IMMUNOLOGY The microanatomic segregation of selection ... · of Microbiology and Immunology, Weill Medical College of Cornell University, New York, NY 10065, USA. 4Howard Hughes Medical

RESEARCH ARTICLE SUMMARY◥

IMMUNOLOGY

The microanatomic segregationof selection by apoptosis in thegerminal centerChristian T. Mayer, Anna Gazumyan, Ervin E. Kara, Alexander D. Gitlin, Jovana Golijanin,Charlotte Viant, Joy Pai, Thiago Y. Oliveira, Qiao Wang, Amelia Escolano,Max Medina-Ramirez, Rogier W. Sanders, Michel C. Nussenzweig*

INTRODUCTION:Germinal centers (GCs) aretransient microanatomic structures that formin lymphoid organs during an immune response.They are the sites of B cell clonal expansion andaffinity maturation, a process that leads to theproduction of high-affinity antibodies. GCs arehighly dynamic and contain activated B cells,specialized T follicular helper (TFH) cells, andantigen-trapping follicular dendritic cells. GCsare organized into two functionally distinctcompartments: a dark zone (DZ) and a lightzone (LZ). The DZ is the site of rapid cell di-vision and random antibody-gene mutation,which is initiated by activation-induced cyti-dine deaminase (AID). The mutation processleads to the accumulation of a large numberof closely related B cells that carry receptorswith distinct antigen-binding properties. Oncethey stop dividing, DZ B cells migrate to the

LZ, where their newly generated B cell recep-tors (BCRs) are tested: GC B cells with rela-tively higher-affinity receptors capture andprocess more antigen, leading to positive se-lection by interaction with TFH cells. The pos-itively selected LZ B cells return to the DZ,where they undergo further cycles of divisionand mutation. Concomitantly, small numbersof memory B cells and antibody-secreting plas-ma cells exit the GC. Together, these processesprovide the mechanistic basis for affinity matu-ration, which is essential for effective vaccina-tion and protection from infections.

RATIONALE: In addition to producing anti-body variants, AID expression is also a threat tothe genome. AID can produce double-strandbreaks that are substrates for chromosometranslocations. It can also produce immuno-

globulin (Ig) gene missense mutations and de-letions or create self-reactive antibodies. Thesedeleterious mutations need to be selectedagainst. Indeed, histologists have long appre-ciated large numbers of apoptotic nuclei inthe specialized tingible body macrophagesfound in GCs. However, beyond histology,little is known about the exact rate of GC B cell

apoptosis and whether itdiffers in the DZ and LZof the GC. Moreover, themechanisms that causeapoptosis, their relativeimportance in each GCcompartment, and their

role in GC B cell selection have not been de-fined. To examine these questions, we createdfluorescent apoptosis-indicator mice and usedthem to enumerate, isolate, and characterizedying cells in the GC.

RESULTS: We found that apoptosis is prev-alent in both the DZ and LZ compartmentsof GCs throughout the immune response: upto 50% of GC B cells undergo programmed celldeath every 6 hours. Single dying GC B cellswere isolated, and their antibody genes werecloned, expressed by transient transfection,and tested for antigen binding and other prop-erties. Apoptotic DZ cells were highly enrichedfor Ig genes damaged by AID, including mis-sense mutations and deletions. By contrast, dy-ing LZ cells primarily expressed intact antibodieswith a range of affinities indistinguishable fromGC B cells in the live LZ compartment. By ex-perimentally blocking positive selection and byusing reporter mice forMyc, a proto-oncogene,as an indicator of positive selection, we foundthat apoptosis is the default fate for LZ GCB cells that are not actively positively selected.Thus, LZ GC B cells carrying low-affinity BCRsdo not preferentially undergo apoptosis. Instead,apoptosis occurs irrespective of BCR-affinity,and LZ B cells carrying high-affinity BCRs aresimply more likely to be positively selected.

CONCLUSION: Apoptosis is a major featureof GC B cell biology and is required to counter-balance the high rate of proliferation and purgeB cells that carry deleterious mutations. Al-though apoptosis occurs in both the DZ andLZ, the underlying mechanisms of apoptosisin each zone are distinct and microanatomicallysegregated. These insights into GC B biologyare relevant for vaccine design, particularly forpathogens that normally evade effective anti-body responses.▪

RESEARCH

Mayer et al., Science 358, 193 (2017) 13 October 2017 1 of 1

The list of author affiliations is available in the full article online.*Corresponding author. Email: [email protected] is an open-access article distributed under theterms of the Creative Commons Attribution license(http://creativecommons.org/licenses/by/4.0/), whichpermits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.Cite this article as C. T. Mayer et al., Science 358, eaao2602(2017). DOI: 10.1126/science.aao2602

Germinal center B cells expressing an apoptosis indicator. Intravital two-photon microscopyof GC B cells in popliteal lymph nodes of immunized mice (GC B cells, yellow and green;follicular dendritic cell networks, red). The fluorescence resonance energy transfer–based INDIAreporter was used to visualize and purify dying GC B cells.

ON OUR WEBSITE◥

Read the full articleat http://dx.doi.org/10.1126/science.aao2602..................................................

on March 23, 2020

http://science.sciencem

ag.org/D

ownloaded from

RESEARCH ARTICLE◥

IMMUNOLOGY

The microanatomic segregationof selection by apoptosis in thegerminal centerChristian T. Mayer,1 Anna Gazumyan,1 Ervin E. Kara,1 Alexander D. Gitlin,1

Jovana Golijanin,1 Charlotte Viant,1 Joy Pai,1 Thiago Y. Oliveira,1 Qiao Wang,1

Amelia Escolano,1 Max Medina-Ramirez,2 Rogier W. Sanders,2,3 Michel C. Nussenzweig1,4*

B cells undergo rapid cell division and affinitymaturation in anatomically distinct sites in lymphoidorgans called germinal centers (GCs). Homeostasis is maintained in part by B cell apoptosis.However, the precise contribution of apoptosis to GC biology and selection is not well defined.Wedeveloped apoptosis-indicator mice and used them to visualize, purify, and characterize dyingGC B cells. Apoptosis is prevalent in the GC, with up to half of all GC B cells dying every 6 hours.Moreover, programmed cell death is differentially regulated in the light zone and the dark zone:Light-zone B cells die by default if they are not positively selected, whereas dark-zone cells diewhen their antigen receptors are damaged by activation-induced cytidine deaminase.

Germinal centers (GCs) are divided into twoanatomic compartments: the light zone(LZ) and the dark zone (DZ). B cells divideand undergo somatic hypermutation (SHM)in the DZ and are positively selected for

affinity-enhancing mutations by interacting withT follicular helper (TFH) cells in the LZ (1–3). Celldivision is a dominant feature of the GC, withrapid cell division rates of 4 to 6 hours and up to30% of cells in cycle at any time (1, 3, 4). Despiteextensive cell division, the size of the GC com-partment can be relatively constant for weeks ormonths (5). Equilibrium is attributed to a combi-nation of cell death by apoptosis (negative selec-tion) and emigration of memory and plasma cellsfrom the GC. Emigration rates are estimated tobe relatively low [<0.1% for plasma cells (6) and<2% for memory cells (7)]. By contrast, cell lossby apoptosis is reported to be high (8, 9), but theprecise rate and causes of apoptosis, its contri-bution to GC B cell selection, and whether it isdifferentially regulated in the LZ and DZ of theGC has not been determined.

Negative selection in the DZ and LZ

We used a monoclonal antibody against activecaspase-3 (aCasp3) to identify apoptotic cells inthe GCs of C57BL/6J mice immunized with either4-hydroxy-3-nitrophenylacetyl (NP)–conjugatedovalbumin (NP-OVA) or an HIV-1 envelope antigen[BG505 SOSIP.v4.1-GT1.1 trimers, (GT1.1)]. Where-

as follicular (FO) B cells expressing aCasp3 werenearly absent, 3.6 to 5.7% of GC B cells were aCasp3+

irrespective of the time of analysis or immunogen(Fig. 1, A and B). Similarly, 3% of B cells in chronicGCs in Peyer’s patches were aCasp3+ (fig. S1, A andB). When compared to nonapoptotic cells, aCasp3+

GC B cells expressed slightly reduced levels of theB cell lineage and activation markers B220, Fas,CD19, CD86, and GL7, but comparable levels ofactivation-induced cytidine deaminase (AID), asindicated by an AID–green fluorescent protein(GFP) knock-in gene [(10), fig. S1C].Analysis of apoptosis among DZ and LZ GC

B cells revealed that 3.7 to 5.7% of the DZ and2.6 to 5.6% of the LZ were aCasp3+ at all timepoints analyzed (Fig. 1, C and D). Similar resultswere also obtained for Peyer’s patch GC B cells(fig. S1E). Thus, the frequency of apoptotic GCB cells is relatively constant over time and nearlyequivalent in LZ and DZ compartments.The size of GCs in Peyer’s patches is relatively

constant over time in mice housed under speci-fic pathogen-free conditions, and thus, the num-ber of dividing cells should equal the number ofdying cells plus a small number that leave the GCto become memory B or plasma cells (6, 7, 11). Toestimate the proportion of dividing cells in GCs,we performed kinetic labeling experiments withthe nucleoside analog 5-ethynyl-2′-deoxyuridine(EdU), which is incorporated into DNA during theS phase of the cell cycle. About 50% of all Peyer’spatch GC B cells were labeled by EdU in 5.3 hours(Fig. 1, E and F), suggesting that a large number ofGC B cells are lost to cell death during this time.

