Tetraspanin CD53 Promotes Lymphocyte Recirculation by Stabilizing L-Selectin Surface Expression

Summary

Tetraspanins regulate key processes in immune cells; however, the function of the leukocyte-restricted tetraspanin CD53 is unknown. Here we show that CD53 is essential for lymphocyte recirculation. Lymph nodes of Cd53^−/− mice were smaller than those of wild-type mice due to a marked reduction in B cells and a 50% decrease in T cells. This reduced cellularity reflected an inability of Cd53^−/− B and T cells to efficiently home to lymph nodes, due to the near absence of L-selectin from Cd53^−/− B cells and reduced stability of L-selectin on Cd53^−/− T cells. Further analyses, including on human lymphocytes, showed that CD53 stabilizes L-selectin surface expression and may restrain L-selectin shedding via both ADAM17-dependent and ADAM17-independent mechanisms. The disruption in lymphocyte recirculation in Cd53^−/− mice led to impaired immune responses dependent on antigen delivery to lymph nodes. Together these findings demonstrate an essential role for CD53 in lymphocyte trafficking and immunity.

Graphical Abstract

Highlights

• •CD53 is essential for lymph node cellularity as Cd53^−/− lymph nodes lack T and B cells
• •CD53 is essential for lymphocyte homing to lymph nodes
• •CD53 stabilizes L-selectin cell surface expression and may restrain shedding
• •Impaired lymphocyte homing leads to diminished immune responses in Cd53^−/− mice

Molecular Mechanism of Behavior; Immunology; Immune Response; Specialized Functions of Cells

Introduction

To initiate an adaptive immune response, antigen selects and clonally expands rare antigen-specific lymphocytes from a vast repertoire of specificities. The frequency of antigen-specific lymphocytes within the naive repertoire is minute, estimated in mice to be as low as 15–20 cells per animal (Moon et al., 2007, Obar et al., 2008), whereas in humans, frequencies can range from 0.6 to 6 cells per million cells (Alanio et al., 2010, Chu et al., 2009). It is therefore not surprising that the immune system has evolved strategies to maximize the chances of a successful encounter between antigen and rare antigen-specific lymphocytes. To this end, naive lymphocytes constantly patrol throughout the body, recirculating through the peripheral lymph nodes via interactions within specialized high endothelial venules (HEVs). Then, upon dwelling in the lymph nodes for 12–24 h, they return to the vasculature via the efferent lymphatics, enabling them to visit further peripheral lymph nodes (Gowans, 1959, Smith and Ford, 1983, Springer, 1994).

Many of the molecular interactions that govern the egress of naive lymphocytes at HEVs are well understood, and the key molecule responsible for lymph node homing is the membrane receptor L-selectin (Grailer et al., 2009, Springer, 1994). L-selectin mediates carbohydrate-based interactions with its HEV-expressed ligands termed peripheral node addressin, thus initiating the tethering and rolling of naive lymphocytes within HEVs (Streeter et al., 1988) and ultimately leading to stable adhesion and transmigration of naive lymphocytes into the lymph nodes (Ivetic, 2013). Essential to the control of lymphocyte trafficking is the tight regulation of L-selectin expression. Upon activation, L-selectin is rapidly cleaved from the lymphocyte surface predominantly by the metalloprotease ADAM17. This loss of L-selectin is thought to prevent effector cells from homing to peripheral lymph nodes thereby indirectly promoting their trafficking to peripheral sites, although this has not yet been shown experimentally (von Andrian and Mackay, 2000). The molecules that regulate L-selectin shedding are still under investigation and include calmodulin, ERM proteins, and protein kinase C (Ivetic, 2013).

