ARTICLE
Cyclin D3 d i es ine ial cell cycling in da k zone
ge minal cen e B cells
Juhee Pae
1
, Jona an E sching
1
, Tiago B.R. Cas o
1
, Ma a Schips
2
, Luka Mesin
1
, Samuel J. Allon
3,4,5
, Jose O do as-Mon anes
5,6,7,8
,
Co aline Mlyna czyk
9
,A iMelnick
9
, Alejo E eyan
10
, Alex K. Shalek
3,4,5,7,8
, Michael Meye -He mann
2,11,12
, and Gab iel D. Vic o a
1

Du ing a ini y ma u a ion, ge minal cen e (GC) B cells al e na e be ween p oli e a ion and soma ic hype mu a ion in he
da k zone (DZ) and a ini y-dependen selec ion in he ligh zone (LZ). This ana omical seg ega ion imposes ha he igo ous
p oli e a ion ha allows clonal expansion o posi i ely selec ed GC B cells akes place os ensibly in he absence o he signals
ha igge ed selec ion in he LZ, as i by “ine ia.”We ind ha such ine ial cycles speci ically equi e he cell cycle egula o
cyclin D3. Cyclin D3 dose-dependen ly con ols he ex en o which B cells p oli e a e in he DZ and is essen ial o e ec i e
clonal expansion o GC B cells in esponse o s ong T ollicula helpe (T h) cell help. In oduc ion in o he Ccnd3 gene o a
Bu ki lymphoma–associa ed gain-o - unc ion mu a ion (T283A) leads o la ge GCs wi h inc eased DZ p oli e a ion and, in
olde mice, clonal B cell lymphop oli e a ion, sugges ing ha he DZ ine ial cell cycle p og am can be coop ed by B cells
unde going malignan ans o ma ion.
In oduc ion
Ge minal cen e s (GCs) a e he si es o a ini y ma u a ion, he
p ocess by which an ibodies imp o e hei a ini y o an igen
o e ime (Cys e and Allen, 2019;De Sil a and Klein, 2015;
Eisen, 2014;Mesin e al., 2016;Rajewsky, 1996;Shlomchik e al.,
2019;Vic o a and Nussenzweig, 2012). Fo e icien a ini y
ma u a ion, GC B cells mus cycle be ween wo majo an-
sc ip ional s a es, associa ed wi h localiza ion o B cells o each o
he wo mic oana omical “zones”o he GC. When in he da k
zone (DZ), B cells p oli e a e igo ously and mu a e hei im-
munoglobulin genes by soma ic hype mu a ion (SHM). A e
ansi ion o he ligh zone (LZ), B cells bea ing ad an ageous
mu a ions a e selec i ely d i en o clonally expand, based a
leas in pa on hei abili y o bind and p esen an igen o GC-
esiden T ollicula helpe (T h) cells (Vic o a e al., 2010).
Successi e cycles o SHM and a ini y-based selec ion ul ima ely
en ich o highe -a ini y cells among he GC B cell popula ion in
a p ocess known as cyclic een y (MacLennan, 1994;Vic o a
and Nussenzweig, 2012).
A unique consequence o he ana omical compa men aliza-
ion o he GC is ha mi ogenic signals a e seg ega ed om he
p oli e a ion hey induce. Upon posi i e selec ion, B cells ypi-
cally ansi ion om G1 o S phase o he cell cycle in he LZ,
mig a e om LZ o DZ while in S phase, and unde go G2 and M
phases in he DZ (Gi lin e al., 2014;Vic o a e al., 2010). A e
his i s di ision, cell cycling con inues in he DZ, wi h mos
B cells unde going on a e age wo addi ional cell cycles be o e
e u ning o he LZ o u he selec ion (Gi lin e al., 2014). GC
B cells ha ecei e s onge signals om T h cells in he LZ,
howe e , can unde go a much g ea e numbe o p oli e a i e
cycles in he DZ, esul ing in exponen ial clonal expansion
(Gi lin e al., 2014;Meye -He mann e al., 2012;Vic o a e al.,
2010). A i s ex eme, his egula ed expansion can lead o
“clonal bu s s,”in which a single B cell can ake o e a 2,000-cell
GC in he cou se o a ew days (Tas e al., 2016). These bu s s a e
associa ed wi h massi e di e si ica ion o gene ally highe -
a ini y SHM a ian s and as such a e likely o play an impo -
an ole in he gene a ion o high-a ini y B cell clones (Ami ai
e al., 2017;Banna d and Cys e , 2017;Mesin e al., 2016).
Despi e he impo ance o DZ p oli e a ion o GC B cell se-
lec ion and a ini y ma u a ion, ou unde s anding o GC B cell
.............................................................................................................................................................................
1
Labo a o y o Lymphocy e Dynamics, The Rocke elle Uni e si y, New Yo k, NY;
2
Depa men o Sys ems Immunology and B aunschweig In eg a ed Cen e o Sys ems
Biology, Helmhol z Cen e o In ec ion Resea ch, B aunschweig, Ge many;
3
Ins i u e o Medical Enginee ing and Science, Depa men o Chemis y, Koch Ins i u e o
In eg a i e Cance Resea ch, Massachuse s Ins i u e o Technology, Camb idge, MA;
4
Ragon Ins i u e o Massachuse s Gene al Hospi al, Massachuse s Ins i u e o
Technology and Ha a d, Camb idge, MA;
5
B oad Ins i u e o Massachuse s Ins i u e o Technology and Ha a d, Camb idge, MA;
6
Di ision o Gas oen e ology, Bos on
Child en’s Hospi al, Bos on, MA;
7
P og am in Immunology Ha a d Medical School, Bos on, MA;
8
Ha a d S em Cell Ins i u e, Camb idge, MA;
9
Depa men o Medicine,
Di ision o Hema ology and Medical Oncology, Weill Co nell Medicine, New Yo k, NY;
10
Spanish Na ional Cance Resea ch Cen e , Mad id, Spain;
11
Ins i u e o
Biochemis y, Bio echnology and Bioin o ma ics, Technische Uni e si ¨
a B aunschweig, B aunschweig, Ge many;
12
Clus e o Excellence RESIST (EXC 2155), Hanno e
Medical School, Hanno e , Ge many.
D . E sching died in Oc obe 2020; Co espondence o Gab iel D. Vic o a: ic o a@ ocke elle .edu.
© 2020 Pae e al. This a icle is dis ibu ed unde he e ms o an A ibu ion–Noncomme cial–Sha e Alike–No Mi o Si es license o he i s six mon hs a e he
publica ion da e (see h p://www. up ess.o g/ e ms/). A e six mon hs i is a ailable unde a C ea i e Commons License (A ibu ion–Noncomme cial–Sha e Alike 4.0
In e na ional license, as desc ibed a h ps://c ea i ecommons.o g/licenses/by-nc-sa/4.0/).
Rocke elle Uni e si y P ess h ps://doi.o g/10.1084/jem.20201699 1o 17
J. Exp. Med. 2020 Vol. 218 No. 4 e20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
selec ion has his o ically ocused on e en s ha ake place in he
LZ (Mesin e al., 2016;Shlomchik e al., 2019). Consequen ly, he
p ecise mechanisms ha allow DZ p oli e a ion o ake place in
he appa en absence o di ec mi ogenic signals a e s ill in-
comple ely unde s ood. Fo ins ance, upon in e ac ion wi h T h
cells, posi i ely selec ed LZ B cells exp ess he ansc ip ion
ac o c-Myc (Calado e al., 2012;Dominguez-Sola e al., 2012;
Finkin e al., 2019), as well as he mechanis ic a ge o apa-
mycin (mTOR)–media ed anabolic p og am (E sching e al.,
2017). While bo h c-Myc and mTOR a e equi ed o LZ B cells
o mig a e o he DZ and en e he p oli e a i e phase, he in-
duc ion o hese pa hways appea o be mos ly es ic ed o he
LZ (Dominguez-Sola e al., 2012;E sching e al., 2017;Finkin
e al., 2019). This sugges s ha DZ B cells e ain a memo y o
he in ensi y o he c-Myc and mTOR complex 1 (mTORC1)–
dependen “cha ge” hey ecei ed p e iously in he LZ and ha
hey subsequen ly ansla e his memo y in o he numbe o cell
cycles hey will unde go in he DZ using a cell-in insic “ ime ”
o “coun e ”(Banna d e al., 2013;Gi lin e al., 2014). The mo-
lecula pa hways ha di ec ly con ol he numbe o cycles a GC
B cell unde goes in he DZ emain uncha ac e ized.
He e, we used a combina ion o unbiased whole- ansc ip ome
single-cell mRNA sequencing (scRNA-seq) analysis and a ge ed
gene ic and pha macological manipula ion o cell cycle egu-
la o s o in es iga e he molecula na u e o DZ cell cycles. We
ind ha DZ B cells adop a dis inc E2F
high
/c-Myc
low
mode o
cell cycling ha allows apid and con inuous p oli e a ion in
he absence o ex e nal mi ogenic signals. We show ha he
cell cycle egula o cyclin D3, p e iously shown o be equi ed
o GC o ma ion and main enance (Ca o e al., 2011;Peled
e al., 2010), is a speci ic, dose-dependen con olle o his
pheno ype and DZ cell cycling and ha loss o cyclin D3 canno
be o e come by LZ cycling induced by s ongly inc eased T h
cell help. In oduc ion in o mice o a gain-o - unc ion mu a-
ion in cyclin D3 de i ed om human Bu ki lymphoma
(Schmi z e al., 2012,2014) leads o exace ba ed B cell p o-
li e a ion speci ically in he GC DZ and de elopmen o clonal
pos -GC B cell expansions, linking he ine ial p oli e a i e
p og am o malignan ans o ma ion.
