nature medicine  volume 16 | number 1 | january 2010 123

t e c H n i c a L r e P O r t S

The B cell lymphoma-6 (Bcl-6) and Bcl-xL proteins are expressed in germinal center B cells and enable them to endure the proliferative and mutagenic environment of the germinal center. By introducing these genes into peripheral blood memory B cells and culturing these cells with two factors produced by follicular helper T cells, CD40 ligand (CD40L) and interleukin-21 (IL-21), we convert them to highly proliferating, cell surface B cell receptor (BCR)–positive, immunoglobulin-secreting B cells with features of germinal center B cells, including expression of activation-induced cytidine deaminase (AID). We generated cloned lines of B cells specific for respiratory syncytial virus and used these cells as a source of antibodies that effectively neutralized this virus in vivo. This method provides a new tool to study B cell biology and signal transduction through antigen-specific B cell receptors and for the rapid generation of high-affinity human monoclonal antibodies.

Stable monoclonal cell lines of immortalized human B cells that express the BCR on their cell surface while secreting antibodies would be attractive tools not only for studying various aspects of BCR signal-ing but also for the generation of human monoclonal antibodies. BCR expression on polyclonal immortalized human B cells would facilitate selection of antigen-specific cells on the basis of binding of antigen to the specific BCR, whereas production of antibodies would enable selection of B cell clones on the basis of the functional activities of the secreted antibodies.

Mature B cells can be cultured in vitro with CD40L and cytokines such as IL-4, IL-10 or IL-21 (refs. 1–3). Whereas B cells cultured with CD40L, IL-2 and IL-4 produce very little immunoglobulin, addition of IL-21 leads to differentiation to plasma cells accompanied by high immunoglobulin secretion4,5. Although this in vitro system has proven useful to study some aspects of B cell differentiation, both naive IgD+

B cells and switched IgD− memory B cells eventually differentiate into terminally differentiated plasma cells, a process accompanied by cell cycle arrest precluding the generation of long-term, antigen-specific BCR+ cell lines. We hypothesized that prevention of terminal differ-entiation would allow activated B cells to expand in vitro for much longer periods.

Recent advances have provided insight into how multiple transcrip-tion factors, including B lymphocyte–induced maturation protein-1, X-box–binding protein-1 and Bcl-6 control development of germinal center B cells into terminally arrested, antibody-producing plasma cells6. The transcriptional repressor Bcl-6 has been shown to pre-vent plasma cell differentiation in germinal center B cells, where it facilitates expansion of B cells by downregulating p53 and prevents premature differentiation of B cells into plasma cell by negatively reg-ulating B lymphocyte–induced maturation protein-1 (refs. 5,7–10).

Recently, we reported that forced expression of constitutively active signal transducer and activator of transcription-5 (STAT5) in B cells induced expression of both Bcl-6 and Bcl-xL and prevented termi-nal differentiation of B cells in vitro11–13. Here we show that ectopi-cally expressed Bcl-6 and Bcl-xL synergize with CD40L and IL-21 to increase the proliferative lifespan of B cells in vitro. The expanded cells have features of germinal center B cells and express a functional BCR while secreting antibodies, providing a new and powerful tool to generate human BCR+, antibody-secreting B cell lines.

RESULTSTransduction of memory B cells with Bcl-6 and Bcl-xLTo test whether Bcl-6 and Bcl-xL synergize to increase the prolifera-tive and survival potential of B cells in vitro, we introduced the genes encoding these proteins into isolated human peripheral blood CD27+ memory B cells by retrovirus-mediated gene transfer. We cultured transduced cells on irradiated L cells expressing CD40L (CD40L-L cells)

Generation of stable monoclonal antibody–producing B cell receptor–positive human memory B cells by genetic programmingMark J Kwakkenbos1,2,8, Sean A Diehl2,7,8, Etsuko Yasuda1,2, Arjen Q Bakker1,2, Caroline M M van Geelen1,2, Michaël V Lukens3, Grada M van Bleek3, Myra N Widjojoatmodjo4, Willy M J M Bogers5, Henrik Mei6, Andreas Radbruch6, Ferenc A Scheeren2,7, Hergen Spits1,2 & Tim Beaumont1,2

1AIMM Therapeutics and 2Department of Cell Biology and Histology, Academic Medical Center, Amsterdam, The Netherlands. 3Department of Pediatrics, The Wilhelmina Children’s Hospital, University Medical Center, Utrecht, The Netherlands. 4Netherlands Vaccine Institute, Bilthoven, The Netherlands. 5Department of Virology, Biomedical Primate Research Centre, Rijswijk, The Netherlands. 6German Rheumatism Research Center, Berlin, Germany. 7Current addresses: Department of Medicine, University of Vermont, Burlington, Vermont, USA (S.A.D.) and Stanford Institute for Stem Cell Biology and Regenerative Medicine, Stanford University, Palo Alto, California, USA (F.A.S.). 8These authors contributed equally to this work. Correspondence should be addressed to T.B. (tbeaumont@aimmtherapeutics.com).