Tetraspanins are four-transmembrane proteins that play a key role in the molecular organization of cellular membranes. There are 34 tetraspanins encoded by the human genome, and these molecules interact with and organize their partner proteins into signal-transducing microdomains (Rubinstein et al., 2013). In immune cells, the tetraspanin molecular partners include integrins, immunoreceptors, and signaling molecules; thus, tetraspanins control key processes such as immune cell adhesion, migration, and activation (Jones et al., 2011, Yeung et al., 2018). The functions of many of the immune cell-expressed tetraspanins are now well documented. For example, in B cells, the tetraspanin CD81 promotes adhesion strengthening of α4β1 integrin (Feigelson et al., 2003) and additionally serves as a key component of the CD19/CD21 co-receptor complex that lowers the threshold of B cell signaling (van Zelm et al., 2010). In dendritic cells, CD81 (Quast et al., 2011) and the tetraspanins CD37 and CD82 (Gartlan et al., 2013, Jones et al., 2016) regulate the cytoskeletal rearrangements that drive migration. CD37 is also required for integrin function, both in neutrophils where it is required for inflammation-associated recruitment (Wee et al., 2015) and in B cells where it is essential for transducing survival signals during the germinal center reaction (van Spriel et al., 2012).

By contrast, the function of the pan-leukocyte marker CD53 remains obscure. Despite being one of the first discovered members of the tetraspanin superfamily (Amiot, 1990), CD53 has not been robustly examined, with a range of in vitro studies weakly suggesting a variety of roles. Cross-linking CD53 at the cell surface can lead to leukocyte activation (Bell et al., 1992, Bosca and Lazo, 1994, Cao et al., 1997, Lagaudriere-Gesbert et al., 1997, Lazo et al., 1997), a result perhaps explained by more recent sophisticated analyses that demonstrate a role for CD53 in PKC signaling (Zuidscherwoude et al., 2017). Transfection and expression studies have suggested a role in the regulation of apoptosis (Kim et al., 2004, Yunta and Lazo, 2003) and T cell development (Puls et al., 2002), whereas genetic and phenotypic analyses suggest a role for mutations in CD53 in immunodeficiency (Mollinedo et al., 1997) or various inflammatory disorders including arthritis, asthma, and Sjögren's syndrome (Bos et al., 2010, Khuder et al., 2015, Lee et al., 2013, Pedersen-Lane et al., 2007, Xu et al., 2015). Here, we analyze CD53 function using a reverse genetics approach. The data show a striking phenotype as CD53-deficient lymphocytes home poorly to lymph nodes, an effect associated with marked reductions in L-selectin expression and stability of these cells. Thus, we demonstrate that CD53 is a key player in the regulation of lymphocyte recirculation.

Results and Discussion

The Cellularity of Cd53^−/− Peripheral Lymph Nodes Is Reduced due to a Striking Defect in Lymphocyte Homing

To investigate the function of the tetraspanin CD53, we first analyzed the lymphoid organs of Cd53^−/− mice. No difference in the gross appearance or mass of spleens was observed; however, Cd53^−/− peripheral lymph nodes (pLN; inguinal and brachial) were smaller in appearance and weighed less than their wild-type (WT) counterparts (Figure 1A). CD53 ablation had no significant effect on the cellularity of the bone marrow, thymus, blood, spleen, and mesenteric lymph nodes. However, pLN from Cd53^−/− mice showed a striking defect in cellularity with a reduction of approximately 60% compared with WT (Figure 1B). This was not a delay in development, as the cellularity in the spleens of WT and Cd53^−/− mice was identical over time, whereas the pLN of Cd53^−/− mice remained smaller even in 52-week-old mice (Figure 1C). To further analyze the impaired cellularity of Cd53^−/− pLN, cells from pLN and spleen were analyzed by flow cytometry and lymphocytes were enumerated. Dot plot analyses showed a slight increase in the frequency of T cells and a striking reduction in the frequency of B cells in the pLN of Cd53^−/− mice (Figure 1D). Quantification of absolute B cell numbers confirmed this reduction (Figure 1F). The number of pLN B cells in Cd53^−/− mice was only ∼14% of that in WT. For pLN T cells, quantification revealed a smaller, ∼50% deficit, applying to both the CD4⁺ and CD8⁺ lineages. In parallel we observed a marginal, although significant, increase in the number of T cells in the spleen (Figure 1E).

Figure 1.

CD53 Ablation Reduces the Cellularity of Peripheral Lymph Nodes

Lymphoid organs from 10-week-old WT and Cd53^−/− mice were harvested and analyzed.

