The tetraspanin CD53 promotes CXCR4 signaling and bone marrow homing in B cells

Abstract

B cell trafficking involves the coordinated activity of multiple adhesive and cytokine-receptor interactions, and the players in this process are not fully understood. Herein, we identified the tetraspanin CD53 as a critical regulator of both normal and malignant B cell trafficking. CXCL12 is a key chemokine in B cell homing to the bone marrow and secondary lymphoid organs, and both normal and malignant B cells from Cd53^−/− mice have reduced migration toward CXCL12 in vitro, as well as impaired marrow homing in vivo. Using proximity ligation studies, we identified the CXCL12 receptor, CXCR4, as novel CD53 binding partner. This interaction promotes receptor function, as Cd53^−/− B cells display reduced signaling and internalization of CXCR4 in response to CXCL12. Together, our data suggest that CD53 interacts with CXCR4 on both normal and malignant B cells to promote CXCL12 signaling, receptor internalization, and marrow homing.

Introduction

Trafficking of B cells involves the coordinated activity of adhesive and cytokine-receptor interactions that allow for their movement throughout the bone marrow and to secondary lymphoid organs. This movement is crucial for B cells to find the appropriate niches to facilitate maturation and optimize their function.(1) Furthermore, malignant B cells utilize similar interactions to normal B cells to gain access to the bone marrow where they receive proliferation and survival signals.(2) Thus, defining the mechanisms that regulate B cell trafficking not only provides essential understanding for normal B cell function, but may also identify novel therapeutic opportunities for disrupting the growth of malignant B cells. Herein, we identify the tetraspanin CD53 as a critical regulator of both normal and malignant B cell trafficking.

CD53 is a member of the tetraspanin family of transmembrane proteins that organize protein interaction networks on the cell surface to regulate cellular processes including proliferation, survival, and migration. CD53 expression is largely confined to the hematopoietic system.(3, 4) A report of CD53 deficiency in humans was associated with recurrent infections and low immunoglobulin levels,(5) and CD53 has subsequently been shown to be important for normal B, T, NK and neutrophil cell development, migration and/or function.(6–13)

Regarding B cell trafficking, prior studies have shown that CD53 regulates lymphocyte homing to lymph nodes via stabilization of L-selectin.(10, 14) We now show that CD53 regulates B cell trafficking via promotion of CXCL12/CXCR4 signaling. The chemokine CXCL12 is a key player in B cell trafficking, and we find that B cells from Cd53^−/− mice have significantly impaired migration toward CXCL12. Further, both normal and malignant B cells have reduced marrow homing in the absence of CD53, as well as reduced signaling and internalization of the CXCL12 receptor, CXCR4, in response to CXCL12. Finally, proximity ligation studies showed a physical interaction between CD53 and CXCR4. Together, our data support a model whereby CD53 interacts with CXCR4 on both normal and malignant B cells to promote CXCL12 signaling, receptor internalization, and marrow homing.

Materials and Methods

Mice

Cd53^−/− mice were generated as previously described.(6) WT littermates were used as controls, and young adult sex-matched (8–10 weeks) mice were used unless otherwise indicated. CD19- Cre (B6.129P2(C)- Cd19^tm1(cre)Cgn/J) and Cxcr4 Flox (B6.129P2-Cxcr4^tm2Yzo/J) mice are from Jackson Laboratory. CD45.1/CD45.2 mice were generated by crossing C57Bl/6J (CD45.2) and B6.SJL-Ptprc^a Pepc^b/BoyJ (CD45.1, Jackson Laboratory) mice. The Eμ-Myc (B6.Cg-Tg(IghMyc)22Bri/J, Jackson Laboratory) mice were crossed to Cd53^−/− mice to produce Eμ-Myc⁺; Cd53^−/− mice. All animal studies were performed in accordance with the guidelines of the Washington University Animal Studies Committee.

Flow cytometry

Blood, bone marrow and spleen cells were processed as previously described (15) and stained using the antibodies listed in Supplemental Table 1. Cell counts were determined using an Element HT5 (Heska Corporation). For phospho-flow, cells were fixed and permeabilized using BD Phosflow FixI/PermIII (BD Biosciences) prior to staining. Stained cells were analyzed on a Cytek Aurora (Cytek Bioscience), or NovoCyte flow cytometer (Agilent Technologies). Data was analyzed with FlowJo software (version 10.5.3; TreeStar).