Received 18 November 2008; accepted 2 July 2009; published online 20 December 2009; doi:10.1038/nm.2071

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

124  volume 16 | number 1 | january 2010 nature medicine

t e c H n i c a L r e P O r t S

in the presence of IL-21, a factor produced by follicular helper T cells in the germinal center that strongly induces proliferation of B cells3. We visualized the transduced cells by their expression of a truncated form of the nerve growth factor receptor (∆NGFR) (for Bcl-6) and GFP (for Bcl-xL) (Fig. 1a). Cells expressing both Bcl-6 and Bcl-xL rapidly increased in frequency to more than 95% of the culture within

a 2-week period (Fig. 1b) and these cells showed a clear expansion advantage compared to cells expressing Bcl-6 or Bcl-xL alone (Fig. 1c). Normal human B cells cultured with CD40L and IL-21 rapidly dif-ferentiate to antibody-producing plasma cells4, a process that is accompanied by a decreased expression of surface BCR and human leukocyte antigen-DR (HLA-DR) and an increased expression of CD38 (ref. 14). In contrast, cells transduced with Bcl-6 alone and cells transduced with Bcl-6 and Bcl-xL both retained BCR expression and were HLA-DRhighCD38intermediate (Fig. 1d), confirming that Bcl-6 inhibits B cell differentiation12,13.

Flow cytometric analysis of the Bcl-6– and Bcl-xL–transduced CD27+ memory B cells expanded with CD40L and IL-21 revealed that they express CD19, CD20, CD21 and CD22, the activation mark-ers CD25, CD30, CD70, CD80, CD86, CD95 and inducible T cell co-stimulator ligand, the cytokine receptors CD132 (γc) and IL-21R, and the BCR (Fig. 2a). In addition, they express activation-induced cytidine deaminase (AICDA) transcripts (Fig. 2b).

When we introduced Bcl-6 and Bcl-xL into B cells from nonhuman primates and mice by retrovirus-mediated gene transfer, we observed rapid expansion for at least 1 month in the presence of IL-21 and CD40L-expressing L cells (Supplementary Fig. 1), similarly to trans-duced human B cells, indicating that Bcl-6 and Bcl-xL allow expansion of B cells of multiple species.

Figure 1 Overexpression of Bcl-6 and Bcl-xL confers a high proliferative capacity and fixed differentiation phenotype to peripheral blood CD27+IgG+ memory B cells. (a) Flow cytometry identification of Bcl-6–transduced (∆NGFR+), Bcl-xL–transduced (GFP+) and Bcl-6– and Bcl-xL–transduced (GFP+∆NGFR+) CD19+ B cells in culture with CD40L-L cells and IL-21 4 d after transduction. (b) Percentages of the individually transduced populations over time within a single unsorted culture maintained with CD40L-L cells and IL-21. (c) Cumulative cell numbers, obtained from cultures with unsorted double-transduced cells (left) or with individual transduced populations that were purified by flow cytometry and cultured separately (right). The cumulative cell number was calculated on the basis of the original number of transduced B cells at input and illustrates the theoretical absolute cell numbers produced in culture. (d) CD38, HLA-DR and surface immuno-globulin light chain κ (IgL-κ) or λ (IgL-λ) expression on CD19+ cells from unsorted double transduced cultures. The gates in the top graphs show the percentage of terminally differentiated cells. The gates in the lower graphs show the percentage of light chain positive cells. Phenotype shown of cells 7 d after transduction. Data represent four separate experiments.

31 50

11

Bcl-6–∆NGFR

Bcl

-xL–

GF

P

8.8

0

4 7 10 13Time after transduction (d)

Time after transduction (d)

16 19 22

25

50P

erce

ntag

etr

ansd

uced

cel

ls

75

100

Bcl-xL onlyBcl-6 only

Bcl-6 + Bcl-xL

Negative

Unsorted

Negative

CD38

IgL-κ

55 47 5.8 1.7

11 14 39 40

41369.86

IgL-

λH

LA-D

R

Bcl-xL only Bcl-6 only Bcl-6 + Bcl-xL

Sorted

Bcl-xL onlyBcl-6 + Bcl-xL Bcl-6 only

Negative

4 8 12 16 4 8 12 16

105

106

Cum

ulat

ive

cell

num

ber

107

108

104

106

105

107

108

a

b

c

d

GC

N/M

CD38

TonsilCD19+CD3– Bcl-6 + Bcl-xL

CD25 (IL-2Rα) CD30 CD70 CD80 CD86

CD95 ICOSL HLA-DR CD71 CD27

IL-21R CD132(γC chain)

BulkBcl-6 + Bcl-xL

transduced

Tonsil GC

IgG memory

0 0.2 0.4 0.6Relative AICDA expression

0.8 1.0 1.2

CXCR4 CD40 CD10

CD19 CD21 CD22

PC6XL

CD

20

a

b

Figure 2 CD27+ memory peripheral blood cells acquire a stable germinal center–like phenotype after transduction with Bcl-6 and Bcl-xL and subsequent culturing. (a) Phenotype of Bcl-6– and Bcl-xL–transduced CD27+ memory B cells (6XL) compared to tonsil germinal center (GC) cells (CD38+CD20+), tonsil naive and memory cells (N/M, CD38−CD20low) and tonsil plasma cells (PC, CD38+CD20low). Bcl-6– and Bcl-xL–transduced monoclonal cell lines showed an identical phenotype (data not shown). ICOSL, inducible T cell co-stimulator ligand. (b) Relative mRNA levels of AICDA in CD19+IgG+CD27+ peripheral blood memory B cells and CD19+CD38+CD20+IgD− tonsillar GC B cells (value set as 1) compared to Bcl-6– and Bcl-xL–transduced bulk CD27+ memory cells, determined by quantitative RT-PCR. Data represent four separate experiments; error bars represent s.d.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nature medicine  volume 16 | number 1 | january 2010 125