(A) Representative images and mass of WT and Cd53^−/− spleens and peripheral lymph nodes (pLN).

(B) Total cell numbers of lymphoid organs (BM, bone marrow; thymus; blood; spleen; mLN, mesenteric lymph nodes; and pLN).

(C) Total cell number of spleens and pLN as determined at time points from 4 to 52 weeks.

(D) Flow cytometry analyses of spleen and pLN identifying B (B220⁺) and T (CD3⁺) cells and CD4 ⁺ and CD8⁺ T cell populations.

(E and F) Quantification of (E) T cell and (F) B cell populations in spleen and pLN.

Data are represented as mean ± SEM, n = 6–17 mice per group pooled from 2–5 independent experiments, ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001, Student's two-tailed unpaired t test.

Given the normal cellularity of the spleen and other lymphoid organs, and the documented roles of tetraspanins in regulating cell migration and leukocyte trafficking (Yeung et al., 2018), we reasoned that a defect in lymphocyte recirculation may underlie the phenotype of poor lymph node cellularity. We therefore evaluated whether CD53 ablation affected lymphocyte homing to lymph nodes. First, to investigate whether CD53 ablation had an effect on lymph node architecture or on HEVs, homing assays were performed where WT splenocytes were labeled with carboxyfluorescein succinimidyl ester (CFSE) and adoptively transferred into WT and Cd53^−/− mice (Figure 2A). Flow cytometric analysis was used to identify the relative frequency of adoptively transferred splenocytes in each organ (Figure 2B), and to distinguish B and T cells (data not shown). The data clearly show that whereas there was an impairment in the ability of adoptively transferred WT B and T cells to home to the spleens of Cd53^−/− mice, homing to peripheral and mesenteric lymph nodes was normal (Figures 2C and 2D). Thus, the reduced cellularity of Cd53^−/− pLN is likely a phenotype intrinsic to Cd53^−/− lymphocytes. Next, we investigated the homing capacity of Cd53^−/− lymphocytes. WT and Cd53^−/− B or T cells were purified, labeled with CFSE, and co-injected into WT recipient mice together with an internal control of WT (B or T) cells differentially labeled with Cell Tracker 670 (Figure 2E). After 48 h immune organs were harvested and homing analyzed by flow cytometry (Figures 2F and 2H). Quantification was calculated using a homing index (Arbones et al., 1994) where the ratio of CFSE⁺ test cells to Cell Tracker 670⁺ internal control cells in the initial injection mixture and also within each organ was determined (Figures 2G and 2I). For B cells, Cd53 deficiency resulted in a striking defect (Figure 2F). Although Cd53^−/− B cells were able to localize to blood normally, they preferentially localized to the spleen, whereas homing to the pLN was almost completely abrogated (Figure 2G). There was also a significant reduction of B cell homing to the mesenteric lymph nodes (25% of WT; Figure 2G). A similar dysregulation in Cd53^−/− T cell lymph node homing was observed, although the phenotype was not as pronounced (Figure 2H). Cd53^−/− T cells also preferentially localized to the spleen, whereas homing to both the pLN and mLN was poor (Figure 2I). In conclusion, CD53 ablation induces reduced cellularity in pLN due to a severe defect in lymphocyte homing, with the phenotype more pronounced in B cells than in T cells.

Figure 2.

CD53 Is Required for Efficient Lymphocyte Homing to Lymph Nodes

(A–D) WT splenocytes were labeled with CFSE and injected intravenously (i.v.) into WT or Cd53^−/− recipient mice. Forty-eight hours later lymphoid organs were harvested and analyzed for CFSE⁺ cells using flow cytometry. (A) Schematic, illustrating experimental design. (B) Representative flow cytometry illustrating the identification of transferred CFSE⁺ cells in the spleens and pLN of recipient mice. (C and D) Enumeration of CD19⁺CFSE⁺ adoptively transferred B cells (C) and CD3⁺CFSE⁺ adoptively transferred T cells (D) in the lymphoid organs of recipient mice. Data are from 8 mice per group pooled from 2 independent experiments.