Resul s
B cell p oli e a ion in he DZ is T h cell independen
To be e unde s and GC B cell cycling in he DZ, we i s sough
o o mally de e mine whe he S phase en y by DZ B cells e-
qui es acu e signals om T h cells in addi ion o hose deli e ed
in he LZ. To his end, we used a synch onized selec ion model
(E sching e al., 2017;Vic o a e al., 2010) o allow kine ically
p ecise blocking o T h cell help o GC B cells a se ime poin s
a e induced posi i e selec ion (Fig. 1 A). In his model, we use
adop i e ans e o 4-hyd oxy-3-ni o-phenylace yl (NP)–
speci ic B1-8
hi
B cells ollowed by immuniza ion wi h NP con-
juga ed o chicken OVA (NP-OVA) o c ea e GCs in which he
majo i y o B cells lack he su ace ecep o DEC-205 (encoded
by he gene Ly75). T ea men o mice wi h ongoing GCs wi h an
an ibody o DEC-205 used o OVA (DEC-OVA) leads o p e-
sen a ion o OVA pep ides by Ly75
+/+
GC B cells only. This
induces Ly75
+/+
cells o p e e en ially in e ac wi h T h cells,
igge ing hei posi i e selec ion (Pasqual e al., 2015;Vic o a
e al., 2010). In his p ocess, GC B cells also become synch o-
nized: 12 h a e DEC-OVA, mos Ly75
+/+
cells a e loca ed in he
LZ, whe e hey in e ac ex ensi ely wi h T h cells (E sching
e al., 2017;Shulman e al., 2014). By 36 h a e DEC-OVA,
mos Ly75
+/+
cells ha e ansi ioned o he DZ, whe e hey begin
hei p oli e a i e bu s (Vic o a e al., 2010). Fig. 1, A–C;and
Fig. S1 p o ide an o e iew o key ime poin s in his kine ics.
En y in o S phase o he cell cycle, e ealed by double-pulsing
mice wi h 5-e hynyl-2-deoxyu idine (EdU) and B dU nucleo-
ides (Gi lin e al., 2014), was g ea ly inc eased in DZ B cells a
36 h a e DEC-OVA, indica ing ha , a his ime poin , GC B cells
p og ess h ough he G1-S checkpoin while in he DZ com-
pa men (Fig. 1, B and C;andFig. S1, E and F).DZSphaseen y
pe sis ed abo e he s eady-s a e le el un il a leas 60 h a e
DEC-OVA ea men , when Ly75
+/+
cells emained p edomi-
nan ly in he DZ and con inued o expand (Fig. 1, B and C;and
Fig. S1, B–D).
To de e mine whe he T cell help is con inually equi ed o
DZ p oli e a ion, we acu ely blocked T h cell–B cell in e ac ion
ei he a he ime o he ini ial T h signal deli e y in he LZ (6 h
a e DEC-OVA) o ollowing he ansi ion o posi i ely selec ed
B cells o he DZ bu be o e he p oli e a i e bu s (30 h a e
DEC-OVA; Fig. 1 D). As expec ed om he ole o T h cell–
media ed signaling in his model (Vic o a e al., 2010), ea ly
blocking o T cell–B cell in e ac ions in he LZ using an ibodies o
MHC class II o CD40 ligand (CD40L) e ec i ely p e en ed clo-
nal expansion o Ly75
+/+
cells by 48 h a e DEC-OVA ea men
(Fig. 1, E and F). By con as , blocking ei he pa hway a e B cells
ansi ioned o he DZ had li le, i any, e ec on he expansion o
Ly75
+/+
cells o e Ly75
−/−
cells (Fig. 1, E and F)o heabili yo
Ly75
+/+
DZ B cells o en e S phase, as e idenced by EdU/B dU
inco po a ion (Fig. 1, G and H). We conclude ha sus ained
p oli e a ion o B cells in he GC DZ upon posi i e selec ion does
no equi e con inuous help om T h cells. We he e o e e e o
DZ cell cycles as “ine ial,”since hey p oceed in a cell-in insic
ashion acco ding o he s eng h o he ini ial “push” om T h
cells, in con as o he “p ima y”cell cycles ha B cells unde go
immedia ely downs eam o selec i e signals in he LZ.
Ine ial cycling is sus ained by p olonged E2F ac i a ion
ollowing a decay in c-Myc ac i i y
To iden i y he ansc ip ional p og ams associa ed wi h ine ial
cycling, we pe o med scRNA-seq on GC B cells a di e en ime
poin s a e o cing posi i e selec ion using DEC-OVA. We
index-so ed single Ly75
+/+
B cells using an ibodies o LZ/DZ
ma ke s, and so ed cells we e sequenced using he Sma -Seq2
p o ocol (T ombe a e al., 2014). We assayed GC B cells i s a
12 h a e DEC-OVA, when Ly75
+/+
B cells a e en iched in he LZ
in he p ocess o ecei ing cogna e help om T h cells, and hen
a 30, 46, and 60 h a e DEC-OVA, as DZ B cells ansi ion om
signal-dependen p oli e a ion o p edominan ly ine ial modes
o cycling be o e e u ning o he LZ be ween 72 and 96 h a e
DEC-OVA (Fig. 1, A–D;andFig. S1;Vic o a e al., 2010). Fo
compa ison, we also included a sample o Ly75
−/−
coun e -
selec ed LZ cells om he 12-h ime poin (Table S1).
Pae e al. Jou nal o Expe imen al Medicine 2o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Figu e 1. Cell cycle en y by DZ B cells does no equi e T h cell help. (A) Expe imen al se up o DEC-OVA–induced Ly75
+/+
GC B cells ollowed by double
labeling wi h EdU/B dU o assay o S phase en y. (B) S phase en y o Ly75
+/+
and Ly75
−/−
B1-8
hi
cells in he DZ, quan i ied in C. (D) Expe imen al se up o
DEC-OVA–induced posi i e selec ion o Ly75
+/+
GC B cells ollowed by inhibi ion o T-B in e ac ion using an i-MHC class II o an i-CD40L be o e (ea ly) o a e
(la e) DZ een y. (E and F) E ec on GC size (le ) and expansion o Ly75
+/+
cells o e Ly75
−/−
(cen e , igh ) upon an i-MHC class II ea men (E) o an i-CD40L
ea men (F). (G and H) S phase en y o Ly75
+/+
B1-8
hi
cells in he DZ upon la e ea men wi h an i-MHC class II (G) o an i-CD40L (H). *, P < 0.05; **, P <
0.01; ***, P < 0.001; n.s., nonsigni ican , a pai ed es compa ison be ween Ly75
+/+
and Ly75
−/−
cells in he same animal (C) o nonpa ame ic Mann–Whi ney
es compa ed wi h he g oup ea ed wi h an iso ype con ol (E–H). Ba s indica e median. Each ci cle ep esen s a mouse. Da a a e pooled om wo (B and C),
h ee (E and G), o wo (F and H) independen expe imen s. Iso, iso ope.
Pae e al. Jou nal o Expe imen al Medicine 3o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
The 1,220 cells ha passed ou quali y con ol h esholds ell
in o se en majo clus e s (Fig. 2 A). These we e de e mined in
la ge pa by cell cycle phase as in e ed om hei ansc ip-
ional p o ile (Ti osh e al., 2016), wi h a lesse con ibu ion o
hei LZ/DZ pheno ype as de ined by su ace s aining (Fig. 2 B).
Clus e s 0 and 4 con ained p ima ily G1 cells; clus e s 2, 3, and 6
we e en iched in S phase cells; and clus e s 1 and 5 we e en-
iched in cells in he G2 and M phases (Fig. 2 C). To ollow he
e olu ion o posi i ely selec ed B cells ac oss hese clus e s as
hey ansi ioned om eac i e o ine ial cell cycles and hen o
quiescence, we compa ed Ly75
+/+
and Ly75
−/−
cells in he LZ a
12 h a e DEC-OVA o Ly75
+/+
cells in he DZ a he 30-, 46-, and
60-h ime poin s. This e ealed a ma ked inc ease in ep e-
sen a ion o cells in clus e 2, as Ly75
+/+
cells a e posi i ely se-
lec ed in he LZ a 12 h (emp y a owheads in Fig. 2 D).
En ichmen in clus e 2 con inued as cells ansi ioned o he DZ
Figu e 2. Single-cell ansc ip omic analysis o GC B cells unde going posi i e selec ion. (A) Uni o m mani old app oxima ion and p ojec ion (UMAP) plo
displaying 1,220 cells colo ed by sha ed nea es neighbo clus e s collec ed (de aul Wilcox es ; see Ma e ials and me hods). Cells we e collec ed om ou
independen expe imen s (see Table S1). (B) Dis ibu ion o cell cycle phase (le ) and LZ/DZ pheno ypes ( igh ). (C) Dis ibu ion o cell cycle phase in clus e s.
(D) Changes in dis ibu ion o clus e s g ouped by cell cycle phase o e he DEC-OVA–induced selec ion ime cou se. (E and F) Exp ession o Myc mRNA,
c-Myc, and E2F a ge gene signa u es in clus e s 2, 3, and 6 (E) o DEC-OVA ime poin s wi h indica ed zonal pheno ypes (F). Do ed line indica es he
h eshold used o quan i ica ion in G. See Fig. S2 C o a comple e lis o P alues. (H) S phase en y o Ly75
+/+
B1-8
hi
cells in he DZ 12 h a e ea men wi h
palbociclib, a CDK4/6 inhibi o , o ehicle (PBS). (I) Quan i ica ion includes all LZ and DZ B1-8
hi
cells, in addi ion o posi i ely selec ed Ly75
+/+
B1-8
hi
cells in he
DZ. **, P < 0.01; ****, P < 0.0001, nonpa ame ic Mann–Whi ney es , compa ed wi h he PBS- ea ed con ol g oup. Ba s indica e median. Each ci cle
ep esen s a mouse (H and I) o a cell (all o he igu es). Da a a e pooled om wo independen expe imen s (H and I).
