PexRAP promotes GC response.

(A) Schematic of immunization with SRBC after inactivation of Dhrs7b in adult mice. Mice (Rosa26-CreERT2; Dhrs7b+/+, or Rosa26-CreERT2; Dhrs7bff) were treated with tamoxifen, immunized with SRBC, and harvested 1 wk after immunization, as described in Materials and Methods. (B) Deletion efficiency of Dhrs7b conditional alleles. B and T lymphocytes were isolated from spleens of Rosa26-CreERT2 or Dhrs7bf/f;Rosa26-CreERT2 mice after in vivo tamoxifen injections followed by immunization with SRBC. WT and Dhrs7bΔ/Δwere analyzed by immunoblotting with antibodies directed against PexRAP (protein product of Dhrs7b) and actin (internal loading control). (C, D) PexRAP acutely regulates B cell numbers. Representative flow plots of viable splenic B cells (C) and aggregate data for three biologically independent replicate experiments (D) (n = 8 WT and 7 cKO). For data on T cells, see supplemental data Fig. S1A-C. (E, F) Effect of PexRAP on GC B cell response. Flow plots of GL7+ CD95+ GC B cells among viable B cells (E), and aggregated frequencies and numbers of GC B cells, as indicated, in the three replicate experiments (F). (G, H) PexRAP impact on GC response. After tamoxifen injections, Dhrs7b deficient mice [Rosa26-CreERT2; Dhrs7b/ff (46)] and WT controls (Rosa26-CreERT2) were immunized with SRBC and analyzed 7 d later, as in Fig. 1A-F. Data on gating strategy, GC counts per unit area and LZ area are in Supplemental data Fig. 1D-F. (G, H) Shown are representative images from immunofluorescence staining of spleens with the indicated Ab in two independent experiments (5 WT vs 4 Dhrs7b cKO), showing a low-power overview with many follicles (G) and a representative higher-magnification image to better delineate primary and secondary follicles (H). (I) Quantitation of number (left panel) and size (right panel) of GC in spleen sections, with GL7-positive, IgD-negative areas that include both CD35+ (LZ) and CD35- (DZ) areas. Mann-Whitney U test was used to calculate p values.

PexRAP promotes proliferation of B cells.

(A) Schematic of Dhrs7b inactivation in mature mice via tamoxifen treatment and a B cell type-specific conditional allele. Mice (huCD20-CreERT2; Dhrs7b+/+, or huCD20-CreERT2; Dhrs7bff) were injected with tamoxifen (d1, 3, 5) and harvested at day 10. (B) Deletion efficiency of conditional Dhrs7b alleles. After in vivo tamoxifen injections, B cells were isolated from spleens of hCD20-CreERT2 or Dhrs7bf/f; hCD20-CreERT2 mice, as indicated, and analyzed by immunoblotting as in Fig. 1B. (C) PexRAP and the maintenance of B cells. Shown are the frequencies and the numbers of CD19+ B220+ B cells among viable lymphocytes in spleen (left and right panels, respectively). Data are pooled from four independent replicate experiments (n = 12 WT and 12 cKO). Shown in feather plots are the means, with whiskers that extend to the minimum and maximum values and boxes that outline upper and lower quartile values with the midline identifying the median.(D-F) PexRAP regulates B cell proliferation in vivo. CTV-labeled B cells were adoptively transferred into µMT recipient mice and analyzed 4 days thereafter. Shown are the frequencies of B220+ CD19+ events among splenocytes in the viable cell gate (D), along with representative flow plots of CTV partitioning and surface IgD in B cell gates (E) and aggregated frequencies of divided B cells from five independent replicate experiments (F) (n =10 WT and 11 cKO). Additional data on the IgDneg population are in supplemental Fig.1J. (G-I) PexRAP promotes B cell proliferation in vitro. After in vivo tamoxifen injections, bead-purified B cells from spleens of CreERT2 mice (WT and Dhrs7bΔ/Δ , i.e., cKO) were stained with Cell Trace Violet (CTV), activated and cultured 4 days in anti-CD40, BAFF, IL-4, IL-5 and 4-hydroxytamoxifen in three biologically independent experiments totaling 6 WT and 7 cKO mice). (G) Representative flow-cytometric analysis of CTV partitioning. (H) Quantified frequencies of divided >3 times. (I) Numbers of B cells recovered at the end of the cultures.

B cell expression of PexRAP is essential for normal concentrations of most ether phospholipids.

