ICAM-1 mediated cell-cell adhesion exerts dual roles on human B cell differentiation and IgG production

Summary Intercellular adhesion molecule 1 (ICAM-1) plays prominent roles in mediating cell-cell adhesion which also facilitates B cell activation and differentiation with the help from CD4+ T cells. Here, we have reported a unique phenomenon that increased ICAM-1 on purified human CD4+ T cells upon anti-CD3/CD28 stimulation enhanced CD4+ T-B cell adhesion whereas induced less B cell differentiation and IgG production. This was largely due to increased PD-1 expression on CD19hi B cells after coculturing with hyperactivated CD4+ T cells. Consequently, ICAM-1 blockade during CD4+ T cell-B cell coculture promoted IgG production with the activation of ERK1/2 and Blimp-1/IRF4 upregulation. Consistently, CD4+ T cells from moderate-to-severe SLE patients with high ICAM-1 expression mediated less IgG production after T-B coculture. Therefore, ICAM-1-mediated human CD4+ T-B cell adhesion provides dual roles on B cell differentiation and IgG production partially depending on expression levels of PD-1 on B cells, supporting cell adhesion and subsequent PD-1 induction as an alternative intrinsic checkpoint for B cell differentiation.


INTRODUCTION
Intercellular adhesion molecule 1 (ICAM-1, also known as CD54) belongs to the family member of cell adhesion molecules characterized as a type I transmembrane glycoprotein.It is constitutively expressed on a variety of hematopoietic cells including B cells, dendritic cells (DCs), and follicular DCs as well as endothelial cells 1 at low levels at the steady state.ICAM-1 interacts with integrin family member lymphocyte functionassociated antigen-1 (LFA-1, CD11a/CD18) to mediate cell-cell adhesion at the early stage of immune cell activation. 2,35][6] Upregulation of ICAM-1 after cell activation is demonstrated to be able to modulate the magnitude of immune responses. 7,8he roles of ICAM-1 in B cell-engaged humoral immunity have been elucidated especially in germinal centers (GCs).Within GCs, ICAM-1/ LFA-1 ligation induces the signals supporting cognate interactions of B cells with follicular T helper (Tfh) cells for the generation and maintenance of humoral immunity. 9ICAM-1/LFA-1 interaction between T cells and B cells lowers the threshold of B cell activation depending on cytohesin-1 and Jun-activating binding protein 1 (JAB-1), which promotes T-B cell adhesion and synapse formation respectively. 4Furthermore, ICAM-1/2 on B cells is essential for long-lasting cognate T-B interactions and selection of low-affinity B cell clones in T cell-dependent antibody immune responses in mice. 10During human T-B cell collaboration, ICAM-1/LFA-1 interactions are also necessary for both the induction of human B cell proliferation and differentiation as well as the induction of IL-2 production in CD4 + T cells. 11LFA-1 on resting B cells and ICAM-1 on activated CD4 + T cells thus play a critical role in initial T cell-dependent B cell activation. 11,124][15][16] For instance, in system lupus erythematosus (SLE) non-synonymous ICAM-1 rs5498 (ICAM Lys469Glu ) is reported to be associated with the increase in soluble ICAM-1 (sICAM-1) levels and the susceptibility to SLE. 17 sICAM-1 is dramatically augmented in the periphery of SLE patients as well. 18

OPEN ACCESS
In the present study, we have observed an unusual phenomenon that purified human CD4 + T cells upon anti-CD3/CD28 at different time points exert dual roles in regulating B cell differentiation and IgG production.While TCR/CD28 stimulation induced ICAM-1 upregulation in a time-dependent manner, IgG production in the supernatants of CD4 + T-B cell co-culture was higher when CD4 + T cells were stimulated for 24 h 19,20 than those for 72 h.Moreover, ICAM-1 blockade on CD4 + T cells upon 24 h' stimulation suppressed IgG production whereas that on CD4 + T cells upon 72 h' stimulation promoted IgG production.The paradoxical effects of ICAM-1 on human B cell activation and IgG production with the help from activated CD4 + T cells for 72 h were in part due to the up-regulation of PD-1 expression on activated CD19 hi B cells after T-B coculture.Addition of PD-L1-Fc protein in 72 hrs-stimulated CD4 + T-B cell co-culture attenuated the increase in IgG production upon ICAM-1 blockade.More interestingly, when co-culturing B cells with autologous CD4 + T cells from moderate-to-severe SLE patients highly expressing ICAM-1, the blockade of ICAM-1 also induced increased IgG production whereas no increase was observed from mild SLE with low-expression of ICAM-1 on CD4 + T cells.Our data thus support that hyperactivation of CD4 + T cells enhances the adhesion with B cells through increased ICAM-1 expression, which is prone to promote the induction of PD-1 on B cells and exert intrinsically negative modulation on B cell differentiation and IgG production.