Pae e al. Jou nal o Expe imen al Medicine 4o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
a 30 h, a e which he dominan S phase pheno ype shi ed o
clus e 6 a 46 h and hen clus e 3 a 60 h a e DEC-OVA ( illed
a owheads in Fig. 2 D). Pai wise compa isons o he h ee S
phase clus e s by gene se en ichmen analysis (GSEA) using he
“hallma k”signa u es om he MSigDB da abase (Moo ha e al.,
2003;Sub amanian e al., 2005) e ealed exp ession o E2F
and c-Myc a ge genes as he mos consis en di e ences be-
ween he h ee clus e s (Fig. S2 A). Clus e 2, en iched in LZ-
pheno ype cells om he 12- and 30-h ime poin s, showed high
exp ession o Myc and o c-Myc–and mTORC1-dependen
ansc ip ional signa u es (Fig. 2 E and Fig. S2 B;Peng e al.,
2002;Schuhmache e al., 2001). Gi en he s ong associa ion
be ween hese signa u es and posi i e selec ion (Calado e al.,
2012;Dominguez-Sola e al., 2012;E sching e al., 2017;Finkin
e al., 2019;Luo e al., 2018;Vic o a e al., 2010), his pa e n
sugges s ha clus e 2 consis s p ima ily o ecen ly selec ed
B cells unde going p ima y S phase, ei he on hei way o o
soon a e DZ een y. Cells in clus e s 3 and 6, on he o he
hand, showed educed le els o Myc and c-Myc–and mTORC1-
dependen gene signa u es while main aining exp ession o E2F
a ge genes (Fig. 2 E and Fig. S2 B), as expec ed i hey we e
en e ing he cell cycle by ine ia while being physically seg e-
ga ed om mi ogenic signals loca ed in he LZ. Kine ic analysis
o Myc and c-Myc and E2F a ge gene signa u es con i med ha
E2F a ge gene exp ession le els emained unal e ed be ween
30 and 46 h a e DEC-OVA, while mRNA exp ession and ac-
i i y o Myc dec eased (Fig. 2, F and G;andFig. S2 C). E2F
signa u es began o subside only a he 60-h ime poin , as in-
e ial cycling de ined by DZ S phase en y (Fig. 1, B and C)is
al eady subsiding. Thus, unlike hei LZ coun e pa s, DZ B cells
appea o engage in a dis inc mode o cell cycling ha does no
equi e con inued exp ession o he Myc gene o i s p o ein
unc ion o main ain E2F ac i i y. This sugges s ha egula o s
o E2F ha a e downs eam o c-Myc and o he selec ion-
dependen signals may be equi ed o sus ain p oli e a ion a -
e posi i ely selec ed GC B cells ansi ion o he DZ.
A majo egula o o E2F ac i i y a e he D- ype cyclins,
which in pa ne ship wi h cyclin-dependen kinases (CDKs) 4
and 6, ac i a e E2F by phospho yla ion o i s nega i e egula o
RB (Musg o e e al., 2011). To es whe he D- ype cyclins could
be esponsible o allowing he p og ession o ine ial cell cycles,
we ea ed mice wi h he inhibi o o CDK4/6 palbociclib 36 h
a e inducing posi i e selec ion o GC B cells wi h DEC-OVA and
analyzed cell cycle p og ession 12 h la e . Unlike blockade o
MHC class II o CD40L (Fig. 1, D–H), palbociclib ea men
s ongly inhibi ed S phase en y in all GC B cells, including hose
unde going ine ial cycling in he DZ (Fig. 2, H and I). Thus,
ine ial S phase en y, while no dependen on T h-media ed
signals, s ill equi es ac i i y o CDK4/6.
Ccnd3, bu no Ccnd2, is equi ed o DZ ine ial cycling
Since B cells exp ess exclusi ely cyclins D2 and D3 (encoded by
Ccnd2 and Ccnd3, espec i ely) upon mi ogenic s imula ion (Reid
and Snow, 1996;Sol ason e al., 1996), we sough o de e mine
he ela i e con ibu ion o hese wo cyclins o he CDK4/6
dependency o ine ial cycles. Ccnd2 mRNA was de ec able p i-
ma ily in he subse o cells unde going S phase in he LZ in he
c-Myc
high
clus e 2 (Fig. S3 A). In ag eemen wi h ou p e ious
epo s, Ccnd2 was highe in Ly75
+/+
cells a 12 h a e DEC-OVA
as well as in he LZ in gene al (Dominguez-Sola e al., 2012;
Vic o a e al., 2010;Fig. S3, B and C). To in es iga e he unc ion
o cyclin D2 in GC B cells, we gene a ed Ccnd2 knockou mice
using CRISPR-Cas9–media ed genome edi ing in zygo es o in-
oduce a 4-bp dele ion/ ameshi in exon 1 o he gene (Fig.
S3 D). Ccnd2
−/−
mice lacked pe i oneal B-1a cells (Fig. S3, E and
F), and emales we e unable o p oduce p ogeny (da a no shown),
as shown p e iously using an independen ly gene a ed knockou
s ain (Sicinski e al., 1996;Sol ason e al., 2000), con i ming ha
ou s is a null allele. We adop i ely ans e ed a 1:1 mix u e o
Ccnd2
+/+
and Ccnd2
−/−
B1-8
hi
cells in o WT hos s, which we hen
immunized wi h NP-OVA o gene a e GCs (Fig. S3 G), and ound
ha zonal dis ibu ion was p ese ed in Ccnd2
−/−
B1-8
hi
B cells
(Fig. S3, H and I). Howe e , a he han educing he abili y o GC
B cells o cycle, loss o cyclin D2 led o a small bu consis en in-
c ease in he p opo ion o B1-8
hi
cells en e ing S phase in bo h he
LZ and DZ (Fig. S3, J–L). Despi e his inc ease, absence o cyclin D2
showed no clea e ec on he compe i i eness o GC B cells o e
ime when compa ed wi h Ccnd2-su icien B cells wi hin he same
GC (Fig. S3, M and N). These expe imen s sugges ha al hough
cyclin D2 can a ec he abili y o GC B cells o en e cell cycle, i is
no equi ed o ine ial cycling in he DZ.
In con as o he dynamic beha io o Ccnd2, ou scRNA-seq
da ase showed ha Ccnd3 mRNA amoun s we e s ably high
h oughou he LZ/DZ cycle and o e ou ime cou se o DEC-
OVA–induced selec ion, wi h only sligh inc eases a he 30-h
ime poin and in he DZ in gene al (Fig. 3, A–C). Despi e hese
modes changes in mRNA exp ession, cyclin D3 p o ein le els
we e subs an ially highe in he DZ (Fig. 3 D), in ag eemen wi h
p e ious epo s based on his ology (Peled e al., 2010). Also
consis en wi h p io epo s (Ca o e al., 2011;Peled e al., 2010),
we ound ha loss o Ccnd3 led o d ama ically educed GC B cell
equency (Fig. 3, E and F). This is unlikely o be due o a gene al
de ec in p oli e a ion, gi en ha Ccnd3
−/−
B cells p oli e a e
no mally in i o and a p e-GC s ages in i o (Ca o e al., 2011;
Peled e al., 2010). Close examina ion o Ccnd3
−/−
GCs showed
ha dec eased GC size was p ima ily caused by a ma ked e-
duc ion in he p opo ion o B cells wi h a DZ pheno ype (Fig. 3,
GandH), sugges ing a s ong block in he abili y o Ccnd3
−/−
B cells o unde go ine ial cycling. This pheno ype was iden ical
when GCs we e gene a ed by adop i e ans e o Ccnd3
−/−
B1-8
hi
cells in o WT hos s (Fig. 3 I), esul ing in a signi ican ly de-
c eased DZ/LZ a io (Fig. 3, J and K) and an almos comple e loss
in S phase ini ia ion in he DZ wi h only a mino dec ease o
i ness in he LZ (Fig. 3, J and L–N). The equi emen o cyclin
D3 is he e o e GC B cell in insic and la gely es ic ed o
he DZ. Ccnd3
−/−
B1-8
hi
cells in he DZ also showed inc eased
o wa d sca e , a su oga e measu e o cell size (Fig. 3, O and P).
Since cell size inc eases be o e he ini ia ion o ine ial cycles
(E sching e al., 2017;Finkin e al., 2019), his inding sugges s
ha he inabili y o Ccnd3
−/−
B cells o unde go ine ial cycling
esul s in hei ailu e o lose cellula mass by di ision. We
conclude ha cyclin D3 is cell-in insically equi ed o GC
B cells o en e S phase speci ically in he DZ while being
dispensable o S phase en y immedia ely downs eam o
Pae e al. Jou nal o Expe imen al Medicine 5o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021

T cell–media ed selec ion in he LZ. This pa e n implica es
cyclin D3 as a non edundan media o o ine ial cell cycles.
Cyclin D3–media ed DZ ine ial cycling links GC posi i e
selec ion wi h clonal expansion
To de e mine he ela i e impo ance o LZ cycles s. ine ial DZ
cycles o clonal expansion, we asked whe he cell p oli e a ion
de ec s seen in Ccnd3
−/−
DZ B cells could be escued by o cing
s ong in e ac ion wi h T h cells using DEC-OVA (Fig. 4 A).
