Cyclin D3 drives inertial cell cycling in dark zone germinal center B cells.

ARTICLE Cyclin D3 drives inertial cell cycling in dark zone germinal center B cells Juhee Pae 1 , Jonatan Ersching 1 , Tiago B.R. Castro 1 , Marta Schips 2 , Luka Mesin 1 , Samuel J. Allon 3,4,5 , Jose Ordovas-Montanes 5,6,7,8 , Coraline Mlynarczyk 9 ,AriMelnick 9 , Alejo Efeyan 10 , Alex K. Shalek 3,4,5,7,8 , Michael Meyer-Hermann 2,11,12 , and Gabriel D. Victora 1  During affinity maturation, germinal center (GC) B cells alternate between proliferation and somatic hypermutation in the dark zone (DZ) and affinity-dependent selection in the light zone (LZ). This anatomical segregation imposes that the vigorous proliferation that allows clonal expansion of positively selected GC B cells takes place ostensibly in the absence of the signals that triggered selection in the LZ, as if by “inertia.”We find that such inertial cycles specifically require the cell cycle regulator cyclin D3. Cyclin D3 dose-dependently controls the extent to which B cells proliferate in the DZ and is essential for effective clonal expansion of GC B cells in response to strong T follicular helper (Tfh) cell help. Introduction into the Ccnd3 gene of a Burkitt lymphoma–associated gain-of-function mutation (T283A) leads to larger GCs with increased DZ proliferation and, in older mice, clonal B cell lymphoproliferation, suggesting that the DZ inertial cell cycle program can be coopted by B cells undergoing malignant transformation. Introduction Germinal centers (GCs) are the sites of affinity maturation, the process by which antibodies improve their affinity for antigen over time (Cyster and Allen, 2019;De Silva and Klein, 2015; Eisen, 2014;Mesin et al., 2016;Rajewsky, 1996;Shlomchik et al., 2019;Victora and Nussenzweig, 2012). For efficient affinity maturation, GC B cells must cycle between two major transcriptional states, associated with localization of B cells to each of the two microanatomical “zones”of the GC. When in the dark zone (DZ), B cells proliferate vigorously and mutate their immunoglobulin genes by somatic hypermutation (SHM). After transition to the light zone (LZ), B cells bearing advantageous mutations are selectively driven to clonally expand, based at least in part on their ability to bind and present antigen to GCresident T follicular helper (Tfh) cells (Victora et al., 2010). Successive cycles of SHM and affinity-based selection ultimately enrich for higher-affinity cells among the GC B cell population in a process known as cyclic reentry (MacLennan, 1994;Victora and Nussenzweig, 2012). A unique consequence of the anatomical compartmentalization of the GC is that mitogenic signals are segregated from the proliferation they induce. Upon positive selection, B cells typically transition from G1 to S phase of the cell cycle in the LZ, migrate from LZ to DZ while in S phase, and undergo G2 and M phases in the DZ (Gitlin et al., 2014;Victora et al., 2010). After this first division, cell cycling continues in the DZ, with most B cells undergoing on average two additional cell cycles before returning to the LZ for further selection (Gitlin et al., 2014). GC B cells that receive stronger signals from Tfh cells in the LZ, however, can undergo a much greater number of proliferative cycles in the DZ, resulting in exponential clonal expansion (Gitlin et al., 2014;Meyer-Hermann et al., 2012;Victora et al., 2010). At its extreme, this regulated expansion can lead to “clonal bursts,”in which a single B cell can take over a 2,000-cell GC in the course of a few days (Tas et al., 2016). These bursts are associated with massive diversification of generally higheraffinity SHM variants and as such are likely to play an important role in the generation of high-affinity B cell clones (Amitai et al., 2017;Bannard and Cyster, 2017;Mesin et al., 2016). Despite the importance of DZ proliferation for GC B cell selection and affinity maturation, our understanding of GC B cell ............................................................................................................................................................................. 1 Laboratory of Lymphocyte Dynamics, The Rockefeller University, New York, NY; 2 Department of Systems Immunology and Braunschweig Integrated Centre of Systems Biology, Helmholtz Centre for Infection Research, Braunschweig, Germany; 3 Institute for Medical Engineering and Science, Department of Chemistry, Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA; 4 Ragon Institute of Massachusetts General Hospital, Massachusetts Institute of Technology and Harvard, Cambridge, MA; 5 Broad Institute of Massachusetts Institute of Technology and Harvard, Cambridge, MA; 6 Division of Gastroenterology, Boston Children’s Hospital, Boston, MA; 7 Program in Immunology Harvard Medical School, Boston, MA; 8 Harvard Stem Cell Institute, Cambridge, MA; 9 Department of Medicine, Division of Hematology and Medical Oncology, Weill Cornell Medicine, New York, NY; 10 Spanish National Cancer Research Center, Madrid, Spain; 11 Institute for Biochemistry, Biotechnology and Bioinformatics, Technische Universit¨ at Braunschweig, Braunschweig, Germany; 12 Cluster of Excellence RESIST (EXC 2155), Hannover Medical School, Hannover, Germany. Dr. Ersching died in October 2020; Correspondence to Gabriel D. Victora: [email protected]. © 2020 Pae et al. This article is distributed under the terms of an Attribution–Noncommercial–Share Alike–No Mirror Sites license for the first six months after the publication date (see http://www.rupress.org/terms/). After six months it is available under a Creative Commons License (Attribution–Noncommercial–Share Alike 4.0 International license, as described at https://creativecommons.org/licenses/by-nc-sa/4.0/). Rockefeller University Press https://doi.org/10.1084/jem.20201699 1of17 J. Exp. Med. 2020 Vol. 218 No. 4 e20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 selection has historically focused on events that take place in the LZ (Mesin et al., 2016;Shlomchik et al., 2019). Consequently, the precise mechanisms that allow DZ proliferation to take place in the apparent absence of direct mitogenic signals are still incompletely understood. For instance, upon interaction with Tfh cells, positively selected LZ B cells express the transcription factor c-Myc (Calado et al., 2012;Dominguez-Sola et al., 2012; Finkin et al., 2019), as well as the mechanistic target of rapamycin (mTOR)–mediated anabolic program (Ersching et al., 2017). While both c-Myc and mTOR are required for LZ B cells to migrate to the DZ and enter the proliferative phase, the induction of these pathways appear to be mostly restricted to the LZ (Dominguez-Sola et al., 2012;Ersching et al., 2017;Finkin et al., 2019). This suggests that DZ B cells retain a memory of the intensity of the c-Myc and mTOR complex 1 (mTORC1)– dependent “charge”they received previously in the LZ and that they subsequently translate this memory into the number of cell cycles they will undergo in the DZ using a cell-intrinsic “timer” or “counter”(Bannard et al., 2013;Gitlin et al., 2014). The molecular pathways that directly control the number of cycles a GC B cell undergoes in the DZ remain uncharacterized. Here, we used a combination of unbiased whole-transcriptome single-cell mRNA sequencing (scRNA-seq) analysis and targeted genetic and pharmacological manipulation of cell cycle regulators to investigate the