RESULTS
IgG production has decreased when B cells are co-cultured with CD4 + T cells upon TCR/CD28 activation for 72 h compared to those for 24 h We have previously established an in vitro CD4 + T-B cell co-culture system to investigate the roles of human CD4 + T cell activation in promoting autologous B cell differentiation. 19,20It was evident that upon anti-CD3/CD28 stimulation for 24 h, CD4 + T cells promoted B cell differentiation and IgG/IgM production efficiently in the in vitro co-culture.Similar to the previous results, we have also observed the increase in IgG levels in the supernatants of CD4 + T-B co-culture when CD4 + T cells were activated for 24 h.Unexpectedly, IgG levels were lower in the supernatants with CD4 + T cells upon 72h 0 stimulation than those with 24 h stimulated CD4 + T cells whereas still higher than those with resting CD4 + T cells (Figure 1A).
We have also reported that CD19 hi B cells, an activated B cell subset associated with IgG/IgM production, could be induced after the in vitro co-culture of activated CD4 + T cells with autologous B cells. 20This B cell subpopulation was detectable in the peripheral of SLE patients as well with high levels of phosphorylated Syk and Erk1/2. 19It was also shown that CD19 hi B cells upregulated Blimp1 (Figures S2A and  S2B), IRF4 (Figures S2C and S2D) and CD138 (Figures S2E and S2F) when compared to CD19 lo B cells, which displayed the phenotypic properties of plasma cells.Therefore, we analyzed the induction of CD19 hi B cells after the co-cultures of B cells with either 24 h or 72 h-stimulated CD4 + T cells.It was apparent that more CD19 hi B cells were induced after 24 h-stimulated CD4 + T-B co-cultures (Figure 1B).Consistent with different patterns of IgG production in CD4 + T-B cell co-cultures, there exhibited increased ratios of CD19 hi /CD19 lo B cells in 24 h-stimulated CD4 + T-B co-culture than resting T cell-engaged co-culture whereas dramatic decrease in the ratios of CD19 hi /CD19 lo B cells were observed in 72 h-stimulated CD4 + T-B co-culture (Figure 1C).Furthermore, we also found that absolute counts of CD19 hi B cell were increased after cocultured with 24 h-stimulated CD4 + T cells whereas decreased with 72 h-stimulated CD4 + T cells (Figure 1D).Collectively, these data reveal an unusual phenomenon that CD4 + T cell activation upon longer TCR/CD28 stimulation (72 h) attenuates B cell differentiation and IgG production when compared to short duration of stimulation (24 h).

Profiles of co-stimulatory molecule expression and cytokine production in human CD4 + T cells upon anti-CD3/CD28 stimulation
To explore the underlying mechanisms that longer activation of human CD4 + T cells dampens functional differentiation of B cells, we firstly examined the profiles of co-stimulatory molecules and cytokine production of CD4 + T cells upon anti-CD3/CD28 stimulation for 24 h and 72 h, respectively.It was obvious that upon TCR/CD3 stimulation, abTCR levels on CD4 + T cells were dramatically decreased upon 24 h' and 72 h' stimulation (Figures 2A and 2B) due to the internalization. 20CD69 expression on CD4 + T cells, an early activation marker, increased dramatically after 24 h and decreased later on at 72 h (Figures 2C and 2D).HLA-DR expression, a late activation indicator, was upregulated along with stimulation duration (Figures 2E and 2F).We also found that ICAM-1 expressions on CD4 + T cells increased gradually with the elongation of stimulation duration (Figures 2G and 2H).Simultaneously, we detected key cytokine production in CD4 + T cells including IL-2 (Figures 2I and  2J), IFN-g (Figures 2K and 2L), IL-4 (Figures 2M and 2N), IL-10 (Figures 2O and 2P), IL-17A (Figures 2Q and 2R) and IL-21 (Figures 2S and 2T).Most of cytokine production at 72 h was higher than that at 24 h upon anti-CD3/CD28 stimulation except IL-2.These results apparently indicate that upon TCR/CD28 stimulation in vitro, CD4 + T cells exhibit more activation status at 72 h than at 24 h with hyperactivation phenotypes.