Dynamics of dying GC B cells

To further characterize these events, we producedtransgenic and knock-in mice that express anapoptosis reporter [indicator of apoptosis (INDIA),

RESEARCH

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 1 of 8

1Laboratory of Molecular Immunology, The RockefellerUniversity, New York, NY 10065, USA. 2Department ofMedical Microbiology, Academic Medical Center, University ofAmsterdam, 1105 AZ Amsterdam, Netherlands. 3Departmentof Microbiology and Immunology, Weill Medical College ofCornell University, New York, NY 10065, USA. 4HowardHughes Medical Institute (HHMI), The Rockefeller University,New York, NY 10065, USA.*Corresponding author. Email: [email protected]

NP-OVAGC B FO B

Time (days)

0

25

50

75

100

% E

dU+

2 4 6 8 100

25

50

75

100

Time (h)

% E

dU+

Exp2

SS

CEdU

0h

2h

4h

6h

8h

10h

GC B cells

0.2%

26.4%

42.3%

52.3%

63.0%

72.7%

% a

Cas

p3+

7 14 21 3502468

10

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

GT1.1GC B FO B

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

NP-OVADZ LZ

GT1.1DZ LZ

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

Time (days)7 14 21 35

% a

Cas

p3+

02468

10

R2=1T50=5.43h

0

Exp1

R2=1T50=5.26h

Fig. 1. Quantitation of cell death in GC B cells.(A to D) C57BL/6J mice were immunizedsubcutaneously with NP-OVA or GT1.1 precipitatedin alum and analyzed after 7, 14, 21, and 35 days.Percentages of aCasp3+ GC B cells and FO B cells(left and right, respectively) after (A) NP-OVAor (B) GT1.1 immunization. Percentages ofaCasp3+ cells in DZ (CXCR4hiCD86lo left) andLZ (CXCR4loCD86hi right) after (C) NP-OVAor (D) GT1.1 immunization. In (A) to (D), each dotrepresents one mouse. (E and F) EdU incorpo-ration into Peyer’s patch GC B cells measuredby flow cytometry at the indicated time pointsafter EdU administration every 2 hours.(E) Representative flow cytometry plots showEdU and side scatter (SSC). (F) Plots showfraction of EdU+ GC B cells versus time. T50

represents the interpolated time (cubic fit) atwhich 50% of GC B cells incorporated EdU. Dataare from at least two independent experiments,each involving two to five mice for every timepoint. R2, coefficient of determination.

on March 23, 2020

http://science.sciencem

ag.org/D

ownloaded from

the complementary DNA of which was em-bedded into immunoglobulin k regulatory ele-ments (Igk INDIA) and the wild-type Rosa26 locus(Rosa26INDIA); fig. S2, A and B]. This reporter con-sists of an optimized fluorescence (Förster) reso-nance energy transfer (FRET) pair, mNeonGreenand mRuby2, linked by a peptide containing anaCasp3-cleavage site (DEVDG, Fig. 2A; D, asparticacid; E, glutamic acid; V, valine; G, glycine). WhenaCasp3 cleaves the linker, mNeonGreen andmRuby2 should be separated, resulting in FRETloss and increased emission from mNeonGreen(Fig. 2A). Both Igk INDIA and Rosa26INDIA B cellsexpressed the INDIA protein (fig. S2C). To vali-date the reporter, transgenic B cells were activatedin vitro and induced to undergo apoptosis by in-cubation with staurosporine (12). Flow cytometryrevealed two distinct populations based on themNeonGreen/FRET ratio (termed “FRET loss”;Fig. 2B, left). Whereas FRET+ B cells were alive,

FRET– B cells were apoptotic, as confirmed byaCasp3, TUNEL, or Annexin V–DAPI staining(TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick endlabeling; DAPI, 4′,6-diamidino-2-phenylindole)(Fig. 2B and fig. S2D).To examine the kinetics of activated B cell death,

we tracked FRET loss in real time in culturedIgk INDIA B cells (Fig. 2C and fig. S2E). On average,the first morphological signs of apoptosis wereobserved within 12.5 min of FRET loss, includ-ing cell shrinkage, bleb formation, and changesin motility (Fig. 2, C and D, fig. S2E, and moviesS1 to S3). Secondary necrosis, as revealed by lossof membrane integrity and leakage (Fig. 2C, fig.S2E, and movies S1 to S3), was observed an av-erage of 68 min after FRET loss (Fig. 2D). Sim-ilar results were obtained in vivo by trackingRosa26INDIA knock-in GC B cell death using two-photon laser scanning microscopy. GC B cell frag-

mentation occurred on average 20.6 min afterFRET loss and was observed in both DZ and LZcompartments (Fig. 2, E to G; Movies 1 to 3; andfig. S3, A and B). Although the precise relation-ship between caspase staining, FRET loss, andcell fragmentation has not been defined, it isclear that the apoptotic compartment in GCsturns over with rapid kinetics. At an apoptosisrate of 3% every 20.6 min (fig. S1, A and B), 46%of GC B cells in Peyer’s patches are estimated tobe lost in 5.3 hours, which agrees with our mea-surements made by EdU labeling (Fig. 1, E andF). Thus, apoptosis is a major feature of the B cellprogram in the GC.

Negative selection against damagedB cell receptors (BCRs) in the DZ

What causes the high level of GC B cell apoptosis?GC B cells express AID, an enzyme that initiatesclass-switch recombination (CSR) and SHM by

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 2 of 8

Fig. 2. Cell death dynamicsand selection against BCRloss. (A) Schematic represen-tation of the apoptosisindicator. (B) Flow cytometryof lipopolysaccharide (LPS)–IL-4–activated IgkINDIA B cellsincubated with staurosporinefor 3 hours. Dot plot showsforward scatter (FSC) and FRETloss of mRuby2+ cells. Histo-grams show aCasp3 or TUNELstaining on purified FRET+

and FRET– cells. FACS,fluorescence-activatedcell sorting.(C) Representative images ofLPS–IL-4–activated IgkINDIA

B cells showing FRET lossas increased green fluorescenceover time after addition ofstaurosporine (scale bar,10 mm). (D) Time from initiationof FRET loss (synchronized to0 min) to signs of apoptosis(Apo) or necrosis (Nec).Red lines indicate means. Apo,n = 70 cells; Nec, n = 82 cells;****P < 0.0001, two-tailedMann-Whitney U rank sum test.(E to G) Intravital imaging ofB1-8hiRosa26INDIA GC B cells inlymph nodes of NP-OVAimmunized mice(B1-8hi, high-affinity NP-specific antibody). (E) CollapsedZ stacks of 75-mm depthshowing FRET loss and dis-integration of a GC B cell overtime. (F) FRET loss ratiostracked over time [red, thedying cell in (E); black, a live GCB cell in the same imaging volume]. (G) Time from FRET loss to GC B cellfragmentation. Red line indicates the mean. (H to J) Paired IgHV1–72-Iglor IgH-Igk sequences from single IgkINDIA live and apoptotic GC LZ andDZ B cells purified from NP-OVA– or GT1.1-immunized mice. (H) Schematicrepresentation of the experiment. cDNA, complementary DNA; PCR,

polymerase chain reaction. (I) and (J) Pie charts show the fraction ofnonfunctional BCRs (red) in live and apoptotic GC B cells (top) or in the LZ andDZ (bottom) after (I) NP-OVA and (J) GT1.1 immunization. The number inthe center indicates the number of Ig pairs analyzed. Data are from at leasttwo independent experiments in all cases. ****P < 0.0001; Fisher’s exact test.

