To explore the molecular mechanism of SLP65-mediated B cell regulation, we first used a split ubiquitin yeast-two-hybrid assay (Y2H) (44). The N-terminus of SLP65 (N-SLP65) was used to identify novel SLP65 interaction partners that might link SLP65 activity to other signaling molecules such as PI3K pathway proteins. The assay showed 10 potential candidates that may interact with N-SLP65 ( Table 1 ). Among these potential partners is the RHOA guanine nucleotide exchange factor GEF-2 (ARHGEF2, also known as GEF-H1). Since small GTPases were previously shown to play an essential role in B cell development (28), we hypothesized that GEF-H1 could also be important in SLP65-mediated signaling. To validate our observation in the Y2H assay, we performed immunoprecipitation (IP) assay to see if the interaction takes place in endogenous protein level. To this end, we used human Burkitt lymphoma derived cell line DG75EB that overexpress HA tagged RHOA (DG75EB/HA-RhoA, see methods for details). The cells were stimulated with anti-human IgM F(ab)’2 antibody for different time points to activate SLP65 and IP was carried out using anti-SLP65 antibody. The interaction of SLP65 with GEF-H1 was detectable at similar extent at all time points including unstimulated condition ( Figure 1A ). As a positive control of IP, we also detected interaction of SLP65 with PLCγ2, a known interactor of SLP65 ( Figure 1A ). This confirms that the interaction between SLP65 and GEF-H1 occurs at the endogenous protein level. In addition to this, we also employed proximity ligation assay (PLA) to investigate the interaction between SLP65 and GEF-H1 in unstimulated DG75EB/HA-RhoA cells ( Figure 1B ). Adjacent binding of the antibodies (red dots) used for the assay suggests that the corresponding proteins are localized at a proximity 10-40 nm in B cells ( Figure 1B ). Moreover, this interaction was also observed in primary B cells derived from mouse bone marrow (BM) and spleen as well as human peripheral blood derived B cells ( Figure 1C ). Thus, using three different methods and also B cells from different sources, we could show that GEF-H1 is a novel interacting partner of SLP65. In this context, one relevant question could be the relative expression level of GEF-H1 and SLP65 in different B cells used. To address this, we performed western blot to detect total protein expression of GEF-H1 and SLP65 in primary mouse B cells and DG75EB/HA-RhoA cells ( Figures S1A, B ). DG75EB/HA-RhoA (DG) cells express considerably higher level of GEF-H1 and SLP65 compared to BM derived pre-B cells while the expression is similar to mature (CD43^-) splenic (SP) B cells ( Figures S1A, B ). Nevertheless, the interaction of SLP65 with GEF-H1 holds true for all sources of B cells analyzed irrespective of the difference in the expression level. Next, to address the functional relevance of this interaction, we checked whether SLP65 activation also leads to activation of GEF-H1. To this end, we used our triple knock out (TKO) cell system with inducible SLP65 ( Figure S1C ). TKO cells are derived from the early pre-B cells from Rag2^-/- λ5^-/- SLP65^-/- mice transduced with an ER^T2-SLP65 fusion protein (45). When transduced with pre-recombined μHC and λ5 LC, the cells express ectopic pre-BCR ( Figure S1D ) and upon treatment with 4-hydroxy tamoxifen (4-OHT), SLP65 is activated leading to calcium mobilization ( Figure S1E ). We also analyzed the relative level of SLP65 expression in TKO-EST cells compared to primary B cells. As TKO-EST cells overexpress ERT2 fused SLP65 protein, the expression level is higher compared to BM derived pre-B cells but similar to mature splenic B cells ( Figure S1F ). Using this system, we found that SLP65 activation leads to increased phosphorylation of GEF-H1 ( Figure 1D ).

Since GEF-H1 activates RHO GTPases, specifically RHOA, RHOB and RHOC, we asked the question if RHOA is also regulated by SLP65. To this end, we employed PLA to determine the association between SLP65 and RHOA in unstimulated DG75EB/HA-RhoA cells. As shown in Figure 2A , red dots indicate close proximity between the two proteins suggesting their interaction. Importantly, stimulating the BCR with anti-IgM F(ab)’2 antibody for 3 min resulted in an increased proximity between phosphorylated SLP65 and RHOA ( Figure 2B ). This interaction was reduced after 5 min indicating that RHOA is recruited to active SLP65 at very early time points. Our data suggests that RHOA can interact with SLP65 in two different modes - one is direct binding, the other is indirect binding through GEF-H1. We also analyzed the relative level of RHOA expression in DG75EB/HA-RhoA cells with that of primary B cells. As depicted in Figure S1A , DG75EB/HA-RhoA cells express higher and similar level of RHOA compared to BM derived pre-B cells and splenic B cells respectively. Next, we further investigated whether inducing SLP65 activity results in activating RHOA. Indeed, activating SLP65 in TKO-EST cells using 4-OHT increased the total cellular activity of RHOA ( Figure 2C ) already after 5 min of antibody stimulation without changing its expression level ( Figure 2D ).