Whe eas Ccnd3
+/+
Ly75
+/+
B1-8
hi
cells p oli e a ed su icien ly o
ou numbe Ly75
−/−
cells by i e old o e a 60-h pe iod, clonal
expansion was much less e icien when Ly75
+/+
B1-8
hi
cells
lacked cyclin D3. Wi h he excep ion o one ou lie mouse, he
a io o Ly75
+/+
o Ly75
−/−
cells inc eased only sligh ly, i a all,
Figu e 3. Ine ial B cell cycling equi es cyclin D3. (A–C) Exp ession o Ccnd3 in UMAP dimension (A), o e ime a e DEC-OVA immuniza ion (B), and in LZ
o DZ (C). P alues in B a e om a K uskal–Wallis es wi h Dunn’s mul iple compa isons es . O he signi ican P alues a e <0.001 (12 × 60 h), 0.013 (Ly75
−/−
s. 30 h), and 0.015 (Ly75
−/−
s. 60 h). (D) Immunoblo s o whole-cell lysa es o LZ o DZ cells so ed om popli eal o mesen e ic LNs (pLN o mLN, e-
spec i ely). Molecula weigh is indica ed in kilodal ons. (E–H) S aining o GC (E) and DZ/LZ (G) in WT (Ccnd3
+/+
) o cyclin D3 mu an (Ccnd3
−/−
) mice ha we e
immunized s.c. in he hind oo pad o pLNs, quan i ied in F. (I) Expe imen al se up o induc ion o GCs con aining WT (Ccnd3
+/+
) o cyclin D3 mu an (Ccnd3
−/−
)
B1-8
hi
cells. (J–P) DZ and LZ s aining o all GC B cells (black) o cells en e ing S phase ( ed; J), S phase en y (L), o wa d sca e (O) o WT (Ccnd3
+/+
), o cyclin
D3 mu an (Ccnd3
−/−
)B1-8
hi
cells, quan i ied in K, M, N, and P. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; nonpa ame ic Mann–Whi ney es . Ba s indica e
median. Each ci cle ep esen s a cell (B and C) o a mouse (all o he igu es). Da a a e pooled om h ee (E–H) o wo (I–P) independen expe imen s. P
adj
,
adjus ed P alue; pos -imm, pos -immuniza ion.
Pae e al. Jou nal o Expe imen al Medicine 6o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Figu e 4. Inc eased T h cell help canno compensa e o loss o cyclin D3. (A) Expe imen al se up o DEC-OVA–induced posi i e selec ion o Ly75
+/+
cells
ha a e ei he WT (Ccnd3
+/+
)o cyclinD3mu an (Ccnd3
−/−
). No e ha due o o lack o compe i i eness in ea ly GC s ages, Ccnd3
−/−
B1-8
hi
cells ha e o be
ans e ed a a highe p opo ion han Ccnd3
+/+
B1-8
hi
cells. (B–E) Clonal expansion (B) and DZ/LZ s aining (D) o WT (Ccnd3
+/+
)o cyclinD3mu an (Ccnd3
−/−
)
Ly75
+/+
cells o e ime a e DEC-OVA immuniza ion, quan i ied in C and E. (F) Immuno luo escence showing he mig a ion o WT (Ccnd3
+/+
)o cyclinD3
mu an (Ccnd3
−/−
)Ly75
+/+
cells 36 h a e DEC-OVA immuniza ion in pLN GCs. Do ed a ea indica es he DZ ma ked by he absence o IgD (Nai e B cell ma ke )
Pae e al. Jou nal o Expe imen al Medicine 7o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
o e inpu le els (Fig. 4, B and C). Failu e o Ccnd3
−/−
B cells o
clonally expand was also accompanied by ailu e o accumula e
in he DZ (Fig. 4, D and E), al hough his ology showed ha
Ccnd3
−/−
cells we e able o a leas access he DZ ana omically a
36 h a e DEC-OVA (Fig. 4 F). E en so, he p opo ion o ea ly S
phase cells among he Ccnd3
−/−
popula ion d opped p ecipi ously
upon ansi ion om he LZ o DZ, as expec ed gi en he in-
abili y o cyclin D3–de icien B cells o sus ain p oli e a ion by
ine ia (Fig. 4, G–K). Thus, he esidual expansion o Ccnd3
−/−
Ly75
+/+
B1-8
hi
seen in some mice can be a ibu ed p ima ily o
cell cycling aking place in he LZ in esponse o con inued
signals deli e ed by T h cells. We conclude ha s ong signaling
om T h cells in esponse o DEC-OVA canno o e come he
equi emen o cyclin D3 in d i ing DZ p oli e a ion and ha
in he absence o ine ial cycling, LZ cell cycles igge ed di ec ly
in eac ion o T h cell–de i ed signals a e no su icien o sus-
ain la ge p oli e a i e bu s s.
Cyclin D3 con ols DZ ine ial cycling in a dose-dependen
manne
A po en ial mechanism o how he s eng h o he ini ial
Tcell–B cell in e ac ions in he LZ de e mines he numbe o
ounds o di ision ha selec ed B cells unde go in he DZ (Gi lin
e al., 2014) is ha DZ B cells ansla e he memo y o hei in-
e ac ions wi h T h cells in o p o ein amoun s o a cell cycle
egula o . I cyclin D3 ollows such a pa e n, hen i could be
p edic ed ha GC B cells wi h a highe capaci y o p oduce cyclin
D3 p o ein would be a a compe i i e ad an age due o inc eased
p oli e a ion in he DZ. To es his, we di ec ly compe ed B1-8
hi
B cells ca ying ei he one o wo in ac alleles o Ccnd3 by
adop i e ans e o a 1:1 mix u e o Ccnd3
+/−
and Ccnd3
+/+
B1-8
hi
B cells in o he same ecipien mice, which we e hen immu-
nized wi h NP-OVA in alum and assayed o ela i e abundance
o hese wo popula ions (Fig. 5 A). Whe eas bo h Ccnd3
+/+
and
Ccnd3
+/−
B1-8
hi
cells we e ound a a simila a io in ea ly GCs a
day 7 a e immuniza ion, he p opo ion o Ccnd3
+/−
B1-8
hi
cells
dec eased g adually o e ime, such ha hese cells we e com-
ple ely elimina ed in h ee ou o se en mice by day 14 a e
immuniza ion (Fig. 5, B and C). Thus, he educ ion in cyclin D3
dosage associa ed wi h he e ozygosis is su icien o impose a
g adual bu clea loss o his popula ion om he GC. Lack o
compe i i eness o he e ozygous B cells was associa ed wi h a
sligh educ ion in he DZ/LZ a io, which was al eady obse -
able a day 7 a e immuniza ion (Fig. 5, D and E), and a dec ease
in he p opo ion o cells en e ing S phase in he DZ, bu no in
he LZ (Fig. 5, F–H). To ex apola e he loss o ine ial p oli e -
a i e capaci y among Ccnd3
+/−
GC B cells om ou di ec com-
pe i ion da a, we simula ed his expe imen in silico using a
p e iously published agen -based model o GC selec ion ha
includes T cell con ol o e he numbe o cell cycles ca ied a
B cell unde goes upon posi i e selec ion (Meye -He mann, 2020
P ep in ;Meye -He mann e al., 2012). We modeled loss o Ccnd3
exp ession as a educ ion in he maximum numbe o di isions a
Ccnd3
+/−
GC B cell can comple e upon posi i e selec ion. We
a ied he ela ionship be ween he deg ee o T cell help
(modeled as in ensi y o c-Myc ac i a ion) and he numbe o
di isions as illus a ed by he cu es shown in Fig. 5 I.Theex-
pe imen ally measu ed kine ics o Ccnd3
+/−
GC B cells we e bes
ep oduced when he maximum numbe o di isions o
Ccnd3
+/−
GC B cells in silico was educed o 72% o WT, wi h a
esidual sum o squa es (RSS) o 0.14 (Fig. 5 J). In silico, a e-
duc ion in esponsi eness o his magni ude in Ccnd3
+/−
GC
B cells was accompanied by a educ ion in he numbe o cell
cycles pe cell o ∼84% o WT le els a day 8 a e immuniza ion,
which is compa ible wi h ou EdU/B dU inco po a ion da a
(Fig. 5 K). We conclude ha cyclin D3 dose-dependen ly con ols
he numbe o ine ial cycles a B cell will unde go in he DZ. The
sensi i i y o GC B cells o loss o e en a single allele o Ccnd3
sugges s ha cyclin D3 p o ein abundance may se e as a mo-
lecula b idge linking he cumula i e signal a B cell ecei es
om T cells in he LZ and he numbe o cycles his cell can
execu e in he DZ.