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

creating base pair mismatches in DNA. The ab-sence of AID in mice and humans is associatedwith enlarged GCs (13, 14) and reduced GC B cellapoptosis, as measured by aCasp3 [fig. S4, A toE, and (15)]. To determine whether AID differ-entially affects cell death in the two GC com-partments, we stained AID-deficient DZ and LZcells for aCasp3. The absence of AID was asso-ciated with a clear reduction in apoptosis primar-ily in the DZ (fig. S4, F to H). Thus, AID activityis a key component of apoptosis in the DZ, andapoptosis appears to be differentially regulatedin the DZ and LZ.AID introduces randommutations in Ig genes

that can increase antibody affinity, but can alsobe deleterious. To determine how Ig mutationimpacts apoptosis, we cloned antibodies fromsingle FRET– IgkINDIA GC B cells that had startedundergoing apoptosis (Fig. 2H and fig. S5A). Igheavy-chain (IgH) and light-chain (Igk and Igl)sequencing revealed that 94 and 83% of live NP-OVA– and GT1.1-elicited GC B cells carried intact

BCRs, respectively (Fig. 2, I and J; top). By con-trast, only 68 and 59% of apoptotic NP-OVA– andGT1.1-elicited GC B cells carried BCR transcriptscapable of producing Ig (Fig. 2, I and J; top). Theloss of BCR expression in the apoptotic compart-ment was confirmed by flow cytometry in NP-OVA–specific GCs and Peyer’s patches and wasAID-dependent (fig. S5, B and C). Apoptotic Bcells with nonfunctional BCRs were highly en-riched in the DZ as compared to the LZ: 43 and58% of apoptotic DZ and 9 and 14% of apoptoticLZ GC B cells in NP-OVA– or GT1.1-immunizedmice, respectively, carried nonproductive Ig tran-scripts (Fig. 2, I and J; bottom). This observationis consistent with reports that AID is expressedat higher levels and accesses DNA in prolifer-ating DZ B cells (5, 16, 17). Although most non-functional apoptotic DZ BCRs carried stop codons(63 and 69% in NP-OVA– and GT1.1-elicited GCs,respectively), a significant fraction (37 and 31%,respectively) was out of frame because of nucle-otide insertions or deletions (fig. S5, D and E).

Thus, apoptotic DZ B cells are enriched for non-functional Ig transcripts as a consequence of AIDactivity. We hypothesize that the small numberof apoptotic LZ GC B cells with nonfunctionalBCRs derive from recent DZ emigrants in whicha delay between aberrant Ig gene mutation, lossof BCR expression, and apoptosis has occurred,as has been reported for naïve B cells (18).

Characterization of monoclonal antibodiescloned from dying GC B cells

In addition to compromising the integrity of theBCR, SHM can also alter antibody affinity, pro-duce autoreactive or polyreactive BCRs, or renderIg heavy and light chains incompatible. To mea-sure the contribution of each of these effects to GCB cell apoptosis, we produced IgHV1–72 Igl a-NPantibodies cloned from GC B cells of NP-OVA–immunized Igk INDIA mice. The relative affinityto NP was measured by enzyme-linked immuno-sorbent assay (ELISA) on NP4– or NP25–bovineserum albumin (BSA). High-affinity antibodies

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 3 of 8

Fig. 3. Binding propertiesof GC-derived IgHV1–72 Iglantibodies. (A) ELISAsshow monoclonal antibodybinding to NP25-BSA (left)or NP4-BSA (right); NP-specific control antibodies(B1-8hi, high affinity, red;B1-8 germline, intermediateaffinity, yellow; B1-8lo, lowaffinity, green) and mGO53(negative control, gray)are included. A, absorbanceat 405 nm; AU, arbitraryunits. (B) Pie charts showthe percentage of high-,low-, and intermediate-affinity antibodies in eachGC compartment (NP4/NP25

ratio = 1, black; <B1-8germline, gray; > B1-8germline < 1, white) or no NPbinding (purple). (C) Sameas (B) but for all LZ and DZcells irrespective of celldeath. (D) Percentage ofantibodies binding to self-antigens in HEp-2 ELISA(cyan). (E) Percentage ofpolyreactive antibodies(green). (F) Percentage ofstructurally compromisedantibodies that cannot besecreted by transfectedhuman embryonic kidney293 (HEK293)–6E cells(yellow). (G) Summaryof negative selection in GCs.Percentage of nonfunctionalBCRs (see Fig. 2I), structurally compromised BCRs (yellow slice), andautoreactive BCRs (cyan slice) in each GC B cell compartment. (B) to (G)The number in the center of the pie charts represents the number ofantibodies tested. (H) Graphs show the fraction of IgG1+ and IgM+ live and

apoptotic Peyer’s patch DZ and LZ GC B cells in AIDCre/CreRosa26lsl-YFPIgH96K/96K mice as determined by aCasp3 staining. Results arecombined from two independent experiments each involving three to fivemice. ****P < 0.0001; paired Student’s t test.

Live Apo

LZ

DZ

****

****Structurally compromised

Total

****

Live LZ

10-4

mAb (µg/ml)10-2 100

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50A

405

(AU

)

Apo LZ

Live DZ

Apo DZ

NP4-BSA

B1-8hi B1-8 B1-8lo

mGO53 GC-derived mAbs

Live Apo

LZ

DZ

43

Total

Live Apo

LZ

DZ

Total 97 69

43 39

54 30

Autoreactive

Live Apo

LZ

DZ

Total 104 97

61 56

NP25-BSA

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50

A40

5 (A

U)

0.00

1.75

3.50

A40

5 (A

U)

10-4

mAb (µg/ml)10-2 100

98 69

43 39

55 30

40

LZ DZ

Total 82 85AIDcre/cre Rosa26lsl-YFP IgH96K/96K

020406080

100

% Ig

M+

020406080

100

% Ig

G1+

LZ

Live Apo0

20406080

100

% Ig

M+

020406080

100

% Ig

G1+

DZ

****

****

****

****

LZ

DZ

98 68

43 39

55 29

Polyreactive Non-functionalStructurally compromisedAutoreactive

Live Apo

LZ

DZ

Total 112 142

44 45

68 97

High Low Int None

Affinity

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

bind better to NP4-BSA than low-affinity anti-bodies, whereas both bind equally well to NP25-BSA[Fig. 3A; (19)]. All but one of the 167 IgHV1–72 IglGC-derived antibodies tested bound to NP25-BSA(Fig. 3, A and B). As might be expected fromthe observation that the DZ contains affinity-selected B cells, high-affinity antibodies were slightlyoverrepresented in the DZ over LZ compartment(60 and 46%, respectively; P = 0.089, Fig. 3C). Thissmall increase in high-affinity reactivity in the DZwas confirmed by increased proportions of theaffinity-enhancing W33L mutation in the DZ (W,tryptophan; L, leucine) (fig. S6, A and B). However,high- and low-affinity antibodies were equally dis-tributed among live and apoptotic B cells in bothGC compartments (Fig. 3, A and B, and fig. S6A).Thus, the observed differences in affinity betweenthe LZ and DZ are likely due to rapid transit ofpositively selected cells from the LZ to the DZ (1),as well as a relatively longer dwell time of higher-affinity cells in the DZ (2), and not apoptosis.To document autoreactivity, we performed

HEp-2 ELISAs, which are used clinically to mea-sure antibody binding to cytoplasmic and nuclearself-antigens. Only 8% of the IgHV1–72 Igl GC an-tibodies cloned from live B cells displayed auto-reactivity (Fig. 3D and fig. S6C). Consistent withthe possibility that autoreactive antibodies can beredeemed by continued SHM (20, 21), there weresimilar numbers of autoreactive cells in live andapoptotic compartments (Fig. 3D and fig. S6C).Thus, autoreactivity accounts for only a smallfraction of the overall death in IgHV1–72 Igl NP-specific GC B cells, and it is not strongly selectedinto the apoptotic compartment.

Memory B cells show increased polyreactivityrelative to naïve B cells (22–25). In agreementwith these observations, 21% of all live GC B cellswere polyreactive, with a slightly higher abun-dance in the DZ (27%) over the LZ (16%) (Fig. 3Eand fig. S6D). Moreover, polyreactive cells wereequally represented in the live and apoptotic com-partments. Thus, polyreactivity is neither posi-tively nor negatively selected into the apoptoticcompartment in the GC.

