Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 11 (1), 723

mTORC1 Coordinates an Immediate Unfolded Protein Response-Related Transcriptome in Activated B Cells Preceding Antibody Secretion

Affiliations

mTORC1 Coordinates an Immediate Unfolded Protein Response-Related Transcriptome in Activated B Cells Preceding Antibody Secretion

Brian T Gaudette et al. Nat Commun.

Abstract

How activated B cells build biosynthetic pathways and organelle structures necessary for subsequent robust antibody secretion is still unclear. The dominant model holds that nascent plasma cells adapt to increased antibody synthesis by activating the unfolded protein response (UPR) under the control of the transcription factor Xbp1. Here, by analyzing gene expression in activated B cells with or without plasma cell-inductive signals, we find that follicular B cells up-regulate a wide array of UPR-affiliated genes before initiating antibody secretion; furthermore, initial transcription of these loci requires the mTORC1 kinase adaptor, Raptor, but not Xbp1. Transcriptomic analyses of resting marginal zone B cells, which generate plasma cells with exceptionally rapid kinetics, reinforce these results by revealing the basal expression of UPR-affiliated mRNA networks without detectable Xbp1 activity. We thus conclude that B cells utilize mTORC1 to prepare for subsequent plasma cell function, before the onset of antibody synthesis.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Control of B cell activation versus differentiation.
a Celltrace violet (CTV) labeled follicular and MZ splenic B cells were purified from B6.Blimp1+/GFP adults by cell sorting, stimulated for the indicated duration with CpG with and without IL-4 and IL-5, and assayed by flow cytometry for GFP expression and CTV dilution. b GFP expression as a function cell division for each cell and condition. c Follicular and MZ B cells from additional B6.Blimp1+/GFP adults were stimulated as in a, then GFP+ and GFP cells that had or had not completed at least one cell division sorted into ELISPOT plates to quantify frequencies of total antibody-secreting cells (ASCs). Data is shown as individual data points, mean(bar) and SEM. d MZ and follicular B cells from a C57BL/6 adult were stimulated with CpG + IL-4,5 as in a for 72 h, then intracellular Xbp1S levels evaluated as a function of CTV dilution. e, f MZ and follicular B cells were stimulated and analyzed as in d except for the additional use of an anti-BiP antibody. Shown are cell fractions back gated based on the indicated division number e, with MFIs graphed as a function of division number in f. g Follicular B cells were stimulated with CpG + IL-4/5 for the indicated times, and expression of the indicated proteins evaluated by flow cytometry and presented as percent of cells increased from baseline expression. Statistic indicates change in consecutive time points. Data is shown as mean and SEM (*p < 0.05, **p < 0.01, ***p < 0.001; repeated measures ANOVA, Tukey HSD post-test). Flow cytometry in a, d and e are representative experiments of at least three experimental replicates. Source data are provided as a Source Data file.
Fig. 2
Fig. 2. Basal UPR gene expression in resting MZ B cells.
a C57BL/6 mouse splenocytes were stained for intracellular Xbp1s. Gating criteria are shown in Supplementary Fig. 1. Shown is a representative histogram and MFI of three animals, which are summarized as the mean (line), 75% CI (box), and 95% CI (whisker) and individual points. One-way ANOVA with Tukey post-test was performed to generate p values. b Intracellular phosphorylated S6 protein staining was performed on follicular (FO) and MZ B cells from mice treated with rapamycin or vehicle control every other day for a total of four treatments. Counts as percent of maximum are displayed with quantification on right. p value is two-tailed Student’s t-test. Summary data for all animals presented as individual