Eμ-myc chimera generation and analysis

Chimeric mice were generated by transplanting a 1:1 mixture (2 ×10⁶ cells total) of whole bone marrow from (pre-malignant) 4-week-old, Eμ-Myc⁺ (CD45.1/CD45.2) and Eμ-Myc⁺; Cd53^−/− (CD45.2) mice into WT (CD45.1) mice conditioned with 2 doses of 550 cGy from a cesium-137 irradiator. Recipient mice were monitored until moribund, then bone marrow, spleen, blood and lymph node cells were spun onto slides and stained using Protocol Hema 3 Wright-Giemsa stain (Fisher Scientific). All slides were reviewed by a hematopathologist (E. Duncavage). Cells were also analyzed by flow cytometry to assess the contribution of Eμ-Myc⁺ (CD45.1/CD45.2) and Eμ-Myc⁺; Cd53^−/− (CD45.2) populations to the tumor origin.

Transwell migration assay

Splenic B cells were isolated using the pan B cell isolation kit (Miltenyi Biotec) following manufacturer’s instructions. For isolation of IgM+ cells, pan-selected cells were further selected using anti-IgM microbeads (Miltenyi Biotec). B cells were placed in the upper compartment of a transwell chamber (Corning Inc.) with media containing 100 ng/mL CXCL12, 100 ng/mL CXCL13, 200 ng/mL CCL19 or 200 ng/mL CCL21 (PeproTech) in the well below. After 2 hours at 37°C, migrated cells were counted by flow cytometry.

In vivo homing assay

Splenic B cells were harvested from WT or Cd53^−/− mice and labeled with CellTracker Green CMFDA dye (ThermoFisher). Two million cells were transferred via tail vein injection into WT mice. After 2 or 18 hours, blood, bone marrow, spleen, lymph node, lung and liver cells were analyzed by flow cytometry for CMFDA fluorescent dye+ cells.

Duolink proximity ligation assay

The Duolink Proximity ligation assay (Duolink In Situ Detection Reagents Green, Millipore-Sigma) was performed following the manufacturer’s instructions. Primary antibodies were conjugated with Probemaker PLUS and MINUS kits (Millipore-Sigma). Fixed B cells on slides were incubated with PLUS-conjugated antibody (45 mins at RT), followed by MINUS-conjugated antibody (45 mins at RT). Washing, ligation, amplification, and detection steps were followed according to manufacturer’s instructions. Slides were mounted with VectaShield with DAPI (Vector Laboratories). Images were acquired using a Leica SP8X with a 63x Apochromat oil objective (N.A. 1.4) and LASXsoftware (v3.6.0; Leica Microsystems).

CXCR4 internalization assay

Splenic B cells were stimulated with 100 ng/mL CXCL12 (PeproTech) for 5,15, 30 and 60 min., then washed, fixed, and stained with anti-mouse CXCR4 antibody (eBioscience).

Statistical analysis

Significance was assessed using an unpaired, two-tailed Student’s t test or two-way analysis of variance (ANOVA) with Bonferroni multiple comparison test, unless otherwise stated. GraphPad Prism (Version 9.0.0) was used for all statistical analyses (GraphPad Software).