Nearly all Igs cloned from live GC B cellsproduced secreted antibodies in transient trans-fection experiments. By contrast, a significantfraction of the antibodies derived from apoptoticIgHV1–72 Igl GC B cells did not, despite func-tional Ig genes (Fig. 3F). This phenomenon wasparticularly prominent among apoptotic DZ cells,where approximately half of the clones did notproduce secreted antibodies in transient trans-fection experiments. Despite the lack of secretedIg, immunoblot analysis of transfected cell pelletsshowed normal levels of IgH and Igl expression,suggesting that structural problems and/or defec-tive Ig pairing interfered with normal antibodysecretion (fig. S7). Thus, a significant fraction ofthe apoptotic cells in NP-specific GCs expressantibodies that are structurally compromised(28%), and this phenomenon is exclusive to DZGC B cells. In summary, selection against SHMsthat introduce nucleotide insertions or deletions,produce stop codons, change the reading frame,or otherwise compromise IgHV1–72 Igl expressionor stability account for 75% of all the apoptoticcells in the DZ of NP-specific GCs (Fig. 3G).

Isotype switching

AID initiates both SHM and CSR. To examinethe role of CSR in regulating apoptosis indepen-dently of SHM, we analyzed Peyer’s patches fromAIDCre/Cre IgH96K/96K Rosa26LSL-YFPmice in which~50% of GC B cells undergo Cre-mediated CSRto IgG1 in the absence of AID and SHM [fig. S8,A and B; (22)]. IgG1+ GC B cells were enrichedamong apoptotic cells in both the LZ and DZ. Bycontrast, IgM+ cells were overrepresented in the

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 4 of 8

Fig. 4. Apoptosis in the GCLZ. (A) Graphs show thefraction of GC B cells amongCD19+ B cells (left column),fraction of EdU+ GC Bcells (second column), andfraction of aCasp3+ cellsin GC LZ and DZ (rightcolumns) in the draininglymph nodes of C57BL/6Jmice 14 days after NP-OVAimmunization. HamsterIgG or a-CD40L was injected24, 48, or 72 hours beforeanalysis. Results are com-bined from two independentexperiments, each with fivemice per condition. Red linesindicate means. *P = 0.037(24 hours, GC size); *P =0.0215 (24 hours, prolifera-tion); **P = 0.0048; ****P <0.0001; NS, P > 0.05 (notstatistically significant); two-tailed Mann-Whitney U ranksum test. (B and C) Analysis of apoptosis in GC B cells in Nur77-GFPor Myc-GFP mice 14 days after immunization with NP-OVA. (B) Represen-tative flow-cytometry plots show the frequency of aCasp3+ cells inNur77-GFP– and Nur77-GFP+ (top) and in Myc-GFP− and Myc-GFP+ GCB cells (bottom). (C) The frequency of aCasp3+ cells in Nur77-GFP–

and Nur77-GFP+ (left graph), or Myc-GFP– and Myc-GFP+ LZ GC B cells(right graph). (B) and (C) Two to three independent experiments eachinvolving three to six Nur77-GFP or Myc-GFP mice with three mice pooledper data set. Nur77-GFP, *P = 0.0355; Myc-GFP, **P = 0.006; pairedStudent’s t test.

Nur77- Nur77+0

2

4

6

% a

Cas

p3+

IgG αCD40L0.0

0.5

1.0

1.5

2.0

% G

C 24h

0.0

0.5

1.0

1.5

2.0%

GC

IgG αCD40L

0.0

0.5

1.0

1.5

2.0

% G

C

IgG αCD40L

0

2

4

6

8

10

% a

Cas

p3+

Apoptosis LZ

0

2

4

6

8

10

% a

Cas

p3+

IgG αCD40L

IgG αCD40L

0

2

4

6

8

10

% a

Cas

p3+

IgG αCD40L

0

2

4

6

8

10

% a

Cas

p3+

IgG αCD40L

IgG αCD40L

IgG αCD40L

0

2

4

6

8

10

% a

Cas

p3+

0

2

4

6

8

10

% a

Cas

p3+

0

10

20

30

% E

dU+

IgG αCD40L

IgG αCD40L

IgG αCD40L

0

10

20

30

% E

dU+

0

10

20

30

% E

dU+

GC size Apoptosis DZProliferation

*

**

****

NS

NS

NS

NS

NS

NS

*

****

*****

LZ GC B cells (NP-OVA)

cMyc- cMyc+0

2

4

6

% a

Cas

p3+ **

SS

C

Nur77- Nur77+

cMyc- cMyc+

SS

C

aCasp3

Total GC B cells (NP-OVA)

48h72h

4.1% 1.3%

4.3% 2.5%

Movie 1. A GC B cell dying in vivo. Collapsedfour-dimensional (4D) images depicting B1-8hiRosa26INDIA GC B cells (yellow, FRET+; green,FRET–; scale bar, 10 μm). A GC B cell inthe image center undergoes FRET loss andfragmentation. Tingible body macrophages areseen as autofluorescent sessile cells.

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

live compartments (Fig. 3H and fig. S8B). Sim-ilar results were also obtained in Peyer’s patchesand in lymph nodes of NP-OVA–immunized AID-sufficient C57BL/6J mice (fig. S8, C and D). Thus,IgG1+ GC B cells are more prone to apoptosis thanIgM-expressing cells irrespective of AID expres-sion or SHM. This effect may be due to alteredIgG1 BCR signaling, as indicated by lower levels ofnuclear receptor Nur77-GFP induction in IgG1+

when compared to IgM+ GC B cells (22).

Apoptosis in the LZ due to the absenceof positive selection

Our data indicate that in contrast to the DZ, 80%of apoptotic LZ cells express intact, nonautoreac-tive BCRs whose antigen-binding properties aresimilar to those found in the live compartment(Fig. 3, A, B, and G). The transition from the DZto the LZ is associated with increased expres-sion of genes that regulate apoptosis (1, 16). Toexamine the possibility that apoptosis may bethe default fate for cells that are not positivelyselected in the LZ, we inhibited positive selec-tion by blocking CD40-CD40L interactions witha-CD40L antibody (fig. S9A). Although GC sizeand the number of positively selected cells un-dergoing proliferation decreased, the rate ofLZ B cell apoptosis remained similar [Fig. 4Aand (26)]. Absence of a measurable increase inapoptosis (26) in a-CD40L antibody–treated miceis consistent with the relatively small number ofLZ B cells (5 to 10%) undergoing positive selec-tion at any time (27).To gain further insights into the mechanisms

responsible for cell death in the LZ, we measuredapoptosis in cells undergoing BCR signaling andpositive selection using Nur77-GFP andMyc-GFP(myelocytomatosis oncogene) reporters, respectively(27–29). A fraction of LZ B cells expressed Nur77,but only a subset of these cells were protectedfrom apoptosis (Fig. 4, B and C, and fig. S9, B andC). By contrast, positive selection, as indicated byMyc-GFP expression, was associated with nearlycomplete protection from apoptosis (Fig. 4, B andC, and fig. S9, B and C). Thus, positive selection,as measured by Myc expression, protects B cellsfrom apoptosis, but BCR cross-linking alone, asmeasured by Nur77, appears to be insufficient.

Conclusions

We have investigated the causes of apoptosis andtheir contribution to cell death in the LZ andDZ of GCs by combining a FRET-indicator ofapoptosis, single-cell sorting, and antibody clon-ing. Only a small fraction of LZ GC B cells arepositively selected to return to the DZ and undergoadditional rounds of division, and this processis, in part, stochastic (1, 30). The element of chanceappears to be introduced by random encountersbetween LZ B cells displaying high levels ofpeptide–major histocompatibility complex IIand cognate T cells. Consistent with this notion,the apoptotic compartment in the LZ containedan arbitrary assortment of B cells, includingthose with high-affinity BCRs (Fig. 3B). Thus,both low- and high-affinity B cells undergo apo-ptosis in the LZ, but high-affinity cells are more

likely to become positively selected after they en-counter cognate TFH cells (1, 31, 32). By contrast,the DZ is the microanatomic site of antibodyquality control by selection against deleteriousmutations introduced by AID. B cells expressingIgs damaged by SHM undergo apoptosis. Thelargest group of apoptotic B cells in the DZ arisesby AID-mediated introduction of stop codons,insertions, or deletions into BCR genes (33).Thus, GC B cells resemble developing and naïveB cells in that they require BCR expression forsurvival (18, 34–36).GC B cells are among the most rapidly dividing

eukaryotic cells with cell-cycle times as short as 4to 6 hours. Despite rapid cell division and limitedexport of memory and plasma cells, the size ofthe GC is relatively stable over periods of weeks.Our experiments indicate that GC homeostasisis maintained by high rates of cell death. More-over, whereas positive selection occurs in theLZ, negative selection by apoptosis occurs inboth GC zones but is differentially regulatedin the LZ and DZ.