Altogether, these data indicate that recruitment of RHOA and GEF-H1 to the SLP65 interactome takes place as an early event and seems to play a role in stimulating the activation of RHOA.

RHOA was previously shown to activate PTEN function (33, 34), and may therefore be involved in inhibiting PI3K activity in pre-B cells. Thus, we hypothesized that GEF-H1/RHOA may provide the molecular platform that mediates SLP65 signals through PI3K pathway.

To investigate the link between PTEN/PI3K and SLP65-dependent RHOA activation, we first studied the effect of RHOA inhibition on PI3K pathway. Interestingly, treatment of pre-B cells with a RHO inhibitor significantly enhanced AKT activation as shown by increased serine 473 (S473) phosphorylation and a small but significant elevation of PTEN expression ( Figure 3A ). Next, we monitored the intracellular localization of PTEN upon induction of SLP65 function in our inducible TKO-EST cell system ( Figure S1C ). SLP65-induction led to PTEN activation and membrane localization ( Figure 3B ). In addition, treating the cells with RHO inhibitor prevented PTEN clustering and subsequent recruitment towards the membrane ( Figure 3B ). We also checked if RHO inhibition changes PTEN expression or AKT activation in these cells. PTEN total protein expression was unchanged upon RHO inhibition ( Figures S2A, B ). This could be explained by the fact that upon RHO inhibition, the subcellular localization of PTEN changes without altering the total protein amount. However, for AKT, we see a strong increase in AKT phosphorylation in TKO-EST cells upon SLP65 activation, which was reduced upon RHO inhibition but still remain higher compared to the control untreated cells ( Figures S2A, B ). This was similar to what we see in the primary WT bone marrow derived B cells.

These data indicate that activation of RHOA by SLP65 leads to induction of the RHOA/PTEN pathway thereby counteracting PI3K signaling.

Next, to analyze the effects of RHOA in B cell development in vivo, we generated B cell specific RhoA deleted mouse line. Mice harboring a “floxed” RhoA allele (RhoA^fl/fl ) were crossed to Mb1^+/hCre transgenic mice to enable RhoA gene deletion at Pro-B cell stage. The animals showed a complete block in B cells development at the pro-B cell stage as indicated by the absence of B220⁺CD25⁺ pre-B cells in bone marrow ( Figure 4A ) and absence of mature B cells in the spleen ( Figure 4B ). These data suggest that RHOA plays an indispensable role in pre-B cell differentiation. This phenotype resembles the phenotype of Pten^fl/fl mice (46) and FoxO1^fl/fl mice (14) crossed to Mb1^+/hCre , which is in line with our hypothesis that RHOA mediates SLP65 function in pre-B cell differentiation through suppressing PI3K signaling (47).

To confirm this observation in vitro, we introduced a 4-OHT-inducible Cre (Cre-ER^T2) retroviral vector, or a control ER^T2, into BM-derived RhoA^fl/fl pre-B cells. Inducible deletion of the RhoA gene led to rapid cell death within 96 hours ( Figures 5A, B , Figures S3A, B ). Interestingly, RHOA knock out shows a small but significant reduction of PTEN expression and a slightly reduced expression of total FOXO1 ( Figure 5C and Figure S3C ). However, the relative level of phosphorylated FOXO1 is increased in the RHOA deleted cells indicating a reduced activity of this transcription factor ( Figure 5C and Figure S3C ). This suggests that RHOA expression is required for proper PTEN expression in pre-B cells and RHOA deficiency interferes with the regulation of PI3K signaling possibly through a negative feedback mechanism by reducing PTEN expression. Of note, the protein expressions were determined 48 hours after RhoA deletion and no cell death was evident at this time point. Thus, the changes in protein expression are not merely due to loss of cells but due to loss of RhoA. In addition, RhoA-deficient pre-B cells show significant decrease in Rag1 and Rag2 gene expression levels ( Figure 5D and Figure S3D ) indicating that they are unable to undergo LC gene recombination. This could be explained by the reduced activation of FOXO1 as demonstrated above. To confirm this observation, BM-derived pre-B cells from either RhoA^fl/fl or RhoA^fl/fl Mb1^+/hCre mice were reconstituted with pMIG-RHOA retroviral vector or with an empty vector (EV, pMIG) ( Figure 5E and Figure S3E ). Whereas ectopically expressed RHOA (GFP+) was not enriched in RhoA^fl/fl pre-B cells ( Figure S3E ), RHOA^pos population was significantly enriched in RHOA-deficient pre-B cells and also led to increased cell viability ( Figures 5E, F ). Importantly, RHOA overexpression restored the ability of these cells to differentiate as represented by the elevated percentage of cells expressing Igκ ( Figures 5E, G ) and by increased levels of Rag1 and Rag2 gene expression ( Figure 5H and Figure S3F ). Next, we investigated whether ectopic expression of either FOXO1 or PTEN can overcome the differentiation block observed in RhoA-deficient pre-B cells. Interestingly, neither of these factors could compensate for RHOA deficiency in driving pre-B cell differentiation ( Figure S3G ). Altogether, these data explain the developmental block observed in vivo and suggest that RHOA is an important factor for Ig gene rearrangement acting downstream of FOXO1/PTEN pathway.