molecular nature of DZ cell cycles. We find that DZ B cells adopt a distinct E2F high /c-Myc low mode of cell cycling that allows rapid and continuous proliferation in the absence of external mitogenic signals. We show that the cell cycle regulator cyclin D3, previously shown to be required for GC formation and maintenance (Cato et al., 2011;Peled et al., 2010), is a specific, dose-dependent controller of this phenotype and DZ cell cycling and that loss of cyclin D3 cannot be overcome by LZ cycling induced by strongly increased Tfh cell help. Introduction into mice of a gain-of-function mutation in cyclin D3 derived from human Burkitt lymphoma (Schmitz et al., 2012,2014) leads to exacerbated B cell proliferation specifically in the GC DZ and development of clonal post-GC B cell expansions, linking the inertial proliferative program to malignant transformation. Results B cell proliferation in the DZ is Tfh cell independent To better understand GC B cell cycling in the DZ, we first sought to formally determine whether S phase entry by DZ B cells requires acute signals from Tfh cells in addition to those delivered in the LZ. To this end, we used a synchronized selection model (Ersching et al., 2017;Victora et al., 2010) to allow kinetically precise blocking of Tfh cell help to GC B cells at set time points after induced positive selection (Fig. 1 A). In this model, we use adoptive transfer of 4-hydroxy-3-nitro-phenylacetyl (NP)– specific B1-8 hi B cells followed by immunization with NP conjugated to chicken OVA (NP-OVA) to create GCs in which the majority of B cells lack the surface receptor DEC-205 (encoded by the gene Ly75). Treatment of mice with ongoing GCs with an antibody to DEC-205 fused to OVA (DEC-OVA) leads to presentation of OVA peptides by Ly75 +/+ GC B cells only. This induces Ly75 +/+ cells to preferentially interact with Tfh cells, triggering their positive selection (Pasqual et al., 2015;Victora et al., 2010). In this process, GC B cells also become synchronized: 12 h after DEC-OVA, most Ly75 +/+ cells are located in the LZ, where they interact extensively with Tfh cells (Ersching et al., 2017;Shulman et al., 2014). By 36 h after DEC-OVA, most Ly75 +/+ cells have transitioned to the DZ, where they begin their proliferative burst (Victora et al., 2010). Fig. 1, A–C;and Fig. S1 provide an overview of key time points in this kinetics. Entry into S phase of the cell cycle, revealed by double-pulsing mice with 5-ethynyl-2-deoxyuridine (EdU) and BrdU nucleotides (Gitlin et al., 2014), was greatly increased in DZ B cells at 36 h after DEC-OVA, indicating that, at this time point, GC B cells progress through the G1-S checkpoint while in the DZ compartment (Fig. 1, B and C;andFig. S1, E and F).DZSphaseentry persisted above the steady-state level until at least 60 h after DEC-OVA treatment, when Ly75 +/+ cells remained predominantly in the DZ and continued to expand (Fig. 1, B and C;and Fig. S1, B–D). To determine whether T cell help is continually required for DZ proliferation, we acutely blocked Tfh cell–B cell interaction either at the time of the initial Tfh signal delivery in the LZ (6 h after DEC-OVA) or following the transition of positively selected B cells to the DZ but before the proliferative burst (30 h after DEC-OVA; Fig. 1 D). As expected from the role of Tfh cell– mediated signaling in this model (Victora et al., 2010), early blocking of T cell–B cell interactions in the LZ using antibodies to MHC class II or CD40 ligand (CD40L) effectively prevented clonal expansion of Ly75 +/+ cells by 48 h after DEC-OVA treatment (Fig. 1, E and F). By contrast, blocking either pathway after B cells transitioned to the DZ had little, if any, effect on the expansion of Ly75 +/+ cells over Ly75 −/− cells (Fig. 1, E and F)ortheabilityof Ly75 +/+ DZ B cells to enter S phase, as evidenced by EdU/BrdU incorporation (Fig. 1, G and H). We conclude that sustained proliferation of B cells in the GC DZ upon positive selection does not require continuous help from Tfh cells. We therefore refer to DZ cell cycles as “inertial,”since they proceed in a cell-intrinsic fashion according to the strength of the initial “push”from Tfh cells, in contrast to the “primary”cell cycles that B cells undergo immediately downstream of selective signals in the LZ. Inertial cycling is sustained by prolonged E2F activation following a decay in c-Myc activity To identify the transcriptional programs associated with inertial cycling, we performed scRNA-seq on GC B cells at different time points after forcing positive selection using DEC-OVA. We index-sorted single Ly75 +/+ B cells using antibodies to LZ/DZ markers, and sorted cells were sequenced using the Smart-Seq2 protocol (Trombetta et al., 2014). We assayed GC B cells first at 12 h after DEC-OVA, when Ly75 +/+ B cells are enriched in the LZ in the process of receiving cognate help from Tfh cells, and then at 30, 46, and 60 h after DEC-OVA, as DZ B cells transition from signal-dependent proliferation to predominantly inertial modes of cycling before returning to the LZ between 72 and 96 h after DEC-OVA (Fig. 1, A–D;andFig. S1;Victora et al., 2010). For comparison, we also included a sample of Ly75 −/− counterselected LZ cells from the 12-h time point (Table S1). Pae et al. Journal of Experimental Medicine 2of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure 1. Cell cycle entry by DZ B cells does not require Tfh cell help. (A) Experimental setup for DEC-OVA–induced Ly75 +/+ GC B cells followed by double labeling with EdU/BrdU to assay for S phase entry. (B) S phase entry of Ly75 +/+ and Ly75 −/− B1-8 hi cells in the DZ, quantified in C. (D) Experimental setup for DEC-OVA–induced positive selection of Ly75 +/+ GC B cells followed by inhibition of T-B interaction using anti-MHC class II or anti-CD40L before (early) or after (late) DZ reentry. (E and F) Effect on GC size (left) and expansion of Ly75 +/+ cells over Ly75 −/− (center, right) upon anti-MHC class II treatment (E) or anti-CD40L treatment (F). (G and H) S phase entry of Ly75 +/+ B1-8 hi cells in the DZ upon late treatment with anti-MHC class II (G) or anti-CD40L (H). *, P < 0.05; **, P < 0.01; ***, P < 0.001; n.s., nonsignificant, a paired ttest comparison between Ly75 +/+ and Ly75 −/− cells in the same animal (C) or nonparametric Mann–Whitney test compared with the group treated with an isotype control (E–H). Bars indicate median. Each circle represents a mouse. Data are pooled from two (B and C), three (E and G), or two (F and H) independent experiments. Iso, isotope. Pae et al. Journal of Experimental Medicine 3of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 The 1,220 cells that passed our quality control thresholds fell into seven major clusters (Fig. 2 A). These were determined in large part by cell cycle phase as inferred from their transcriptional profile (Tirosh et al., 2016), with a lesser contribution of their LZ/DZ phenotype as defined by surface staining (Fig. 2 B). Clusters 0 and 4 contained primarily G1 cells; clusters 2, 3, and 6 were enriched in S phase cells; and clusters 1 and 5 were enriched in cells in the G2 and M phases (Fig. 2 C). To follow the evolution of positively selected B cells across these clusters as they transitioned from reactive to inertial cell cycles and then to quiescence, we compared Ly75 +/+ and Ly75 −/− cells in the LZ at 12 h after DEC-OVA to Ly75 +/+ cells in the DZ at the 30-, 46-, and 60-h time