Materials and methodsGeneration of IgkINDIA transgenic miceand Rosa26INDIA knock-in mice

Indicator of apoptosis (INDIA) cDNA was as-sembled from sequence-optimized N-terminallyMyc-tagged mNeonGreen (37), an 18–amino acidlinker bearing the aCasp3 cleavage site DEVD[SSSELSGDEVDGTSGSEF; (38)] and mRuby2 (39)by gBlocks Gene fragment synthesis (IntegratedDNA Technologies), overlap PCR, and standardcloning procedures.To produce IgkINDIA transgenic mice, INDIA

was placed under transcriptional control of mouseIgk regulatory elements as previously described(40, 41), with the modification that the entireopen reading frame was placed after the non-coding Vk exon followed by the Igk polyadenyl-ation signal. After sequence and vector integrity

confirmation, the vector backbone was eliminatedby NotI/MluI digestion, and the resulting 7.2-kbfragment was injected into the pronuclei of fer-tilized C57BL/6J oocytes. Transgenic foundermice were identified by PCR (table S1; anneal-ing temperature 55°C) on tail DNA. The trans-genic founder line IgkINDIA was selected andmaintained by mating to C57BL/6J mice or byintercrossing. Genotyping was performed by flowcytometry on peripheral blood.To generate Rosa26INDIA knock-in mice, INDIA

was cloned into the AscI site of the CTV targetingvector (42) that was a gift from Klaus Rajewsky(Addgene plasmid #15912). The frt-flanked IRES-eGFP sequence was deleted by L-arabinose–inducible Flp recombineering in SW105 bacteria(43). In brief, Flp expression was induced by 0.09%(w/v) L-arabinose for 1 hour at 32°C, followed byelectroporation of 1 ng vector and growth on LBagar plates containing ampicillin over night at32°C. Vector integrity and deletion of IRES-eGFP in the resulting clones were confirmed byrestriction-enzyme digestion and sequencing.CY2.4 albino C57BL/6J-Tyrc-2J–derived ES cellswere targeted with the modified CTV vector at theGene Targeting Resource Center (The RockefellerUniversity) and homologous recombination wasverified by Southern blot and PCR. Chimeric malesobtained after blastocyst injection were crossedto B6(Cg)-Tyrc-2J/J females. Rosa26LSL-INDIA miceexhibiting germline transmission were identifiedby PCR and crossed to B6.C-Tg(CMV-cre)1Cgn/Jmice to induce germline deletion of the LSL cas-sette. Rosa26LSL-INDIA CMV-cre+ offspring werecrossed to C57BL/6J mice. The resulting CMV-cre− offspring ubiquitously expressing INDIAwere intercrossed to establish Rosa26INDIA mice.Genotyping was performed by flow cytometryon peripheral blood.

Mice

B6.C-Tg(CMV-cre)1Cgn/J, B6(Cg)-Tyrc-2J/J, B6.SJLand C57BL/6J mice were purchased from JacksonLaboratories. AIDCre/CreIgH96K/96KRosa26LSL-YFP,AID–/–, AID-GFP, Nur77-GFP,Myc-GFP and B1-8hi

mice were described previously (10, 13, 22, 28, 29,44, 45). Bone marrow chimeras were generatedas described (46). All animal experiments wereapproved by the Institutional Review Board andthe IACUC at The Rockefeller University.

Immunizations and treatments

Primary GCs were elicited by immunizing miceof the indicated genotypes subcutaneously with25 ml of PBS containing 12.5 mg of NP15-OVA(Biosearch Technologies) or 4 mg of HIV enve-lope antigen (BG505 SOSIP.v4.1-GT1.1 trimers,(GT1.1); provided by Rogier W. Sanders, WeillMedical College of Cornell University, New York)precipitated in alum (Imject alum, ThermoFisherScientific) at a 2:1 ratio. Igk INDIA mice were alsoimmunized intraperitoneally with 100 ml of PBScontaining 50 mg of NP15-OVA precipitated inalum. For blocking CD40L/CD40 interactions,immunized mice were injected intravenouslywith 300 mg of a-CD40L (MR-1, Bio X Cell) or300 mg of Armenian Hamster IgG (Bio X Cell). To

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 5 of 8

Movie 2. Example of a GC B cell dying in theDZ.Collapsed4D images depictingB1-8hiRosa26INDIA

GC B cells (yellow, FRET+; green, FRET−; scale bar,10 μm). AGC B cell in the image center undergoesFRET loss and apoptosis in the DZ. Follicular dendriticcells (FDCs) were labeled with immune complexesconsisting of the fluorescent protein B-phycoerythrin(B-PE) and a-B-PE antibodies to discriminate DZand LZ.Tingible body macrophages are seen asautofluorescent sessile cells.

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

estimate dynamic GC B cell loss by apoptosis,mice were injected intraperitoneally with 1 mgof 5-ethynyl-2′-deoxyuridine (EdU; ThermoFisherScientific) every 2 hours for up to 10 hours. Foridentifying GC B cells in S phase, mice weregiven a single intravenous pulse of 1 mg of EdU1.0 to 2.5 hours prior to sacrifice. To elicit second-ary GCs, mice were immunized intraperitoneallywith 100 ml of PBS containing 50 mg of OVA(Grade V, Sigma) precipitated in alum. Twoweeks later, mice received 5 × 106 B1-8hi B cellsintravenously (~5 × 105 Igl+ NP-specific B cells;5% Rosa26INDIA / 95% Rosa26WT) followed oneday later by a subcutaneous boost with 25 ml ofPBS containing 25 mg of NP15-OVA (BiosearchTechnologies). Follicular dendritic cells (FDC)were labeled by intravenous injection of 0.5 to1.0 mg of polyclonal rabbit a-B-Phycoerythrin(a-B-PE, Rockland) 4 days after boost immuniza-tion, followed 1 day later by subcutaneous injec-tion of 0.05 mg B-PE (Thermo Fisher Scientific).

B cell isolation and culture

B cells were purified from spleens and subcuta-neous lymph nodes as previously described (2).IgkINDIA B cells were stimulated for 4 days with25 mg/ml of LPS (L-2630, Sigma) and 5 ng/mlof IL-4 (I1020, Sigma) in vitro as described (47).Activated B cells were harvested, washed andcultured at 2 × 106 cells/ml with LPS/IL-4-freemedia for 3 hours in the presence of 1 mMstaurosporine (1285, Tocris) to induce apopto-sis. GC B cells were enriched from immunizedIgkINDIAmice by incubating single-cell suspensionswith 1.25 mg/ml of biotinylated a-IgD for 10 minon ice followed by incubation with a-Biotin- andmouse CD43MicroBeads (Miltenyi Biotech). Cellswere passed through a magnetized LS column(Miltenyi Biotec) and enriched GC B cells werecollected in the flow-through.

Live imaging of cultured B cells

On day 4, LPS/IL-4-activated IgkINDIA B cells werewashed and placed in a 35-mm m-Dish (Ibidi). Af-ter the addition of 1 mM staurosporine, cells wereimaged in an environmental chamber set to 37°Cusing a DeltaVision Inverted Olympus IX-71 ImageRestoration Microscope (GE Healthcare) with anInsight SSI 7 color solid state illumination sys-tem and a 20× dry objective. Separate excitationand emission filter wheels were employed to col-lect data for mNeonGreen (488-nm excitation;525/50 BP), FRET (488-nm excitation; 632/60 BP)and standard brightfield images for each timepoint. 512 × 512 pixel images were taken every60 s for 3 hours using Ultimate Focus and a PriorSYZ piezo stage for multiple point visiting. Imageswere processed and analyzed with ImageJ 1.48q(National Institutes of Health) and videos weregenerated at a frame rate of 10 fps.

Intravital imaging and image analysis

Intravital imaging of popliteal lymph nodes andimage acquisition were essentially performed asdescribed previously (31). Mice were anesthetizedby the inhalation of 4% isofluorane in pure ox-ygen, placed on a stage warmer set to 37°C and

maintained on anesthesia by inhalation of 1.25%isofluorane in pure oxygen. Popliteal lymph nodesin shaved hind legs were exposed by microsur-gery and animals were placed under the heatedOlympus 25× 1.05 NA Plan objective of an OlympusBX61 upright microscope fitted with a CoherentChameleon Vision II IR laser. The femtosecond-pulsed multiphoton laser was tuned to 900 nm.A filter cube containing a 690LP mirror followedby a 495LP mirror was used to split the emissionto either 2 GaAsp detectors (with a 500-550-nmfilter for mNeonGreen fluorescence and a 575- to630-nm filter for FRET fluorescence) and a PMTdetector (with a 460- to 500-nm filter for CFP/autofluorescence). Images were acquired every30 s as 75-mm Z-stacks (5-mm steps) with 1.4×zoom and with 512 × 512 X-Y resolution. Imarissoftware (Bitplane) was used to process data.Collapsed Z-stacks were exported as TIFF seriesfrom Imaris and videos were generated in ImageJ(National Institutes of Health) at a frame rate of7 fps. The colocalization tool in Imaris was usedto detect B1-8hi Rosa26INDIA GC B cells that weremNeonGreen+ and lacked CFP/autofluorescence.The surface tool in Imaris was used to trackmNeonGreen+ GC B cells and the mean fluo-rescence intensities of mNeonGreen and FRETchannels over time. The FRET loss ratio wascalculated by dividing mNeonGreen and FRETfluorescence.