Since RHOA is indispensable for pre-B cell development, we investigated if it has any role in the development of pre-B cell malignancies such as in BCP-ALL. Here, we generated BCR-ABL1 transformed bone marrow derived pre-B cells from RhoA^fl/fl mice. For inducible deletion of the RhoA gene, we introduced a 4-OHT-inducible Cre-ER^T2 vector into the BCR-ABL1-transformed RhoA^fl/fl cells. Inducible deletion of the RhoA gene led to cell death of the BCR-ABL1-transformed pre-B cells ( Figure 6A ). Interestingly small GTPases, like RHOA, often function downstream of G-protein coupled receptors, such as CXC family chemokine receptor 4 (CXCR4) or integrins, which are important co-receptors regulating B cell development and migration (28). Previously, we showed that the signaling pathways of IL7R and CXCR4 are tightly regulated by the activity of the oncogenic kinase BCR-ABL1 (48). Here we show that the inhibition or deletion of RHOA significantly reduced cell migration toward CXCL12, the ligand of the chemokine receptor CXCR4 ( Figure 6B ).

Next, we studied whether RhoA is also required for the proliferation and survival of transformed B cells in vivo. Since BCR-ABL1 transformed pre-B cells die after RhoA deletion in vitro ( Figure 6A ), we isolated bone marrow derived pro- B cells from RhoA^fl/fl Mb1^+/hCre mice, transformed them with BCR-ABL1 and injected them intravenously into immunodeficient Bl6 Rag2 ^-/- γc ^-/- mice. RhoA^fl/fl Mb1^+/+ pro-B cells transformed with BCR-ABL1 were used as control group. The mice were then sacrificed after two weeks when the control mice showed leukemia symptoms such as hind limb paralysis. Mice injected with RhoA-deficient BCR-ABL1 cells did not develop leukemia when compared to mice injected with control cells as evidenced by reduced spleen size and total number of blasts in both spleen and bone marrow ( Figures 6C, D ). Consequently, significantly less CNS infiltration was observed for mice injected with RhoA-deficient BCR-ABL1 cells compared to mice injected with control RhoA-proficient BCR-ABL1 cells ( Figure 6E ) (49). In order to exclude that RhoA deficient BCR-ABL1 transformed cells failed to engraft in vivo, the bone marrow from both mouse groups was recovered and cultured in vitro in medium without IL7 which enables the growth of transformed cells only. After one week, a pure population of transformed cells was generated from both groups (RhoA-competent and RhoA-deficient) indicating that all mice were successfully engrafted with BCR-ABL cells. Next, we performed an in vitro proliferation assay with recovered BCR-ABL cells. Cells were labelled with fluorescent cell proliferation dye and the rate of proliferation was measured by flow cytometry after 3 days. RhoA deficient transformed cells were shown to proliferate to a similar extent as transformed cells with intact RhoA ( Figure 6F ). These data suggest that RhoA deficiency does not affect the proliferation of the transformed cells in vitro. Nevertheless, RhoA deficiency prevents leukemia development and disease progression in vivo ( Figures 6C–E ).