Splenic B cells (CreERT2+ or CreERT2+, Dhrs7b Δ/Δ) from tamoxifen-injected mice were analyzed by LC-MS-MS either directly ex vivo (“Naive”) or after activation and culture (48 h) (“Activated”) as in Fig. 2. Shown are z-scored heat maps for the relative abundance of subsets of ether and plasmalogen phospholipids and lysophospholipids identified by exact mass and secondary fragmentation in (A) positive and/or (B) negative ion modes. More species are displayed in Supplemental Fig. 2. (C) Quantitated peak areas for activation-induced increases (’fold-induction’ ratios of activated/resting, plotted on log10 scale) for the indicated ether, plasmalogen, and diacyl (non-ether) phospholipids in the PexRAP-sufficient and -depleted B cells. * denotes species for which P<0.05 for the effect of Dhrs7b inactivation. (D) Shown are the raw values of mean peak areas for the indicated species in the freshly purified and the activated B cells of the indicated genotypes, as indicated. (Three independent samples from individual mice of each genotype were analyzed.) * p<0.05, ** p<0.01 by unpaired Student’s. t-test. Additional data including controls comparing B cells of unmanipulated B6 mice to those expressing CreERT2 and injected with tamoxifen are in Supplemental Fig. 2.

PexRAP is essential for normal concentrations and distributions of some ether phospholipids in splenic follicles.

(A) Schematic of immunization for IMS analyses. Mice of the indicated genotypes were treated with tamoxifen, immunized with SRBC, and harvested 1 wk after immunization. (B) Schematic diagram illustrating the 2D-IMS analysis work flow, image merging and quantitation. Harvested spleens were used to generate serial tissue sections followed by immunofluorescence (IF) staining of one section and IMS analysis with the adjacent one. IF and IMS images were aligned to map ion intensity distributions to microanatomic regions (B cell follicle and GC). The intensities of specific ions on B cell follicles and GC regions were quantitated as described in Materials and Methods. (C) Identification of ether lipid species localizing to lymphoid follicles. Representative ion images of two ions [m/z 752.5545, and m/z 776.5556] with spleens from immunized mice (WT and Dhrs7bΔ/Δ). (D) Identification of ether lipid species localizing to lymphoid follicles. Representative ion images of three ions [m/z 752.5545, m/z 872.5749, and m/z 700.4958] in mass spectrometry imaging of spleens from WT and Dhrs7bΔ/Δ−B mice. Immunofluorescent images at higher magnification, delineating LZ and DZ marked by CD35 staining, along with quantification of sizes of GC and their LZ and DZ, are shown in supplemental data Fig. S4A, B. (E) Shown in the bar graphs are the mean (±SEM) ion intensities in primary lymphoid follicles (B cell zones) and GC from spleens of WT and Dhrs7bΔ/Δ-B mice, immunized or not (“UI”) as indicated. Median intensity of each ion was obtained from 3 follicles / spleen and 3 GC / spleen from WT and Dhrs7bΔ/Δ-B mice (three biological replications comprising 7 WT and 6 cKO spleens). P values were calculated by Mann-Whitney U test.

Immunization-induced increases of selected lipids in GC

Impact of PexRAP on relative quantities of selected lipid species in primary and secondary follicles (GC).

B cell intrinsic role of PexRAP in GC response.

(A) Schematic of immunization with NP-OVA in alum after inactivation of Dhrs7b in B lineage cells. Tamoxifen-treated WT (hCD20-CreERT2; Dhrs7b+/+) or Dhrs7bΔ/Δ-B mice (hCD20-CreERT2; Dhrs7bff) were immunized with NP-OVA, boosted with NP-OVA 3 wk thereafter, and harvested 1 wk after the 2nd immunization. (B, C) Representative flow plots of splenic IgDneg B cells (B) and aggregate data (C) for two replicate experiments (n = 5 WT and 5 cKO). The cocktail of reagents for the dump channel included anti-IgD. (D, E) B cell-intrinsic function of PexRAP in GC B cell response. Representative flow plots of GL7+ CD95+ GC B cells among viable IgDneg B cells (D), and aggregated frequencies and numbers of GC B cells for two replicate experiments (E). Mann-Whitney U test was used to calculate p values.

Ab response and affinity increase promoted by PexRAP in B cells.