A lymphoma-associa ed mu a ion ha s abilizes cyclin D3
p omo es DZ ine ial cycling and d i es clonal B cell
lymphop oli e a ion
A pa allel line o e idence poin ing o he impo ance o cyclin
D3 in GC p oli e a ion is he p esence o a se ies o Ccnd3 mu-
a ions ha s abilize cyclin D3 p o ein in oughly 40% o cases o
spo adic Bu ki lymphoma (Casano as e al., 2004;Schmi z
e al., 2012,2014;Sonoki e al., 2001). Because Bu ki lym-
phoma cells closely esemble DZ B cells in gene exp ession
(Ca on e al., 2009;Vic o a and Mouque , 2018), we hypo he-
sized ha s abiliza ion o cyclin D3 by means o a lymphoma-
associa ed mu a ion may inc ease he p opensi y o GC B cells o
unde go ine ial cycles in he DZ. To es his, we used CRISPR-
Cas9–media ed genome edi ing o gene a e a Ccnd3 allele en-
coding a e sion o cyclin D3 p o ein ha is s abilized by he
eplacemen o a phospho yla able h eonine (T283A) in i s
C- e minal domain (Fig. 6 A). This mu a ion p e en s phos-
pho yla ion o T283, which would o he wise p omo e nuclea
expo and p o easomal deg ada ion o cyclin D3 (Casano as
e al., 2004;Ca o e al., 2011). When in oduced in o he ge-
nome, he T283A mu a ion alone did no cause any o e
anomaly in B cell de elopmen , wi h he possible excep ion o a
sligh inc ease in he pe cen age o p e/p o-B cells in he bone
ma ow (BM), no did i lead o spon aneous GC o ma ion in
he spleens o young adul mice (Fig. S4, A and D). None heless,
we ound ha upon immuniza ion wi h NP-OVA, Ccnd3
T283A/+
mice gene a ed GCs ha we e app oxima ely wice as la ge as
WT GCs and had a ma kedly highe DZ/LZ a io, sugges ing
inc eased DZ p oli e a ion (Fig. 6, B–E). Acco dingly, low
and CD35 (FDC ma ke ). Pe cen age o Ly75
+/+
cells wi hin he DZ ou o he en i e GC is quan i ied. Scale ba = 100 μm. (G–K) S phase en y o posi i ely
selec ed Ly75
+/+
B1-8
hi
cells ha a e ei he WT (Ccnd3
+/+
)o cyclinD3mu an (Ccnd3
−/−
) in he LZ (G) o DZ (H), quan i ied in I–K. *, P < 0.05; **, P < 0.01; ***,
P < 0.001; ****, P < 0.0001; ns, nonsigni ican , nonpa ame ic Mann–Whi ney es , compa ed wi h WT (Ccnd3
+/+
). Ba s indica e median. Each ci cle ep esen s
a mouse. Da a pooled om ou independen expe imen s.
Pae e al. Jou nal o Expe imen al Medicine 8o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
cy ome ic measu emen o EdU/B dU inco po a ion showed a
∼50% inc ease in GC B cells en e ing S phase in he DZ bu no
inc ease in S phase en y in he LZ (Fig. 6, F–H). To con i m ha
he p oli e a i e e ec o cyclin D3 on he GC is B cell in insic,
we gene a ed mixed chime as in which BM cells om ei he
WT o Ccnd3
T283A/+
mice we e mixed wi h hose om B cell–
de icien J
H
T mice a a 20:80 a io. Consis en wi h he analysis
o Ccnd3
T283A/+
mice, chime as wi h Ccnd3
T283A/+
B cells showed
inc eased GC size, DZ expansion, and S phase en y in he DZ,
bu no in he LZ, upon immuniza ion wi h NP-OVA in alum
(Fig. 6, C, E, G, and H). This pheno ype was ecapi ula ed upon
immuniza ion wi h KLH, uling ou any po en ial e ec s on o
he hypomo phic immunoglobulin λligh chain o he SJL
s ain in which Ccnd3
T283A
mice we e o iginally gene a ed on
he no mally Igλ-domina ed esponse o NP-OVA (Fig. S4, E–G).
Mo eo e , Ccnd3
T283A/+
DZ B cells showed a dec ease in size as
measu ed by o wa d sca e , mi o ing he pheno ype ound
in Ccnd3
−/−
B cells (Fig. 6, I and J). This indica es ha he in-
c eased p oli e a ion o Ccnd3
T283A/+
B cells in he DZ ex ends o
he ex eme o wha hese cells a e me abolically p og ammed
Figu e 5. Cyclin D3 con ols ine ial cell cycling in a dose-dependen manne . (A) Expe imen al se up o induc ion o GCs con aining mix u es o B1-8
hi
cells wi h a ull (Ccnd3
+/+
)o educed(Ccnd3
+/−
) dose o cyclin D3. (B and C) Clonal expansion o B1-8
hi
cells wi h a ull (Ccnd3
+/+
)o educed(Ccnd3
+/−
)doseo
cyclin D3 o e ime, quan i ied in C ela i e o day 7. (D and E) DZ and LZ s aining in GC B1-8hi cells 7 d a e NP-OVA immuniza ion, quan i ied in E. Do ed line
indica es a e aged DZ/LZ a io o o al GC B cells. (F–H) S phase en y o B1-8
hi
cells wi h a ull (Ccnd3
+/+
)o educed(Ccnd3
+/−
)doseo cyclinD3inLZo DZ
8 d a e NP-OVA immuniza ion, quan i ied in G and H. (I) Rela ionship be ween s eng h o T cell help (modeled as c-Myc signal in ensi y) and numbe o B cell
di isions in he di e en models used o es ima e loss o unc ion in Ccnd3
+/−
GC B cells. The maximum numbe o di isions allowed o Ccnd3
+/−
cells co -
esponds o 50% ( ed) o 72% (blue), o 80% (g een) o he maximum numbe o di isions o he WT. (J) Ccnd3
+/−
kine ics in silico. Loss o Ccnd3
+/−
cells as a
ac ion o he GC popula ion o e ime in silico. The maximum numbe o di isions o Ccnd3
+/−
GC B cells was ixed o 50% ( ed line), 72% (blue line), o 80%
(g een line) o he WT alue. Mean (solid lines) and SD (shaded a ea) o e 60 GC simula ions. Expe imen al mean ± SD is shown as whi e ci cles. (K) Ccnd3
+/−
di isions in silico. The mean numbe o di isions o Ccnd3
+/−
cells is shown as a pe cen age o WT a he indica ed days. Repo ed da a esul om simula ions
wi h he maximum numbe o di isions o he Ccnd3
+/−
GC B cells ixed o 72% o he WT maximum numbe o di isions. Mean and SD o e 60 GC simula ions.
Whi e ci cle ep esen s expe imen al da a. *, P < 0.05; **, P < 0.01; n.s., no signi ican , pai ed es . Ba s indica e median (C). Each ci cle ep esen s a mouse.
Da a a e pooled om a wo independen expe imen s.
Pae e al. Jou nal o Expe imen al Medicine 9o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Vic o a). J. Pae is a Damon Runyon Cance Resea ch Founda ion
Be ge Founda ion Fellow (DRG-2353-19). J. E sching was a Can-
ce Resea ch Ins i u e-I ing on pos doc o al ellow. M. Schips
was suppo ed by he Eu opean Union’s Ho izon 2020 esea ch
and inno a ion p og am unde Ma ie Skłodowska-Cu ie g an
ag eemen no. 765158. J. O do as-Mon anes was suppo ed by he
Damon Runyon Cance Resea ch Founda ion (DRG-2274-16) and
he Richa d and Susan Smi h Family Founda ion. C. Mlyna czyk
was suppo ed by a Lymphoma Resea ch Founda ion pos doc o al
ellowship, a Leukemia and Lymphoma Socie y Special Fellow
Awa d, and he Ame ican Socie y o Hema ology Resea ch Res a
Awa d o ea ly ca ee in es iga o s in hema ology. A. Melnick
was suppo ed by Na ional Cance Ins i u e g an R35CA220499.
A. E eyan is a Minis e io de Ciencia, Inno ación y Uni e sidades/
Agencia Es a al de In es igación Ramon y Cajal Awa dee (RYC-
2013-13546). A.K. Shalek was suppo ed by he Sea le Schola s
P og am, he Beckman Young In es iga o P og am, a Sloan Fel-
lowship in Chemis y, and Na ional Ins i u es o Heal h g an s
1DP2GM119419 and 5U24AI118672. G.D. Vic o a is a Sea le Schola ,
a Bu oughs Wellcome In es iga o in he pa hogenesis o in ec-
ious disease, a Pew-S ewa Schola , and a MacA hu Fellow.
Au ho con ibu ions: J. Pae, J. E sching, A. E eyan, and G.D.
Vic o a designed and pe o med all in i o expe imen s. T.B.R.
Cas o ca ied ou he bioin o ma ic analysis o scRNA-seq
da a. L. Mesin, S.J. Allon, and J. O do as-Mon anes con ib-
u ed addi ional scRNA-seq lib a ies, unde supe ision o A.K.
Shalek. M. Schips and M. Meye -He mann pe o med ma h-
ema ical modeling. C. Mlyna czyk and A. Melnick con ibu ed
o he desc ip ion o he malignan pheno ype o Ccnd3
T283A
mice,includinghis ologicalanalysis.J.Pae,T.B.R.Cas o,and
G.D. Vic o a w o e he manusc ip wi h inpu om all au-
ho s. G.D. Vic o a supe ised he wo k.
Disclosu es: A. Melnick epo s g an s om Janssen, Sano i, and
Daiichi Sankyo, and pe sonal ees om KDAC, Epizyme, Con-
s ella ion, Jubilan , and ExoThe apeu ics ou side he submi ed
wo k. J. O do as-Mon anes epo s pe sonal ees om Cella i y
ou side he submi ed wo k. No o he disclosu es we e epo ed.
Submi ed: 7 Augus 2020
Re ised: 16 No embe 2020
Accep ed: 30 No embe 2020
Re e ences
Ami ai, A., L. Mesin, G.D. Vic o a, M. Ka da , and A.K. Chak abo y. 2017. A
Popula ion Dynamics Model o Clonal Di e si y in a Ge minal Cen e .