points. This revealed a marked increase in representation of cells in cluster 2, as Ly75 +/+ cells are positively selected in the LZ at 12 h (empty arrowheads in Fig. 2 D). Enrichment in cluster 2 continued as cells transitioned to the DZ Figure 2. Single-cell transcriptomic analysis of GC B cells undergoing positive selection. (A) Uniform manifold approximation and projection (UMAP) plot displaying 1,220 cells colored by shared nearest neighbor clusters collected (default Wilcox test; see Materials and methods). Cells were collected from four independent experiments (see Table S1). (B) Distribution of cell cycle phase (left) and LZ/DZ phenotypes (right). (C) Distribution of cell cycle phase in clusters. (D) Changes in distribution of clusters grouped by cell cycle phase over the DEC-OVA–induced selection time course. (E and F) Expression of Myc mRNA, c-Myc, and E2F target gene signatures in clusters 2, 3, and 6 (E) or DEC-OVA time points with indicated zonal phenotypes (F). Dotted line indicates the threshold used for quantification in G. See Fig. S2 C for a complete list of P values. (H) S phase entry of Ly75 +/+ B1-8 hi cells in the DZ 12 h after treatment with palbociclib, a CDK4/6 inhibitor, or vehicle (PBS). (I) Quantification includes all LZ and DZ B1-8 hi cells, in addition to positively selected Ly75 +/+ B1-8 hi cells in the DZ. **, P < 0.01; ****, P < 0.0001, nonparametric Mann–Whitney test, compared with the PBS-treated control group. Bars indicate median. Each circle represents a mouse (H and I) or a cell (all other figures). Data are pooled from two independent experiments (H and I). Pae et al. Journal of Experimental Medicine 4of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 at 30 h, after which the dominant S phase phenotype shifted to cluster 6 at 46 h and then cluster 3 at 60 h after DEC-OVA (filled arrowheads in Fig. 2 D). Pairwise comparisons of the three S phase clusters by gene set enrichment analysis (GSEA) using the “hallmark”signatures from the MSigDB database (Mootha et al., 2003;Subramanian et al., 2005) revealed expression of E2F and c-Myc target genes as the most consistent differences between the three clusters (Fig. S2 A). Cluster 2, enriched in LZphenotype cells from the 12and 30-h time points, showed high expression of Myc and of c-Myc–and mTORC1-dependent transcriptional signatures (Fig. 2 E and Fig. S2 B;Peng et al., 2002;Schuhmacher et al., 2001). Given the strong association between these signatures and positive selection (Calado et al., 2012;Dominguez-Sola et al., 2012;Ersching et al., 2017;Finkin et al., 2019;Luo et al., 2018;Victora et al., 2010), this pattern suggests that cluster 2 consists primarily of recently selected B cells undergoing primary S phase, either on their way to or soon after DZ reentry. Cells in clusters 3 and 6, on the other hand, showed reduced levels of Myc and c-Myc–and mTORC1dependent gene signatures while maintaining expression of E2F target genes (Fig. 2 E and Fig. S2 B), as expected if they were entering the cell cycle by inertia while being physically segregated from mitogenic signals located in the LZ. Kinetic analysis of Myc and c-Myc and E2F target gene signatures confirmed that E2F target gene expression levels remained unaltered between 30 and 46 h after DEC-OVA, while mRNA expression and activity of Myc decreased (Fig. 2, F and G;andFig. S2 C). E2F signatures began to subside only at the 60-h time point, as inertial cycling defined by DZ S phase entry (Fig. 1, B and C)is already subsiding. Thus, unlike their LZ counterparts, DZ B cells appear to engage in a distinct mode of cell cycling that does not require continued expression of the Myc gene or its protein function to maintain E2F activity. This suggests that regulators of E2F that are downstream of c-Myc and other selectiondependent signals may be required to sustain proliferation after positively selected GC B cells transition to the DZ. A major regulator of E2F activity are the D-type cyclins, which in partnership with cyclin-dependent kinases (CDKs) 4 and 6, activate E2F by phosphorylation of its negative regulator RB (Musgrove et al., 2011). To test whether D-type cyclins could be responsible for allowing the progression of inertial cell cycles, we treated mice with the inhibitor of CDK4/6 palbociclib 36 h after inducing positive selection of GC B cells with DEC-OVA and analyzed cell cycle progression 12 h later. Unlike blockade of MHC class II or CD40L (Fig. 1, D–H), palbociclib treatment strongly inhibited S phase entry in all GC B cells, including those undergoing inertial cycling in the DZ (Fig. 2, H and I). Thus, inertial S phase entry, while not dependent on Tfh-mediated signals, still requires activity of CDK4/6. Ccnd3, but not Ccnd2, is required for DZ inertial cycling Since B cells express exclusively cyclins D2 and D3 (encoded by Ccnd2 and Ccnd3, respectively) upon mitogenic stimulation (Reid and Snow, 1996;Solvason et al., 1996), we sought to determine the relative contribution of these two cyclins to the CDK4/6 dependency of inertial cycles. Ccnd2 mRNA was detectable primarily in the subset of cells undergoing S phase in the LZ in the c-Myc high cluster 2 (Fig. S3 A). In agreement with our previous reports, Ccnd2 was higher in Ly75 +/+ cells at 12 h after DEC-OVA as well as in the LZ in general (Dominguez-Sola et al., 2012; Victora et al., 2010;Fig. S3, B and C). To investigate the function of cyclin D2 in GC B cells, we generated Ccnd2 knockout mice using CRISPR-Cas9–mediated genome editing in zygotes to introduce a 4-bp deletion/frameshift in exon 1 of the gene (Fig. S3 D). Ccnd2 −/− mice lacked peritoneal B-1a cells (Fig. S3, E and F), and females were unable to produce progeny (data not shown), as shown previously using an independently generated knockout strain (Sicinski et al., 1996;Solvason et al., 2000), confirming that ours is a null allele. We adoptively transferred a 1:1 mixture of Ccnd2 +/+ and Ccnd2 −/− B1-8 hi cells into WT hosts, which we then immunized with NP-OVA to generate GCs (Fig. S3 G), and found that zonal distribution was preserved in Ccnd2 −/− B1-8 hi B cells (Fig. S3, H and I). However, rather than reducing the ability of GC B cells to cycle, loss of cyclin D2 led to a small but consistent increase in the proportion of B1-8 hi cells entering S phase in both the LZ and DZ (Fig. S3, J–L). Despite this increase, absence of cyclin D2 showed no clear effect on the competitiveness of GC B cells over time when compared with Ccnd2-sufficient B cells within the same GC (Fig. S3, M and N). These experiments suggest that although cyclin D2 can affect the ability of GC B cells to enter cell cycle, it is not required for inertial cycling in the DZ. In contrast to the dynamic behavior of Ccnd2, our scRNA-seq dataset showed that Ccnd3 mRNA amounts were stably high throughout the LZ/DZ cycle and over our time course of DECOVA–induced selection, with only slight increases at the 30-h time point and in the DZ in general (Fig. 3, A–C). Despite these modest changes in mRNA expression, cyclin D3 protein levels were substantially higher in the DZ (Fig. 3 D), in agreement with previous reports based on histology (Peled et al., 2010). Also consistent with prior reports (Cato et al., 2011;Peled et al., 2010), we found that loss of Ccnd3 led to dramatically reduced GC B cell frequency (Fig. 3, E and F). This is unlikely to be due to a general defect in proliferation, given that Ccnd3 −/− B cells proliferate normally in vitro and at pre-GC stages in vivo (Cato et al., 2011; Peled et al., 2010). Closer examination of Ccnd3 −/− GCs showed that decreased GC size was primarily caused by a marked reduction in the proportion of B cells with a DZ phenotype (Fig. 3, GandH), suggesting a strong block in the ability of Ccnd3 −/− B cells to undergo inertial cycling. This phenotype was identical when GCs were generated by adoptive transfer of Ccnd3 −/− B1-8 hi cells into WT hosts (Fig. 3 I), resulting in a significantly decreased DZ/LZ ratio (Fig. 3, J and K) and an almost complete loss in S phase initiation in the DZ with only a minor decrease of fitness in the LZ (Fig. 3, J and L–N). The requirement for cyclin D3 is therefore GC B cell intrinsic and largely restricted to the DZ. Ccnd3 −/− B1-8 hi cells in the DZ also showed increased forward scatter, a surrogate measure of cell size (Fig. 3, O and P). Since cell size increases before the initiation of inertial cycles (Ersching et al., 2017;Finkin et al., 2019), this finding suggests that the inability of Ccnd3 −/− B cells to undergo inertial cycling results in their failure to lose cellular mass by division. We conclude that cyclin D3 is cell-intrinsically required for GC B cells to enter S phase specifically in the DZ while being dispensable for S phase entry immediately downstream of Pae et al. Journal of Experimental Medicine 5of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 T cell–mediated selection in the LZ. This pattern implicates cyclin D3 as a nonredundant mediator of inertial cell cycles. Cyclin D3–mediated DZ inertial cycling links GC positive selection with clonal expansion To determine the relative importance of LZ cycles vs. inertial DZ cycles for clonal expansion, we asked whether cell proliferation defects seen in Ccnd3 −/− DZ B cells could be rescued by forcing strong interaction with Tfh cells using DEC-OVA (Fig. 4 A). Whereas Ccnd3 +/+ Ly75 +/+ B1-8 hi cells proliferated sufficiently to outnumber Ly75 −/− cells by fivefold over a 60-h period, clonal expansion was much less efficient when Ly75 +/+ B1-8 hi cells lacked cyclin D3. With the exception of one outlier mouse, the ratio of Ly75 +/+ to Ly75 −/− cells increased only slightly, if at all, Figure 3. Inertial B cell cycling requires cyclin D3. (A–C) Expression of Ccnd3 in UMAP dimension (A), over time after DEC-OVA immunization (B), and in LZ or DZ (C). P values in B are from a Kruskal–Wallis test with Dunn’s multiple comparisons test. Other significant P values are <0.001 (12 × 60 h), 0.013 (Ly75 −/− vs. 30 h), and 0.015 (Ly75 −/− vs. 60 h). (D) Immunoblots of whole-cell lysates of LZ or DZ cells sorted from popliteal or mesenteric LNs (pLN or mLN, respectively). Molecular weight is indicated in kilodaltons. (E–H) Staining for GC (E) and DZ/LZ (G) in WT (Ccnd3 +/+ ) or cyclin D3 mutant (Ccnd3 −/− ) mice that were immunized s.c. in the hind footpad for pLNs, quantified in F. (I) Experimental setup for induction of GCs containing WT (Ccnd3 +/+ ) or cyclin D3 mutant (Ccnd3 −/− ) B1-8 hi cells. (J–P) DZ and LZ staining of all GC B cells (black) or cells entering S phase (red; J), S phase entry (L), forward scatter (O) of WT (Ccnd3 +/+ ), or cyclin D3 mutant (Ccnd3 −/− )B1-8 hi cells, quantified in K, M, N, and P. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; nonparametric Mann–Whitney test. Bars indicate median. Each circle represents a cell (B and C) or a mouse (all other figures). Data are pooled from three (E–H) or two (I–P) independent experiments. P adj , adjusted P value; post-imm, post-immunization. Pae et al. Journal of Experimental Medicine 6of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure 4. Increased Tfh cell help cannot compensate for loss of cyclin D3. (A) Experimental setup for DEC-OVA–induced positive selection of Ly75 +/+ cells that are either WT (Ccnd3 +/+ )orcyclinD3mutant(Ccnd3 −/− ). Note that due to of lack of competitiveness in early GC stages, Ccnd3 −/− B1-8 hi cells have to be transferred at a higher proportion than Ccnd3 +/+ B1-8 hi cells. (B–E) Clonal expansion (B) and DZ/LZ staining (D) of WT (Ccnd3 +/+ )orcyclinD3mutant(Ccnd3 −/− ) Ly75 +/+ cells over time after DEC-OVA immunization, quantified in C and E. (F) Immunofluorescence showing the migration of WT (Ccnd3 +/+ )orcyclinD3 mutant (Ccnd3 −/− )Ly75 +/+ cells 36 h after DEC-OVA immunization in pLN GCs. Dotted area indicates the DZ marked by the absence of IgD (Naive B cell marker) Pae et al. Journal of Experimental Medicine 7of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 over input levels (Fig. 4, B and C). Failure of Ccnd3 −/− B cells to clonally expand was also accompanied by failure to accumulate in the DZ (Fig. 4, D and E), although histology showed that Ccnd3 −/− cells were able to at least access the DZ anatomically at 36 h after DEC-OVA (Fig. 4 F). Even so, the proportion of early S phase cells among the Ccnd3 −/− population dropped precipitously upon transition from the LZ to DZ, as expected given the inability of cyclin D3–deficient B cells to sustain proliferation by inertia (Fig. 4, G–K). Thus, the residual expansion of Ccnd3 −/− Ly75 +/+ B1-8 hi seen in some mice can be attributed primarily to cell cycling taking place in the LZ in response to continued signals delivered by Tfh cells. We conclude that strong signaling from Tfh cells in response to DEC-OVA cannot overcome the requirement for cyclin D3 in driving DZ proliferation and that in the absence of inertial cycling, LZ cell cycles triggered directly in reaction to Tfh cell–derived signals are not sufficient to sustain large proliferative bursts. Cyclin D3 controls DZ inertial cycling in a dose-dependent manner A potential mechanism for how the strength of the initial Tcell–B cell interactions in the LZ determines the number of rounds of division that selected B cells undergo in the DZ (Gitlin et al., 2014) is that DZ B cells translate the memory of their interactions with Tfh cells into protein amounts of a cell cycle regulator. If cyclin D3 follows such a pattern, then it could be predicted that GC B cells with a higher capacity to produce cyclin D3 protein would be at a competitive advantage due to increased proliferation in the DZ. To test this, we directly competed B1-8 hi B cells carrying either one or two intact alleles of Ccnd3 by adoptive transfer of a 1:1 mixture of Ccnd3 +/− and Ccnd3 +/+ B1-8 hi B cells into the same recipient mice, which were then immunized with NP-OVA in alum and assayed for relative abundance of these two populations (Fig. 5 A). Whereas both Ccnd3 +/+ and Ccnd3 +/− B1-8 hi cells were found at a similar ratio in early GCs at day 7 after immunization, the proportion of Ccnd3 +/− B1-8 hi cells decreased gradually over time, such that these cells were completely eliminated in three out of seven mice by day 14 after immunization (Fig. 5, B and C). Thus, the reduction in cyclin D3 dosage associated with heterozygosis is sufficient to impose a gradual but clear loss of this population from the GC. Lack of competitiveness of heterozygous B cells was associated with a slight reduction in the DZ/LZ ratio, which was already observable at day 7 after immunization (Fig. 5, D and E), and a decrease in the proportion of cells entering S phase in the DZ, but not in the LZ (Fig. 5, F–H). To extrapolate the loss of inertial proliferative capacity among Ccnd3 +/− GC B cells from our direct competition data, we simulated