t e c H n i c a L r e P O r t S

Selection of B cell clones on the basis of antigen bindingThe expression of the BCR should allow for isolation of antigen-specific B cells. Therefore, we isolated tetanus toxin–specific B cells from peripheral blood by flow cytometry using phycoerythrin-labeled tetanus toxoid (Fig. 3a) and seeded the sorted cells at limiting cell numbers at one cell per two wells and one cell per one well into a 96-well plate. Within 2 weeks, 39 (frequency of 68%) tetanus toxin–specific B cell clones, positive for either IgGκ or IgGλ light chain, arose and were expanded further (Fig. 3a,b). Cloned cells cultured with CD40L and IL-21 maintained a very short doubling time of 35–40 h (Fig. 3c), and could be maintained for extended periods of time, during which they remained tetanus toxin specific (>4 months, data not shown).

To examine whether the BCR on B cells transduced with Bcl-6 and Bcl-xL was functional, we cultured IgG+ tetanus toxin–specific and non-specific B cell lines in the absence of CD40L-L cells or IL-21 for 4 h. We then stimulated the rested cells with tetanus toxin antigen or a human

IgG-specific F(ab)2 antibody and analyzed them for phosphorylation (p) of extracellular signal–regulated kinase (ERK)15. Within 30 s after addition of tetanus toxin, we detected pERK in the tetanus toxin–specific cells but not in the control cells that did not bind tetanus toxin (Fig. 3d), indicating the presence of antigen-specific, functional, BCR-coupled sig-nal transduction machinery in these cells. We detected pERK in both cell lines stimulated with the IgG-specific F(ab)2 antibody (Fig. 3d). Addition of tetanus toxin or IgG-specific F(ab)2 antibody to cells labeled with the calcium-sensing dye Indo-1 induced a rapid shift in the ratio of bound versus unbound Indo-1, indicating that tetanus toxin and IgG-specific F(ab)2 antibody induce Ca2+ flux (Fig. 3e). These results show that the BCR expressed on the cell surface of memory B cells transduced with Bcl-6 and Bcl-xL is signaling competent.

Selection of B cell clones secreting neutralizing antibodiesThe IL-21 used to expand CD27+ B cells transduced with Bcl-6 and Bcl-xL is a strong inducer of differentiation into immunoglobu-lin-secreting cells4. The transduced cells indeed secreted immu-noglobulin (Fig. 4a), but the average production of IgG by cells transduced with Bcl-6 and Bcl-xL was lower than that of plasma cells, although it was higher than that of memory B cells (Supplementary Fig. 2). Analysis of the IgG and IgM amounts produced by mono-clonal cell lines in a 3-d culture revealed that all clones produced

Before TT sort After TT sort

97.60.76

TT

CD

19

Kappa+

clone

IgL-κ

IgL-λ

Rel

ativ

e ce

ll nu

mbe

r

slgG

Lambda+

clone

TT

CD

19

TT clone 1TT clone 2

104

0 10 20Time (d)

30 40 50

105106107108109

1010

Cum

ulat

ive

cell

num

ber

1011101210131014

Stimulation

200

170

140

Rat

io b

ound

ver

sus

unbo

und

Indo

-1

110

800 100

Time (s)200 300

F(ab)2 anti-IgG

Control IgGTT+

TT–TT antigen

0 0.5 0.5 30 Time (min)pERK

STAT3

STAT3

pERK

1

TT antigen

TT+

TT–

12 2

F(ab)2 anti-IgG

5 5

a c

d

e

b

Figure 3 Generation and characterization of tetanus toxin (TT)-specific monoclonal human B cell lines. (a) Cell sorting of TT-specific B cells from total Bcl-6– and Bcl-xL–transduced peripheral blood CD19+CD27+ memory B cells using TT-phycoerythrin staining. (b) Expression of either surface IgG, IgL-λ or IgL-κ light chain on monoclonal cell lines. Shaded histograms are CD19+ tonsil B cells. (c) The cumulative cell number (as in Fig. 1c) of two representative B cell clones in time. (d) Immunoblot analysis of pERK in TT-specific and control TT negative cell lines incubated with TT or stimulated with a human IgG–specific F(ab)2. STAT3 blotting was used to verify equal loading. Results are representative of two experiments (e) Calcium flux in TT-specific IgG+ B cell lines. Cells were loaded with Indo-1 AM and were subsequently stimulated with TT, a human IgG specific F(ab)2 or control IgG. The Indo-1 response of a TT-negative clone stimulated with TT antigen is shown as an antigen specificity control. Shown is the ratio of bound versus unbound Indo-1. Data represent three separate experiments.