To better understand the role of RHOA in late B cell developmental stages, we crossed the RhoA^fl/fl mice to Cd21 ^+/Cre transgenic mice (50). CD21 is expressed at the time when immature/transitional B cells differentiate into mature long-lived peripheral B cells (51). Results show that B cell development in the bone marrow was not affected in RhoA^fl/fl Cd21 ^+/Cre mice ( Figure S4A ). A strong reduction in RhoA gene expression and a concomitant increase in RhoB and RhoC genes were observed in the splenic B cells from RhoA^fl/fl Cd21 ^+/Cre mice ( Figure S4B ) indicating a compensatory mechanism in the cells. Interestingly, Rock1 and Rock2, the two RHO activated kinases, are reduced in the RhoA^fl/fl Cd21 ^+/Cre mice ( Figure S4B ). In the spleen, although there were no significant changes in the follicular (FO) B cells percentages or total cell number ( Figures 7A, B ), the marginal zone (MZ) B cell population was strongly reduced in RhoA-deficient mice compared to control mice ( Figures 7A, B and Figure S4C ). Thus, we investigated whether this phenotype was correlated with altered functional responses. We stimulated splenic B cells from RhoA^fl/fl Cd21 ^+/Cre or RhoA^fl/fl Cd21 ^+/+ mice with anti-BCR (anti-µHC), CpG, LPS and anti-CD40 antibody and measured the surface expression of early activation markers CD86 and CD69 after 24 hours of stimulation. ( Figure S4D ). RhoA-deficient B cells showed some reduction in CD86 surface levels, but not CD69, in response to the stimulation in comparison to wildtype B cells ( Figures S4D–F ). Nevertheless, in response to LPS stimulation, RhoA-deficient B cells showed impaired ability to proliferate and become CD138⁺/B220^lo plasma cells ( Figures 7C, D ) and expressed lower levels of BLIMP-1 ( Figures 7E, F ) after 4 days of stimulation. Interestingly, we found that under stimulation with CpG, RhoA-deficient B cell proliferated in similar pattern as control cells ( Figures 7C, D ). These results suggest that while RHOA may be dispensable for proliferation through TLR9, it is absolutely required for TLR4. The reduced expression of BLIMP-1 suggests that RhoA-deficient B cells cannot be committed into terminal differentiation, most likely due to deregulated PI3K signaling (19). Altogether, the data suggest that functional RHOA is required for proper selection and regulation of peripheral B cells.

Next, we measured the serum level of IgM and IgG antibodies in RhoA^fl/fl Cd21 ^+/Cre mice compared to the control group. Serum IgM did not show any alteration ( Figure S5A ) while IgG titre was reduced in RhoA^fl/fl Cd21 ^+/Cre mice specifically at the older age ( Figure S5A ). Importantly, RhoA-deficient mice had an increased amount of autoreactive anti-double stranded DNA (dsDNA) antibodies in sera compared to control mice ( Figure S5A ). Thus we investigated whether RhoA^fl/fl Cd21 ^+/Cre mice develop autoimmune diseases such as kidney failure with age. Histological analysis of HE and PAS stained kidney sections revealed no pathology of glomerulonephritis ( Figure S5B ). In addition, proteinuria levels were similar in both groups ( Figure S5C ) indicating no autoimmune phenotype in these mice.

To assess the ability of RhoA-deficient B cells to undergo GC reaction and mount immune responses, we immunized RhoA^fl/fl Cd21 ^+/Cre and RhoA^fl/fl Cd21 ^+/+ control mice with 4-hydroxy-3-nitrophenylacetyl (24)-keyhole limpet hemocyanin (NP(24)KLH) and analyzed the serum titre of NP specific IgM and IgG antibodies. Although IgM did not show an alteration (except for a reduction at day 7, Figure 8A ), RhoA^fl/fl Cd21 ^+/Cre mice showed significantly reduced IgG titres at days 7,14 and 21 after the first immunization compared to control mice ( Figure 8B ). After a second injection at day 21, NP specific IgG titer increased in RhoA^fl/fl Cd21 ^+/Cre mice at day 28 but to a lesser extent than control mice ( Figure 8B ). Also, it declines at a much faster rate in RhoA^fl/fl Cd21 ^+/Cre than control mice ( Figure 8C ) indicating that RhoA-deficient mice fail to mount long-lived humoral immunity.

So far, we showed that RHOA interacts with SLP65 and is directly involved in B cell differentiation at earlier as well as later developmental stages. RHOA has been shown to promote the reorganization of the actin cytoskeleton thereby regulating cell shape, attachment, and motility (22). It has previously shown that a constitutive activity of the BCR as well as altered cytoskeleton structure is associated with B cell-derived neoplasms, such as the activated B cell subtype of diffuse large B cell lymphoma (DLBCL, ABC) or chronic lymphocytic leukemia (CLL) (52). Interestingly, increased levels of RHOA were reported for CD5⁺ B cells (53) and therefore we hypothesized that RHOA may play a role in the development and survival of CLL B cells.