this experiment in silico using a previously published agent-based model of GC selection that includes T cell control over the number of cell cycles carried a B cell undergoes upon positive selection (Meyer-Hermann, 2020 Preprint;Meyer-Hermann et al., 2012). We modeled loss of Ccnd3 expression as a reduction in the maximum number of divisions a Ccnd3 +/− GC B cell can complete upon positive selection. We varied the relationship between the degree of T cell help (modeled as intensity of c-Myc activation) and the number of divisions as illustrated by the curves shown in Fig. 5 I.Theexperimentally measured kinetics of Ccnd3 +/− GC B cells were best reproduced when the maximum number of divisions for Ccnd3 +/− GC B cells in silico was reduced to 72% of WT, with a residual sum of squares (RSS) of 0.14 (Fig. 5 J). In silico, a reduction in responsiveness of this magnitude in Ccnd3 +/− GC B cells was accompanied by a reduction in the number of cell cycles per cell to ∼84% of WT levels at day 8 after immunization, which is compatible with our EdU/BrdU incorporation data (Fig. 5 K). We conclude that cyclin D3 dose-dependently controls the number of inertial cycles a B cell will undergo in the DZ. The sensitivity of GC B cells to loss of even a single allele of Ccnd3 suggests that cyclin D3 protein abundance may serve as a molecular bridge linking the cumulative signal a B cell receives from T cells in the LZ and the number of cycles this cell can execute in the DZ. A lymphoma-associated mutation that stabilizes cyclin D3 promotes DZ inertial cycling and drives clonal B cell lymphoproliferation A parallel line of evidence pointing to the importance of cyclin D3 in GC proliferation is the presence of a series of Ccnd3 mutations that stabilize cyclin D3 protein in roughly 40% of cases of sporadic Burkitt lymphoma (Casanovas et al., 2004;Schmitz et al., 2012,2014;Sonoki et al., 2001). Because Burkitt lymphoma cells closely resemble DZ B cells in gene expression (Caron et al., 2009;Victora and Mouquet, 2018), we hypothesized that stabilization of cyclin D3 by means of a lymphomaassociated mutation may increase the propensity of GC B cells to undergo inertial cycles in the DZ. To test this, we used CRISPRCas9–mediated genome editing to generate a Ccnd3 allele encoding a version of cyclin D3 protein that is stabilized by the replacement of a phosphorylatable threonine (T283A) in its C-terminal domain (Fig. 6 A). This mutation prevents phosphorylation of T283, which would otherwise promote nuclear export and proteasomal degradation of cyclin D3 (Casanovas et al., 2004;Cato et al., 2011). When introduced into the genome, the T283A mutation alone did not cause any overt anomaly in B cell development, with the possible exception of a slight increase in the percentage of pre/pro-B cells in the bone marrow (BM), nor did it lead to spontaneous GC formation in the spleens of young adult mice (Fig. S4, A and D). Nonetheless, we found that upon immunization with NP-OVA, Ccnd3 T283A/+ mice generated GCs that were approximately twice as large as WT GCs and had a markedly higher DZ/LZ ratio, suggesting increased DZ proliferation (Fig. 6, B–E). Accordingly, flow and CD35 (FDC marker). Percentage of Ly75 +/+ cells within the DZ out of the entire GC is quantified. Scale bar = 100 μm. (G–K) S phase entry of positively selected Ly75 +/+ B1-8 hi cells that are either WT (Ccnd3 +/+ )orcyclinD3mutant(Ccnd3 −/− ) in the LZ (G) or DZ (H), quantified in I–K. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, nonsignificant, nonparametric Mann–Whitney test, compared with WT (Ccnd3 +/+ ). Bars indicate median. Each circle represents a mouse. Data pooled from four independent experiments. Pae et al. Journal of Experimental Medicine 8of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 cytometric measurement of EdU/BrdU incorporation showed a ∼50% increase in GC B cells entering S phase in the DZ but no increase in S phase entry in the LZ (Fig. 6, F–H). To confirm that the proliferative effect of cyclin D3 on the GC is B cell intrinsic, we generated mixed chimeras in which BM cells from either WT or Ccnd3 T283A/+ mice were mixed with those from B cell– deficient J H T mice at a 20:80 ratio. Consistent with the analysis of Ccnd3 T283A/+ mice, chimeras with Ccnd3 T283A/+ B cells showed increased GC size, DZ expansion, and S phase entry in the DZ, but not in the LZ, upon immunization with NP-OVA in alum (Fig. 6, C, E, G, and H). This phenotype was recapitulated upon immunization with KLH, ruling out any potential effects on of the hypomorphic immunoglobulin λlight chain of the SJL strain in which Ccnd3 T283A mice were originally generated on the normally Igλ-dominated response to NP-OVA (Fig. S4, E–G). Moreover, Ccnd3 T283A/+ DZ B cells showed a decrease in size as measured by forward scatter, mirroring the phenotype found in Ccnd3 −/− B cells (Fig. 6, I and J). This indicates that the increased proliferation of Ccnd3 T283A/+ B cells in the DZ extends to the extreme of what these cells are metabolically programmed Figure 5. Cyclin D3 controls inertial cell cycling in a dose-dependent manner. (A) Experimental setup for induction of GCs containing mixtures of B1-8 hi cells with a full (Ccnd3 +/+ )orreduced(Ccnd3 +/− ) dose of cyclin D3. (B and C) Clonal expansion of B1-8 hi cells with a full (Ccnd3 +/+ )orreduced(Ccnd3 +/− )doseof cyclin D3 over time, quantified in C relative to day 7. (D and E) DZ and LZ staining in GC B1-8hi cells 7 d after NP-OVA immunization, quantified in E. Dotted line indicates averaged DZ/LZ ratio of total GC B cells. (F–H) S phase entry of B1-8 hi cells with a full (Ccnd3 +/+ )orreduced(Ccnd3 +/− )doseofcyclinD3inLZorDZ 8 d after NP-OVA immunization, quantified in G and H. (I) Relationship between strength of T cell help (modeled as c-Myc signal intensity) and number of B cell divisions in the different models used to estimate loss of function in Ccnd3 +/− GC B cells. The maximum number of divisions allowed to Ccnd3 +/− cells corresponds to 50% (red) or 72% (blue), or 80% (green) of the maximum number of divisions of the WT. (J) Ccnd3 +/− kinetics in silico. Loss of Ccnd3 +/− cells as a fraction of the GC population over time in silico. The maximum number of divisions of Ccnd3 +/− GC B cells was fixed to 50% (red line), 72% (blue line), or 80% (green line) of the WT value. Mean (solid lines) and SD (shaded area) over 60 GC simulations. Experimental mean ± SD is shown as white circles. (K) Ccnd3 +/− divisions in silico. The mean number of divisions of Ccnd3 +/− cells is shown as a percentage of WT at the indicated days. Reported data result from simulations with the maximum number of divisions of the Ccnd3 +/− GC B cells fixed to 72% of the WT maximum number of divisions. Mean and SD over 60 GC simulations. White circle represents experimental data. *, P < 0.05; **, P < 0.01; n.s., not significant, paired ttest. Bars indicate median (C). Each circle represents a mouse. Data are pooled from at two independent experiments. Pae et al. Journal of Experimental Medicine 9of17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Victora). J. Pae is a Damon Runyon Cancer Research Foundation Berger Foundation Fellow (DRG-2353-19). J. Ersching was a Cancer Research Institute-Irvington postdoctoral fellow. M. Schips was supported by the European Union’s Horizon 2020 research and innovation program under Marie Skłodowska-Curie grant agreement no. 765158. J. Ordovas-Montanes