Results and Discussion

Loss of CD53 impairs B cell migration and homing

We previously showed that CD53 deficient pre-B cells have impaired migration toward CXCL12 in a transwell assay.(16) To determine if this impaired migration is specific to CXCL12 or reflective of a more generalized motility defect, we assessed the transwell migration of isolated splenic B cells from Cd53^−/− mice or their WT littermates toward several chemokines important for B cell trafficking, including CXCL12, CXCL13, CCL19, and CCL21. We found that loss of CD53 significantly impairs B cell migration towards CXCL12, but not toward the other chemokines (Fig. 1A). We previously reported that CD53 promotes normal B cell development, with Cd53^−/− mice having reduced bone marrow pre-B and immature B cells, as well as fewer transitional 1 and follicular splenic B cells than WT mice.(6) Thus, to circumvent potential migration differences related to population differences, we repeated the CXCL12-directed assay using a more select population of IgM+ splenic cells. Again, Cd53^−/− cells had reduced migration compared to WT (Fig. 1B). In addition, we stained for B cell subsets and found the relative frequencies of migrated splenic B cell populations were similar to the input frequencies for both groups, suggesting that migratory defects are not specific to a single population and are not a result of altered population frequencies (Supplemental Fig. 1A). We then assessed in vivo homing of isolated B cells from the spleens of Cd53^−/− and WT mice transferred into WT recipients. WT and Cd53^−/− B cells were found at similar frequencies in the marrow and spleen at 2 hrs post-transfer (as well as the blood, lymph nodes, liver, and lungs, Fig. 1C and Supplemental Fig. 1B), and the greatest numbers of homed cells were detected in the spleen. However, there was a significant reduction in transferred Cd53^−/− B cells compared to WT in the bone marrow and spleen at 18 hours (Fig. 1C). Importantly, while the frequency of WT bone marrow B cells increased between 2 to 18 hours, the frequency of Cd53^−/− B cells declined, suggesting that CD53 is important for B cell retention in the marrow. (Fig. 1C). While we did not observe differences in DAPI+ (dying) cells between WT and Cd53^−/− transferred B cells (Supplemental Fig. 1B), further assessments are needed to formally exclude a contribution of impaired cell survival to the reduction in Cd53^−/− marrow B cells. Finally, we observed reduced Cd53^−/− B cells in the lymph nodes (Supplemental Fig. 1B), consistent with a prior report.(10)

Fig 1. Loss of Cd53 impairs B cell migration and homing.

(A) Splenic B cells from Cd53^−/− mice or WT littermates were assessed for migration toward the indicated chemokines in a transwell over 2 hours. Shown are the percentage of the input cells that migrated. (B) IgM enriched splenic B cells from WT and Cd53^−/− mice were tested for transwell migration in the presence and absence of CXCL12. (C) Isolated splenic B cells were transferred into the blood of WT recipient mice, and donor cells were enumerated from the bone marrow, spleen, and blood, after 2 hrs or 18 hrs (see Supplemental Fig. 1B for lymph nodes, lung, and liver). Data represent the mean ± SEM from two to three independent experiments. *p<0.05, ****p < 0.0001 by two-way ANOVA.

Loss of CD53 reduces the marrow tropism of malignant B cells

Next, we asked whether malignant B cell homing is similarly reduced by the loss of CD53, as CD53 is highly expressed in many B cell malignancies,(17) and malignant B cells utilize similar interactions (including CXCL12/CXCR4) as normal B cells to gain access to the bone marrow.(2, 18) To this end, we crossed the Cd53^−/− mice to the Eμ-myc transgenic strain. These mice express the Myc oncogene under the control of the B-cell specific IgH promoter, and develop heterogeneous B cell leukemias and lymphomas by about 12 weeks of life.(19) Of note, like many malignant human B cells, CD53 surface expression is upregulated on Eμ-myc CD19+ cells compared to WT CD19+ cells, particularly cells residing in the bone marrow (Fig. 2A and Supplemental Fig. 1D). While loss of CD53 did not affect survival of the Eμ-myc mice (Supplemental Fig. 1C), we noted a striking difference in the tropism of the malignant cells. Specifically, while Eμ-Myc;Cd53^−/− mice develop lymphomas, they rarely have significant marrow involvement (Fig. 2B; significant marrow involvement defined as >20% blasts). In contrast, the marrow was almost ubiquitously involved in the Eμ-myc mice. Furthermore, in a chimeric model where Eμ-Myc and Eμ-Myc;Cd53^−/− cells were transplanted together into WT recipients (Fig. 2C–E), we found that the tumors, when originating from the Eμ-myc;Cd53^−/− donor population, nearly always spared the bone marrow, but when they originated from the Eμ-Myc population with normal CD53 expression, they almost always involved the marrow. Finally, isolated splenic B cells from Eμ-Myc;Cd53^−/− mice had impaired migration towards CXCL12 in a transwell assay, demonstrating the importance of CD53 for malignant B cell migration (Fig. 2F).

Fig 2. Loss of Cd53 reduces the marrow tropism of malignant B cells.