ICAM-1 is engaged in increased human CD4 + T cell-B cells adhesion upon TCR/CD28 stimulation
T-B cell interaction is essential for T cell-dependent antibody responses.][23] Our previous study indicated that during human CD4 + T-B cell interaction, ICOS and CD40L exerted distinct roles in cell-cell adhesion and B cell differentiation. 19ICAM-1 is one of key molecules to mediate cell-cell adhesion which should be engaged in the first step of T-B interaction.We therefore investigated how ICAM-1 regulated CD4 + T cell-B cell adhesion.To address this question, micropipette adhesion frequency assay (Figure 3A) 19,24 has been adapted.Consistent with our previous study, when CD4 + T cells were activated with anti-CD3/CD28 beads, the adhesion frequencies (Pa values) (Figure 3B, red line) were significantly higher than those between resting CD4 + T cells and B cells (both p < 0.001) from the contacting time at 3 s (Figure 3B, blue line).When pre-incubating anti-ICAM-1 blocking mAb with activated CD4 + T cells, the Pa values decreased dramatically that was close to those of resting CD4 + T cells.When we performed flow cytometry to determine the molecular density of ICAM-1 on activated CD4 + T cells at different time point (Figure S1), it was indicated that ICAM-1 expression was upregulated on activated CD4 + T cells especially after 72 h stimulation (Figure S1D).These results thus indicate that there exhibits increased adhesion between B cells and CD4 + T cells once CD4 + T cells are activated upon TCR/CD28 signaling in which ICAM-1 is engaged in increased human CD4 + T-B cell adhesion.

ICAM-1 blockade leads to increased IgG production by B cells when co-culturing with 72 h-stimulated CD4 + T cells
Since cell-cell adhesion is mostly the first step for T cell-dependent B cell differentiation, we further investigated whether ICAM-1 played certain roles in IgG production through adding anti-ICAM-1 blocking mAb during CD4 + T-B cocultures.Human CD4 + T cells stimulated with anti-CD3/CD28 beads either for 24 h or 72 h were co-cultured with autologous B cells with or without anti-ICAM-1 mAb.IgG levels in the supernatants were determined after 12 days.It was found that IgG levels were decreased in the supernatants of 24 h-stimulated CD4 + T-B cell cultures with the blockade of ICAM-1 (Figure 4A) when compared to isotype controls, which suggests that ICAM-1-mediated adhesion facilitates IgG production.Surprisingly, addition of anti-ICAM-1 blocking mAb during the coculture led to the increase in IgG production after 12 days (Figure 4B), which is contrary to the results from CD4 + T cells with the activation time for 24 h.The blockade of cell-cell adhesion with 72 h-stimulated CD4 + T cells was inclined to facilitate IgG production.We also analyzed the induction of CD19 hi B cells after the co-cultures of 24 h or 72 h-stimulated CD4 + T cells and B cells with or without the blockade of ICAM-1.It was found that addition of anti-ICAM-1 blocking mAb in the 24 h-stimulated CD4 + T-B cell co-culture led to the dramatic decrease in CD19 hi /CD19 lo B cell ratios after 12 days.However, CD19 hi /CD19 lo B cell ratios after 72 h-stimulated CD4 + T-B cell co-culture increased significantly when compared to isotype controls (Figures 4C and 4D).These results were extremely consistent with IgG production we detected in the supernatants of CD4 + T-B co-cultures with two types of activated CD4 + T cells.Our data therefore strongly implies that ICAM-1 on activated CD4 + T cells might exert diverse roles in B cell differentiation and IgG production depending on activation times.

Increased IgG production upon ICAM-1 blockade in 72 h
It is out-of-expectation that although CD4 + T cells upon 72 h activation displays more activation features including higher ICAM-1 expression, interruption of ICAM-1-mediated adhesion seems to promote B cell differentiation and IgG production.To validate this phenomenon from the viewpoint of intracellular signals, we first determined whether Erk1/2 activation was involved in ICAM-1 blockade-mediated IgG over-production in 72 h-stimulated CD4 + T-B co-culture.Erk1/2 activation is reported to be involved in B cell activation and proliferation. 25,26When an Erk1/2 inhibitor SCH772984 was added in 72 h-stimulated CD4 + T-B co-culture at a final concentration of 2 mM in the presence of anti-ICAM-1 blocking mAb, it was found that addition of SCH772984 dramatically diminished the increment of IgG production (Figure 5A) as well as the ratios of CD19 hi /CD19 lo B cells when compared to anti-ICAM-1 mAb alone group (Figures 5B and 5C).Addition of SCH772984 slightly reduced the absolute number of CD19 hi B cells after the coculture (Figure 5D).Meanwhile, Erk1/2 inhibitor also dramatically suppressed the phosphorylation of NF-kB (p65) (Figure 5E) and Erk1/2 (pT202/pY204) (Figure 5F) in CD19 hi B cell subset.Both of them are reported to be involved in B cell activation and differentiation. 27,28However, phosphorylation of p65 was not altered in CD19 lo B cells.Blimp1 and IRF4 are key transcriptional factors to promote the differentiation of plasma cells. 29,30Erk1/2 inhibition also led to dramatic decrease in the expressions of Blimp1 and IRF4 in CD19 hi B cell subpopulation after the co-culture (Figures 5G and 5H).Taken together, our data demonstrate that ICAM-1 blockade-induced IgG hyper-production is Erk1/2 dependent, which is accompanied by increased phosphorylation of NF-kB and over-expression of Blimp1 and IRF4 in CD19 hi B cell subsets.