F on . Mic obiol. 8:1693. h ps://doi.o g/10.3389/ micb.2017.01693
Banna d, O., and J.G. Cys e . 2017. Ge minal cen e s: p og ammed o a ini y
ma u a ion and an ibody di e si ica ion. Cu . Opin. Immunol. 45:21–30.
h ps://doi.o g/10.1016/j.coi.2016.12.004
Banna d, O., R.M. Ho on, C.D. Allen, J. An, T. Nagasawa, and J.G. Cys e . 2013.
Ge minal cen e cen oblas s ansi ion o a cen ocy e pheno ype acco ding
o a imed p og am and depend on he da k zone o e ec i e selec ion.
Immuni y. 39:912–924. h ps://doi.o g/10.1016/j.immuni.2013.08.038
Calado, D.P., Y. Sasaki, S.A. Godinho, A. Pelle in, K. K¨
oche , B.P. Sleckman,
I.M. de Albo ´
an, M. Janz, S. Rodig, and K. Rajewsky. 2012. The cell-cycle
egula o c-Myc is essen ial o he o ma ion and main enance o
ge minal cen e s. Na . Immunol. 13:1092–1100. h ps://doi.o g/10.1038/
ni.2418
Ca on, G., S. Le Gallou, T. Lamy, K. Ta e, and T. Fes . 2009. CXCR4 ex-
p ession unc ionally disc imina es cen oblas s e sus cen ocy es
wi hin human ge minal cen e B cells. J. Immunol. 182:7595–7602.
h ps://doi.o g/10.4049/jimmunol.0804272
Casano as,O.,M.Jaumo ,A.B.Paules,N.Agell,andO.Bachs.2004.P38SAPK2
phospho yla es cyclin D3 a Th -283 and a ge s i o p o easomal deg a-
da ion. Oncogene. 23:7537–7544. h ps://doi.o g/10.1038/sj.onc.1208040
Ca o, M.H., S.K. Chin alapa i, I.W. Yau, S.A. Omo i, and R.C. Ricke . 2011.
Cyclin D3 is selec i ely equi ed o p oli e a i e expansion o ge minal
cen e B cells. Mol. Cell. Biol. 31:127–137. h ps://doi.o g/10.1128/MCB
.00650-10
Clijs e s, L., C. Hoencamp, J.J.A. Calis, A. Ma zio, S.M. Handg aa , M.C.
Cui ino, B.R. Rosenbe g, G. Leone, and M. Pagano. 2019. Cyclin F Con-
ols Cell-Cycle T ansc ip ional Ou pu s by Di ec ing he Deg ada ion o
he Th ee Ac i a o E2Fs. Mol. Cell. 74:1264–1277.e7. h ps://doi.o g/10
.1016/j.molcel.2019.04.010
Cys e , J.G., and C.D.C. Allen. 2019. B Cell Responses: Cell In e ac ion Dy-
namics and Decisions. Cell. 177:524–540. h ps://doi.o g/10.1016/j.cell
.2019.03.016
De Sil a, N.S., and U. Klein. 2015. Dynamics o B cells in ge minal cen es.
Na . Re . Immunol. 15:137–148. h ps://doi.o g/10.1038/n i3804
Dobin,A.,C.A.Da is,F.Schlesinge ,J.D enkow,C.Zaleski,S.Jha,P.
Ba u , M. Chaisson, and T.R. Ginge as. 2013. STAR: ul a as uni-
e sal RNA-seq aligne . Bioin o ma ics. 29:15–21. h ps://doi.o g/10
.1093/bioin o ma ics/b s635
Dominguez-Sola, D., G.D. Vic o a, C.Y. Ying, R.T. Phan, M. Sai o, M.C. Nus-
senzweig, and R. Dalla-Fa e a. 2012. The p o o-oncogene MYC is e-
qui ed o selec ion in he ge minal cen e and cyclic een y. Na .
Immunol. 13:1083–1091. h ps://doi.o g/10.1038/ni.2428
Eisen, H.N. 2014. A ini y enhancemen o an ibodies: how low-a ini y an-
ibodies p oduced ea ly in immune esponses a e ollowed by high-
a ini y an ibodies la e and in memo y B-cell esponses. Cance Im-
munol. Res. 2:381–392. h ps://doi.o g/10.1158/2326-6066.CIR-14-0029
E sching, J., A. E eyan, L. Mesin, J.T. Jacobsen, G. Pasqual, B.C. G abine , D.
Dominguez-Sola, D.M. Saba ini, and G.D. Vic o a. 2017. Ge minal
Cen e Selec ion and A ini y Ma u a ion Requi e Dynamic Regula ion
o mTORC1 Kinase. Immuni y. 46:1045–1058.e6. h ps://doi.o g/10.1016/
j.immuni.2017.06.005
Finkin, S., H. Ha wege , T.Y. Oli ei a, E.E. Ka a, and M.C. Nussenzweig.
2019. P o ein Amoun s o he MYC T ansc ip ion Fac o De e mine
Ge minal Cen e B Cell Di ision Capaci y. Immuni y. 51:324–336.e5.
h ps://doi.o g/10.1016/j.immuni.2019.06.013
Gi lin, A.D., Z. Shulman, and M.C. Nussenzweig. 2014. Clonal selec ion in he
ge minal cen e by egula ed p oli e a ion and hype mu a ion. Na u e.
509:637–640. h ps://doi.o g/10.1038/na u e13300
Gi lin, A.D., C.T. Maye , T.Y. Oli ei a, Z. Shulman, M.J. Jones, A. Ko en, and
M.C. Nussenzweig. 2015. HUMORAL IMMUNITY. T cell help con ols
he speed o he cell cycle in ge minal cen e B cells. Science. 349:
643–646. h ps://doi.o g/10.1126/science.aac4919
Ha emeis e , C., and R. Sa ija. 2019. No maliza ion and a iance s abiliza ion
o single-cell RNA-seq da a using egula ized nega i e binomial e-
g ession. Genome Biol. 20:296. h ps://doi.o g/10.1186/s13059-019-1874-1
Inaba, K., W.J. Swigga d, M. Inaba, J. Mel ze , A. Mi za, T. Sasagawa, M.C.
Nussenzweig, and R.M. S einman. 1995. Tissue dis ibu ion o he DEC-
205 p o ein ha is de ec ed by he monoclonal an ibody NLDC-145. I.
Exp ession on dend i ic cells and o he subse s o mouse leukocy es.
Cell. Immunol. 163:148–156. h ps://doi.o g/10.1006/cimm.1995.1109
Koza , K., M.A. Cieme ych, V.I. Rebel, H. Shigema su, A. Zagozdzon, E. Si-
cinska, Y. Geng, Q. Yu, S. Bha acha ya, R.T. B onson, e al. 2004.
Mouse de elopmen and cell p oli e a ion in he absence o D-cyclins.
Cell. 118:477–491. h ps://doi.o g/10.1016/j.cell.2004.07.025
Li, B., and C.N. Dewey. 2011. RSEM: accu a e ansc ip quan i ica ion om
RNA-Seq da a wi h o wi hou a e e ence genome. BMC Bioin o ma ics.
12:323. h ps://doi.o g/10.1186/1471-2105-12-323
Luo, W., F. Weisel, and M.J. Shlomchik. 2018. B Cell Recep o and CD40
Signaling A e Rewi ed o Syne gis ic Induc ion o he c-Myc T an-
sc ip ion Fac o in Ge minal Cen e B Cells. Immuni y. 48:313–326.e5.
h ps://doi.o g/10.1016/j.immuni.2018.01.008
MacLennan, I.C. 1994. Ge minal cen e s. Annu. Re . Immunol. 12:117–139.
h ps://doi.o g/10.1146/annu e .iy.12.040194.001001
Mesin, L., J. E sching, and G.D. Vic o a. 2016. Ge minal Cen e B Cell Dy-
namics. Immuni y. 45:471–482. h ps://doi.o g/10.1016/j.immuni.2016
.09.001
Mesin, L., A. Schiepe s, J. E sching, A. Ba bulescu, C.B. Ca azzoni, A. An-
gelini, T. Okada, T. Ku osaki, and G.D. Vic o a. 2020. Res ic ed
Pae e al. Jou nal o Expe imen al Medicine 16 o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021

Clonali y and Limi ed Ge minal Cen e Reen y Cha ac e ize Memo y
B Cell Reac i a ion by Boos ing. Cell. 180:92–106.e11. h ps://doi.o g/10
.1016/j.cell.2019.11.032
Meye -He mann, M. 2020. Ge minal cen e B cells sepa a e signals o se-
lec ion and di ision like compac discs. bioRxi . h ps://doi.o g/10.1101/
2020.01.29.925479 (P ep in pos ed Janua y 30, 2020)
Meye -He mann, M., E. Moh , N. Pelle ie , Y. Zhang, G.D. Vic o a, and K.M.
Toellne . 2012. A heo y o ge minal cen e B cell selec ion, di ision,
and exi . Cell Rep. 2:162–174. h ps://doi.o g/10.1016/j.cel ep.2012.05
.010
Min, M., Y. Rong, C. Tian, and S.L. Spence . 2020. Tempo al in eg a ion o
mi ogen his o y in mo he cells con ols p oli e a ion o daugh e cells.
Science. 368:1261–1265. h ps://doi.o g/10.1126/science.aay8241
Moo ha, V.K., C.M. Lindg en, K.F. E iksson, A. Sub amanian, S. Sihag, J.
Leha , P. Puigse e , E. Ca lsson, M. Ridde s ˚
ale, E. Lau ila, e al. 2003.
PGC-1alpha- esponsi e genes in ol ed in oxida i e phospho yla ion a e
coo dina ely down egula ed in human diabe es. Na . Gene . 34:267–273.
h ps://doi.o g/10.1038/ng1180
Musg o e, E.A., C.E. Caldon, J. Ba aclough, A. S one, and R.L. Su he land.