data points, mean and SEM. cg RNA-seq was performed on splenic follicular and MZ B cells and short-lived (B220+) and long-lived (B220) BM PCs from five individual B6.Blimp+/GFP adults. c Differentially expressed genes were averaged by group and gene co-regulation was determined by hierarchical clustering by Pearson correlation with a grouping cutoff (k) of six chosen using best of 26 indices by NbClust. Gene expression is averaged by group (n = 5) for clarity and displayed as z score across each row. d Gene ontology clustering enrichment analysis of selected co-expression clusters is shown. Indicated is the founder term for each GO term cluster followed by gene numbers for that term. Bar length indicates the enrichment score for the GO cluster. e Volcano plots showing genes differentially expressed in MZ B cells over follicular B cells for all genes (first panel), UPR hallmark genes, mTORC1-signaling hallmark genes, the top 250 upregulated genes in B220+ BM PCs versus follicular B cells and the top 250 genes upregulated in B220 BM PCs versus follicular B cells. Genes are color coded by adjusted p-value (red = p < 0.05). Adjusted p-value is BH-adjusted (eBayes method—Limma). f Log2 TPM expression magnitude of selected genes is shown as mean (line), 75%CI (box) and 95%CI (whisker) as well as individual data points (jitter). (*p < 0.05, **p < 0.01, ***p < 0.001; differential expression statistic: BH-adjusted p-value—eBayes method—Limma). g GSEA was performed using the BROAD UPR and mTORC1-signaling hallmark gene sets as well as B220+ and B220 upregulated gene sets described in e comparing MZ B and follicular B cells. FDR-q values computed using 1000 geneset permutations. Source data are provided as a Source Data file.
Fig. 3
Fig. 3. Development of a plasma cell-inductive UPR signature.
CTV-labeled splenic follicular B cells were sorted from five B6.Blimp1+/GFP adults, and cultured with CpG with and without IL-4 + IL-5 for 72 h. RNA-seq was performed on cDNA prepared from 4th division GFP/Blimp1 cells treated with CpG or CpG plus IL-4/5, 5th division GFP/Blimp1+ CpG plus IL-4/5 cells, as well as freshly isolated follicular and MZ B cells (Supplementary Fig. 3). a Volcano plots display differential expression of UPR and mTORC1-signaling hallmark gene sets comparing the indicated treatment groups to freshly isolated follicular B cells (group n = 5 animals). Genes are color coded by the differential expression statistic p < 0.05, log2 fold change > 1 (BH-adjusted p-value—eBayes method—Limma). b GSEA was performed using the BROAD UPR and mTORC1-signaling hallmark gene sets comparing indicated treatment groups and freshly isolated follicular B cells. FDR-q values computed using 1000 geneset permutations (group n = 5 animals). c RNA-seq comparing UPR hallmark gene expression in BM B220 plasma cells (PCs) to splenic follicular B cells was used to identify PC-specific UPR genes. The expression of these genes was then assessed across in vitro-activated and differentiated follicular B cells to determine those genes induced by B cell activation (magenta) and those requiring PC-inductive stimulus (cyan). Indeterminate genes (orange) were omitted. Those genes printed in red are contained within both the BROAD UPR and mTORC1 signaling gene sets. d, e Volcano plots display differential expression of gene sets defined in c for indicated treatment groups compared to freshly isolated follicular B cells (group n = 5 animals). Genes are color coded by the differential expression statistic p < 0.05, log2 fold change > 1 (BH-adjusted p-value—eBayes method—Limma).
Fig. 4
Fig. 4. Qualitative differences in activation induced and plasma cell-differentiation induced UPR gene sets.