(A) Representative flow plot of CD53 expression on bone marrow CD19+ cells of an Eμ-myc mouse compared to a WT littermate. FMO= full minus one (no CD53 antibody). (B) Eμ-Myc;Cd53^−/− mice were generated by breeding Cd53^−/− mice with the Eμ-Myc strain. Shown are the frequency of lymph node and/or spleen and bone marrow involvement of malignant cells from the indicated groups (**p<0.01 by Fisher’s exact test; n=9–10 mice per group) (C) Chimeras were generated by transplanting a mixture of Eμ-Myc (CD45.1/CD45.2) and Eμ-Myc;Cd53^−/− (CD45.2) bone marrow cells into WT (CD45.1) mice. Recipient mice were followed until moribund and then bone marrow, blood, spleen and lymph nodes were harvested for analysis of tumor involvement, which could originate primarily (>50% of the tumor cells as determined by the presence of CD45.1 and CD45.2) from either the Eμ-myc or Eμ-Myc; Cd53^−/− donor cells. (D) Representative Wright-Giemsa stains from the indicated tissues from recipient mice with tumors that originated from Eμ-myc vs Eμ-Myc; Cd53^−/− donor cells. Quantification of bone marrow involvement is shown in (E). *p<0.05 by Fisher’s exact test. N= 14 total chimeras analyzed, with 9 having primarily Eμ-Myc;Cd53^−/− involved tumors and 5 having Eμ-myc involved tumors. (F) Isolated splenic Eμ-Myc and Eμ-Myc;Cd53^−/− B cells were tested for their ability to migrate in a transwell toward CXCL12; shown are the percentage of the input cells that migrated (**p<0.01 by two-way ANOVA). Data represent the mean ± SEM from two to three independent experiments.

CD53 is necessary for normal CXCL12/CXCR4 signaling and receptor internalization

We next asked whether the loss of CXCL12-directed migration of CD53-deficient B cells is due to reduced receptor expression or function. Surface CXCR4 expression was assessed on splenic B cell populations (see reference #6 for gating strategy)(6) using flow cytometry, and showed that levels were actually increased on transitional type 2- marginal zone progenitors (T2-MZP) and marginal zone (MZ) B cells in the Cd53^−/− mice compared to controls (Fig. 3A–B). Additionally, we assessed expression of CXCR7, CCR7, and CXCR5 to evaluate whether CD53 loss influenced other chemokine receptors related to B cell migration. We did not observe any alterations in surface expression of CXCR7, CCR7, and CXCR5 in any populations of splenic B cells (Supplemental Fig. 2A–C).

Fig 3. CD53 is necessary for normal CXCL12 signaling.

(A) CXCR4 surface expression as determined by flow cytometry on splenic B cell populations from Cd53^−/− and WT mice, including transitional 1 (T1), follicular (FO), transitional 2- marginal zone precursor (T2-MZP), and marginal zone (MZ) B cells. (B) Representative flow plots of CXCR4 surface expression on T2-MZP and MZ B cells. (C-D) Phosphorylation of AKT (C) and ERK (D) was assessed on splenic B220+ cells from WT and Cd53^−/− mice following ex-vivo stimulation with CXCL12 for 1 hour (see Supplemental Fig. 2 for subpopulation analyses). Shown are the median fluorescent intensity (MFI) of expression by flow cytometry. (E) Isolated splenic B cell from the indicated populations were assessed for migration toward CXCL12 in a transwell assay. Shown are the percentage of the input cells that migrated. Data represent the mean ± SEM from 2–3 independent experiments. **p <0.01, ***p <0.001, ****p<0.0001 as determined by two-way ANOVA.

Ligand binding to CXCR4 leads to the activation of multiple signaling pathways, including Jak/Stat, PI3K-Akt and Mek1/2-Erk1/2.(20) To assess CXCR4 signaling, we stimulated splenic B cells from WT and Cd53^−/− mice ex-vivo with CXCL12 for 1 hour and assessed for phosphorylation of AKT and ERK via flow cytometry. As shown in Figure 3C–D and Supplemental Fig. 2D, both pERK and pAKT were increased with CXCL12 in the WT, but not Cd53^−/− B cells across multiple subpopulations.