was supported by the Damon Runyon Cancer Research Foundation (DRG-2274-16) and the Richard and Susan Smith Family Foundation. C. Mlynarczyk was supported by a Lymphoma Research Foundation postdoctoral fellowship, a Leukemia and Lymphoma Society Special Fellow Award, and the American Society of Hematology Research Restart Award for early career investigators in hematology. A. Melnick was supported by National Cancer Institute grant R35CA220499. A. Efeyan is a Ministerio de Ciencia, Innovación y Universidades/ Agencia Estatal de Investigación Ramon y Cajal Awardee (RYC2013-13546). A.K. Shalek was supported by the Searle Scholars Program, the Beckman Young Investigator Program, a Sloan Fellowship in Chemistry, and National Institutes of Health grants 1DP2GM119419 and 5U24AI118672. G.D. Victora is a Searle Scholar, a Burroughs Wellcome Investigator in the pathogenesis of infectious disease, a Pew-Stewart Scholar, and a MacArthur Fellow. Author contributions: J. Pae, J. Ersching, A. Efeyan, and G.D. Victora designed and performed all in vivo experiments. T.B.R. Castro carried out the bioinformatic analysis of scRNA-seq data. L. Mesin, S.J. Allon, and J. Ordovas-Montanes contributed additional scRNA-seq libraries, under supervision of A.K. Shalek. M. Schips and M. Meyer-Hermann performed mathematical modeling. C. Mlynarczyk and A. Melnick contributed to the description of the malignant phenotype of Ccnd3 T283A mice,includinghistologicalanalysis.J.Pae,T.B.R.Castro,and G.D. Victora wrote the manuscript with input from all authors. G.D. Victora supervised the work. Disclosures: A. Melnick reports grants from Janssen, Sanofi, and Daiichi Sankyo, and personal fees from KDAC, Epizyme, Constellation, Jubilant, and ExoTherapeutics outside the submitted work. J. Ordovas-Montanes reports personal fees from Cellarity outside the submitted work. No other disclosures were reported. Submitted: 7 August 2020 Revised: 16 November 2020 Accepted: 30 November 2020 References Amitai, A., L. Mesin, G.D. Victora, M. Kardar, and A.K. Chakraborty. 2017. A Population Dynamics Model for Clonal Diversity in a Germinal Center. Front. Microbiol. 8:1693. https://doi.org/10.3389/fmicb.2017.01693 Bannard, O., and J.G. Cyster. 2017. Germinal centers: programmed for affinity maturation and antibody diversification. Curr. Opin. Immunol. 45:21–30. https://doi.org/10.1016/j.coi.2016.12.004 Bannard, O., R.M. Horton, C.D. Allen, J. An, T. Nagasawa, and J.G. Cyster. 2013. Germinal center centroblasts transition to a centrocyte phenotype according to a timed program and depend on the dark zone for effective selection. Immunity. 39:912–924. https://doi.org/10.1016/j.immuni.2013.08.038 Calado, D.P., Y. Sasaki, S.A. Godinho, A. Pellerin, K. K¨ ochert, B.P. Sleckman, I.M. de Albor´ an, M. Janz, S. Rodig, and K. Rajewsky. 2012. The cell-cycle regulator c-Myc is essential for the formation and maintenance of germinal centers. Nat. Immunol. 13:1092–1100. https://doi.org/10.1038/ ni.2418 Caron, G., S. Le Gallou, T. Lamy, K. Tarte, and T. Fest. 2009. CXCR4 expression functionally discriminates centroblasts versus centrocytes within human germinal center B cells. J. Immunol. 182:7595–7602. https://doi.org/10.4049/jimmunol.0804272 Casanovas,O.,M.Jaumot,A.B.Paules,N.Agell,andO.Bachs.2004.P38SAPK2 phosphorylates cyclin D3 at Thr-283 and targets it for proteasomal degradation. Oncogene. 23:7537–7544. https://doi.org/10.1038/sj.onc.1208040 Cato, M.H., S.K. Chintalapati, I.W. Yau, S.A. Omori, and R.C. Rickert. 2011. Cyclin D3 is selectively required for proliferative expansion of germinal center B cells. Mol. Cell. Biol. 31:127–137. https://doi.org/10.1128/MCB .00650-10 Clijsters, L., C. Hoencamp, J.J.A. Calis, A. Marzio, S.M. Handgraaf, M.C. Cuitino, B.R. Rosenberg, G. Leone, and M. Pagano. 2019. Cyclin F Controls Cell-Cycle Transcriptional Outputs by Directing the Degradation of the Three Activator E2Fs. Mol. Cell. 74:1264–1277.e7. https://doi.org/10 .1016/j.molcel.2019.04.010 Cyster, J.G., and C.D.C. Allen. 2019. B Cell Responses: Cell Interaction Dynamics and Decisions. Cell. 177:524–540. https://doi.org/10.1016/j.cell .2019.03.016 De Silva, N.S., and U. Klein. 2015. Dynamics of B cells in germinal centres. Nat. Rev. Immunol. 15:137–148. https://doi.org/10.1038/nri3804 Dobin,A.,C.A.Davis,F.Schlesinger,J.Drenkow,C.Zaleski,S.Jha,P. Batut, M. Chaisson, and T.R. Gingeras. 2013. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 29:15–21. https://doi.org/10 .1093/bioinformatics/bts635 Dominguez-Sola, D., G.D. Victora, C.Y. Ying, R.T. Phan, M. Saito, M.C. Nussenzweig, and R. Dalla-Favera. 2012. The proto-oncogene MYC is required for selection in the germinal center and cyclic reentry. Nat. Immunol. 13:1083–1091. https://doi.org/10.1038/ni.2428 Eisen, H.N. 2014. Affinity enhancement of antibodies: how low-affinity antibodies produced early in immune responses are followed by highaffinity antibodies later and in memory B-cell responses. Cancer Immunol. Res. 2:381–392. https://doi.org/10.1158/2326-6066.CIR-14-0029 Ersching, J., A. Efeyan, L. Mesin, J.T. Jacobsen, G. Pasqual, B.C. Grabiner, D. Dominguez-Sola, D.M. Sabatini, and G.D. Victora. 2017. Germinal Center Selection and Affinity Maturation Require Dynamic Regulation of mTORC1 Kinase. Immunity. 46:1045–1058.e6. https://doi.org/10.1016/ j.immuni.2017.06.005 Finkin, S., H. Hartweger, T.Y. Oliveira, E.E. Kara, and M.C. Nussenzweig. 2019. Protein Amounts of the MYC Transcription Factor Determine Germinal Center B Cell Division Capacity. Immunity. 51:324–336.e5. https://doi.org/10.1016/j.immuni.2019.06.013 Gitlin, A.D., Z. Shulman, and M.C. Nussenzweig. 2014. Clonal selection in the germinal centre by regulated proliferation and hypermutation. Nature. 509:637–640. https://doi.org/10.1038/nature13300 Gitlin, A.D., C.T. Mayer, T.Y. Oliveira, Z. Shulman, M.J. Jones, A. Koren, and M.C. Nussenzweig. 2015. HUMORAL IMMUNITY. T cell help controls the speed of the cell cycle in germinal center B cells. Science. 349: 643–646. https://doi.org/10.1126/science.aac4919 Hafemeister, C., and R. Satija. 2019. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol. 20:296. https://doi.org/10.1186/s13059-019-1874-1 Inaba, K., W.J. Swiggard, M. Inaba, J. Meltzer, A. Mirza, T. Sasagawa, M.C. Nussenzweig, and R.M. Steinman. 1995. Tissue distribution of the DEC205 protein that is detected by the monoclonal antibody NLDC-145. I. Expression on dendritic cells and other subsets of mouse leukocytes. Cell. Immunol. 163:148–156. https://doi.org/10.1006/cimm.1995.1109 Kozar, K., M.A. Ciemerych, V.I. Rebel, H. Shigematsu, A. Zagozdzon, E. Sicinska, Y. Geng, Q. Yu, S. Bhattacharya, R.T. Bronson, et al. 2004. Mouse development and cell proliferation in the absence of D-cyclins. Cell. 118:477–491. https://doi.org/10.1016/j.cell.2004.07.025 Li, B., and C.N. Dewey. 2011. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 12:323. https://doi.org/10.1186/1471-2105-12-323 Luo, W., F. Weisel, and M.J. Shlomchik. 2018. B Cell Receptor and CD40 Signaling Are Rewired for Synergistic Induction of the c-Myc Transcription Factor in Germinal Center B Cells. Immunity. 48:313–326.e5. https://doi.org/10.1016/j.immuni.2018.01.008 MacLennan, I.C. 1994. Germinal centers. Annu. Rev. Immunol. 12:117–139. https://doi.org/10.1146/annurev.iy.12.040194.001001 Mesin, L., J. Ersching, and G.D. Victora. 2016. Germinal Center B Cell Dynamics. Immunity. 