(E–I) WT and Cd53^−/− B or T cells were purified, CFSE-labeled, and co-injected i.v. into WT recipient mice together with an internal control of WT (B or T) cells differentially labeled with Cell Tracker 670. (E) Schematic, illustrating experimental design. (F and H) Representative flow cytometry comparing the homing of WT-Cell tracker 670⁺, WT-CFSE⁺, and Cd53^−/−CFSE⁺ CD19⁺ B cells (F) and CD3⁺ T cells (H) to peripheral lymph nodes. (G and I) Homing indices of transferred B cells (G) and T cells (I) to each organ.

Data are represented as mean + SEM, 8 mice per group pooled from 2 independent experiments, ∗p ≤ 0.05, ∗∗∗p ≤ 0.001, Student's two-tailed unpaired t test (C and D) or Mann-Whitney test (G and I).

CD53 Ablation Impairs Expression of L-Selectin on Lymphocytes

Given that CD53 ablation induces a defect in lymph node homing, we hypothesized that CD53 might regulate the key lymph node homing receptor, L-selectin. We therefore analyzed L-selectin expression by flow cytometry and identified that the residual B cell (B220⁺) population found in Cd53^−/− pLN had significantly reduced expression of L-selectin protein (Figure 3A). Indeed, across spleen, pLN, and mLN there was a major reduction in the surface expression of L-selectin on Cd53^−/− B cells of up to two orders of magnitude (Figure 3B). To investigate whether our findings extended to human cells, we generated human Cd53^−/− B cell lines (BJAB) using CRISPR/Cas9 technology. We observed impaired cell surface L-selectin expression in two independent human Cd53^−/− B cell lines, in accord with our mouse data (Figure 3C). In comparison, the expression of L-selectin on Cd53^−/− T cells revealed a small, but significant, reduction of L-selectin expression of 8% and 20% in splenic and pLN T cells, respectively (Figures 3D and 3E). This diminution of L-selectin expression was exaggerated in T cells from aged mice where there was an approximately 60% reduction of L-selectin expression (Figure 3F).

Figure 3.

CD53 Stabilizes L-Selectin Expression on the Lymphocyte Surface

(A and B) (A) Representative flow cytometry histograms (splenic B cells), and quantification of L-selectin protein expression (B; mean fluorescent intensity, MFI) on B220⁺ WT and Cd53^−/− B cells harvested from spleen, peripheral lymph nodes (pLN), and mesenteric lymph nodes (mLN). Data are represented as mean + SEM, 3 mice per group, representative of ≥3 independent experiments.

(C) Expression of WT human L-selectin on WT and hCd53^−/− human BJAB cells.

(D and E) (D) Representative flow cytometry (splenic T cells) and quantification (E, MFI) of L-selectin expression on CD3⁺ WT and Cd53^−/− T cells harvested from spleen and pLNs from 10-week-old mice. Data are represented as mean + SEM, 10 mice per group pooled from 2 independent experiments.

(F) Quantification of L-selectin expression (MFI) on CD3⁺ T cells harvested from aged (52-week-old) WT and Cd53^−/− mice. Data are represented as mean + SEM, n = 6 mice per group representative of 2 independent experiments.

(G–J) WT and Cd53^−/− splenocytes were cultured at 37°C with or without the broad metalloprotease inhibitor GM6001 (GM; 100μM). Flow cytometry was used to distinguish B (B220⁺) and T (CD3⁺) cells and monitor L-selectin expression. (G and I) Representative flow cytometry histograms displaying L-selectin expression on B cells (G) and T cells (I) after 2 h at 37°C. (H and J) The kinetics of L-selectin expression on cultured B (H) and T (J) cells relative to cells held at 4°C.

(K–N) Purified WT and Cd53^−/− T cells were stimulated with PMA (50 ng/mL) to induce shedding in the presence or absence of metalloprotease inhibitors. (K) Representative histograms illustrating L-selectin expression in unstimulated (unstim), PMA stimulated (5′), and PMA stimulated in the presence of the metalloprotease inhibitor GM6001 (PMA 5′ + GM). (L) The kinetics of L-selectin surface expression (relative to unstimulated cells) and (M) amount of soluble L-selectin shed into supernatants of WT and Cd53^−/− T cells stimulated with PMA. (N) The percentage of L-selectin expression (relative to unstimulated cells) remaining after 15′ PMA stimulation in the presence or absence of the metalloprotease inhibitors GM6001 (100 μM, GM), GI 254023X (20 μM, GI ), TAPI-1 (200 μM), and A9 (5 μM).