Tamoxifen-treated mice (hCD20-CreERT2; Dhrs7b+/+ or hCD20-CreERT2; Dhrs7bff, i.e., Dhrs7bΔ/Δ-B) were immunized as in Fig. 5, with venous blood collected to measure Ag-specific Ab just prior to a second immunization, followed by harvesting a week thereafter. (A-D) Ag-specific Ab in primary response sera, prior to the boost. Shown are the all-(NP20) and high- (NP2) -affinity anti-NP IgM (A, B) and IgG1 (C, D), as indicated, with NP20 a high hapten density to detect both low- and high-affinity Ab and NP2 a low hapten density selective for high-affinity Ab. (E, F) Levels of anti-NP IgM detected using NP20 and NP2 for ELISpot (F) and ELISA (G) as described in Materials and Methods. Graphs show mean (± SEM) number of Ab-secreting cells in spleen (E) and (F) mean (± SEM) OD450nm in measurement of NP-specific IgM in serial dilutions of sera. P values were calculated by students’ t-test. * indicates p<0.05, and ** indicates p<0.01. (G) PexRAP regulates Ab affinity maturation. The bar graph shows the mean (± SEM) ratios of high-affinity to all-affinity NP-specific IgM Ab in sera of individual mice (each dot representing one subject) using OD450nm values at the 1:2000 dilution, with data from three independent experiments comprising 8 mice of each type (WT; Dhrs7bΔ/Δ−B). P values were calculated by Mann-Whitney U test. (H, I) PexRAP promotes CD138+ cell differentiation in vivo. B cells were adoptively transferred into µMT recipient mice and analyzed at 4 days after transfer. Representative flow plot of CD138 and cD19 expression (H) and aggregated frequencies of CD138+ CD19+ cells in the viable cell gate (I) from four independent replicate experiments. To test for a potential distortion arising from outlier values, statistical testing was performed both with and without their inclusion.

Function of PexRAP in GC response.

(A) Schematic of the time line of immunization followed later by tamoxifen injections into S1pr2-CreERT2 mice (Dhrs7b+/+or Dhrs7bf/f). Note that deletion is initiated after the germinal center reaction starts (∼ 3.5 d post-immunization). Mice (S1pr2-CreERT2; Dhrs7b+/+ or S1pr2-CreERT2; Dhrs7bf/f) were immunized with NP-OVA, treated with tamoxifen on days 3, 5, and 7 after NP-OVA immunization (36) and harvested at day 9. (B) Representative flow plots of GL7+CD95+ GC B cells among viable B cells and (C) aggregated mean (±SEM) frequencies of GC B cells from three replicate experiments (7 WT; 9 cKO mice). Mann-Whitney U test was used to calculate p values. (D, E) Effect of PexRAP on the level of plasmablasts in spleen. Representative flow plots of B220+ CD19+ plasmablasts among the events in a dump- CD138+ TACI+ gate (D) and aggregated mean (±SEM) frequencies of these plasmablasts (E).

PexRAP contributes to ROS homeostasis and B cell population growth in vitro.

(A-D) PexRAP is critical for maintenance of normal ROS levels. Bead-purified B cells from spleens of tamoxifen-treated huCD20-CreERT2 mice (Dhrs7bΔ/Δ-B and Dhrs7b+/+) were cultured 3 days in anti-CD40, BAFF, IL-4, IL-5 and 4-hydroxytamoxifen. Total cellular (A, B) and mitochondrial ROS, mtROS (C, D) in B lymphoblasts were determined by flow cytometry after staining with surface markers and H2DCFDA and MitoSOX, as described in the Methods. Representative histogram image of H2DCFDA (A) and MitoSOX (C) in the B cell gate, and aggregated mean (± SEM) geometric MFI of H2DCFDA (B) and MitoSOX (D) from 3 independent experiments, each using two mice of each type (6 WT; 6 cKO). P values were calculated by Mann-Whitney U test. (E) PexRAP restrains lipid peroxidation. B cells were activated and cultured as in (A). A representative result of flow cytometric analyses of lipid peroxidation assayed using C11-Bodipy is shown, based on three independent experiments. (F, G) PexRAP promotes B cell survival. (F) Shown are the mean (±SEM) frequencies of total viable ’events’ (by FSC, SSC, and 7-AAD exclusion) in flow cytometry (filled circles) and MFI of Bodipy-C11 (open squares) after exposure to H2O2 (200 µM). (*, ** -p=0.03 and 0.003, respectively). (G-I) WT and Dhrs7bΔ/Δ B cells were cultured (2 d) in anti-CD40, BAFF, IL-4, IL-5 and 4-hydroxytamoxifen. (G) Increased cleaved caspase 3 (CC3) in PexRAP-deficient cells generated in vitro. Cleaved caspase 3 in B cells was detected by intracellular staining and flow cytometry. Shown are a representative pair of histograms for WT and Dhrs7bΔ/Δ B cells as indicated (left panel) and a dot graph aggregating all experiments’ outcomes (right panel). (H) Frequencies of apoptotic B cells in cultures as in (A-D) were scored for annexin V and 7AAD as described (84). Shown are representative data of flow plots in the lymphocyte gate (left panel) and a dot graph aggregating all experiments’ outcomes (right panel). (I, J) PexRAP and ROS control promote B cell population growth in vitro. WT and Dhrs7bΔ/Δ B cells were cultured 5 days in anti-CD40, BAFF, IL-4, IL-5 and 4-hydroxytamoxifen in the presence or absence of ROS scavenger NAC (1 mM vs 5 mM) (H). (I) B cells activated as in (A-G) were cultured 5 d in the presence or absence of NAC (1 mM) or thioridazine (100 nM). The bar graphs show the mean (± SEM) recovered cell number from three independent experiments, each with two independent samples of each genotype (WT; Dhrs7bΔ/Δ).

