an inflammatory reaction. It is known that the gut-associated immune system disposes on a high level of regulatory elements to avoid persisting activation of lymphocytes [19 20 21 22 ]. We would like to propose that the gut also uses a fail-safe mechanism in the sense that apoptosis protects a minority of antigen-specific TH cells. The expression profile of CD44v7 would be compatible with CD44v7 contributing to the protection of activated TH cells.
Such a physiological function of CD44v7 would also be in line with our observations regarding the impact of CD44v7 on TNBS-induced colitis, i.e., that mice with a targeted deletion of the CD44v7 exon product are resistant toward experimental colitis and that application of a CD44v7-specific antibody prevents the development of chronic colitis. CD44v7-/- mice as well as anti-CD44v7-treated mice develop an unimpaired, initial inflammatory reaction [25
, 26
]. Furthermore, the inflammatory reaction observed during the initial 3 days after TNBS application is not aggravated in CD44v4-v7 TG mice (data not shown). The observations support our finding that CD44v7 does not contribute to T cell activation. As CD44v4-v7 TG mice express CD44v4-v7 constitutively, it also becomes unlikely that the failure to detect an impact of CD44v7/anti-CD44v7 on early inflammatory events is a result of the fact that CD44v7 expression is low in unstimulated lymphocytes. At 5 days after TNBS application, when an anti-CD44v7 treatment or a CD44v7 deficiency becomes effective, a reasonable number of clonally expanded, antigen-specific TH cells will have infiltrated the gut. Finally, it has already been demonstrated that a blockade of CD44v7 [26
] is therapeutically very effective. According to the presented data and our proposed hypothesis, anti-CD44v7 should neither affect primary responsiveness nor be accompanied by an overshooting down-regulation of response.
Taken together, we could demonstrate that CD44v7 is important in preventing AICD by stabilizing expression of antiapoptotic molecules. CD44v7 preferentially associates with ERM proteins, which exclude the molecule from the immunological synapse. The signaling cascade initiated by the CD44v7–ERM complex, which results in pBAD and the sustained expression of BcL-Xl and Bcl-2, remains to be explored.
Received December 19, 2002; revised February 26, 2003; accepted March 19, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
plus IFN gamma producing CD4+ T cells that respond to bacterial antigens J. Immunol. 162,1884-1888This article has been cited by other articles:
 |
|
 |
|
  M. Rajasagi, M. Vitacolonna, B. Benjak, R. Marhaba, and M. Zoller CD44 promotes progenitor homing into the thymus and T cell maturation J. Leukoc. Biol., February 1, 2009; 85(2): 251 - 261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Klingbeil, R. Marhaba, T. Jung, R. Kirmse, T. Ludwig, and M. Zoller CD44 Variant Isoforms Promote Metastasis Formation by a Tumor Cell-Matrix Cross-talk That Supports Adhesion and Apoptosis Resistance Mol. Cancer Res., February 1, 2009; 7(2): 168 - 179. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ori, M. Nardini, P. Casini, R. Perris, and I. Nardi XHas2 activity is required during somitogenesis and precursor cell migration in Xenopus development Development, February 15, 2006; 133(4): 631 - 640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Hauptschein, K. E. Sloan, C. Torella, R. Moezzifard, M. Giel-Moloney, C. Zehetmeier, C. Unger, L. L. Ilag, and D. G. Jay Functional Proteomic Screen Identifies a Modulating Role for CD44 in Death Receptor-Mediated Apoptosis Cancer Res., March 1, 2005; 65(5): 1887 - 1896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Krettek, G. K. Sukhova, U. Schonbeck, and P. Libby Enhanced Expression of CD44 Variants in Human Atheroma and Abdominal Aortic Aneurysm: Possible Role for a Feedback Loop in Endothelial Cells Am. J. Pathol., November 1, 2004; 165(5): 1571 - 1581. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