350

300

250

200

150

Imm

unog

lobu

linpr

oduc

tion

(ng

per

10,0

00 c

ells

)

100

50

0IgG 0.01 0.1 1

IgG (ng ml–1)10 100 1,000

rD25AM14

AM16AM23

Palivizumab

rD25AM14AM16AM23Palivizumab

PalivizumabrD25

IgG1 ctrl mAb

IgM

100

80

60

40

Per

cent

age

RS

V-A

2 in

hibi

tion

20

0

0.1 1IgG (ng ml–1)

10

0.04IgG1 dose (mg per kg body weight)

0.15 0.6 2

100 1,000 10,000

100

80

60

40

Per

cent

age

RS

V-X

inhi

bitio

n

20

0

6

5

4

Lung

RS

V ti

ters

(log 10

TC

ID50

per

g)

3

< 2.1

a b

c

dFigure 4 Isolation of high-affinity, broadly RSV-neutralizing anti-bodies from RSV-specific B cell clones. (a) IgG or IgM production of individually cloned Bcl-6– and Bcl-xL–immortalized cell lines, from the CD19+CD27+IgA−IgM− and CD19+CD27+IgA−IgG− memory B cell pool, respectively, maintained on CD40L-L cells and IL-21 in a 3-d culture. (b) Neutralization dose-response curve against the RSV A2 virus for the newly generated RSV antibodies D25, AM14, AM16 and AM23 compared to palivizumab. (c) Neutralization dose-response curve of the RSV-specific antibodies D25, AM14, AM16, AM23 and palivizumab against the primary RSV A isolate X. (d) Effects of rD25, palivizumab and a control IgG1 on RSV replication in cotton rat lungs.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

126  volume 16 | number 1 | january 2010 nature medicine

t e c H n i c a L r e P O r t S

immunoglobulin (Fig. 4a). The IgM-producing clones secreted five to six times more immunoglobulin than the IgG-producing clones (Fig. 4a), which has also been observed in cultures with nontrans-duced memory B cells16.

Given the relatively high amounts of secreted antibodies by B cells transduced with Bcl-6 and Bcl-xL, we examined whether we could select respiratory syncytial virus (RSV)-specific B cells on the basis of secretion of specific antibody. RSV is the most common cause of bronchiolitis and pneumonia among infants and children under 1 year of age and is a serious health problem for elderly people17. We transduced memory B cells from a healthy donor with Bcl-6 and Bcl-xL, seeded them at 100 cells per well and expanded them with CD40L-L cells and IL-21. After 2 weeks of culture, we collected the supernatants and screened them for the presence of RSV-neutralizing antibodies in a microneutralization experiment18. Of 384 cultures (100 cells per well), 31 prevented RSV infection of HEp2 cells. We subcloned the four microcultures with the highest neutralizing activity by limiting dilution. We then selected and characterized four monoclonal cell lines (D25, AM14, AM16 and AM23). We observed median half-maximal inhibitory concentrations (IC50) against the RSV-A2 virus ranging from 2.1 ng ml−1 for D25 and AM14 and 4.3 ng ml−1 for AM23 to 145 ng ml−1 for AM16 (Fig. 4b). Three of the four antibodies showed substantially lower IC50 values compared to palivizumab (IC50 209 ng ml−1), a humanized monoclonal antibody that is currently used as prophylaxis in infants who are at high risk of developing pathology after infection with RSV19. In addition, all RSV-specific antibodies efficiently neutralized a primary RSV isolate X (RSV-X) (Fig. 4c).

D25 antibody inhibits RSV infection in cotton ratsTo test the in vivo neutralizing activity of D25 in a prophylactic set-ting, we infected cotton rats, the standard model to test RSV infectiv-ity and pathogenicity with the primary isolate RSV-X20. We injected rats intramuscularly (i.m.) with various quantities of recombinant D25 (0.6, 0.15 and 0.04 mg per kg body weight), palivizumab (2.0 and 0.6 mg per kg body weight) or control IgG1 (2.0 mg per kg body weight) 1 d before intranasal inoculation with RSV-X (1 × 106 half-maximal tissue culture infectious dose (TCID50) per rat). Whereas palivizumab completely prevented in vivo viral replication at 2.0 mg per kg body weight, D25 showed the same efficacy already at a dose of 0.6 mg per kg body weight and was partially neutralizing at a dose of 0.15 mg per kg body weight (Fig. 4d). These data show that D25 is functional and potent in reducing RSV replication also in vivo.

AID is active in cells transduced with Bcl-6 and Bcl-xLAnalysis of AICDA expression in 23 monoclonal cell lines by real-time PCR revealed a variable expression in different clonal cell lines (data not shown). Several clones expressed levels similar to those in germi-nal center tonsil B cells, whereas others expressed low levels of AICDA (Fig. 5a). To determine whether AID is functional in the B cell clones, we subcloned the RSV-specific monoclonal B-cell line D25 by single-cell sorting and analyzed the genes encoding the immunoglobulin variable heavy chain region for the presence of mutations. Expression of AID did not result in genetic instability leading to growth arrest and cell death, as 63% of the wells that were seeded showed robust expansion (data not shown). After 3 weeks, we collected the culture supernatant of 108 subclones and isolated the RNA from the cells. Sequence analysis revealed a total of 184 IGVH mutations (107 unique mutations), resulting in an estimated mutation rate between 8.85 × 10−5 and 5.14 × 10−5 mutations per base pair per cell division, which is at the lower end of the estimated AID-mediated mutation rate in vivo (1 × 10−3 to 1 × 10−5)21. The IGVH genes of the individual subclones showed a variable number of mutations, with 65% of subclones harboring one to three mutations (of which 23% were silent), 11% harboring more than three mutations and 24% without mutations (Fig. 5b). The 372–base pair IGVH gene of D25 contains 26 AID mutational hotspots (RGYW/WRCY; the underlined nucleotide is preferentially mutated by AID) that account for 30% of the total muta-tions. We observed mutations predominantly in the complementarity-determining regions and framework region-3 (Fig. 5c).

Although the supernatants of the majority of D25 subclones bound RSV-infected HEp2 cells similarly to recombinant D25, supernatants from some clones bound either less or more than D25 (Fig. 5d). The differences in binding activities were associated with mutations in the VH regions.