In addition to CXCR4, CXCL12 also binds to CXCR7.(21) Therefore, to confirm the specific role of CD53 for CXCL12- induced CXCR4 signaling, we generated mice with B-cell specific loss of CXCR4 (CXCR4^F/F;Cd19-Cre⁺) and Cd53 deletion. Splenic B cells isolated from CXCR4^F/F;Cd19-Cre⁺ mice with Cd53 deletion completely abrogated CXCL12 induced migration similar to CXCR4^F/F;Cd19-Cre⁺ mice with normal CD53 expression (Fig. 3E), suggesting that the loss of CXCL12-directed migration in the absence of CD53 is dependent on CXCR4.

Following activation, CXCR4 is rapidly internalized,(22) and the regulation of CXCL12/CXCR4 signaling is largely dependent upon this receptor internalization rate. Thus, we assessed whether loss of CD53 impairs CXCR4 internalization by analyzing the surface expression of CXCR4 in isolated splenic B cells in a regular time interval for an hour after CXCL12 stimulation. As shown in Figure 4A–B, internalization was reduced in the Cd53^−/− B cells compared to WT.

Fig 4. CD53 interacts with CXCR4 and promotes CXCR4 internalization.

Splenic B cells from WT and Cd53^−/− mice were stimulated ex-vivo with CXCL12, and CXCR4 internalization was assessed in 15 min intervals by analyzing CXCR4 surface expression. (A) Shown are the mean fluorescent intensity (MFI) of the surface expression of CXCR4 at the indicated timepoints from the indicated groups in one representative experiment. (B) CXCR4 internalization rate per minute was calculated from the mean fluorescence intensity of CXCR4 expression. Data represent the mean ± SEM from four independent experiments. **p < 0.01 by unpaired student t test. (C) Shown are representative maximum intensity projection confocal images of Duolink foci (green) and DAPI (blue), as well as DIC images, of splenic B cells from WT and Cd53^−/− mice. Data are representative of two independent experiments. See Supplemental Fig. 3 for no-antibody controls.

CD53 interacts with CXCR4

Finally, we asked whether CD53 physically interacts with CXCR4. To answer this, we used the Duolink proximity ligation assay (PLA) system, which can identify interacting proteins within 40nm of one another using oligonucleotide-conjugated antibodies and rolling-circle amplification. Indeed, proximity labeling of enriched splenic B cells using oligonucleotide-conjugated antibodies to CD53 and CXCR4 identified an interaction between the two proteins (Fig. 4C and Supplemental Fig. 3).

Together, these data suggest that CD53 interacts with CXCR4, and promotes CXCR4 signaling in both normal and malignant B cells. While several other tetraspanins have been shown to regulate CXCR4,(23–28) this is, to our knowledge, the first report of an interaction between CD53 and CXCR4.

As discussed above, CD53 is highly expressed on malignant B cells, and our data show that loss of CD53 impairs CXCL12-directed migration and bone marrow residence of these cells. Notably, CXCR4 has been considered an attractive molecular target for the treatment of hematopoietic malignancies, as the CXCR4-CXCL12 interaction contributes to the retention of leukemic cells in the bone marrow where they may receive protective, pro-survival and growth signals, and evade the effects of chemotherapy.(2) Numerous early-phase studies combining CXCR4 or CXCL12 inhibitors with chemotherapy are in clinical trials for both acute myeloid and lymphoid leukemias.(29) Our data thus suggest CD53 as a potential new target for the treatment of leukemias, and further studies are needed to determine whether CD53 loss sensitizes leukemic cells to conventional cytotoxic chemotherapies. As our current study focused on the relationship between CD53 and CXCR4 in B cells, further studies are also needed to determine whether CD53 influences CXCR4 signaling in other hematopoietic cell populations as well. In addition, the reduction in splenic homing of Cd53^−/− B cells in our in vivo transfer experiments suggest that CD53 may promote CXCR4- independent homing as well, which requires further study.

Key Points:

Normal and malignant B cells lacking CD53 have impaired CXCL12-directed migration.

CD53 is a novel binding partner of CXCR4 in B cells.

CD53 promotes CXCL12/CXCR4 signaling and marrow homing in B cells.