To check this, we studied the effect of RhoA deletion in DG75EB/HA-RhoA cell line by inactivating the RhoA gene using CRISPR/Cas9 genome editing ( Figures S6A, B ). Induction of RhoA deletion in these cells ( Figure S6C ) by removing doxycycline from the culture media resulted in an increased cell size ( Figure S6D ). Interestingly, RhoA deletion resulted in up-regulation of RAC1 and RAC2 indicating that the expression of RAC1/2 proteins is regulated by the amount of RHOA in the cells ( Figure S6C ).

Next, we studied the effect of inducible RhoA deletion in CLL mouse model. Eμ-TCL1 mice (54) were crossed with Mb1^+/Cre-ERT2 strain (55) which expresses a tamoxifen-inducible form of the Cre recombinase under the control of the Mb1 promoter region. Then, the mice were crossed with RhoA^fl/fl mice to generate Eµ-TCL1 RhoA^fl/fl Mb1^+/Cre-ERT2 strain. The mice were maintained for 10 months until they developed the disease which is characterized by accumulation of CLL B cells (CD19⁺CD5^low). After 3 weeks of tamoxifen-induced RhoA deletion, TCL1-transgenic mice had significantly smaller spleens as compared to untreated mice ( Figure 9A ). In addition to this, RhoA deletion also led to the elimination of CLL-B cells in spleen and bone marrow ( Figures 9B, C ) and significantly improved the survival rate of the mice ( Figure 9D ). This suggests that RHOA is essential for the development and maintenance of CLL-B cells.

Freshly isolated bone marrow cells from WT, RhoA ^fl/fl and RhoA ^fl/fl mb1 +/hcre mice were cultured in Iscove’s medium (Biochrom AG) containing 10% heat-inactivated fetal calf serum (FCS; Sigma-Aldrich) while triple knock out (TKO) pro B cell line reconstituted with an expression vector for the pre-BCR and inducible form of SLP-65 cells were cultured in Iscove’s medium containing 5% heat-inactivated FCS. The media were supplemented with a supernatant of J558L plasmacytoma cells stably transfected with a vector encoding murine IL7. Phoenix cells and mature splenic cells were cultured in Iscove’s medium containing 10% heat-inactivated FCS without IL-7. All the above-mentioned media were supplemented with 100 U/ml penicillin (Gibco), 2mM L-glutamine, 100 U/ml streptomycin (Gibco), and 50 μM 2-mercaptoethanol (Gibco).

DG75EB/HA-RhoA cells are derived from human Burkitt lymphoma cell line DG75 (64). The cells were stably transfected to express the murine cationic amino acid transporter 1 (SLC7A1) to make them susceptible for infection with MMLV-based retrovirus particles (65). These modified cells are termed as DG75EB. To obtain HA-RhoA-expressing DG75EB cells, we retrovirally infected the cells with pRetroX-OneTet-Pruo (Clontech) encoding HA tagged RHOA. Upon treatment with Doxycycline, the cells express HA-RHOA. These cells were cultured in RPMI medium containing 10% heat-inactivated FCS (PAN Biotech) and 1 mg/ml Doxycycline (Sigma Aldrich).

For cloning of CRISPR/Cas-based genome editing constructs, the pSpCas9(BB)-2A-GFP vector was used (Addgene Cataloug number 48138). Design of guide RNA was performed by using the CRISPR/Cas Design software (http://crispr.mit.edu/) against human RHOA Exon2, (sgRNA: GAA CTA TGT GGC AGA TAT CGA GG, Score: 85). Oligos for cloning were: 5’-CAC CGG AAC TAT GTG GCA GAT ATC G-3’ and 5’-AAA CCG ATA TCT GCC ACA TAG TTC C-3’. EcoRV was used for activity screening on genomic DNA of transfected cells. Primer sequences used to construct HA-RHOA were: forward 5’- AAT TAC GTT GCT GAC ATA GAA GTG GAT GGA AAG CAG GTA GAG TTG GCT-3’ and reverse 5’- CCA TCC ACT TCT ATG TCA GCA ACG TAA TTC TCA AAC ACT GTG GGC ACA TAC ACC TC-3’.

For our study, we used cells cultured in presence of doxycycline to have intact RhoA expression (referred to as: DG75EB/HA-RhoA). For experiment where we induced RhoA deletion by removal of Dox, the cells are referred to as: DG75EB/HA-RhoA/RhoA-KO.