For some applications, it would be desirable to inhibit AID to pre-vent accumulation of mutations in the immunoglobulin genes of B cell clones transduced with Bcl-6 and Bcl-xL. To achieve this, we took advantage of the fact that AID is regulated by the basic helix- loop-helix transcription factor E47 (ref. 22). We overexpressed the helix-loop-helix factor inhibitor of DNA binding-3 (Id-3), which is known to form transcriptionally inactive complexes with E47, thereby inhibiting AID expression in D25 cells22. Overexpression of Id-3 strongly reduced AICDA levels in the D25 cell line (Fig. 5e) without influencing proliferation of the D25 cells (data not shown). Thus, modulation of Id-3 abundance provides a method to prevent AID-induced mutations in B cells transduced with Bcl-6 and Bcl-xL.

Figure 5 Expression and activity of AID in Bcl-6– and Bcl-xL–transduced cells. (a) mRNA levels of AICDA in CD19+CD38+CD20+IgD− tonsillar GC B cells and CD19+IgG+CD27+ peripheral blood memory B cells compared to 23 Bcl-6– and Bcl-xL–transduced monoclonal cell lines and monoclonal cell line D25, as determined by quantitative RT-PCR. (b) Percentage of subclones with the indicated number of VH mutations, as percentage of total number of subclones sequenced. (c) Location of VH mutations, percentage of mutations per VH region per base pair. (d) Binding of D25 subclone immunoglobulin to RSV-infected HEp2 cells. Blue circles indicate clones with deviating binding activity. (e) mRNA levels of AICDA in the monoclonal cell line D25 transduced with control-YFP or Id-3–YFP, as determined by quantitative RT-PCR. Similar results were obtained with overexpression of Id-2 (data not shown).

Tonsil GC

IgGmemory

Clonalcell lines

D25

0 0.2 0.4 0.6Relative AICDA expression

0.8 1.0 1.2 0 1 2 3 4 5Number of IGVH mutations

0

5

10

15

20

Per

cent

age

of D

25 s

ubcl

ones

25

30

6 7 8 9 10

1.2

0.9

0.6

Rel

ativ

e AICDA

exp

ress

ion

0.3

0Id3 Ctrl

a b

FR105

101520

Per

cent

age

mut

atio

ns p

erIG

VH

reg

ion

per

base

pair

2530354045

CDR1FR2

CDR2FR3

CDR3FR4 0

0

1,000

2,000

Mea

n flu

ores

cenc

e in

tens

ity

3,000

4,000 D25 subclonesrD25

500IgG (ng ml–1)

1,000 1,500

c d e

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nature medicine  volume 16 | number 1 | january 2010 127

t e c H n i c a L r e P O r t S

DISCUSSIONHere we have described an efficient and rapid method for the genera-tion of antigen-specific, BCR+ B cell lines. By retroviral introduction of Bcl-6 and Bcl-xL, we converted peripheral blood memory cells into germinal center–like B cells that can be maintained for prolonged periods in culture. The cells transduced with Bcl-6 and Bcl-xL express markers for human germinal center B cells including AID, CD38, CXCR4 and inducible T-cell co-stimulator ligand; the latter has been shown to have a key role in germinal center formation23.

AID was expressed in all clones that were derived from CD27+ peripheral blood memory cells transduced with Bcl-6 and Bcl-xL, but the level was variable. We did not observe isotype switching in these cells over time, indicating that AICDA expression by itself is insufficient to drive this process at least in the time frame and culture conditions we applied. However, we found somatic hypermutations at a relatively low frequency in the IGH genes of subclones of one RSV-specific B cell clone. These findings indicate that the require-ments for AID-mediated somatic hypermutation and class-switch recombination are different24,25.

Screening of the immunoglobulin-containing culture supernatant of D25 subclones on RSV-infected cells revealed variations in bind-ing capacity, suggesting that the VH gene mutations of some sub-clones resulted in changes in the antigen-binding affinity. These data raise the possibility that ongoing AID-mediated mutational activ-ity in B cell clones can be used to select higher (or lower) affinity clones, obviating the need for deliberate molecular engineering of the relevant immunoglobulin-encoding genes. It is conceivable that, in some cases, mutations in the immunoglobulin-encoding genes of established cloned B cell lines need to be avoided. Introduction of Id-3 into B cells that already overexpress Bcl-6 and Bcl-xL resulted in inhibition of AICDA expression, consistent with previously reported findings22. Thus overexpression of Id-3 should prevent ongoing AID-mediated mutational activity, if changes in the IGVH and IGVL genes are undesirable.

We previously showed that IL-21 was able to elicit immunoglobulin secretion from polyclonal Bcl-6–expressing B cells13. Whereas stable BCR expression allows for selection of antigen-specific B cells by binding of fluorescent-labeled antigen, the secreted immunoglobu-lin can be screened for functional activity. Both methods provide a tool for rapid isolation of potential therapeutic antibodies. Indeed, we were able to generate unique RSV-neutralizing antibodies within 3 months after initiation of the B cell cultures by direct functional screens after immortalization of B cells obtained from an RSV-exposed individual.