PD-1/PD-L1 ligation attenuates ICAM-1 blockade-induced IgG over-production in 72 h-stimulated CD4 + T-B co-culture
Since blockade of ICAM-1 might interrupt CD4 + T-B cell adhesion (Figure 3), we speculated that interruption of CD4 + T-B adhesion might also affect certain signals negatively regulating B cell differentiation and IgG production.We therefore determined the expression of PD-1 on B cells after co-culturing with either 24 h or 72 h-stimulated CD4 + T cells.It was shown that while no difference in PD-1 expression on CD19 lo B cells was observed after the co-culture with 24 h or 72 hrs-stimulated CD4 + T cells, PD-1 expression on CD19 hi B cells was increased dramatically when co-culturing with 72 h-stimulated CD4 + T cells compared to those with 24 h-stimulated CD4 + T cells (Figures 6A and 6B).However, PD-1 expression on CD4 + T cells was comparable after the co-cultures (Figures 6C and 6D).To validate whether PD-1/PD-L1 ligation is involved in regulating IgG production upon ICAM-1 blockade, we added PD-L1-Fc proteins at the concentrations of 5, 10 and 20 mg/mL respectively in 72 h-stimulated CD4 + T-B co-cultures with anti-ICAM-1 blocking mAb.After 12 days, it was found that addition of PD-L1-Fc protein dramatically reduced IgG production in the co-culture supernatants in a dose dependent manner (Figure 6F).However, it did not significantly affect the ratios of CD19 hi /CD19 lo B cells after the co-culture in the presence of anti-ICAM-1 blocking mAb (Figures 6E and  6G).Noteworthy, PD-L1-Fc protein alone only had slightly decreased IgG production (Figure 6F) with no effects on the ratios of CD19 hi / CD19 lo B cells (Figures 6E and 6G).
Consistently, addition of PD-L1-Fc protein in the ICAM-1 blockade co-culture led to the decrease in the phosphorylation of NF-kB (p65) (Figure 6H) and Erk1/2 (pT202/pY204) (Figure 6I) in CD19 hi B cell subpopulation after the co-culture.We also detected phosphorylation levels of SHP-2, a protein phosphatase to dephosphorylate NF-kB (p65) and Erk1/2 (pT202/pY204), 31,32 in CD19 hi B cell subpopulation.It was found that phospho-SHP-2 (Y542) levels were increased in CD19 hi B cell subpopulation with the addition of PD-L1-Fc protein in a dose dependent manner (Figure 6J).Collectively, our data support that PD-1 expression is up-regulated on activated B cells after coculturing with 72 h-stimulated CD4 + T cells with more extent.ICAM-1 blockade likely interrupts PD-1-PD-L1 ligation, which in turn attenuates negative signaling for B cell differentiation and IgG production and leads to the increased IgG production we observed in 72 h-stimulated CD4 + T-B co-culture upon ICAM-1 blockade.
High expression of ICAM-1 on CD4 + T cells from moderate/severe SLE patients also mediates suppression on IgG production after T-B coculture SLE is an autoimmune disease characterized by autoantibody-driven tissue and organ damages.Emerging evidence reveals that exaggerated B cell immune responses play central roles in the pathogenesis of SLE.In fact, we have also observed dramatic increase in ICAM-1 (Figures 7A  and 7B) and ICOS (Figures 7C and 7D) expressions on peripheral CD4 + T cells from moderate/severe SLE patients when compared to HCs and mild SLE, which indicated the hyper-activation of CD4 + T cells in moderate/severe SLE.By using the micropipette adhesion frequency assay, we have observed increased adhesion between freshly isolated CD4 + T cells and B cells from SLE patients (Figure 7E, red line) when compared to that from HC counterparts (Figure 7E, black line).When pre-incubating CD4 + T cells with anti-ICAM-1 blocking mAb, there exhibited the decreased Pa values in the CD4 + T cell-B cell adhesion assay (Figure 7E, blue line), demonstrating the involvement of ICAM-1 in increased cell-cell adhesion under SLE pathogenesis.
We further performed ICAM-1 blockade in CD4 + T-B co-cultures from either mild or moderate/severe SLE patients whose ICAM-1 expression levels were different.Consistent with the results from healthy donors, blockade of ICAM-1 in CD4 + T-B cell coculture from moderate/severe SLE patients also led to the increase in IgG production (Figure 7F, right) whereas no difference existed in those from mild SLE patients with lower expression of ICAM-1 on CD4 + T cells (Figure 7F, left).These results to some extent recapitulated the phenomenon we observed in healthy donors.Furthermore, we also found that the percentages of peripheral ICAM-1 + CD4 + T cells had significantly negative correlations with IgG contents in SLE patients (Figure 7G).The data from SLE patients further support that high expression of ICAM-1 on over-activated CD4 + T cells promotes T-B cell adhesion whereas mediates negative signals for B cell differentiation and IgG production.