2011. Cyclin D as a he apeu ic a ge in cance . Na . Re . Cance . 11:
558–572. h ps://doi.o g/10.1038/n c3090
Pasqual, G., A. Angelini, and G.D. Vic o a. 2015. T igge ing posi i e selec ion
o ge minal cen e B cells by an igen a ge ing o DEC-205. Me hods Mol.
Biol. 1291:125–134. h ps://doi.o g/10.1007/978-1-4939-2498-1_10
Peled, J.U., J.J. Yu, J. Venka esh, E. Bi, B.B. Ding, M. K upski-Downs, R.
Shakno ich, P. Sicinski, B. Diamond, M.D. Scha , and B.H. Ye. 2010.
Requi emen o cyclin D3 in ge minal cen e o ma ion and unc ion.
Cell Res. 20:631–646. h ps://doi.o g/10.1038/c .2010.55
Peng, T., T.R. Golub, and D.M. Saba ini. 2002. The immunosupp essan a-
pamycin mimics a s a a ion-like signal dis inc om amino acid and
glucose dep i a ion. Mol. Cell. Biol. 22:5575–5584. h ps://doi.o g/10
.1128/MCB.22.15.5575-5584.2002
Quad os, R.M., H. Miu a, D.W. Ha ms, H. Aka suka, T. Sa o, T. Aida, R.
Redde , G.P. Richa dson, Y. Inagaki, D. Sakai, e al. 2017. Easi-CRISPR: a
obus me hod o one-s ep gene a ion o mice ca ying condi ional and
inse ion alleles using long ssDNA dono s and CRISPR ibonucleop o-
eins. Genome Biol. 18:92. h ps://doi.o g/10.1186/s13059-017-1220-4
Rajewsky, K. 1996. Clonal selec ion and lea ning in he an ibody sys em.
Na u e. 381:751–758. h ps://doi.o g/10.1038/381751a0
Reid, S., and E.C. Snow. 1996. The egula ed exp ession o cell cycle- ela ed
p o eins as B-lymphocy es en e and p og ess h ough he G1 cell cycle
s age ollowing deli e y o comple e e sus pa ial ac i a ion s imuli. Mol.
Immunol. 33:1139–1151. h ps://doi.o g/10.1016/S0161-5890(96)00065-X
Robbiani, D.F., S. De oubaix, N. Feldhahn, T.Y. Oli ei a, E. Callen, Q. Wang,
M. Janko ic, I.T. Sil a, P.C. Rommel, D. Bosque, e al. 2015. Plasmodium
In ec ion P omo es Genomic Ins abili y and AID-Dependen B Cell
Lymphoma. Cell. 162:727–737. h ps://doi.o g/10.1016/j.cell.2015.07.019
Sakaue-Sawano, A., H. Ku okawa, T. Mo imu a, A. Hanyu, H. Hama, H.
Osawa, S. Kashiwagi, K. Fukami, T. Miya a, H. Miyoshi, e al. 2008.
Visualizing spa io empo al dynamics o mul icellula cell-cycle p o-
g ession. Cell. 132:487–498. h ps://doi.o g/10.1016/j.cell.2007.12.033
Schmi z, R., R.M. Young, M. Ce ibelli, S. Jha a , W. Xiao, M. Zhang, G. W igh ,
A.L. Sha e , D.J. Hodson, E. Bu as, e al. 2012. Bu ki lymphoma
pa hogenesis and he apeu ic a ge s om s uc u al and unc ional
genomics. Na u e. 490:116–120. h ps://doi.o g/10.1038/na u e11378
Schmi z, R., M. Ce ibelli, S. Pi aluga, G. W igh , and L.M. S aud . 2014.
Oncogenic mechanisms in Bu ki lymphoma. Cold Sp ing Ha b. Pe -
spec . Med. 4:a014282. h ps://doi.o g/10.1101/cshpe spec .a014282
Schuhmache , M., F. Kohlhube , M. H¨
olzel,C. Kaise , H. Bu sche , M. Ja sch,
G.W. Bo nkamm, G. Laux, A. Polack, U.H. Weidle, and D. Eick. 2001.
The ansc ip ional p og am o a human B cell line in esponse o Myc.
Nucleic Acids Res. 29:397–406. h ps://doi.o g/10.1093/na /29.2.397
Shih, T.A., E. Me e, M. Roede e , and M.C. Nussenzweig. 2002. Role o BCR
a ini y in T cell dependen an ibody esponses in i o. Na . Immunol. 3:
570–575. h ps://doi.o g/10.1038/ni803
Shlomchik, M.J., W. Luo, and F. Weisel. 2019. Linking signaling and selec ion
in he ge minal cen e . Immunol. Re . 288:49–63. h ps://doi.o g/10.1111/
im .12744
Shulman, Z., A.D. Gi lin, J.S. Weins ein, B. Lainez, E. Esplugues, R.A. Fla ell,
J.E. C a , and M.C. Nussenzweig. 2014. Dynamic signaling by T ollic-
ula helpe cells du ing ge minal cen e B cell selec ion. Science. 345:
1058–1062. h ps://doi.o g/10.1126/science.1257861
Sicinski, P., J.L. Donahe , Y. Geng, S.B. Pa ke , H. Ga dne , M.Y. Pa k, R.L.
Robke , J.S. Richa ds, L.K. McGinnis, J.D. Bigge s, e al. 1996. Cyclin D2
is an FSH- esponsi e gene in ol ed in gonadal cell p oli e a ion and
oncogenesis. Na u e. 384:470–474. h ps://doi.o g/10.1038/384470a0
Sigl, R., C. Wandke, V. Rauch, J. Ki k, T. Hun , and S. Geley. 2009. Loss o he
mammalian APC/C ac i a o FZR1 sho ens G1 and leng hens S phase
bu has li le e ec on exi om mi osis. J. Cell Sci. 122:4208–4217.
h ps://doi.o g/10.1242/jcs.054197
Sol ason, N., W.W. Wu, N. Kab a, X. Wu, E. Lees, and M.C. Howa d. 1996.
Induc ion o cell cycle egula o y p o eins in an i-immunoglobulin-
s imula ed ma u e B lymphocy es. J. Exp. Med. 184:407–417. h ps://doi
.o g/10.1084/jem.184.2.407
Sol ason, N., W.W. Wu, D. Pa y, D. Mahony, E.W. Lam, J. Glass o d, G.G.
Klaus, P. Sicinski, R. Weinbe g, Y.J. Liu, e al. 2000. Cyclin D2 is es-
sen ial o BCR-media ed p oli e a ion and CD5 B cell de elopmen . In .
Immunol. 12:631–638. h ps://doi.o g/10.1093/in imm/12.5.631
Sonoki, T., L. Ha de , D.E. Ho sman, L. Ka an, I. Taniguchi, T.G. Willis, S.
Gesk, D. S einemann, E. Zucca, B. Schlegelbe ge , e al. 2001. Cyclin D3
is a a ge gene o (6;14)(p21.1;q32.3) o ma u e B-cell malignancies.
Blood. 98:2837–2844. h ps://doi.o g/10.1182/blood.V98.9.2837
Spence , S.L., S.D. Cappell, F.C. Tsai, K.W. O e on, C.L. Wang, and T. Meye .
2013. The p oli e a ion-quiescence decision is con olled by a bi u ca-
ion in CDK2 ac i i y a mi o ic exi . Cell. 155:369–383. h ps://doi.o g/
10.1016/j.cell.2013.08.062
S ua , T., A. Bu le , P. Ho man, C. Ha emeis e , E. Papalexi, W.M. Mauck III,
Y. Hao, M. S oeckius, P. Smibe , and R. Sa ija. 2019. Comp ehensi e
In eg a ion o Single-Cell Da a. Cell. 177:1888–1902.e21. h ps://doi.o g/
10.1016/j.cell.2019.05.031
Sub amanian,A.,P.Tamayo,V.K.Moo ha,S.Mukhe jee,B.L.Ebe ,M.A.Gil-
le e, A. Paulo ich, S.L. Pome oy, T.R. Golub, E.S. Lande , and J.P. Mesi o .
2005. Gene se en ichmen analysis: a knowledge-based app oach o
in e p e ing genome-wide exp ession p o iles. P oc. Na l. Acad. Sci. USA.
102:15545–15550. h ps://doi.o g/10.1073/pnas.0506580102
Tas, J.M., L. Mesin, G. Pasqual, S. Ta g, J.T. Jacobsen, Y.M. Mano, C.S. Chen,
J.C. Weill, C.A. Reynaud, E.P. B owne, e al. 2016. Visualizing an ibody
a ini y ma u a ion in ge minal cen e s. Science. 351:1048–1054. h ps://
doi.o g/10.1126/science.aad3439
Ti osh,I.,B.Iza ,S.M.P akadan,M.H.Wadswo hII,D.T eacy,J.J.
T ombe a, A. Ro em, C. Rodman, C. Lian, G. Mu phy, e al. 2016.
Dissec ing he mul icellula ecosys em o me as a ic melanoma by
single-cell RNA-seq. Science. 352:189–196. h ps://doi.o g/10.1126/
science.aad0501
T ombe a, J.J., D. Genne , D. Lu, R. Sa ija, A.K. Shalek, and A. Rege . 2014.
P epa a ion o Single-Cell RNA-Seq Lib a ies o Nex Gene a ion Se-
quencing. Cu . P o oc. Mol. Biol. 107:1–17. h ps://doi.o g/10.1002/
0471142727.mb0422s107
Vic o a, G.D., and H. Mouque . 2018. Wha A e he P ima y Limi a ions in
B-Cell A ini y Ma u a ion, and How Much A ini y Ma u a ion Can We
D i e wi h Vaccina ion? Lessons om he An ibody Response o HIV-1.