PexRAP deficiency in activated B cells reduces mitochondrial metabolism and ER mass.

(A-G) Purified B cells were activated with and cultured (2 d) in αCD40, BAFF, IL-4, IL-5, and 4-OHT, then divided and analyzed using a Seahorse XFe96. (A) Shown are aggregated results from three independent experiments measuring oxygen consumption rate (OCR) quantified via metabolic flux analysis during mitochondrial stress testing. (B) Basal OCR, (C) maximal OCR, (D) ATP-production coupled respiration, (E) proton leak, and (F) spare respiratory capacity of WT and Dhrs7bΔ/Δ B cells assayed in (A). Data points of individual samples are show (each individual dot), as well as bars to display mean (±SEM) values. T-tests with Welch’s correction were used to calculate p-values. *, p<0.05; ***, p<0.001. (G) Extracellular acidification rate (ECAR) of B cells cultured as described previously (36) and assessed using the glycolytic stress test. (H, I) PexRAP influences ER mass. WT and Dhrs7bΔ/Δ B cells were cultured 5 days in anti-CD40, BAFF, IL-4, IL-5 and 4-hydroxytamoxifen followed by the staining with ER-Tracker Green. Representative histogram of ER-Tracker Green in viable cells (H) and aggregated MFI (±SEM) of ER-Tracker Green from three replicate experiments (I). Mann-Whitney U test was used to calculate p values.

PexRAP is dispensable for T cell homeostasis and does not affect balance among pre-immune B cell subsets.

Mice (Rosa26-CreERT2; Dhrs7b+/+, or Rosa26-CreERT2; Dhrs7bf/f) were treated with tamoxifen and immunized with SRBC as described in Materials and Methods, and 1 wk after immunization. (A-C) PexRAP is dispensable for homeostatic maintenance of T cells. Shown are the representative flow plots of splenic T cells (A), mean (± SEM) frequencies of TCRβ+ CD4+ T cells (B), and CD8 (TCRβ+ CD4-) T cells (C) from WT and cKO mice in three replicate experiments totaling 8 WT and 7 cKO mice. (D) A representative gating scheme illustrating the flow-cytometric determination of the GC B cell population in mice. (E) Numbers of GC identified by microscopy of immunofluorescently stained sections of immunized mice as in Fig. 1. (F) A representative gating scheme illustrating the flow-cytometric estimation of the MZ, FO, and B1 B cell populations in mice. (G) Shown are the mean (±SEM) frequencies of CD19+ IgM+ CD23- MZB cells (left panel) and CD19+ IgM-CD23+ FOB cells (righthand panel) in spleen, and (H) CD19+ CD23lo CD43+ CD5+ B1B cells in peritoneal cavity. (I) In vivo generation of IgD- progeny from four independent replicate experiments (n =8 WT and 9 cKO). Shown are mean (±SEM) frequencies of IgD- B cells as aggregate data for the frequencies of IgD-events in the gate for viable B cells recovered from µMT recipient mice after transfer of WT or PexRAP-depleted B cells (experiments of Fig. 2D-F), as indicated, with representative flow plots in Fig. 2E.