Data are represented as mean ± SEM, n = 6–7 mice per group pooled from 3 independent experiments, ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001, Student's two-tailed unpaired t test (B, C, E, F, M, N) or Mann-Whitney test (H, J, L). See also Figures S1–S3.

To investigate the mechanism by which CD53 regulates L-selectin we first analyzed mRNA expression in lymphocytes and could detect no reduction in L-selectin mRNA expression associated with CD53 ablation in both resting and activated cells (Figure S1). L-selectin is cleaved from the surface of lymphocytes by metalloproteases. In activated lymphocytes, the predominant metalloprotease that drives shedding is ADAM17, whereas the metalloprotease responsible for the constitutive shedding observed in unactivated lymphocytes has not been identified (Mohammed et al., 2019). There is growing evidence that tetraspanins can regulate this enzyme family. For example, the TspanC8 subgroup of tetraspanins molecularly interacts with and regulates the activity of ADAM10 (Matthews et al., 2017). We therefore investigated whether CD53 regulates L-selectin expression via effects on metalloproteases. We first examined L-selectin expression on unstimulated WT and Cd53^−/− B cells incubated at 37°C in the presence of the broad-spectrum matrix metalloprotease inhibitor, GM6001. Incubation of WT B cells at 37°C induced a spontaneous reduction in L-selectin expression, which was completely inhibited by GM6001 (Figure 3H). L-selectin expression on Cd53^−/− B cells was much lower than on WT cells, and although incubation of Cd53^−/− B cells with GM6001 (Figures 3G and 3H) increased L-selectin expression more than 3-fold, it could not restore expression to WT levels. This suggests that the reduced L-selectin expression on Cd53^−/− B cells may be only partially accounted for by metalloproteinase activity. Incubation of WT T cells at 37°C also induced a spontaneous reduction in L-selectin expression, and this reduction was significantly exaggerated in Cd53^−/− T cells (Figures 3I and 3J). In marked contrast to B cells, the reduced levels of L-selectin on CD53^−/− T cells were substantially restored up to ∼80% of WT levels by GM6001. This finding suggests that the lower levels of L-selectin in CD53^−/− T cells are largely due to spontaneous shedding. To study the impact of CD53 on ADAM-17-dependent L-selectin shedding in T cells, we used phorbol 12-myrsitate 13-acetate (PMA) activation. We observed that the reduction of L-selectin surface expression from Cd53^−/− T cells was significantly accelerated (Figures 3K and 3L), with an increase in shed soluble L-selectin levels observed in the supernatant of Cd53^−/− when compared with WT cells (Figure 3M). GM6001 and TAPI-1 (another small molecule metalloprotease inhibitor that can inhibit ADAM17) were able to largely prevent the loss of L-selectin surface expression in WT T cells and maintain L-selectin at ∼70% of unstimulated levels, whereas these inhibitors were all less effective in PMA-stimulated Cd53^−/− T cells and could only maintain L-selectin expression at ∼25% of unstimulated levels. Similarly, the ADAM-17 blocking monoclonal antibody A9 (Kwok et al., 2014) was also less effective in inhibiting PMA-induced reduction of L-selectin surface expression on Cd53^−/− T cells compared with WT T cells. The ADAM-10-specific inhibitor GI254023X had no effect in these assays (Figure 3N), demonstrating that ADAM10 did not substitute for ADAM-17 in PMA-activated T cells as shown here and by others (Mohammed et al., 2019).