DISCUSSION
T-B cell interactions are critical for T cell-dependent B cell activation and terminal differentiation as well as IgG generation mostly through the ligations of co-stimulatory molecules. 33ICAM-1 is a key adhesion molecule to mediate cell-cell adhesion.Previous studies have revealed that ICAM-1 was engaged in long-lasting cognate T-B interactions necessary for the selection of low-affinity B cell clones in T cell-dependent antibody responses. 10In this study, we have reported an extraordinary phenomenon that ICAM-1 expression on hyper-activated CD4 + T cells (anti-CD3/CD28 activation for 72 h) mediated strong cell-cell adhesion whereas unexpectedly dampened B cell differentiation and IgG production.This is largely due to inducible expression of PD-1 on CD19 hi B cells after T-B co-culture mediating negative signals to suppress IgG production.We therefore intend to propose dual modules of ICAM-1-engaged adhesion between activated CD4 + T cells and B cells for B cells differentiation and IgG production probably relying on inducible expression of certain negative regulatory molecules such as PD-1 on B cells.Since PD-1 expression is induced along with CD4 + T-B interactions, our results imply a novel intrinsic mechanism to regulate B cell differentiation and IgG production through cell-cell adhesion.
We observed this phenomenon unexpectedly in the longitude investigations to decipher the effects of cell-cell adhesion on B cell differentiation.In fact previous investigations revealed the critical roles of cell-cell adhesion in promoting T cell activation, 34 cytotoxicity of NK cells 35 as well as B cell differentiation in GC. 30 Cell-cell adhesion occurs at the early stage of immune responses.Formation of immune cell conjugates not only establishes direct liaisons between targeted immune cells, but also delivers key signals to determine whether immune cells undergo activation or not through the ligations between multiple molecule pairs. 36In most cases immune responses will not be triggered without direct cell-cell contact.Being one of the key adhesion molecules, ICAM-1 is demonstrated to be engaged in cell-cell adhesion mainly. 37Herein, we have observed dramatic upregulation of ICAM-1 on human CD4 + T cells once activated by anti-CD3/CD28 in vitro with more adhesion frequencies between CD4 + T cells and B cells (Figure 3).However, increased adhesions between activated CD4 + T cells and B cells are not aligned with increased IgG production in our study (Figure 1A).Longer activated duration CD4 + T cells undergo, less IgG B cells produce.We therefore deduced that cell-cell adhesion of B cells with hyper-activated CD4 + T cells might delivery certain negative signals to restrain B cell activation and differentiation.
We have further demonstrated that inducible expression of PD-1 on B cells after encountering activated CD4 + T cells was one of the mechanisms to regulate human B cell differentiation and IgG production negatively.PD-1 is inducibly expressed on T cells and exerts co-inhibitory effects to restrain T cell immunity especially in anti-tumor immunity. 38,39It can recruit protein diphosphatase SHP-2 to dephosphotate activated signal molecules downstream TCR signal including CD3, ZAP-70 et al. to suppress T cell activation. 40Several studies have reported the roles of PD-1 in B cell activation and differentiation in mouse models.For instance, through inducing more cell death in GCs and less cytokine production by Tfh cells, PD-1 is dedicated to optimal formation of long-lived plasma cells by regulating the survival and selection of B cells within GCs. 41 PD-1 is also required for optimal IL-21 production by Tfh cells. 41,42We have previous reported that injection of PD-1 blocking antibodies in OVA immunized mice led to the augmentation of total IgG and OVA-specific antibodies, 43 supporting negative regulation of PD-1 on B cell differentiation.It is also reported that depletion of PD-1 promoted Tfh cell expansion in vaccine-immunized mice via promoting cytokines secretion and enhanced humoral immunity to Malaria infection. 44,45In mechanism exploration, Shi et al. reported that PD-1 suppressed the recruitment of Tfh cells but promoted the numbers of Tfh cells in the GC territory, which helped to maintain the stringency of GC affinity selection. 42These data have outlined the direct roles of PD-1 in regulating B cell differentiation and IgG production in mouse GCs.In our study, based on the in vitro human T-B coculture assay, we have observed the up-regulation of PD-1 expression on CD19 hi B cells, a subpopulation responsible for IgG production, 19,20 when coculturing with 72 h-stimulated CD4 + T cells.The upregulation of PD-1 on activated B cells could interact with PD-L1 to transmit a negative signal and attenuate B cell differentiation and IgG production where ICAM-1 was largely responsible for cell-cell contact to bridge PD-1 and PD-L1 ligation.