Flow cytometry

Spleens and lymph nodes were collected in RPMImedia containing 6% of serum on ice. Single-cellsuspensions were obtained by forcing the tissuethrough a 70-mm cell strainer (BD Biosciences).All centrifugation steps were performed at 4°Cand cells were otherwise handled on ice to min-imize apoptosis. Erythrocytes were lysed with1 ml of ACK lysing buffer (Gibco) for 3 min onice. After incubation with 5 mg/ml of a-CD16/32 (rat mAb 2.4G2, Bio X Cell) for 20 min at 4°C,biotinylated a-CXCR4 was incubated for 45 minat 4°C. Biotin and additional surface antigenswere detected by staining for 30 min at 4°C. The

Fixation/Permeabilization Solution Kit (BD Bio-sciences) was used for intracellular staining.a-CD19 (rat mAb eBio1D3), a-CD38 (rat mAb 90),a-CD45R/B220 (rat mAb RA3-6B2) and a-IgM(rat mAb II/41) were from eBioscience. a-CD19(rat mAb 6D5) and a-CD86 (rat mAb 6D5) werefrom Biolegend. a-aCasp3 (Alexa Fluor 647-conjugated; rabbit mAb C92-605), a-CXCR4 (ratmAb 2B11), a-Fas (hamster mAb Jo2), a-Igk (ratmAb 187.1), a-Igl1-3 (rat mAb R26-46), a-IgG1(rat mAb A85-1), streptavidin conjugated to V500,and a-T- and B cell activation antigen (rat mAbGL7) were from BD Biosciences. Alexa Fluor 647-conjugated polyclonal rabbit IgG was from Cellsignaling. EdU was detected with the Click-iTPlus EdU Alexa Fluor 488 Flow Cytometry AssayKit (ThermoFisher Scientific). Alexa Fluor 647-conjugated Annexin-V and 5× Annexin BindingBuffer for flow cytometry were from ThermoFisherScientific. Annexin-V staining was performed for15 min at room temperature (1:50 dilution) andDAPI (Sigma) was added at 0.04 mg/ml prior toacquisition. TUNEL staining was carried outwith the Apo-BrdU Apoptosis Detection Kit andAlexa Fluor 647-conjugated a-BrdU (mouse mAbMoBU-1, both from ThermoFisher Scientific) ac-cording to the manufacturer’s instructions.Flow cytometry data were acquired on a BD

Fortessa (BD Biosciences) and data were analyzedwith Flowjo (Tristar). Intact cells and singlets wereidentified by their FSC/SSC profiles and in thecase of IgkINDIA B cells additionally by mRuby2expression (561-nm excitation; 582/15 BP). FOB cells were gated CD19+CD38+Fas− and GC Bcells were gated CD19+CD38−Fas+ and addition-ally GL7+ where indicated. Live and apoptoticfractions were discriminated by aCasp3 staining,or by mNeonGreen (488-nm excitation; 505LPand 530/30 BP) and FRET (488-nm excitation;600LP and 610/20 BP) for IgkINDIA mice. FRETloss was derived as a separate parameter inFlowjo defined as the ratio of mNeonGreen andFRET fluorescence among intact mRuby2+ B cells.GC B cell fractions were differentiated into LZ(CXCR4loCD86hi) and DZ (CXCR4hiCD86lo) whereindicated. Due to lower expression of CD86 inapoptotic compared to live GC B cells (see fig. S1),DZ and LZ were separately gated for live andapoptotic GC compartments. Apoptosis rates inDZ and LZ were calculated as: (% aCasp3+ ofGC) × (% DZ or LZ of aCasp3+ GC) / (% DZ orLZ of total GC).

Cell sorting

Cell sorting was carried out on a FACS Aria II (BDBiosciences). For bulk sorting, cultured Igk INDIA

B cells were washed and directly re-suspended inPBS containing 1% serum and 2 mM EDTA. Forsingle-cell sorting, Igk INDIA GC B cells were identi-fied as B220+DUMP−CD38−Fas+ (DUMP = TCRb/F4/80/Gr-1/NK1.1/CD11c). Intact cells and singletswere identified by their FSC/SSC profiles and bymRuby2 expression (561-nm excitation; 570LP,585/15 BP). Live and apoptotic fractions were dis-criminated based on mNeonGreen (488-nm ex-citation; 505LP and 530/30 BP), FRET (488-nmexcitation; 595LP and 610/20 BP) and B220 signals

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 6 of 8

Movie 3. Example of a GC B cell dying in theLZ-DZ border. Collapsed 4D images depictingB1-8hiRosa26INDIA GC B cells (yellow, FRET+;green, FRET–; scale bar, 10 μm). A GC B cell inthe image center undergoes FRET loss andapoptosis at the border between LZ and DZ.FDCs were labeled with B-PE and α-B-PEimmune complexes.

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

(Live: FRET+B220+; apoptotic: FRET−B220lo). Liveand apoptotic GC B cell fractions were furtherdifferentiated into LZ (CXCR4loCD86hi) and DZ(CXCR4hiCD86lo). Single GC B cells of each com-partment were sorted into 96-well PCR platescontaining 4 ml of lysis buffer (48) and were eitherdirectly processed or stored at –80°C.

Single-cell cloning and recombinantantibody expression

Reverse transcription, nested PCR amplification,sequencing, and ligation-independent cloningof IgG1, Igl, and Igk were as described (48, 49)with minor modifications. IgM sequences wereamplified in the same reaction as IgG1 by addingspecific reverse primers (First PCR: 5′-AGGGGG-CTCTCGCAGGAGACGAGG-3′; sequencing PCR:5′-AGGGGGAAGACATTTGGGAAGGAC-3′). A spe-cific forward primer (5′-CTAGTAGCAACTGCAACC-GGTGTACATTCTCAGGTGCAGCTGCAGGAGTC-3′)was used to amplify and clone IgHV1–72. Sequenceswere analyzed with IMGT/V-QUEST and IgBlast.Some Ig sequences were directly ordered fromGenscript. The VDJ sequences of NP-specific anti-bodies B1-8, B1-8hi and B1-8lo (3C52) were previ-ously described (45, 50, 51) and synthesized byGenscript. Ig sequences were cloned into hu-man IgG1 and human Igl expression vectors, andIgHV1–72 Igl monoclonal antibodies were expressedby transient transfection of HEK293-6E cells andpurified with Protein G Sepharose 4 Fast Flow(GE Healthcare) as described (48). Some anti-bodies that appeared to be instable in PBS pH7.0 were instead buffer exchanged into 20 mMacetic acid/sodium acetate buffer containing 9%sucrose (pH 5.2).

Immunoblot analysis

Transfected HEK293-6E cell cultures were col-lected on day 7, centrifuged and the superna-tant was harvested. The cell pellet was lysed(1% SDS, 10 mM EDTA, 50 mM Tris/HCl, pH 8)and sonicated for 10 min. Antibodies were ana-lyzed in the matched supernatants and cellpellets by SDS-PAGE and immunoblot. HRP-conjugated goat a-human IgG (H+L) antibody(Jackson ImmunoResearch, cat# 109-036-088)was detected with HyGLO Quick Spray (DenvilleScientific).

ELISAs

Autoreactivity against nuclear and cytoplasmicself-antigens was determined with QUANTA LiteANA ELISA (Inova Diagnostics) as described (52).Polyreactivity against ssDNA, dsDNA, Keyholelimpet hemocyanin (KLH), human insulin, andlipopolysaccharide (LPS) was determined as de-scribed (22). To measure NP binding, high-binding96-well plates (Corning) were coated overnightwith 50 ml of PBS containing 10 mg/ml of NP4-BSAor NP25-BSA (Biosearch Technologies). After wash-ing with PBS containing 0.05% of Tween 20 (Sigma),wells were blocked with PBS containing 1% ofBSA for 2 hours at room temperature. Monoclonalantibodies were incubated at 4 mg/ml or 7 con-secutive 1:4 dilutions in PBS for 2 hours at roomtemperature. Afterwashing,HRP-conjugatedgoat

a-human IgG (Jackson ImmunoResearch) wasadded at 0.16 mg/ml for 1 hour at room temper-ature. After additional washing, HRPwas revealedwith1-StepABTSSubstrateSolution(ThermoFisherScientific). Absorbance was measured at 405 nmafter incubation for 20min at room temperature.