Full-length cDNA encoding murine RhoA was subcloned into the retroviral vector backbone of pMIG to generate a RhoA-IRES-GFP expression vector. Plasmids for the expression of Foxo1, Pten and tamoxifen-inducible form of Cre (Cre-ERT2) are described previously (13, 46). Viral supernatants were generated using the Phoenix retroviral producer cell line as described in the manufacturer’s instructions. In summary Phoenix, cells were cultured in Iscove’s culture medium + 10% FCS. Cells were plated at a density of 0.25*10⁶ cells/mL to generate supernatants by using the transfection reagent GeneJuice (Merck Millipore). Retroviral supernatants were harvested after 48 hr. For the subsequent transduction, the respective cells were mixed with supernatants and centrifuged for 3 hours at 300g at 37°C. Transduced cells were cultured for 2-3 days and analyzed by flow cytometry. For reconstitution experiments, single-cell suspensions were prepared from the freshly isolated bone marrow of RhoA ^fl/fl and RhoA ^fl/fl mb1 ^+/hcre mice. CD19⁺ cells were isolated from this single-cell suspension using the CD19 Microbeads based on the positive selection method (Miltenyi biotech) in an AutoMACS Pro separator. Sorted cells were cultured in Iscove’s medium containing 10% heat-inactivated FCS and IL7. After 3 days, cells were retrovirally transduced with an expression vector for RhoA, Foxo1, Pten or empty vector alone as the control. Stable growing cell lines of RhoA ^fl/fl bm-derived pre-B cells were retrovirally transduced with tamoxifen-inducible Cre-recombinase (ERT2-Cre) or empty control vectors and subsequently selected by addition of puromycin. Cre-recombinase was activated by the addition of 1- 2 μM 4-hydroxytamoxifen (4-OHT, Sigma Aldrich). As control cells were treated with EtOH (solvent of 4-OHT). RhoA deletion was analyzed 72 hours after 4-OHT inductions by FACS analysis, PCR or immunoblotting. The following primers were used to detect RhoA by PCR after deletion; m-RhoA-F TGC CAT CAG GAA GAA ACT CGT and m-RhoA-R CAA GAT GAG GCA CCC AGA CTT. The size of the respective band is 574 bp. Actin with a size of 400 bp was used as the loading control and detected using the following primers; m-Actin-F GATCACTATTGGCAACGAGC, m-Actin-R ACGCAGCTCAGTAACAGTCC

BM cells from RhoA^fl/fl Mb1^fl/fl or RhoA^fl/fl Mb1^Cre/Cre mice were cultured for 3-7 days in Iscove’s medium containing 10% heat-inactivated FCS, 2 mM L-glutamine, 100 U/ml penicillin/streptomycin, and 50µM 2-mercaptoethanol. The medium was also supplemented with the supernatant of J558L plasmacytoma cells stably transfected with a vector encoding murine IL7. The pro-/pre-B-cells were retrovirally transformed with a pMIG vector expressing BCR-ABL1. Transformed cells were selected by IL7 withdrawal (48).

Measurement of Ca²⁺ mobilization was previously described (45). Briefly, a total of 1 x10⁶ ER^T2-SLP65, pre-BCR positive TKO cells were loaded with Indo-1 AM (Invitrogen) using Pluronic F27 (Invitrogen). ER^T2-SLP65 function was induced by the addition of 2 µM 4-OHT or the respective amount of its solvent Ethanol (Et) as control. Pre-BCR expression and Ca2+ mobilization was acquired at a FACS LSR Fortessa flow cytometer (BD).

All mouse housing, breeding, and surgical procedures were approved by the governmental institutions of Baden-Württemberg (Regierungspräsidium Tübingen).

For PLA experiments (68) with SLP65, pSLP65, RHOA and GEF-H1 the corresponding antibodies were used (RHOA (26C4) santa cruz sc-418; BLNK (2B11) santa cruz sc-8003; Phospho-BLNK (Thr152) (E4P2P) Cell Signaling #62144; GEF-H1 (55B6) Cell Signaling #4076). The antibodies for RHOA and BLNK were labeled directly using the Duolink in situ Probemaker plus/minus reagent (Sigma-Aldrich). GEF-H1 and pSLP65 were detected by the corresponding secondary antibody Duolink in situ PLA Probe Anti-Rabbit plus/minus (Sigma-Aldrich).

The PLA probes were then subjected to ligation and polymerization reactions using the Duolink in situ detection reagents orange (Sigma-Aldrich). In the end, the cells were examined for the frequency of signals per cell under the fluorescence microscope (Leica). Pictures were taken and quantified by Image J software.

TKO cells reconstituted with an expression vector for the pre-BCR and inducible form of SLP65 was used for the study. Cells were serum starved for 48 hours and were either stimulated with 1µM 4-OHT or its control EtOH for different time points. After stimulation cells were then snap-frozen in liquid nitrogen and GTP bound RHOA level was measured using GLISA-Kit (Cytoskeleton, Inc) according to manufacture’s protocol.