The results from molecular mechanism exploration are also consistent with the observations on B cell differentiation and IgG production upon ICAM-1 blockade.Collectively, Erk1/2 and NF-kB activation were detectable upon ICAM-1 blockade together with Blimp1 and IRF4 upregulation in CD19 hi B cell subset in 72 h-stimulated CD4 + T-B cell coculture.They are key contributors to B cell differentiation and antibody production. 46,47When PD-L1-Fc fusion protein was added in the blocking assays, the activation status was apparently reversed whereas SHP-2 phosphorylation was up-regulated.Since no dramatic difference of PD-1 expression on CD4 + T cells is observed after the co-culture with or without ICAM-1 blockade (Figures 6C and 6D), the inducible expression of PD-1 on CD19 hi B cells is probably an intrinsic key event after B cells contact with activated CD4 + T cells to regulate B cell activation and differentiation.When cell-cell adhesion is interrupted by ICAM-1 blockade, this negative feedback is impaired which leads to over-activation of B cells and subsequent IgG hyper-production that we observed in 72 h-stimulated CD4 + T-B cell coculture.
It is well-known that ICAM-1 is also expressed on B cells.However, the expression patterns of ICAM-1 on B cells and T cells are different according to our results.B cells express ICAM-1 constitutively.Even after 12 days of T-B co-culture, the expression levels of ICAM-1 on B cells  are comparable (Figure S2G).However, we have observed dramatic upregulation of ICAM-1 on CD4 + T cells upon anti-CD3/CD28 stimulation, which could represent the activation status as well as functional implementation of ICAM-1 for CD4 + T cell activation to some extent.Therefore, the effects of ICAM-1 blockade might be performed on both B cells and CD4 + T cells where it is more dramatic on CD4 + T cells.We herein propose that up-regulated ICAM-1 on CD4 + T cells should serve as a bridge mediating T-B adhesion, which is critical for both co-stimulatory and co-inhibitory molecules to deliver diverse signals for B cell differentiation in our system.What is more, in our in vitro co-culture system B cell activation and differentiation are non-antigen specific.We therefore would like to propose that current CD4 + T-B co-culture model is similar to T-B interactions in extrafollicular GC responses in mouse models with less antigen-dependence. 48,49t might represent the situation at the early stage of T cell-dependent B cell activation when activated CD4 + T cells initiate B cells differentiation.Although BCR signal is not involved, increased adhesion through ICAM-1 could be more feasible for cytokine and co-stimulatory or co-inhibitory signaling to regulate B cell differentiation and IgG production.The direct relationship between ICAM-1-mediated ligations and differentiation phenotypes of B cells needs to be further investigation.
Since either membrane or soluble ICAM-1 is prominent in SLE patients from our investigations and others, 50 we also explored whether activated CD4 + T cells from SLE patients functioned similarly to affect B cell differentiation and IgG production.CD4 + T cells from moderate-to-severe SLE patients with high ICAM-1 expression triggered comparable IgG production to those from mild SLE patients.However, blockade of ICAM-1 with the decrease in cell-cell adhesion promoted IgG production from moderate-to-severe SLE patients, which could be mirrored with the results from 72 h-stimulated CD4 + T cells from healthy donors.The significance might lie in the fact that although auto-reactive CD4 + T cells are hyper-activated under pathogenic status to promote B cell activation and auto-antibody production, they still possess intrinsic feedback mechanisms to control autoimmune responses to a less degree.The results we obtained may provide new clues for clinic to exaggerate PD-1 mediated inhibition as a new therapeutic strategy for the suppression of B cell over-activation in SLE.
Taken together, by using a well-established in vitro T-B co-culture system, we have illustrated that ICAM-1-mediated cell-cell adhesion could exert either positive or negative regulations during human CD4 + T cell and B cell interactions.Inducible expression of PD-1 on CD19 hi B cells accounts for negative feedback during B cell differentiation and IgG production after co-culturing with hyper-activated CD4 + T cells (Figure 6A).Based on our results, we would deduce that moderate activation of CD4 + T cells by the antigens is prone to introduce optimal B cell activation and differentiation whereas over-activation of CD4 + T cells will retard B cell activation.This might be helpful in vaccine designing to screen antigenic proteins or peptides with moderate immunogenicity for optimal humoral immunity.Among this process, ICAM-1 mediated T-B adhesion probably becomes a unique mechanism to regulate CD4 + T cell-guided B cell differentiation under given situations.The exact mechanisms of how activation status of CD4 + T cell regulating PD-1 expression on B cells needs to be investigated in the future.