Recently, another method of generating monoclonal antibod-ies from human B cells was described on the basis of Epstein-Barr virus (EBV) transformation of B cells activated with Toll-like recep-tor 9 antagonists26. Both that method26 and ours use human B cells from pathogen-exposed individuals to select monoclonal antibodies against a variety of pathogens without the need for recent vaccination of the donors. However, in contrast to B cells transformed by EBV27, all B cell lines obtained with Bcl-6– and Bcl-xL–mediated genetic programming stably express high levels of a signaling-competent, membrane-bound BCR. This high level of cell surface BCR facilitates isolation of B cells on the basis of antigen binding, which greatly increases the frequency of antigen-specific cells and abrogates the need for extensive subcloning.

The fact that rare B cells can easily be cloned compares our method favorably with a recently described single-cell cloning approach used to isolate influenza-specific antibodies28. The feasibility of this

technique depends on a relatively high frequency of antigen-specific B cells that can be obtained by vaccination or very recent exposure to a pathogen28,29. As discussed above, expression of AID in the geneti-cally modified clones may be used to select antibodies with higher affinities, providing an advantage compared to the EBV transforma-tion and single-cell PCR methods as platforms to generate human monoclonal antibodies.

Another advantage of the method described here is that it can be used to obtain stable lines of B cells from species other than humans. We have shown here that Bcl-6– and Bcl-xL–transduced B cells from mice and nonhuman primates proliferate for at least 1 month in the presence of CD40L and IL-21. We also generated a series of cloned lines of mouse B cells (data not shown). It is, therefore, feasible to assume that our method will allow for expansion of antigen-specific B cells of multiple species, including rabbits and goats. Such lines cannot be obtained with EBV, although EBV can transform B cells from a variety of monkey species30.

The monoclonal cell lines that can be generated with this tech-nology are very useful not only for generation of monoclonal anti-bodies but also for detailed mechanistic studies of BCR-mediated signal transduction induced by specific antigens. They may be used to study signaling of BCR variants that differ in affinity for antigens. Furthermore, these cell lines can be useful tools to study the effects of a variety of external signals such as antigen or cytokines on the func-tion of AID and the induction of somatic hypermutation.

METhODSMethods and any associated references are available in the online version of the paper at http://www.nature.com/naturemedicine/.

Note: Supplementary information is available on the Nature Medicine website.

ACKNoWLEDGMENTSWe thank B. Hooijbrink for his excellent help with FACS sorting and maintenance of the flow cytometry facility, J. Boes for her excellent work on virus titrations at the Netherlands Vaccine Institute and R. Molenkamp and D. Pajkrt of the Department of Clinical Virology of the Academic Medical Center for helpful discussions. Human cDNA encoding Bcl-xL was kindly provided by S. Korsmeyer (Howard Hughes Medical Institute, Harvard Medical School). Id-3 was a gift from C. Murre (University of California, San Diego). CD40L-L cells were from J. Banchereau (Baylor University). S.A.D. is supported by US National Institutes of Health–National Institute of Allergy and Infectious Diseases grant F32AI063846. M.V.L. is supported by grant OZF-02-004 from the Wilhelmina Research Fund.

AUTHoR CoNTRIBUTIoNSM.J.K., S.A.D. and T.B. performed experiments, analyzed data and wrote the paper. H.S. and T.B. organized the research and wrote the paper. E.Y., A.Q.B. and C.M.M.v.G. performed experiments. M.V.L., G.M.v.B., H.M., A.R. and W.M.J.M.B. contributed valuable reagents and developed assays. M.N.W. performed cotton rat experiments. F.A.S. analyzed data and provided valuable suggestions.

CoMPETING INTERESTS STATEMENTThe authors declare competing financial interests: details accompany the full-text HTML version of the paper at http://www.nature.com/naturemedicine/.

Published online at http://www.nature.com/naturemedicine/. Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/.

1. Banchereau, J., de Paoli, P., Valle, A., Garcia, E. & Rousset, F. Long-term human B cell lines dependent on interleukin-4 and antibody to CD40. Science 251, 70–72 (1991).

2. Kim, C.H. et al. Subspecialization of CXCR5+ T cells: B helper activity is focused in a germinal center-localized subset of CXCR5+ T cells. J. Exp. Med. 193, 1373–1381 (2001).

3. Good, K.L., Bryant, V.L. & Tangye, S.G. Kinetics of human B cell behavior and amplification of proliferative responses following stimulation with IL-21. J. Immunol. 177, 5236–5247 (2006).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

128  volume 16 | number 1 | january 2010 nature medicine

t e c H n i c a L r e P O r t S

4. Ettinger, R. et al. IL-21 induces differentiation of human naive and memory B cells into antibody-secreting plasma cells. J. Immunol. 175, 7867–7879 (2005).

5. Kuo, T.C. et al. Repression of Bcl-6 is required for the formation of human memory B cells in vitro. J. Exp. Med. 204, 819–830 (2007).

6. Shapiro-Shelef, M. et al. Blimp-1 is required for the formation of immunoglobulin secreting plasma cells and pre-plasma memory B cells. Immunity 19, 607–620 (2003).

7. Dent, A.L., Shaffer, A.L., Yu, X., Allman, D. & Staudt, L.M. Control of inflammation, cytokine expression, and germinal center formation by Bcl-6. Science 276, 589–592 (1997).

8. Fukuda, T. et al. Disruption of the Bcl6 gene results in an impaired germinal center formation. J. Exp. Med. 186, 439–448 (1997).

9. Phan, R.T. & Dalla-Favera, R. The BCL6 proto-oncogene suppresses p53 expression in germinal-centre B cells. Nature 432, 635–639 (2004).