Bone marrow cells derived from wild type mice were cultured in Iscove’s medium supplemented with IL7 for 5 days. Cells were serum starved for 3 hours before treating with different concentrations of RHO Inhibitor (CT04, Cytoskeleton, Inc). After 16 hours cells were collected, washed and lysed with RIPA buffer containing Protease and Phosphatase inhibitor cocktail. The amounts of protein in the lysates were determined using Bradford assay and proteins of interest were detected using immunoblotting.

DG75EB/HA-RhoA cells were serum starved for 30 mins and stimulated for 0,1,3,5 and 10 mins with 10 µg/ml anti-human IgM F(ab)’2 (Southern biotech) at 37°C in a thermomixer. Reactions were stopped by adding ice-cold PBS and cells were lysed using co-immunoprecipitation lysis buffer (I mM EDTA, 50 mM Tris pH-7.5, 150 mM NaCl, 0.1% Triton X-100 together with Protease and Phosphatase inhibitor cocktail). Immunoprecipitation of cell lysates was performed using Protein G Sepharose 4 Fast Flow 50% gel suspension (GE health care) with an anti-BLNK antibody according to the manufacture’s instructions. Precipitated proteins were detected using immunoblotting.

Equal amounts of denatured protein were separated on a 10% gel by SDS-electrophoresis and transferred onto a PVDF membrane. The membrane was incubated for 1 hour with blocking buffer (TBS/0.1% Tween-20; Sigma) with 5% BSA (Serva) and incubated overnight with the primary antibody diluted in blocking solution. Antibodies used for immunoblotting are listed in Table 2 .

After washing steps, the membrane was incubated for 1 hour at room temperature (RT) with horseradish peroxidase (HRP)-coupled secondary antibodies diluted in blocking buffer. Membranes were washed off to remove excess antibody and stained proteins were detected with an ECL Ultra Solution kit (Lumigen) on a Fusion SL advanced imaging system (Vilber Lourmat).

SLP65-deficient pre-B cells were retrovirally reconstituted with ER^T2-SLP65 and were then induced with 4-OHT for 30 min with or without Rho-specific inhibitor C3-toxin (RHOA I). Cells treated with vehicle were used as a control. Cells were then fixed with 2% paraformaldehyde and labeled with DyLight 488 conjugated anti-PTEN antibody and DAPI staining for immunofluorescent microscopy. Images were captured using a Zeiss LSM 780 with GaAsP detectors laser scanning confocal microscope with a 63× 1.4-NA Plan-Apochromat oil immersion objective. Images were taken at z-sections (15–20 sections) of 0.5μm intervals by using the 488nm Argon laser and 405nm diode laser for Alexa 488 and DAPI visualization, respectively. To avoid bleed-through effects in double triple-staining experiments, each dye was scanned independently in a multitracking mode, with the emission of DAPI being collected between 410-490nm and of Alexa 488 between 500-550nm. Clustered PTEN and 3D surface rendering, was determined using Imaris software (Bitplane, Inc.).

For flow cytometric analysis, cell suspensions were pretreated with anti-CD16/CD32 Fc-Block (2,4G2; BD Biosciences) and stained by standard procedures. Dead cells were excluded by staining with Fixable Viability Dye eFluor780 (eBioscience). Cells were stained by using antibodies enlisted in Table 3 .

Intracellular staining for flow cytometry was performed after surface staining by using the True-Nuclear Transcription Buffer Staining Set (BioLegend) according to the manufacturer’s instructions.

Cells were acquired at a FACSCantoII flow cytometer (BD Biosciences). If not indicated otherwise, numbers in the dot plots indicate the percentages of cells in the respective gates, numbers in the histograms indicate the mean fluorescence intensity (MFI).

Total RNA was isolated using ReliaPrep™ RNA Cell Miniprep System (Promega), and PicopureTM RNA isolation kit (Arcturus). cDNA was synthesized using High capacity RNA to cDNA kit (Applied Biosystems) and RevertAid Reverse Transcriptase kit (Thermo Fischer). Quantitative real-time PCR (qRT-PCR) analyses were performed using TaqMan PCR probe (GAPDH: Mm99999915_g1, RAG1: Mm01270936_m1, RAG2: Mm01270938_m1, Applied Biosystems) with TaqMan gene expression mastermix (Applied Biosystems). qRT-PCR data were acquired on a StepOnePlus Real-Time thermocycler (Applied Biosystems) and analyzed with the StepOne Software version 2.3. Relative quantification (RQ) was calculated using the 2^-ΔCCT equation.