Signifinance
CD4 + T cell-B cell adhesion facilitates signal transduction to promote B cell activation and differentiation.Herein we have reported that although activated human CD4 + T cells upregulated ICAM-1 expression to promote cell-cell adhesion with autologous B cells, ICAM-1engaged cell-cell adhesion also acted as one of attenuating mechanisms for B cell differentiation and IgG production partially through inducible expression of PD-1 on B cells and downstream negative signal transduction once CD4 + T cells were over-activated.Therefore, proper activation of CD4 + T cells would be beneficial for optimal B-cell terminal differentiation and IgG secretion, which might be helpful for long-term efficacy of humoral immunity.

Limitations of the study
Our findings have revealed the engagement of ICAM-1 in CD4 + T-B adhesion in negative signal transduction during B-cell differentiation and IgG secretion.Nevertheless, our study has several limitations.We did not clarified that ICAM-1 on CD4 + T cells or B cells exerts more dominant roles to mediate human T-B adhesion and subsequent regulation, which could use CRISPR-Cas9 system to knock out the expressions of ICAM-1 on human CD4 + T cell or B cell seperately.Othewise, we can also validate the in vivo regulatory modes of ICAM-1 on T-B adhesion and B cell differentiation in GCs via using CD4 Cre ICAM-1 f l /fl mice.Moreover, we can not exclude the possibility that other negative co-inhibitory molecules may cooperate (or even synergize) with PD-1 signal to regulate IgG production, warranting additional mechanistic investigation.Finally, we could provide more evidence to address whether PD-1 agonist would inhibit pathological B cell differentiation and autoantibody production under SLE pathogenesis in the future.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: In the blockade assay, function-grade mouse anti-human ICAM-1(eBioscience, San Diego, CA, USA) mAb was added at the concentration of 5 mg/mL in the co-culture system.Erk1/2 inhibitor (SCH772984) (Selleck Chemicals, Houston, TX, USA) was added at the final concentration of 2 mM.PD-L1-Fc protein (Sino Biological, Beijing, China) was added at the final concentrations of 5, 10 and 20 mg/mL in the co-culture.

Micropipette adhesion frequency assay
The adhesion between CD4 + T and B cells was also measured by using a micropipette adhesion frequency assay modified from the protocol described previously. 19,51Briefly, one B cell and one CD4 + T cell were aspirated by two apposite micropipettes with the diameters of 3-5 mm in an isotonic chamber containing RPIM 1640 medium plus 5% FBS.The adhesion between one CD4 + T cell and one B cell was detected by placing the cells into controlled contact via driving in and out of the contact with controlled area and duration.The presence of adhesion at the end of a given contact time was detected by observing the deflection of the membrane of B cell or T cell when CD4 + T cell was retracted.This approach-contact-retraction cycle was repeated 100 times for one pair of B cell and CD4 + T cell at a given contact time.For each subject at an indicated given contact time, at least 3-5 pairs of one CD4 + T cell and one B cell were tested to calculate an adhesion probability (Pa) represented by the ratio of adhesion events to total events.In blocking experiments, mouse anti-human ICAM-1 mAb (5 mg/mL) (eBioscience) were pre-incubated with CD4 + T cells at 37 C for 30 min before the adhesion assay.

Detection of molecular density
Molecular density of ICAM-1 on human CD4 + T cells was determined by flow cytometry as described previously. 24In brief, CD4 + T cells were incubated with anti-human ICAM-1-PE Ab (BD Biosciences) in FACS buffer (PBS containing 1% FBS and 1 mM EDTA) at 4 C for 30 min.Cell mixtures were washed once with PBS and resuspended in 200 mL FACS buffer.Cells were acquired by BD LSR Fortessa flow cytometry (BD Biosciences) along with four standard calibration beads (Quantibrite PE beads, from BD Biosciences).A calibration curve of PE molecules/ beads (provided by manufacturer) (Figure S1A) vs mean fluorescence intensity (MFI) of anti-human ICAM-1-PE (Figure S1B) was plotted based on the curve of Quantibrite PE beads (filled circles, Figure S1C).The density of ICAM-1 on human CD4 + T cells was calculated by comparing the MFI of the sample (open square, Figure S1C) with the calibration curve.
For cytokine staining, CD4 + T cells were fixed and permeabilized by using Fix/Perm Solution for 20 min.Cells were washed once with Perm/ Wash Buffer, and were incubated with mouse anti-human IL-2-AF700, anti-human IL-4-APC Abs (all from Biolegend), mouse anti-human IL-10-PE-Cy7 Ab (eBioscience), anti-human IFN-g-PE, anti-human IL-17A-BV786 and anti-human IL-21-AF647 Abs (all from BD Biosciences) at 4 C for 45 min.After washing once with Perm/Wash Buffer, CD4 + T cells were resuspended in 200mL FACS buffer and acquired on BD LSR Fortessaä (BD Biosciences) and data analysis was performed by using FlowJo 10.5.0 software (Tree Star Inc.)

Enzyme-linked immunosorbent assay (ELISA)
The co-culturing supernatants were subjected to IgG quantification according to the manufacture's instruction (Sen Xiong Biotech., Shanghai, China).The absorbance at 450 nm was detected within 5 min by the Power Wave XS2 microplate spectrophotometer (BioTek, VT, USA).The concentrations of IgG in the supernatants were calculated based on the standard curves.