10. Shapiro-Shelef, M. & Calame, K. Regulation of plasma-cell development. Nat. Rev. Immunol. 5, 230–242 (2005).

11. Shvarts, A. et al. A senescence rescue screen identifies BCL6 as an inhibitor of anti-proliferative p19(ARF)-p53 signaling. Genes Dev. 16, 681–686 (2002).

12. Scheeren, F.A. et al. STAT5 regulates the self-renewal capacity and differentiation of human memory B cells and controls Bcl-6 expression. Nat. Immunol. 6, 303–313 (2005).

13. Diehl, S.A. et al. STAT3-mediated up-regulation of BLIMP1 is coordinated with BCL6 down-regulation to control human plasma cell differentiation. J. Immunol. 180, 4805–4815 (2008).

14. Liu, Y.J. & Arpin, C. Germinal center development. Immunol. Rev. 156, 111–126 (1997).

15. Niiro, H. & Clark, E.A. Regulation of B cell fate by antigen-receptor signals. Nat. Rev. Immunol. 2, 945–956 (2002).

16. Bryant, V.L. et al. Cytokine-mediated regulation of human B cell differentiation into Ig-secreting cells: predominant role of IL-21 produced by CXCR5+ T follicular helper cells. J. Immunol. 179, 8180–8190 (2007).

17. Hall, C.B. et al. The burden of respiratory syncytial virus infection in young children. N. Engl. J. Med. 360, 588–598 (2009).

18. Johnson, S. et al. A direct comparison of the activities of two humanized respiratory syncytial virus monoclonal antibodies: MEDI-493 and RSHZl9. J. Infect. Dis. 180, 35–40 (1999).

19. Sidwell, R.W. & Barnard, D.L. Respiratory syncytial virus infections: recent prospects for control. Antiviral Res. 71, 379–390 (2006).

20. Domachowske, J.B., Bonville, C.A. & Rosenberg, H.F. Animal models for studying respiratory syncytial virus infection and its long term effects on lung function. Pediatr. Infect. Dis. J. 23, S228–S234 (2004).

21. Peled, J.U. et al. The biochemistry of somatic hypermutation. Annu. Rev. Immunol. 26, 481–511 (2008).

22. Sayegh, C.E., Quong, M.W., Agata, Y. & Murre, C. E-proteins directly regulate expression of activation-induced deaminase in mature B cells. Nat. Immunol. 4, 586–593 (2003).

23. Dong, C., Temann, U.A. & Flavell, R.A. Cutting edge: critical role of inducible costimulator in germinal center reactions. J. Immunol. 166, 3659–3662 (2001).

24. Shinkura, R. et al. Separate domains of AID are required for somatic hypermutation and class-switch recombination. Nat. Immunol. 5, 707–712 (2004).

25. McBride, K.M. et al. Regulation of class switch recombination and somatic mutation by AID phosphorylation. J. Exp. Med. 205, 2585–2594 (2008).

26. Traggiai, E. et al. An efficient method to make human monoclonal antibodies from memory B cells: potent neutralization of SARS coronavirus. Nat. Med. 10, 871–875 (2004).

27. Dykstra, M.L., Longnecker, R. & Pierce, S.K. Epstein-Barr virus coopts lipid rafts to block the signaling and antigen transport functions of the BCR. Immunity 14, 57–67 (2001).

28. Wrammert, J. et al. Rapid cloning of high-affinity human monoclonal antibodies against influenza virus. Nature 453, 667–671 (2008).

29. Zwick, M.B., Gach, J.S. & Burton, D.R. A welcome burst of human antibodies. Nat. Biotechnol. 26, 886–887 (2008).

30. Deinhardt, F., Falk, L.A. & Wolfe, L.G. Transformation of nonhuman primate lymphocytes by Epstein-Barr virus. Cancer Res. 34, 1241–1244 (1974).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nature medicinedoi:10.1038/nm.2071

ONLINE METhODSB cell isolation. We obtained B cells from peripheral blood (Sanquin) by Ficoll separation and CD22 MACS microbeads (Miltenyi Biotech). Subsequently, we sorted these cells for the CD19+CD3−CD27+IgM−IgA− phenotype (IgG memory cells) or the CD19+CD3−CD27+IgG−IgA− phenotype (IgM memory cells) on a FACSAria (Becton Dickinson). We obtained tonsil B cells from routine tonsillectomies performed at the Department of Otolaryngology at the Academic Medical Center. The use of these tissues was approved by the Medical Ethical Committee of the Academic Medical Center and was contin-gent on informed consent.

Cell culture. We maintained B cells (2 × 105 cells ml−1) in Iscove’s modified Dulbecco’s medium (Gibco) containing 8% FBS (HyClone) and penicillin- streptomycin (Roche) supplemented with recombinant mouse IL-21 (25 ng ml−1, R&D systems) and cultured them on γ irradiated (50 Gy) mouse L cell fibroblasts stably expressing CD40L (obtained from J. Banchereau) (CD40L-L cells, 105 cells ml−1). We used recombinant human IL-4 (R&D Systems) at 50 ng ml−1 and recombinant human IL-2 (R&D Systems) at 100 U ml−1. We tested cells routinely by PCR for the presence of mycoplasma and EBV (data not shown).

Retroviral transduction. The Bcl-6 and Id-3 retroviral constructs have been described previously11,31. We cloned cDNA encoding human Bcl-xL into the LZRS retroviral vector that was described previously32. We performed trans-fections and produced amphotropic and ecotropic virus by Phoenix packag-ing cells as described previously11,12 and concurrently transduced memory B cells by retroviruses after activation on CD40L-L cells in the presence of recombinant mouse IL-21 for 36 h as before13, but then we centrifuged cells and virus at 20 °C for 60 min at 360g.