Statistical analyses

Statistical significance was determined withGraphpad Prism Version 6.0 using the tests in-dicated in each figure.

REFERENCES AND NOTES

1. G. D. Victora et al., Germinal center dynamics revealed bymultiphoton microscopy with a photoactivatable fluorescentreporter. Cell 143, 592–605 (2010). doi: 10.1016/j.cell.2010.10.032; pmid: 21074050

2. A. D. Gitlin, Z. Shulman, M. C. Nussenzweig, Clonal selection inthe germinal centre by regulated proliferation andhypermutation. Nature 509, 637–640 (2014). doi: 10.1038/nature13300; pmid: 24805232

3. C. D. Allen, T. Okada, H. L. Tang, J. G. Cyster, Imaging ofgerminal center selection events during affinity maturation.Science 315, 528–531 (2007). doi: 10.1126/science.1136736;pmid: 17185562

4. A. E. Hauser et al., Definition of germinal-center B cellmigration in vivo reveals predominant intrazonal circulationpatterns. Immunity 26, 655–667 (2007). doi: 10.1016/j.immuni.2007.04.008; pmid: 17509908

5. D. F. Robbiani et al., Plasmodium infection promotesgenomic instability and AID-dependent B cell lymphoma.Cell 162, 727–737 (2015). doi: 10.1016/j.cell.2015.07.019;pmid: 26276629

6. D. R. Fooksman et al., Development and migration ofplasma cells in the mouse lymph node. Immunity 33,118–127 (2010). doi: 10.1016/j.immuni.2010.06.015;pmid: 20619695

7. B. J. Laidlaw et al., The Eph-related tyrosine kinase ligandEphrin-B1 marks germinal center and memory precursor Bcells. J. Exp. Med. 214, 639–649 (2017). pmid: 28143955

8. Y. J. Liu et al., Mechanism of antigen-driven selection ingerminal centres. Nature 342, 929–931 (1989). doi: 10.1038/342929a0; pmid: 2594086

9. Y. J. Liu et al., Germinal center cells express bcl-2 protein afteractivation by signals which prevent their entry into apoptosis.Eur. J. Immunol. 21, 1905–1910 (1991). doi: 10.1002/eji.1830210819; pmid: 1868875

10. E. E. Crouch et al., Regulation of AID expression in the immuneresponse. J. Exp. Med. 204, 1145–1156 (2007). doi: 10.1084/jem.20061952; pmid: 17452520

11. E. C. Butcher et al., Surface phenotype of Peyer’s patchgerminal center cells: Implications for the role of germinalcenters in B cell differentiation. J. Immunol. 129, 2698–2707(1982). pmid: 6982940

12. H. J. Chae et al., Molecular mechanism of staurosporine-induced apoptosis in osteoblasts. Pharmacol. Res. 42, 373–381(2000). doi: 10.1006/phrs.2000.0700; pmid: 10987998

13. M. Muramatsu et al., Class switch recombination andhypermutation require activation-induced cytidine deaminase(AID), a potential RNA editing enzyme. Cell 102, 553–563(2000). doi: 10.1016/S0092-8674(00)00078-7;pmid: 11007474

14. P. Revy et al., Activation-induced cytidine deaminase (AID)deficiency causes the autosomal recessive form of theHyper-IgM syndrome (HIGM2). Cell 102, 565–575 (2000).doi: 10.1016/S0092-8674(00)00079-9; pmid: 11007475

15. A. Zaheen et al., AID constrains germinal center size byrendering B cells susceptible to apoptosis. Blood 114,547–554 (2009). doi: 10.1182/blood-2009-03-211763;pmid: 19478044

16. G. D. Victora et al., Identification of human germinal centerlight and dark zone cells and their relationship to humanB-cell lymphomas. Blood 120, 2240–2248 (2012). doi: 10.1182/blood-2012-03-415380; pmid: 22740445

17. Q. Wang et al., The cell cycle restricts activation-inducedcytidine deaminase activity to early G1. J. Exp. Med. 214,49–58 (2017). doi: 10.1084/jem.20161649;pmid: 27998928

18. M. Kraus, M. B. Alimzhanov, N. Rajewsky, K. Rajewsky,Survival of resting mature B lymphocytes depends on BCRsignaling via the Iga/b heterodimer. Cell 117, 787–800 (2004).doi: 10.1016/j.cell.2004.05.014; pmid: 15186779

19. Y. Takahashi, P. R. Dutta, D. M. Cerasoli, G. Kelsoe, In situstudies of the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl. V. Affinity maturation develops in twostages of clonal selection. J. Exp. Med. 187, 885–895 (1998).doi: 10.1084/jem.187.6.885; pmid: 9500791

20. J. Tan et al., A LAIR1 insertion generates broadly reactiveantibodies against malaria variant antigens. Nature 529,105–109 (2016). doi: 10.1038/nature16450; pmid: 26700814

21. J. H. Reed, J. Jackson, D. Christ, C. C. Goodnow, Clonalredemption of autoantibodies by somatic hypermutationaway from self-reactivity during human immunization.J. Exp. Med. 213, 1255–1265 (2016). doi: 10.1084/jem.20151978; pmid: 27298445

22. A. D. Gitlin et al., Independent roles of switching andhypermutation in the development and persistence ofB lymphocyte memory. Immunity 44, 769–781 (2016).doi: 10.1016/j.immuni.2016.01.011; pmid: 26944202

23. T. Tiller et al., Autoreactivity in human IgG+ memoryB cells. Immunity 26, 205–213 (2007). doi: 10.1016/j.immuni.2007.01.009; pmid: 17306569

24. J. F. Scheid et al., Differential regulation of self-reactivitydiscriminates between IgG+ human circulating memory B cellsand bone marrow plasma cells. Proc. Natl. Acad. Sci. U.S.A.108, 18044–18048 (2011). doi: 10.1073/pnas.1113395108;pmid: 22025722

25. B. Mietzner et al., Autoreactive IgG memory antibodies inpatients with systemic lupus erythematosus arise fromnonreactive and polyreactive precursors. Proc. Natl. Acad.Sci. U.S.A. 105, 9727–9732 (2008). doi: 10.1073/pnas.0803644105; pmid: 18621685

26. S. Han, B. Zheng, J. Dal Porto, G. Kelsoe, In situ studiesof the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl. IV. Affinity-dependent, antigen-driven B cell apoptosisin germinal centers as a mechanism for maintaining self-tolerance. J. Exp. Med. 182, 1635–1644 (1995). doi: 10.1084/jem.182.6.1635; pmid: 7500008

27. D. Dominguez-Sola et al., The proto-oncogene MYC is requiredfor selection in the germinal center and cyclic reentry.Nat. Immunol. 13, 1083–1091 (2012). doi: 10.1038/ni.2428;pmid: 23001145

28. C. Y. Huang, A. L. Bredemeyer, L. M. Walker, C. H. Bassing,B. P. Sleckman, Dynamic regulation of c-Myc proto-oncogeneexpression during lymphocyte development revealed by aGFP-c-Myc knock-in mouse. Eur. J. Immunol. 38, 342–349(2008). doi: 10.1002/eji.200737972; pmid: 18196519

29. J. Mueller, M. Matloubian, J. Zikherman, Cutting edge:An in vivo reporter reveals active B cell receptor signaling inthe germinal center. J. Immunol. 194, 2993–2997 (2015).doi: 10.4049/jimmunol.1403086; pmid: 25725108

30. J. M. Tas et al., Visualizing antibody affinity maturationin germinal centers. Science 351, 1048–1054 (2016).doi: 10.1126/science.aad3439; pmid: 26912368

31. Z. Shulman et al., Dynamic signaling by T follicular helpercells during germinal center B cell selection. Science345, 1058–1062 (2014). doi: 10.1126/science.1257861;pmid: 25170154

32. Z. Shulman et al., T follicular helper cell dynamics in germinalcenters. Science 341, 673–677 (2013). doi: 10.1126/science.1241680; pmid: 23887872

33. L. S. Yeap et al., Sequence-intrinsic mechanisms thattarget AID mutational outcomes on antibody genes. Cell163, 1124–1137 (2015). doi: 10.1016/j.cell.2015.10.042;pmid: 26582132

34. K. P. Lam, R. Kühn, K. Rajewsky, In vivo ablation of surfaceimmunoglobulin on mature B cells by inducible genetargeting results in rapid cell death. Cell 90, 1073–1083 (1997).doi: 10.1016/S0092-8674(00)80373-6; pmid: 9323135