45:471–482. https://doi.org/10.1016/j.immuni.2016 .09.001 Mesin, L., A. Schiepers, J. Ersching, A. Barbulescu, C.B. Cavazzoni, A. Angelini, T. Okada, T. Kurosaki, and G.D. Victora. 2020. Restricted Pae et al. Journal of Experimental Medicine 16 of 17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Clonality and Limited Germinal Center Reentry Characterize Memory B Cell Reactivation by Boosting. Cell. 180:92–106.e11. https://doi.org/10 .1016/j.cell.2019.11.032 Meyer-Hermann, M. 2020. Germinal center B cells separate signals for selection and division like compact discs. bioRxiv. https://doi.org/10.1101/ 2020.01.29.925479 (Preprint posted January 30, 2020) Meyer-Hermann, M., E. Mohr, N. Pelletier, Y. Zhang, G.D. Victora, and K.M. Toellner. 2012. A theory of germinal center B cell selection, division, and exit. Cell Rep. 2:162–174. https://doi.org/10.1016/j.celrep.2012.05 .010 Min, M., Y. Rong, C. Tian, and S.L. Spencer. 2020. Temporal integration of mitogen history in mother cells controls proliferation of daughter cells. Science. 368:1261–1265. https://doi.org/10.1126/science.aay8241 Mootha, V.K., C.M. Lindgren, K.F. Eriksson, A. Subramanian, S. Sihag, J. Lehar, P. Puigserver, E. Carlsson, M. Ridderstr˚ ale, E. Laurila, et al. 2003. PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat. Genet. 34:267–273. https://doi.org/10.1038/ng1180 Musgrove, E.A., C.E. Caldon, J. Barraclough, A. Stone, and R.L. Sutherland. 2011. Cyclin D as a therapeutic target in cancer. Nat. Rev. Cancer. 11: 558–572. https://doi.org/10.1038/nrc3090 Pasqual, G., A. Angelini, and G.D. Victora. 2015. Triggering positive selection of germinal center B cells by antigen targeting to DEC-205. Methods Mol. Biol. 1291:125–134. https://doi.org/10.1007/978-1-4939-2498-1_10 Peled, J.U., J.J. Yu, J. Venkatesh, E. Bi, B.B. Ding, M. Krupski-Downs, R. Shaknovich, P. Sicinski, B. Diamond, M.D. Scharff, and B.H. Ye. 2010. Requirement for cyclin D3 in germinal center formation and function. Cell Res. 20:631–646. https://doi.org/10.1038/cr.2010.55 Peng, T., T.R. Golub, and D.M. Sabatini. 2002. The immunosuppressant rapamycin mimics a starvation-like signal distinct from amino acid and glucose deprivation. Mol. Cell. Biol. 22:5575–5584. https://doi.org/10 .1128/MCB.22.15.5575-5584.2002 Quadros, R.M., H. Miura, D.W. Harms, H. Akatsuka, T. Sato, T. Aida, R. Redder, G.P. Richardson, Y. Inagaki, D. Sakai, et al. 2017. Easi-CRISPR: a robust method for one-step generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR ribonucleoproteins. Genome Biol. 18:92. https://doi.org/10.1186/s13059-017-1220-4 Rajewsky, K. 1996. Clonal selection and learning in the antibody system. Nature. 381:751–758. https://doi.org/10.1038/381751a0 Reid, S., and E.C. Snow. 1996. The regulated expression of cell cycle-related proteins as B-lymphocytes enter and progress through the G1 cell cycle stage following delivery of complete versus partial activation stimuli. Mol. Immunol. 33:1139–1151. https://doi.org/10.1016/S0161-5890(96)00065-X Robbiani, D.F., S. Deroubaix, N. Feldhahn, T.Y. Oliveira, E. Callen, Q. Wang, M. Jankovic, I.T. Silva, P.C. Rommel, D. Bosque, et al. 2015. Plasmodium Infection Promotes Genomic Instability and AID-Dependent B Cell Lymphoma. Cell. 162:727–737. https://doi.org/10.1016/j.cell.2015.07.019 Sakaue-Sawano, A., H. Kurokawa, T. Morimura, A. Hanyu, H. Hama, H. Osawa, S. Kashiwagi, K. Fukami, T. Miyata, H. Miyoshi, et al. 2008. Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell. 132:487–498. https://doi.org/10.1016/j.cell.2007.12.033 Schmitz, R., R.M. Young, M. Ceribelli, S. Jhavar, W. Xiao, M. Zhang, G. Wright, A.L. Shaffer, D.J. Hodson, E. Buras, et al. 2012. Burkitt lymphoma pathogenesis and therapeutic targets from structural and functional genomics. Nature. 490:116–120. https://doi.org/10.1038/nature11378 Schmitz, R., M. Ceribelli, S. Pittaluga, G. Wright, and L.M. Staudt. 2014. Oncogenic mechanisms in Burkitt lymphoma. Cold Spring Harb. Perspect. Med. 4:a014282. https://doi.org/10.1101/cshperspect.a014282 Schuhmacher, M., F. Kohlhuber, M. H¨ olzel,C. Kaiser, H. Burtscher, M. Jarsch, G.W. Bornkamm, G. Laux, A. Polack, U.H. Weidle, and D. Eick. 2001. The transcriptional program of a human B cell line in response to Myc. Nucleic Acids Res. 29:397–406. https://doi.org/10.1093/nar/29.2.397 Shih, T.A., E. Meffre, M. Roederer, and M.C. Nussenzweig. 2002. Role of BCR affinity in T cell dependent antibody responses in vivo. Nat. Immunol. 3: 570–575. https://doi.org/10.1038/ni803 Shlomchik, M.J., W. Luo, and F. Weisel. 2019. Linking signaling and selection in the germinal center. Immunol. Rev. 288:49–63. https://doi.org/10.1111/ imr.12744 Shulman, Z., A.D. Gitlin, J.S. Weinstein, B. Lainez, E. Esplugues, R.A. Flavell, J.E. Craft, and M.C. Nussenzweig. 2014. Dynamic signaling by T follicular helper cells during germinal center B cell selection. Science. 345: 1058–1062. https://doi.org/10.1126/science.1257861 Sicinski, P., J.L. Donaher, Y. Geng, S.B. Parker, H. Gardner, M.Y. Park, R.L. Robker, J.S. Richards, L.K. McGinnis, J.D. Biggers, et al. 1996. Cyclin D2 is an FSH-responsive gene involved in gonadal cell proliferation and oncogenesis. Nature. 384:470–474. https://doi.org/10.1038/384470a0 Sigl, R., C. Wandke, V. Rauch, J. Kirk, T. Hunt, and S. Geley. 2009. Loss of the mammalian APC/C activator FZR1 shortens G1 and lengthens S phase but has little effect on exit from mitosis. J. Cell Sci. 122:4208–4217. https://doi.org/10.1242/jcs.054197 Solvason, N., W.W. Wu, N. Kabra, X. Wu, E. Lees, and M.C. Howard. 1996. Induction of cell cycle regulatory proteins in anti-immunoglobulinstimulated mature B lymphocytes. J. Exp. Med. 184:407–417. https://doi .org/10.1084/jem.184.2.407 Solvason, N., W.W. Wu, D. Parry, D. Mahony, E.W. Lam, J. Glassford, G.G. Klaus, P. Sicinski, R. Weinberg, Y.J. Liu, et al. 2000. Cyclin D2 is essential for BCR-mediated proliferation and CD5 B cell development. Int. Immunol. 12:631–638. https://doi.org/10.1093/intimm/12.5.631 Sonoki, T., L. Harder, D.E. Horsman, L. Karran, I. Taniguchi, T.G. Willis, S. Gesk, D. Steinemann, E. Zucca, B. Schlegelberger, et al. 2001. Cyclin D3 is a target gene of t(6;14)(p21.1;q32.3) of mature B-cell malignancies. Blood. 98:2837–2844. https://doi.org/10.1182/blood.V98.9.2837 Spencer, S.L., S.D. Cappell, F.C. Tsai, K.W. Overton, C.L. Wang, and T. Meyer. 2013. The proliferation-quiescence decision is controlled by a bifurcation in CDK2 activity at mitotic exit. Cell. 155:369–383. https://doi.org/ 10.1016/j.cell.2013.08.062 Stuart, T., A. Butler, P. Hoffman, C. Hafemeister, E. Papalexi, W.M. Mauck III, Y. Hao, M. Stoeckius, P. Smibert, and R. Satija. 2019. Comprehensive Integration of Single-Cell Data. Cell. 177:1888–1902.e21. https://doi.org/ 10.1016/j.cell.2019.05.031 Subramanian,A.,P.Tamayo,V.K.Mootha,S.Mukherjee,B.L.Ebert,M.A.Gillette, A. Paulovich, S.L. Pomeroy, T.R. Golub, E.S. Lander, and J.P. Mesirov. 2005. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA. 102:15545–15550. https://doi.org/10.1073/pnas.0506580102 Tas, J.M., L. Mesin, G. Pasqual, S. Targ, J.T. Jacobsen, Y.M. Mano, C.S. Chen, J.C. Weill, C.A. Reynaud, E.P. Browne, et al. 2016. Visualizing antibody affinity maturation in germinal centers. Science. 351:1048–1054. https:// doi.org/10.1126/science.aad3439 Tirosh,I.,B.Izar,S.M.Prakadan,M.H.WadsworthII,D.Treacy,J.J. Trombetta, A. Rotem, C. Rodman, C. Lian, G. Murphy, et al. 2016. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science. 