In summary, treatment with metalloprotease inhibitors: increases the expression of L-selectin on Cd53^−/− B cells (Figures 3G and 3H), prevents the instability of L-selectin expression on Cd53^−/− T cells (Figures 3I and 3J), and partially restores L-selectin expression in PMA-treated Cd53^−/− T cells (Figures 3K and 3N). Taken with the exaggerated accumulation of L-selectin in the supernatant of PMA-activated Cd53^−/− T cells (at early time points, Figure 3M), the data support a role for CD53 in inhibiting metalloprotease-mediated shedding of L-selectin from the lymphocyte surface. However, analyses of ADAM17-deficient cells and mice indicate that there are additional mechanisms of regulating L-selectin expression and shedding that are independent of ADAM17 (Le Gall et al., 2009, Walcheck et al., 2003, Wang et al., 2010). Given that metalloprotease inhibitors, or a blocking antibody targeting ADAM17, diminished but did not abrogate the excessive loss of L-selectin induced by PMA stimulation on the surface of Cd53^−/− T cells, our data indicate that CD53 also inhibits ADAM17-independent mechanisms, some of which may be metalloprotease-independent. One possibility is that CD53 might have a role in L-selectin trafficking. We therefore distinguished intracellular and extracellular L-selectin expression in resting and PMA-activated permeabilized lymphocytes, but could detect no evidence of an intracellular accumulation of L-selectin in Cd53^−/− cells, arguing against a role for CD53 in either trafficking L-selectin to the cell surface or internalizing L-selectin after activation (Figure S2). Taken together, we conclude that CD53 is required to stabilize L-selectin expression at the lymphocyte surface, and further studies will be required to identify the additional metalloprotease(s) or mechanisms responsible.

Cd53^−/− B cells express very low levels of L-selectin, and consequently are unable to home to peripheral lymph nodes. By contrast, L-selectin expression in Cd53^−/− T cells is only marginally reduced (Figure 3) but is unstable, as shown by the relatively poor expression of L-selectin on T cells from aged Cd53^−/− mice (Figure 3F), and the spontaneous loss of L-selectin surface expression induced by incubation of Cd53^−/− T cells at 37°C (Figures 3I and 3J). This L-selectin instability likely explains the impairment of Cd53^−/− T cell homing to peripheral lymph nodes, although we cannot rule out other mechanisms such as a defect in signal transduction. In line with this, we have recently shown that CD53 can directly interact with PKC and is required for PKC activation in B cells (Zuidscherwoude et al., 2017), suggesting that CD53 interacts with PKC and L-selectin in the same molecular complex (Kilian et al., 2004). The difference in the biology of CD53 ablation on B and T cells likely reflects the ratio of CD53:L-selectin expression on the two cell types. CD53 is expressed at higher levels on B cells than on T cells (Tomlinson et al., 1995), whereas T cells express approximately 50% more L-selectin than do B cells (Gauguet et al., 2004). Thus, on B cells there is relatively more CD53 available to stabilize lower levels of L-selectin (and vice versa). Alternatively, there may be another tetraspanin that compensates for CD53 on T cells, but is not expressed in B cells. One candidate is CD9, as in vitro studies have shown that it can inhibit ADAM17-mediated shedding of another substrate, tumor necrosis factor (Gutierrez-Lopez et al., 2011), and it is strongly expressed on T cells but not on most B cells. However, we think this is unlikely as lymph node cellularity in Cd9^−/− mice is normal (Iwasaki et al., 2013) and CD53 ablation does not significantly affect CD9 expression in either Cd53^−/− T cells or hCd53^−/− BJAB cells (Figure S3).

Defective Lymphocyte Homing Delays Adaptive Immune Responses when Antigen Is Delivered to Lymph Nodes