35. K. Rajewsky, Clonal selection and learning in the antibodysystem. Nature 381, 751–758 (1996). doi: 10.1038/381751a0;pmid: 8657279

36. L. Srinivasan et al., PI3 kinase signals BCR-dependentmature B cell survival. Cell 139, 573–586 (2009). doi: 10.1016/j.cell.2009.08.041; pmid: 19879843

37. N. C. Shaner et al., A bright monomeric green fluorescentprotein derived from Branchiostoma lanceolatum. Nat. Methods10, 407–409 (2013). doi: 10.1038/nmeth.2413; pmid: 23524392

38. L. Tyas, V. A. Brophy, A. Pope, A. J. Rivett, J. M. Tavaré, Rapidcaspase-3 activation during apoptosis revealed usingfluorescence-resonance energy transfer. EMBO Rep. 1,

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 7 of 8

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

266–270 (2000). doi: 10.1093/embo-reports/kvd050;pmid: 11256610

39. A. J. Lam et al., Improving FRET dynamic range with brightgreen and red fluorescent proteins. Nat. Methods 9, 1005–1012(2012). doi: 10.1038/nmeth.2171; pmid: 22961245

40. D. F. Robbiani et al., AID produces DNA double-strandbreaks in non-Ig genes and mature B cell lymphomaswith reciprocal chromosome translocations. Mol. Cell 36,631–641 (2009). doi: 10.1016/j.molcel.2009.11.007;pmid: 19941823

41. D. F. Robbiani, K. Colon, M. Affer, M. Chesi, P. L. Bergsagel,Maintained rules of development in a mouse B-cell tumor.Leukemia 19, 1278–1280 (2005). doi: 10.1038/sj.leu.2403774;pmid: 15843816

42. T. H. Thai et al., Regulation of the germinal center responseby microRNA-155. Science 316, 604–608 (2007). doi: 10.1126/science.1141229; pmid: 17463289

43. S. Warming, N. Costantino, D. L. Court, N. A. Jenkins,N. G. Copeland, Simple and highly efficient BACrecombineering using galK selection. Nucleic Acids Res. 33,e36 (2005). doi: 10.1093/nar/gni035; pmid: 15731329

44. M. Guo et al., A monoclonal antibody to the DEC-205endocytosis receptor on human dendritic cells. Hum. Immunol.61, 729–738 (2000). doi: 10.1016/S0198-8859(00)00144-0;pmid: 10980384

45. T. A. Shih, M. Roederer, M. C. Nussenzweig, Role of antigenreceptor affinity in T cell-independent antibody responsesin vivo. Nat. Immunol. 3, 399–406 (2002). doi: 10.1038/ni776;pmid: 11896394

46. M. M. Meredith et al., Expression of the zinc fingertranscription factor zDC (Zbtb46, Btbd4) defines the classicaldendritic cell lineage. J. Exp. Med. 209, 1153–1165 (2012).doi: 10.1084/jem.20112675; pmid: 22615130

47. D. F. Robbiani et al., AID is required for the chromosomalbreaks in c-myc that lead to c-myc/IgH translocations.

Cell 135, 1028–1038 (2008). doi: 10.1016/j.cell.2008.09.062;pmid: 19070574

48. L. von Boehmer et al., Sequencing and cloning of antigen-specific antibodies from mouse memory B cells. Nat. Protoc.11, 1908–1923 (2016). doi: 10.1038/nprot.2016.102;pmid: 27658009

49. P. Dosenovic et al., Immunization for HIV-1 broadly neutralizingantibodies in human Ig knockin mice. Cell 161, 1505–1515(2015). doi: 10.1016/j.cell.2015.06.003; pmid: 26091035

50. D. Allen, T. Simon, F. Sablitzky, K. Rajewsky, A. Cumano, Antibodyengineering for the analysis of affinity maturation of an anti-hapten response. EMBO J. 7, 1995–2001 (1988). pmid: 3138111

51. A. Cumano, K. Rajewsky, Clonal recruitment and somaticmutation in the generation of immunological memory to thehapten NP. EMBO J. 5, 2459–2468 (1986). pmid: 2430792

52. H. Wardemann et al., Predominant autoantibody production byearly human B cell precursors. Science 301, 1374–1377(2003). doi: 10.1126/science.1086907; pmid: 12920303

ACKNOWLEDGMENTS

We thank T. Eisenreich for help with mouse colony management;K.-H. Yao for technical help; K. M. Gordon and N. M. Thomas forcell sorting; J. Moore, D. Robbiani, M. Jankovic, and all members ofthe Nussenzweig laboratory for discussion. We further thankC. Yang, the Gene Targeting Resource Center, and A. North andK. Thomas from the Bio-imaging Resource Center of The RockefellerUniversity with support by the Empire State Stem Cell Fundthrough New York State Department of Health (NYSDOH) contract#C023046. Opinions expressed here are solely those of theauthors and do not necessarily reflect those of the Empire StateStem Cell Fund, the NYSDOH, or the state of New York. Thedata presented in this manuscript are tabulated in the main paperand in the supplementary materials. C.T.M. was supported by aEuropean Molecular Biology Organization (EMBO) long-term

fellowship (ALTF 456-2014) and by the European CommissionSeventh Framework Programme (FP7) (Marie Curie Actions,EMBOCOFUND2012, GA-2012-600394). A.D.G. was supported bythe NIH Medical Scientist Training Program grant T32GM07739and the National Institute of Allergy and Infectious Diseases (NAID)of the NIH, grant F30-AI109903-03. M.M.-R. and R.W.S. weresupported by Aidsfonds Netherlands. M.C.N. was supported bythe Bill and Melinda Gates Foundation Collaboration for AIDSVaccine Discovery (OPP1033115 and OPP1124068), the NIH Centerfor HIV/AIDS Vaccine Immunology and Immunogen Discovery(CHAVI-ID, 1UM1 AI100663), the NAID of the NIH (AI100148,AI037526), the Robertson Foundation, and the RockefellerUniversity. M.C.N. is an HHMI investigator. M.M.-R. and R.W.S. areinventors on a patent application submitted by the Universityof Amsterdam that covers GT1.1. This work is licensed under aCreative Commons Attribution 4.0 International (CC BY 4.0)license, which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properlycited. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. This license does not apply to figures/photos/artwork or other content included in the article that iscredited to a third party; obtain authorization from the rightsholder before using such material.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/358/6360/eaao2602/suppl/DC1Figs. S1 to S9Table S1ReferencesMovies S1 to S3

30 June 2017; accepted 29 August 2017Published online 21 September 201710.1126/science.aao2602

Mayer et al., Science 358, eaao2602 (2017) 13 October 2017 8 of 8

RESEARCH | RESEARCH ARTICLEon M

arch 23, 2020

http://science.sciencemag.org/

Dow

nloaded from

The microanatomic segregation of selection by apoptosis in the germinal center

Oliveira, Qiao Wang, Amelia Escolano, Max Medina-Ramirez, Rogier W. Sanders and Michel C. NussenzweigChristian T. Mayer, Anna Gazumyan, Ervin E. Kara, Alexander D. Gitlin, Jovana Golijanin, Charlotte Viant, Joy Pai, Thiago Y.

originally published online September 21, 2017DOI: 10.1126/science.aao2602 (6360), eaao2602.358Science 

, this issue p. eaao2602; see also p. 171Sciencegenes damaged by random antibody-gene mutations.they were rescued by positive selection. In contrast, apoptotic dark-zone B cells were highly enriched among cells with mechanisms were distinct and microanatomically segregated. Light-zo ne B cells underwent apoptosis by default unlessexperience very high rates of apoptosis (see the Perspective by Bryant and Hodgkin). However, the underlying

studied apoptosis reporter mice and found that both GC zoneset al.interacting with T follicular helper cells. Mayer undergo somatic hypermutation, and the light zone, where they are selected for affinity-enhancing mutations after

andnormal immune responses. GCs are separated into two anatomic compartments: the dark zone, where B cells divide Germinal centers (GCs) are areas within lymphoid organs where mature B cells expand and differentiate during

Light- and dark-zone death dynamics

ARTICLE TOOLS http://science.sciencemag.org/content/358/6360/eaao2602

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2017/09/20/science.aao2602.DC1

CONTENTRELATED

http://stke.sciencemag.org/content/sigtrans/11/542/eaai9192.fullhttp://science.sciencemag.org/content/sci/358/6360/171.full

REFERENCES

http://science.sciencemag.org/content/358/6360/eaao2602#BIBLThis article cites 52 articles, 20 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2017, American Association for the Advancement of Science

on March 23, 2020

http://science.sciencem

ag.org/D

ownloaded from