352:189–196. https://doi.org/10.1126/ science.aad0501 Trombetta, J.J., D. Gennert, D. Lu, R. Satija, A.K. Shalek, and A. Regev. 2014. Preparation of Single-Cell RNA-Seq Libraries for Next Generation Sequencing. Curr. Protoc. Mol. Biol. 107:1–17. https://doi.org/10.1002/ 0471142727.mb0422s107 Victora, G.D., and H. Mouquet. 2018. What Are the Primary Limitations in B-Cell Affinity Maturation, and How Much Affinity Maturation Can We Drive with Vaccination? Lessons from the Antibody Response to HIV-1. Cold Spring Harb. Perspect. Biol. 10:a029389. https://doi.org/10.1101/ cshperspect.a029389 Victora, G.D., and M.C. Nussenzweig. 2012. Germinal centers. Annu.Rev.Immunol. 30:429–457. https://doi.org/10.1146/annurev-immunol-020711-075032 Victora, G.D., T.A. Schwickert, D.R. Fooksman, A.O. Kamphorst, M. MeyerHermann, M.L. Dustin, and M.C. Nussenzweig. 2010. Germinal center dynamics revealed by multiphoton microscopy with a photoactivatable fluorescent reporter. Cell. 143:592–605. https://doi.org/10.1016/j.cell .2010.10.032 Victora, G.D., D. Dominguez-Sola, A.B. Holmes, S. Deroubaix, R. Dalla-Favera, and M.C. Nussenzweig. 2012. Identification of human germinal center light and dark zone cells and their relationship to human B-cell lymphomas. Blood. 120:2240–2248. https://doi.org/10.1182/blood-2012-03 -415380 Wang, H., H. Yang, C.S. Shivalila, M.M. Dawlaty, A.W. Cheng, F. Zhang, and R. Jaenisch. 2013. One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering. Cell. 153: 910–918. https://doi.org/10.1016/j.cell.2013.04.025 Yang, H., H. Wang, C.S. Shivalila, A.W. Cheng, L. Shi, and R. Jaenisch. 2013. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell. 154:1370–1379. https://doi.org/10.1016/j.cell.2013.08.022 Pae et al. Journal of Experimental Medicine 17 of 17 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Supplemental material Figure S1. Synchronization and enrichment of Ly75 +/+ cells with DEC-OVA immunization (related to Fig. 1). (A) Experimental setup for DECOVA–induced positive selection of Ly75 +/+ GC B cells followed by double labeling with EdU/BrdU to assay for S phase entry. (B) Expansion of Ly75 +/+ cells over Ly75 −/− over the time course. (C and D) DZ and LZ staining of Ly75 +/+ and Ly75 −/− cells over the time course (C), quantified in D. (E and F) S phase entry of Ly75 +/+ and Ly75 −/− B1-8 hi cells in the LZ (E), quantified in F. **, P < 0.01; ***, P < 0.001; n.s., not significant; paired ttest comparison between Ly75 +/+ and Ly75 −/− cells in the same animal. Bars indicate median. Each circle represents a mouse. Data are pooled from two independent experiments. Untr., untreated. Pae et al. Journal of Experimental Medicine S1 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure S2. GSEA between the S phase clusters (related to Fig. 2). (A) Top 10 most significantly varied “hallmark”signatures from the MSigDB database using GSEA to compare clusters 2, 3, and 6 (C2, C3, and C6). Pathways that are common among three sets of comparisons are indicated in bold. (B) Expression of mTORC1 signature in clusters 2, 3, and 6 (left) or in DEC-OVA time points with indicated zonal phenotypes (right). (C) Summary of P values using Mann–Whitney Utest. ***, P < 0.001; n.s., not significant; P adj , adjusted P value. Pae et al. Journal of Experimental Medicine S2 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure S3. Cyclin D2 is dispensable for inertial B cell cycling (related to Fig. 3). (A–C) Expression of Ccnd2 in UMAP dimension (A), over time after DECOVA immunization (B), and in LZ or DZ (C). P values in B are from a Kruskal–Wallis test with Dunn’s multiple comparisons test. Other significant P values are <0.001 (12 × 60 h), 0.0013 (12 vs. 30 h), and 0.037 (Ly75 −/− vs. 12 h). (D) CRISPR/Cas9-mediated gene targeting strategy to introduce a 4-bp deletion and premature stop codon in Ccnd2.(E and F) Staining for B1 (left) and B1-a (right) cells isolated from peritoneal cavities of WT (Ccnd2 +/+ )orcyclinD2–deficient (Ccnd2 −/− ; E) mice, quantified in F. (G) Experimental setup for induction of GCs containing WT (Ccnd2 +/+ )orcyclinD2mutant(Ccnd2 −/− )B1-8 hi cells. (H–L) DZ and LZ staining (H) and S phase entry (J) in WT (Ccnd2 +/+ )orcyclinD2–deficient (Ccnd2 −/− )B1-8 hi GC cells 12 d after NP-OVA immunization, quantified in I, K, and L. (M) Experimental setup for induction of GCs containing mixtures of WT (Ccnd2 +/+ ) or cyclin D2 mutant (Ccnd2 −/− )B1-8 hi cells. (N) Clonal expansion of WT (Ccnd2 +/+ ) or cyclin D2 mutant (Ccnd2 −/− )B1-8 hi cells over time, relative to day 7 (represented by a dotted line). *, P < 0.05; **, P < 0.01; n.s., nonsignificant, nonparametric Mann–Whitney test, compared with WT (Ccnd2 +/+ ; G, I, and J) or day 7 (L). Bars indicate median. Each circle representsa mouse. Data are pooled from two independent experiments. P adj , adjusted P value. Pae et al. Journal of Experimental Medicine S3 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure S4. Analysis of Ccnd3 T283A/+ mice (related to Fig. 6). (A and B) Bone marrow (A) and splenic (B) B cells development in distribution WT (Ccnd3 +/+ )or Ccnd3 T283A/+ mice (left) and quantification (right). FO, follicular; MZ, marginal zone. Bars indicate SD. (C and D) Staining (left) and quantification (right) of splenic B cells (C) and GC (D) in WT (Ccnd3 +/+ )orCcnd3 T283A/+ mice. Bars indicate median. SSC-A, side scatter area. For A–D, 6-wk-old mice (n= 4 for each genotype) were sacrificed. Ccnd3 T283A/+ mice used are generation N8. (E–G) Staining (left) and quantification (right) for GC (E), DZ/LZ (F), and S phase entry (G) of GC B cells induced upon immunization with KLH in pLNs from WT (Ccnd3 +/+ )orCcnd3 T283A/+ mice. **, P < 0.01; n.s., nonsignificant, nonparametric Mann–Whitney test. Bars indicate median. Each circle represents a mouse. Data pooled from two independent experiments. Ccnd3 T283A/+ mice used in this experiment are from generations N6 and N7. Pae et al. Journal of Experimental Medicine S4 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Figure S5. Analysis of Ccnd3 T283A/+ mice (related to Fig. 7). (A) Spleens from WT (Ccnd3 +/+ )orCcnd3 T283A/+ animals (right) and quantification of spleen to body weight ratios (left). Bars indicate median. (B) BandTcellstaininginWT(Ccnd3 +/+ )orCcnd3 T283A/+ mice revealed expansion of abnormal lymphocyte populations with diminished or no expression of TCRb/B220. (C) H&E and peanut agglutinin (PNA) staining of WT (Ccnd3 +/+ )orCcnd3 T283A/+ spleens revealed disrupted morphology. (D) H&E staining of nonlymphoid tissues from WT (Ccnd3 +/+ )orCcnd3 T283A/+ revealed lymphocytic infiltration. Pae et al. Journal of Experimental Medicine S5 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021 Table S1 and Table S2 are provided online as separate Word files. Table S1 lists samples included in scRNA-seq analysis. Table S2 lists antibodies used in the study. Data S1 and Data S2 are provided online as separate Excel files. Data S1 shows a list of cell cycle genes, Myc signature genes, E2F target genes, and mTOCR1 target genes used in scRNA-seq analysis. Data S2 shows Igh sequences obtained from Ccnd3 T283A/+ mice. Pae et al. Journal of Experimental Medicine S6 Cyclin D3 drives inertial cell cycling in germinal centers https://doi.org/10.1084/jem.20201699 Downloaded from http://rupress.org/jem/article-pdf/218/4/e20201699/1406609/jem_20201699.pdf by Helmholtz-Zentrum Fuer Infektionsforschung - Bibliothekuser on 07 January 2021