To examine whether the impairment of lymphocyte recirculation in Cd53^−/− mice had an effect on immunity, mice were first challenged with classical model antigens (namely the type I T-independent antigen NP-LPS, the type II T-independent antigen NP-FICOLL, and the classical T-dependent antigen NP-KLH) to induce humoral responses. Antibody responses of Cd53^−/− mice were largely normal to all antigens administered systemically (intraperitoneally, Figures 4A, 4C, and 4E). Similarly, responses to intradermal NP-LPS were also normal (Figure 4D). Collectively, these data indicate that CD53 ablation does not significantly affect the germinal center reaction, antigen presentation, T cell help, or the ability of B cells to proliferate and generate plasma cells. However, when NP-KLH was delivered to lymph nodes via intradermal injection, Cd53^−/− antibody responses were significantly delayed (Figure 4F). Similarly, with regard to T cell responses, the clinical responses of Cd53^−/− mice in experimental autoimmune encephalomyelitis, a T cell-dependent model of autoimmunity initiated by subcutaneous immunization with MOG peptide, were also delayed (Figure 4G). We conclude that when antigen is delivered to the lymph nodes of Cd53^−/− mice adaptive immune responses are delayed. This pattern of normal responses to systemically administered antigen versus delayed immunity when antigen is delivered to pLN via the subcutaneous and intradermal routes is strikingly similar to that reported for L-selectin knockout mice (Catalina et al., 1996, Steeber et al., 1996). This illustrates the importance of lymphocyte recirculation in immunity; presumably, the poor cellularity of lymphocytes, particularly B cells, in the lymph node delays immune responses. The one exception here was the exaggerated immune responses of Cd53^−/− mice to NP-FICOLL when administered intradermally (Figure 4B). It is difficult to explain this result, although this was also observed in L-selectin-deficient mice (Steeber et al., 1996). T-independent type II antigen responses arise in the marginal zone of the spleen where marginal zone B cells, B1 cells, and splenic macrophages are all thought to play a critical role (Vinuesa and Chang, 2013). The numbers of these cell populations are normal in Cd53^−/− mice, and we can only speculate either that a deficiency in the CD53/L-selectin axis induces subtle changes in trafficking or localization in these populations or that L-selectin expressed in these populations has a hitherto undiscovered function in negatively regulating type II T-independent responses.

Figure 4.

Adaptive Immune Responses in Cd53^−/− Mice Are Delayed when Antigen Is Delivered to Lymph Nodes

(A–F) WT and Cd53^−/− mice were immunized intraperitoneally (IP; A, C, and E) or intradermally (ID; B, D, and F) with (A and B) 50 μg NP-FICOLL, (C and D) 50 μg NP-LPS, or (E and F) 50 μg NP-KLH. NP-specific antibody responses were determined by ELISA for IgM, IgG1, IgG2b, and IgG3. Data are represented as mean ± SEM, 9–17 mice per group pooled from 2 independent experiments.

(G) Autoimmune encephalomyelitis was induced in WT and Cd53^−/− mice where mice were immunized subcutaneously with MOG peptide. Disease progression was measured daily and expressed as a clinical score as mean ± SEM, 10 mice per group from 2 independent experiments. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p ≤ 0.0001, Student's two-tailed unpaired t test (A–F) or two-way ANOVA (G).

In summary, these analyses identify an indispensable role for CD53 in maintaining lymph node cellularity (Figure 1). A very recent analysis of Cd53^−/− mice attributed the poor lymph node cellularity to defects in B cell development and showed evidence that CD53 is required for optimal interleukin-7R signaling (Greenberg et al., 2020). Here we identify another mechanism: CD53 is required for lymphocyte homing to lymph nodes (Figure 2). The central role of L-selectin in lymph node homing is well established. However, the molecular mechanisms underlying L-selectin stabilization at the plasma membrane are less well understood. This article demonstrates that L-selectin expression on lymphocytes is tightly controlled by the leukocyte-specific tetraspanin CD53 and that CD53, at least partially, mediates this effect by inhibiting L-selectin shedding (Figure 3). Consequently, CD53 is required for efficient and timely immune responses (Figure 4).

Limitations of This Study

This article analyzes the development and immune responses of Cd53^−/− mice and the immunobiology of Cd53^−/− lymphocytes in in vivo and in vitro assays. The molecular interactions of CD53 and the mechanisms by which it stabilizes L-selectin surface expression await further analyses.

Resource Availability

Lead Contact

Further information and requests should be directed to and will be fulfilled by the Lead Contact, Mark Wright (mark.wright@monash.edu).

Materials Availability

All materials are available from the lead contact upon reasonable request, but we may require a materials transfer agreement.

Data and Code Availability

The data that support the findings of this paper are available from the lead contact upon reasonable request.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Published: May 22, 2020