QUANTIFICATION AND STATISTICAL ANALYSIS
All statistical analyses were performed with the GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA, USA).Graphs were created using the GraphPad Prism 8.0 and edited using Adobe Illustrator CS6 (Adobe Inc., San Jose, CA, USA).Data were presented as means G standard deviation (S.D).The normality of the data was evaluated by the Shapiro-Wilk normality test.If the data did not follow the normal distribution or the normality test indicated that the number was too small, the Mann-Whitney U or Wilcoxon matched paired test, or the Kruskal-Wallis test with subsequent Duuns multiple-comparison test was used to calculate the statistics difference.Otherwise, the unpaired or paired Student's t-test, or one-way ANOVA with subsequent Tukey's post-tests, was performed.Unless stated, p < 0.05 was considered statistically significant.

Figure 1 .
Figure 1.Decreased IgG content and CD19 hi B cell subpopulation after B cells are co-cultured with CD4 + T cells upon TCR/CD28 activation for 72 h when compared to those for 24 h Purified human CD4 + T cells from health donors were stimulated with human T-Activator CD3/CD28 Dynabeads for 24 h or 72 h.(A) IgG content in the supernatants after B cells were co-cultured with autologous activated or resting CD4 + T cells for 12 days (n = 10).(B) Presence of CD19 lo and CD19 hi B cell subpopulations in total B cell subset after the co-culture of resting (NS T + B) or activated CD4 + T cells (24h-CD4 + T+B and 72h-CD4 + T+B) at Day 12 where B cell alone was blank control.(C) Comparison of the ratios of CD19 hi /CD19 lo B cells (n = 10) in different CD4 + T-B cell co-culture groups.CD19 hi /CD19 lo B cell ratios of the blank controls of each individual HC was used for normalization.(D) The absolute counts of CD19 hi B cells after CD4 + T-B cell coculture.The data were representative of at least three independent experiments.Data were represented by mean G SD.In Figures 1A, 1C and 1D, the Kruskal-Wallis test with subsequent Duuns multiple-comparison test was used.*: p < 0.05; ***: p < 0.001.

Figure 3 .
Figure 3. ICAM-1 is engaged in increased human CD4 + T cell-B cells adhesion upon TCR/CD28 stimulation (A) Scheme of micropipette adhesion frequency assay in vitro.(B) Adhesions between B cells and resting or activated CD4 + T cells from HCs with or without anti-ICAM-1 blocking mAb (5 mg/mL) by the micropipette adhesion frequency assay.The data were representative of at least three independent experiments.Data were represented by mean G SD.The Kruskal-Wallis test with subsequent Duuns multiple-comparison test was used.*: p < 0.05; **: p < 0.01; ***: p < 0.001.

Figure 4 .
Figure 4. ICAM-1 blockade leads to increased IgG production when B cells are co-cultured with 72 hrs-stimulated CD4 + T cells (A and B) IgG levels in the supernatants of the co-cultures of B cells with 24 hrs-(A) or 72 hrs-(B) stimulated CD4 + T cells with isotype Ab or anti-ICAM-1 blocking mAb.(C) CD19 lo and CD19 hi B cell subpopulations after the co-culture of 24 hrs-or 72 hrs-stimulated CD4 + T cells at Day 12. (D) Comparisons on the ratios of CD19 hi /CD19 lo B cells after the co-culture.The data were representative of at least three independent experiments.Data were represented by mean G SD.In Figures 4A and 4B the paired Student's t test was used.In Figure 4D, the Kruskal-Wallis test with subsequent Duuns multiplecomparison test was used.*: p < 0.05; **: p < 0.01.

Figure 5 .
Figure 5. Increased IgG production upon ICAM-1 blockade in 72 hrs-stimulated CD4 + T-B co-culture is Erk1/2 dependent (A) IgG content in the supernatants of 72 hrs-stimulated CD4 + T-B cell co-culture upon ICAM-1 blockade with or without an Erk1/2 inhibitor.(B) Detection of CD19 lo and CD19 hi B cell subsets in the co-cultures at Day 12. (C) Comparisons on the ratios of CD19 hi /CD19 lo B cells after the co-culture.(D) Absolute counts of CD19 lo and CD19 hi B cells after the co-culture.(E-H) The mean fluorescence intensity of phosphorylation of p-NF-kB p65 (pS529) (E), p-Erk1/2 (pT202/pY204) (F), Blimp1 (G) and IRF4 (H) in CD19 lo and CD19 hi B cell subsets after the co-culture for 12 days.The data were representative of at least three independent experiments.Data were represented by mean G SD.The Kruskal-Wallis test with subsequent Duuns multiple-comparison test was used.*: p < 0.05; **: p < 0.01.