Flow cytometry. We analyzed stained cells on an LSRII (BD) and processed flow cytometry data with FlowJo software (Tree Star). The antibodies are described in the Supplementary Methods.

Reverse transcription PCR. We carried out quantitative RT-PCR with a BioRad iCycler and used the 2−(∆∆CT) method to calculate relative mRNA expression levels normalized to actin. Primers for AICDA (encoding AID) were described previously33.

Enzyme-linked immunosorbent assay. We coated plates with either antibody to human IgG or IgM Fc fragment (Jackson ImmunoResearch Laboratories) at 5 µg ml−1 in PBS for 1 h at 37 °C or overnight at 4 °C and washed them in ELISA wash buffer (PBS, 0.5% Tween-20). We used 4% milk in PBS as a blocking agent before we added serial dilutions of cell culture supernatants and horseradish peroxidase (HRP)-conjugated detection antibodies (dilutions: 1 in 2,500 for HRP-conjugated IgG-specific antibody (Jackson); 1 in 5,000 for HRP-conjugated IgM-specific antibody (Jackson)). We used tetramethyl-benzidine substrate solution (Biosource) for development of the ELISAs.

Immunoblotting. We lysed cells in Triton Lysis Buffer supplemented with HALT protease inhibitor (Roche) and 1 µM Na2VO4. We separated equal amounts of protein by 10% SDS-PAGE, transferred them to nitrocellulose and probed for them using antibody to phospho-ERK1/2 (on Thr202 and Tyr204) (Cell Signaling Technologies) and STAT3-specific antibody (Santa Cruz).

We used enhanced chemiluminescence (Pierce Biotechnology) and X-ray film (Hyperfilm, GE Healthcare Life Sciences) for detection.

Calcium flux. We cultured transduced B cells for 2 d without CD40L-L cells in the presence of IL-21. Subsequently, we loaded the cells with 2 µM Indo-1 AM (Sigma-Aldrich). We measured fluorescence ratios of Indo-1 emission at 405 nm and 485 nm by flow cytometry on a FACSVantage SE (BD).

Respiratory syncytial virus culture and neutralization assay. We seeded 1 × 104 HEp2 cells in flat-bottom 96-well plates (Costar). The next day, we pre-incubated 100 TCID50 of RSV A2 or X virus mixed with B cell culture super-natant for 30 min at 37 °C before adding it in triplicate to HEp2 cells for 1 h at 37 °C. After 2 d, we fixed the cells with 80% acetone and stained them with polyclonal HRP-conjugated antibody to RSV (Biodesign) and 3-amino-9-ethylcarbazole for detection and visualization of RSV plaques by light micro-scopy (plaques were counted).

Cloning of D25. We isolated total RNA with the RNeasy mini kit (Qiagen), generated cDNA, performed PCR and cloned the heavy and light chain varia-ble regions into the pCR2.1 TA cloning vector (Invitrogen). To rule out reverse transcriptase– or DNA polymerase–induced mutations, we performed several independent cloning experiments. To produce recombinant D25 monoclonal antibody, we cloned D25 heavy and light variable regions in frame with human IgG1 and κ constant regions into a pcDNA3.1 (Invitrogen)-based vector and transiently transfected 293T cells (American Type Culture Collection). We purified recombinant D25 from the culture supernatant with Protein A (Sigma-Aldrich).

Cotton rat experiments. We anesthetized 7- to 9-week-old cotton rats (Sigmodon hispidus, Charles River) with isoflurane, bled them and gave them 0.1 ml of purified monoclonal antibody per 100 g body weight by i.m. injec-tion at doses of 2.0 or 0.6 mg per k body weight for palivizumab, 0.6, 0.15 or 0.04 mg per k body weight for D25 and 2.0 mg per k body weight for the control antibody. Twenty-four hours later, we anesthetized the rats and chal-lenged them by intranasal instillation of 1 × 106 TCID50 RSV-X (100 µl). Five days34 later, we killed the rats and collected their lungs. We calculated lung virus titers per gram of lung tissue. We determined lung virus titers by standard TCID50 culture assay; control prophylaxis reached a titer of 4.7 log TCID50 g−1 whereas the lower limit of detection was 2.1 log TCID50 g−1. The Animal Experiment Committee of the Netherlands Vaccine Institute approved all procedures involving cotton rats.

31. Jaleco, A.C. et al. Genetic modification of human B cell development: B cell development is inhibited by the dominant negative helix loop helix factor Id3. Blood 94, 2637–2646 (1999).

32. Heemskerk, M.H. et al. Enrichment of an antigen-specific T cell response by retrovirally transduced human dendritic cells. Cell. Immunol. 195, 10–17 (1999).

33. Smit, L.A. et al. Expression of activation-induced cytidine deaminase is confined to B cell non-Hodgkin′s lymphomas of germinal-center phenotype. Cancer Res. 63, 3894–3898 (2003).

34. Prince, G.A., Prieels, J.P., Slaoui, M. & Porter, D.D. Pulmonary lesions in primary respiratory syncytial virus infection, reinfection, and vaccine-enhanced disease in the cotton rat (Sigmodon hispidus). Lab. Invest. 79, 1385–1392 (1999).

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.