a Gene sets defined in Fig. 3c were subjected to gene ontology (GO) cluster-enrichment analysis. Biological process, cellular compartment, and molecular function annotation was used to determine clusters of enriched terms. Only those terms with a p value < 0.05 were used in clustering. The term with maximum number of contained genes was chosen as the founder term and the number of genes is shown in the wedge. b Each gene in the gene sets defined in Fig. 3c was annotated by functional category. Wedge indicates the number of genes in the term as part of the total genes in the gene set. c From the RNA-seq dataset described in Fig. 3, the Log2 TPM magnitude of expression of selected genes in each gene set is shown as mean (line), 75%CI (box) and 95%CI (whisker) as well as individual data points (jitter). (*p < .05, **p < .01, ***p < .001; differential expression statistic: BH-adjusted p-value—eBayes method—Limma). PC plasma cell. Source data are provided as a Source Data file.
Fig. 5
Fig. 5. Induction of UPR-related genes in activated B cells in vivo.
B6.Blimp1+/GFP mice were immunized with 50 μg NP-LPS intraperitoneally. Naïve (IgDhigh, Dump[CD4, CD8α, F4/80, Ter199], B220+, NP), antigen-specific activated B cells (IgD, Dump[CD4, CD8α, F4/80, Ter199], GFP, CD138, B220+, NP+), and antigen-specific plasma cells (PCs) (IgD, Dump[CD4, CD8α, F4/80, Ter199], GFP+, CD138+, NP+) were twice sorted for RNA and cDNA preparation and RNA-sequencing. a 1840 differentially expressed genes between the three groups were defined by BH-adjusted p-value < 0.01 and log2 fold change ≥ 3 and hierarchically clustered by Pearson correlation and broken into six clusters as determined by within sum of squares (elbow) method. b Gene ontology clustering enrichment analysis of selected co-expression clusters is shown. Indicated is the founder term for each GO term cluster followed by the number of genes represented in that term. Bar length is an indication of the enrichment score for the GO cluster. c GSEA was performed using the BROAD UPR and mTORC1-signaling hallmark gene sets as well as the B220+ PC gene signature defined in Fig. 2e comparing indicated groups. FDR-q values computed using 1000 geneset permutations (n = 3 animals). d Volcano plots display the differential gene expression between NP-specific B cells and PCs versus IgD+ B cells. Genes in red indicate adjusted p < 0.05 and Log2 fold change at least 0.5. Indicated genesets are overlayed on all genes in gray. e Gene expression for the PC-specific UPR genesets defined in Fig. 3 are shown. Indicated in red are genes found in both the mTORC1 signaling and UPR hallmark genesets.
Fig. 6
Fig. 6. Xbp1-independent early activation of UPR gene expression.
Follicular B cells were prepared from either hCD20-TAM-Cre (n = 3) or Xbp1fl/fl;hCD20-TAM-Cre (n = 5) animals after 10 days of tamoxifen diet. Sorting was performed as in Supplementary Fig. 1e. These cells were then in vitro activated and differentiated as in Fig. 3 using CD138 staining to identify bona fide plasma cells (PCs). a CTV-labeled follicular B cells from the mice described in a were stimulated with CpG + IL-4/5 for 72 h, stained for CD138 expression, and analyzed. All plots show viable cells only. (b, upper panel) CD20-TAM-Cre;Xbp1fl/fl (1–5) and CD20-TAM-Cre;Xbp1+/+ (6–9) female adults from above. Genomic DNA from purified follicular B cells (CD19+, AA4.1, CD23+, CD21/35int) and CD19, TCRβ+ T cells was used to detect WT and floxed Xbp1 by PCR. Genotypes shown at right. Note loss of PCR product in CD19+ CD20-TAM-Cre;Xbp1fl/fl cells. Lanes 10 and 11 are PCR products from control tail DNA from Xbp1f and C57BL6 mice, respectively. (b, lower panel) Alignment tracks of RNA-seq reads at the Xbp1 locus from Blimp CpG + IL-4,5-stimulated follicular B cells from the indicated individual mice displaying the selective loss of the floxed exon 2. c Volcano plots display differential expression of UPR hallmark genes and the gene sets defined in c for indicated treatment groups compared to freshly isolated follicular B cells (group n = 3–5 animals). Genes are color coded by the differential expression statistic p < 0.05, log2 fold change > 1 (BH-adjusted p-value—eBayes method—Limma). d Expression of PC-inductive UPR genes as defined in Fig. 3c are shown for each animal and in vitro-treatment group as the z score across each row. e Log2 TPM magnitude of expression of selected genes in each gene set is shown as mean (line), 75%CI (box) and 95%CI (whisker) as well as individual data points (jitter) (*p < 0.05, **p < 0.01, ***p < 0.001; differential expression statistic: BH-adjusted P-value—eBayes method—Limma). Source data are provided as a Source Data file.