Activation of B cells induces PexRAP-dependent increases in their ether and plasmalogen P-lipids.

As in Fig. 3, B lymphocytes of the indicated Dhrs7b genotypes were prepared from individual tamoxifen-injected mice and divided to analyze without activation (“naive”) or after activation and culture (48 h) (“activated”). Shown are heat maps with more extensive lists of ether phospholipids and plasmalogens identified in the LC-MS-MS analyses for both genotypes in positive (A) and negative (B) ion modes. (C, D) LC-MS-MS results for lipids in B cells, activated or not, as indicated, of conventional non-transgenic B6 mice not injected with tamoxifen and those of tamoxifen-injected CreERT2+ mice (the “+/+“ samples of panels A, B and of Figure 3). Bar graphs in (C) and (D) show results of these control comparisons for some of the phospholipids analyzed in prior figures panels (Fig. 3A, B).

Immunization induces increases in P-lipids of both primary and secondary follicles dependent on PexRAP in B cells.

(A, B) Selected representative m/z features Spleens of AID-GFP were harvested at 7 days after immunization with SRBC and analyzed for exact mass signatures of ether lipid species in (A) negative and (B) positive ion modes. (C) Spleens of immunized mice (tamoxifen-treated hCD20-CreERT2; Dhrs7bf/f and hCD20-CreERT2 controls) were analyzed by IMS. Shown are the positive ions presented in (B) for GC vs follicles of immunized AID-GFP mice. The bar graph shows the mean (± SEM) ion intensities in lymphoid follicles and GC from WT and Dhrs7bΔ/Δ-B spleens. Median intensity of each ion was obtained from three follicles per spleen and three GC per spleen from WT and Dhrs7bΔ/Δ-B mice (three biological replication experiments totaling seven WT and six cKO subjects). P values were calculated by Mann-Whitney U test.

A subset of ether phospholipids in primary follicles and GC (secondary follicles) depend on PexRAP in B cells.

(A) As in Fig. 4, Dhrs7b was inactivated by tamoxifen injections into mice bearing the hCD20-CreERT2 transgene, followed by immunization with SRBC and analysis. Shown are representative higher-magnification images defining the primary (IgD+) and secondary (IgDlo/neg, GL7+) follicles and the identification of light (CD35+) and dark (CD35neg) zones. (B) The sizes of GC and their light zones (LZ) after immunization were quantified. Spleens of immunized mice (tamoxifen-treated hCD20-CreERT2; Dhrs7bf/fand hCD20-CreERT2 controls, as indicated) were analyzed by immunofluorescence staining as in (A). Each dot represents one follicle in a given mouse. Totals of three GC per spleen from immunized mice were used were used, with seven WT and five Dhrs7bΔ/Δ-B. (C, D) Mean (±SEM) absorbances of ELISA performed across the indicated dilutions of sera from the tamoxifen-treated mice (hCD20-CreERT2; Dhrs7b+/+or hCD20-CreERT2; Dhrs7bf/f, i.e., Dhrs7bΔ/Δ-B) of Fig. 6. Shown are the results for (C) all- (NP20 captured) and (D) high- (NP2 captured) -affinity anti-NP Ab of the IgG1 and IgG2c isotypes, as indicated. (E) Proliferation and development of CD138+ progeny. Shown are representative flow plots from three independent experiments with six WT and seven Dhrs7bΔ/Δ-B mice.

Distinct outcomes of heightened ROS elicited by H2O2 versus menadione.

(A-B) Induction of ROS independent from B cell activation. Shown are representative flow plots of DCFDA in the viable B cell gate after exposure (1 h) (A) to H2O2 (200 μM) or (B) menadione (8 μM), as in (87). (C-D) Representative flow plots of annexin V vs 7-AAD for identification of early-apoptotic B cells. Cultured B cells were analyzed 3 hr after treatment with (C) 200 μM H2O2 or (D) 8 μM menadione. Inset numbers represent the fraction of B cells that are Annexin V+ 7-AADneg or Annexin Vneg 7-AAD+. (E-G) Shown are the results of three independent experiments to measure (E) total cell viability, (F) BODIPY C11 MFI and (G) frequencies of Annexin V+ 7-AADneg frequency in B cells treated (8 μM) menadione overnight (E) or for 3 h (D). P-values were calculated using the unpaired Student’s t-test.