For cryosections, spleens were embedded in O.C.T.-compound (SAKURA) and frozen at -80°C. Five μm sections were prepared using a cryo-microtome (Reichert-Jung 2800 Frigocut) with a S35 knife (Feather) and fixed on SuperfrostPLUS slides (Thermo Scientific) by treatment with pure acetone. Prior to staining, the sections were rehydrated with PBS + 2% BSA + 0.1% sodium azide and blocked with Fc-Block (anti-CD16/32; BD Biosciences). The section was mounted with Fluoromount-G (Southern Biotech) after staining with FITC conjugated anti-CD169 (MOMA1, AbD Serotec) and Cy5 conjugated anti-IgM (Jackson Immunoresearch) antibodies. The staining was detected using a DMi8 fluorescence microscope (Leica).

Kidneys from RhoA^fl/fl or RhoA^+/fl (N=3) and RhoA^fl/fl Cd21^+/Cre (N=6) mice of age 33-49 were embedded in O.C.T.-compound (SAKURA), shock frozen in liquid nitrogen and stored at -80°C. Two µm sections were prepared using a cryo-microtome (Leica cryostat CM1950) with a C35 microtome blade (Feather). Sections were mounted on SuperfrostPLUS slides (Thermo Scientific) and fixed within pure acetone (Sigma Aldrich) for 9 min. Sections were stained with HE stain (Waldeck GmbH & Co.) and PAS stain (Sigma Aldrich). The slides were blinded and the sections were examined independently by two senior pathologists for pathological changes in the general cellularity and architectural features of the kidney.

Urine samples were collected from mice (age 23-48 weeks) of the indicated genotype and proteinuria levels were measured using Combur-Test ^® strips (Roche).

One million cells/animal were injected intravenously into black6 γc ^-/- Rag2 ^-/- mice and leukemia engraftment was followed by detection of GFP⁺ cells in the peripheral blood via flow cytometry analysis. Animals were sacrificed after 14 days upon showing clinical signs of leukemia including loss of activity, organomegaly, and hindlimb paralysis. Leukemic infiltration of the murine CNS was assessed using histological sections in blinded experiments (49).

RhoA^fl/fl (N=7) and RhoA^fl/fl Cd21^+/Cre (N=8) mice of 6-8 weeks old were immunized with 100 µg NP(24)KLH (Biosearch Technologies California, N-5060-5) + 100µg CpG-ODN1826 (Biomer, custom design) in PBS or with 100µg CpG-ODN1826 (control immunization). Sera were collected at days 7, 14 and 21 post-primary immunization (day 0) and after 7 and 35 days after booster immunization (day 21). Then IgM and IgG antibodies to NP were measured by NP-BSA (Biosearch Technologies California, N-5050L-10) ELISA fitted concentration in arbitrary units (AU) according to a standard (SouthernBoitech). Mice from which no blood were withdrawn were excluded from the analysis. The decline of anti-NP-IgG titers was calculated as the ratio of IgG titers at day 28 relative to day 56.

100 µl of mouse blood samples were collected from living mice by a transverse incision through the major tail vein, and incubated in an Eppendorf tube containing heparin, the samples were centrifuged for 15min at 1400 rpm at 4°C. For the anti-IgM and anti-IgG ELISAs, 96-well plates (NUNC, maxisorp) were coated with polyclonal anti-mouse IgM or IgG antibody (SouthernBiotech) and blocked with buffer containing 1% BSA. Dilutions of anti-mouse IgM/IgG (SouthernBiotech) were used as standard. The concentration of IgM and IgG in the supernatants was determined by detection with alkaline-phosphatase-labeled anti-mouse IgM or IgG (Southern Biotech), respectively. P-nitrophenylphosphate (Genaxxon) in diethylamine buffer was added and data were acquired at 405 nm using a Multiskan FC ELISA plate reader (Thermo Scientific). To determine the content of autoreactive anti-dsDNA specific antibodies, concentration-adjusted cell culture supernatants were applied to plates coated with calf thymus DNA (Rockland Immunochemicals).

5 × 10⁵ cells were cultured on the top chamber of a transwell culture insert (Corning, pore size 5 μm) and were allowed to migrate toward media containing 100 ng/ml CXCL12 (ImmunoTools) for 16 h. The total cell number in the lower chamber was determined using hematocytometer.

The proliferation dye (eFluor 670; eBioscience) was used to label cells which were cultured in optimum conditions as described previously (69). The percentage of proliferating cells were determined by flow cytometry after 72 hours.

Statistical tests are indicated in the figure legends. Results were analyzed for statistical significance with GraphPad Prism 8.3.0 software or SPSS (v 24.0.0.2). A p value of <0.0500 was considered significant (*p<0.005, **p < 0.001, ***p < 0.001, ****p < 0.001). In vitro panels are representative of at least 3 independent experiments, unless mentioned otherwise.