Fig. 7
Fig. 7. mTORC1-driven UPR gene expression in activation and differentiation.
CTV-labeled follicular B cells were prepared from either CD20-TAM-Cre (Rptor+/+) (n = 3) or Rptor1fl/fl;CD20-TAM-Cre (Rptor−/−) (n = 3) animals after 10 days of tamoxifen diet. Sorting was performed as in Supplementary Fig. 1e. These cells were stimulated in vitro with CpG + IL-4,5 for 24, 48, or 72 h and live cells were sorted by CD138 staining. RNA-seq libraries were prepared from cDNA extracted from the resulting populations. a Volcano plots display differential expression of UPR hallmark genes and the gene sets defined in Fig. 3c for indicated treatment time points comparing Rptor−/− populations to Rptor+/+ controls (group n = 3 animals). Genes are color coded by the differential expression statistic p < 0.05, log2 fold change > 1 (BH-adjusted p-value—eBayes method—Limma). b Expression of B-cell activating and plasma cell-inductive UPR genes as defined in Fig. 3c are shown for each animal and in vitro-treatment time point as the z score across each row. c, Log2 TPM magnitude of expression of selected genes in each gene set is shown as mean (line), 75%CI (box) and 95%CI (whisker), as well as individual data points (jitter) (*p < .05, **p < .01, ***p < .001; differential expression statistic: BH-adjusted p-value—eBayes method—Limma). Source data are provided as a Source Data file.
Fig. 8
Fig. 8. Model for mTORC1 priming of the ER preceding plasma cell differentiation.
a The conventional model for ER remodeling during plasma cell differentiation holds that Blimp1-mediated increases in immunoglobulin expression causes the chaperone BiP to release from the luminal tails of IRE1α, allowing it to dimerize and mediate Xbp1 mRNA splicing. In this model Xbp1s is the main effector of ER remodeling. Paradoxically, with this scenario activation of PERK and ATF6 would occur by the same mechanism, causing translation inhibition and increased chance of apoptosis downstream of PERK activation. b Revised model wherein, preceding Blimp1 induction, mTORC1 signaling facilitates the upregulation of plasma cell-affiliated UPR genes, including chaperones, ERAD machinery, and enzymes necessary for enhancing protein secretion. The resulting “primed” ER is then better equipped to handle increased protein load following Blimp1 induction. At this point the UPR is tuned so that we see activation of the Xbp1-splicing activity of IRE1α without RIDD activity and activation of PERK.

Similar articles

See all similar articles

References

    1. Helmreich E, Kern M, Eisen HN. The secretion of antibody by isolated lymph node cells. J. Biol. Chem. 1961;236:464–473. - PubMed
    1. Hibi T, Dosch HM. Limiting dilution analysis of the B cell compartment in human bone marrow. Eur. J. Immunol. 1986;16:139–145. doi: 10.1002/eji.1830160206. - DOI - PubMed
    1. Wiest DL, et al. Membrane biogenesis during B cell differentiation: most endoplasmic reticulum proteins are expressed coordinately. J. Cell Biol. 1990;110:1501–1511. doi: 10.1083/jcb.110.5.1501. - DOI - PMC - PubMed
    1. Walter P, Ron D. The unfolded protein response: from stress pathway to homeostatic regulation. Science. 2011;334:1081–1086. doi: 10.1126/science.1209038. - DOI - PubMed
    1. Bettigole SE, Glimcher LH. Endoplasmic reticulum stress in immunity. Annu. Rev. Immunol. 2015;33:107–138. doi: 10.1146/annurev-immunol-032414-112116. - DOI - PubMed
Feedback