Cold Sp ing Ha b. Pe spec . Biol. 10:a029389. h ps://doi.o g/10.1101/
cshpe spec .a029389
Vic o a, G.D., and M.C. Nussenzweig. 2012. Ge minal cen e s. Annu.Re .Immunol.
30:429–457. h ps://doi.o g/10.1146/annu e -immunol-020711-075032
Vic o a, G.D., T.A. Schwicke , D.R. Fooksman, A.O. Kampho s , M. Meye -
He mann, M.L. Dus in, and M.C. Nussenzweig. 2010. Ge minal cen e
dynamics e ealed by mul ipho on mic oscopy wi h a pho oac i a able
luo escen epo e . Cell. 143:592–605. h ps://doi.o g/10.1016/j.cell
.2010.10.032
Vic o a, G.D., D. Dominguez-Sola, A.B. Holmes, S. De oubaix, R. Dalla-Fa e a,
and M.C. Nussenzweig. 2012. Iden i ica ion o human ge minal cen e
ligh and da k zone cells and hei ela ionship o human B-cell lym-
phomas. Blood. 120:2240–2248. h ps://doi.o g/10.1182/blood-2012-03
-415380
Wang, H., H. Yang, C.S. Shi alila, M.M. Dawla y, A.W. Cheng, F. Zhang, and
R. Jaenisch. 2013. One-s ep gene a ion o mice ca ying mu a ions in
mul iple genes by CRISPR/Cas-media ed genome enginee ing. Cell. 153:
910–918. h ps://doi.o g/10.1016/j.cell.2013.04.025
Yang, H., H. Wang, C.S. Shi alila, A.W. Cheng, L. Shi, and R. Jaenisch. 2013.
One-s ep gene a ion o mice ca ying epo e and condi ional alleles
by CRISPR/Cas-media ed genome enginee ing. Cell. 154:1370–1379.
h ps://doi.o g/10.1016/j.cell.2013.08.022
Pae e al. Jou nal o Expe imen al Medicine 17 o 17
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Supplemen al ma e ial
Figu e S1. Synch oniza ion and en ichmen o Ly75
+/+
cells wi h DEC-OVA immuniza ion ( ela ed o Fig. 1). (A) Expe imen al se up o DEC-
OVA–induced posi i e selec ion o Ly75
+/+
GC B cells ollowed by double labeling wi h EdU/B dU o assay o S phase en y. (B) Expansion o Ly75
+/+
cells
o e Ly75
−/−
o e he ime cou se. (C and D) DZ and LZ s aining o Ly75
+/+
and Ly75
−/−
cells o e he ime cou se (C), quan i ied in D. (E and F) S phase en y o
Ly75
+/+
and Ly75
−/−
B1-8
hi
cells in he LZ (E), quan i ied in F. **, P < 0.01; ***, P < 0.001; n.s., no signi ican ; pai ed es compa ison be ween Ly75
+/+
and
Ly75
−/−
cells in he same animal. Ba s indica e median. Each ci cle ep esen s a mouse. Da a a e pooled om wo independen expe imen s. Un ., un ea ed.
Pae e al. Jou nal o Expe imen al Medicine S1
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Figu e S2. GSEA be ween he S phase clus e s ( ela ed o Fig. 2). (A) Top 10 mos signi ican ly a ied “hallma k”signa u es om he MSigDB da abase
using GSEA o compa e clus e s 2, 3, and 6 (C2, C3, and C6). Pa hways ha a e common among h ee se s o compa isons a e indica ed in bold. (B) Exp ession
o mTORC1 signa u e in clus e s 2, 3, and 6 (le ) o in DEC-OVA ime poin s wi h indica ed zonal pheno ypes ( igh ). (C) Summa y o P alues using
Mann–Whi ney U es . ***, P < 0.001; n.s., no signi ican ; P
adj
, adjus ed P alue.
Pae e al. Jou nal o Expe imen al Medicine S2
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Figu e S3. Cyclin D2 is dispensable o ine ial B cell cycling ( ela ed o Fig. 3). (A–C) Exp ession o Ccnd2 in UMAP dimension (A), o e ime a e DEC-
OVA immuniza ion (B), and in LZ o DZ (C). P alues in B a e om a K uskal–Wallis es wi h Dunn’s mul iple compa isons es . O he signi ican P alues a e
<0.001 (12 × 60 h), 0.0013 (12 s. 30 h), and 0.037 (Ly75
−/−
s. 12 h). (D) CRISPR/Cas9-media ed gene a ge ing s a egy o in oduce a 4-bp dele ion and
p ema u e s op codon in Ccnd2.(E and F) S aining o B1 (le ) and B1-a ( igh ) cells isola ed om pe i oneal ca i ies o WT (Ccnd2
+/+
)o cyclinD2–de icien
(Ccnd2
−/−
; E) mice, quan i ied in F. (G) Expe imen al se up o induc ion o GCs con aining WT (Ccnd2
+/+
)o cyclinD2mu an (Ccnd2
−/−
)B1-8
hi
cells. (H–L) DZ
and LZ s aining (H) and S phase en y (J) in WT (Ccnd2
+/+
)o cyclinD2–de icien (Ccnd2
−/−
)B1-8
hi
GC cells 12 d a e NP-OVA immuniza ion, quan i ied in I, K,
and L. (M) Expe imen al se up o induc ion o GCs con aining mix u es o WT (Ccnd2
+/+
) o cyclin D2 mu an (Ccnd2
−/−
)B1-8
hi
cells. (N) Clonal expansion o WT
(Ccnd2
+/+
) o cyclin D2 mu an (Ccnd2
−/−
)B1-8
hi
cells o e ime, ela i e o day 7 ( ep esen ed by a do ed line). *, P < 0.05; **, P < 0.01; n.s., nonsigni ican ,
nonpa ame ic Mann–Whi ney es , compa ed wi h WT (Ccnd2
+/+
; G, I, and J) o day 7 (L). Ba s indica e median. Each ci cle ep esen sa mouse. Da a a e pooled
om wo independen expe imen s. P
adj
, adjus ed P alue.
Pae e al. Jou nal o Expe imen al Medicine S3
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Figu e S4. Analysis o Ccnd3
T283A/+
mice ( ela ed o Fig. 6). (A and B) Bone ma ow (A) and splenic (B) B cells de elopmen in dis ibu ion WT (Ccnd3
+/+
)o
Ccnd3
T283A/+
mice (le ) and quan i ica ion ( igh ). FO, ollicula ; MZ, ma ginal zone. Ba s indica e SD. (C and D) S aining (le ) and quan i ica ion ( igh ) o
splenic B cells (C) and GC (D) in WT (Ccnd3
+/+
)o Ccnd3
T283A/+
mice. Ba s indica e median. SSC-A, side sca e a ea. Fo A–D, 6-wk-old mice (n= 4 o each
geno ype) we e sac i iced. Ccnd3
T283A/+
mice used a e gene a ion N8. (E–G) S aining (le ) and quan i ica ion ( igh ) o GC (E), DZ/LZ (F), and S phase en y (G)
o GC B cells induced upon immuniza ion wi h KLH in pLNs om WT (Ccnd3
+/+
)o Ccnd3
T283A/+
mice. **, P < 0.01; n.s., nonsigni ican , nonpa ame ic
Mann–Whi ney es . Ba s indica e median. Each ci cle ep esen s a mouse. Da a pooled om wo independen expe imen s. Ccnd3
T283A/+
mice used in his
expe imen a e om gene a ions N6 and N7.
Pae e al. Jou nal o Expe imen al Medicine S4
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021

Figu e S5. Analysis o Ccnd3
T283A/+
mice ( ela ed o Fig. 7). (A) Spleens om WT (Ccnd3
+/+
)o Ccnd3
T283A/+
animals ( igh ) and quan i ica ion o spleen o
body weigh a ios (le ). Ba s indica e median. (B) BandTcells aininginWT(Ccnd3
+/+
)o Ccnd3
T283A/+
mice e ealed expansion o abno mal lymphocy e
popula ions wi h diminished o no exp ession o TCRb/B220. (C) H&E and peanu agglu inin (PNA) s aining o WT (Ccnd3
+/+
)o Ccnd3
T283A/+
spleens e ealed
dis up ed mo phology. (D) H&E s aining o nonlymphoid issues om WT (Ccnd3
+/+
)o Ccnd3
T283A/+
e ealed lymphocy ic in il a ion.
Pae e al. Jou nal o Expe imen al Medicine S5
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021
Table S1 and Table S2 a e p o ided online as sepa a e Wo d iles. Table S1 lis s samples included in scRNA-seq analysis. Table S2
lis s an ibodies used in he s udy. Da a S1 and Da a S2 a e p o ided online as sepa a e Excel iles. Da a S1 shows a lis o cell cycle
genes, Myc signa u e genes, E2F a ge genes, and mTOCR1 a ge genes used in scRNA-seq analysis. Da a S2 shows Igh sequences
ob ained om Ccnd3
T283A/+
mice.
Pae e al. Jou nal o Expe imen al Medicine S6
Cyclin D3 d i es ine ial cell cycling in ge minal cen e s h ps://doi.o g/10.1084/jem.20201699
Downloaded om h p:// up ess.o g/jem/a icle-pd /218/4/e20201699/1406609/jem_20201699.pd by Helmhol z-Zen um Fue In ek ions o schung - Biblio hek- use on 07 Janua y 2021