The A-Myb Transcription Factor Is a Marker of Centroblasts In Vivo Josée Golay, Vania Broccoli, Giuseppe Lamorte, Carlo Bifulco, Carlo Parravicini, Arnold Pizzey, N. Shaun B. Thomas, Domenico This information is current as Delia, Paola Ferrauti, Domenico Vitolo and Martino Introna of October 4, 2021. J Immunol 1998; 160:2786-2793; ; http://www.jimmunol.org/content/160/6/2786 Downloaded from References This article cites 47 articles, 18 of which you can access for free at: http://www.jimmunol.org/content/160/6/2786.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 4, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The A-Myb Transcription Factor Is a Marker of Centroblasts In Vivo1

Jose´e Golay,2* Vania Broccoli,* Giuseppe Lamorte,† Carlo Bifulco,‡ Carlo Parravicini,‡ Arnold Pizzey,§ N. Shaun B. Thomas,§ Domenico Delia,¶ Paola Ferrauti,ʈ Domenico Vitolo,ʈ and Martino Introna*

The A-Myb transcription factor is structurally related to the c-myb proto-oncogene and is involved in the control of proliferation and/or differentiation of mature B lymphocytes. We have shown previously by PCR analysis that A-myb is preferentially expressed in CD38؉CD39؊sIgM؊ mature B cells. We demonstrate here, using in situ hybridization, that A-myb expression is restricted to the dark zone of human tonsils and lymph nodes. Furthermore, we show that A-Myb expression is cell cycle regulated both in tonsillar B cells and in Burkitt’s lymphoma cell lines, being detectable only in the S and G2/M phases of the cell cycle and not in ␮ Downloaded from G0/G1 phase. Strong proliferation of resting human B cells induced in vitro by a variety of physiologic signals, including anti- , CD40 ligand, IL-2, IL-4, IL-6, IL-13, IFN-␥, TNF-␣, anti-CD19, and anti-CD20, failed to induce A-myb expression, suggesting that proliferation alone is not sufficient for A-myb expression in the absence of induction of a true centroblast phenotype. Finally, we show that differentiation of germinal center B cells in vitro toward either memory or plasma cells is accompanied by rapid down-regulation of A-myb expression. We conclude that A-myb is a marker of centroblasts generated in vivo. The Journal of Immunology, 1998, 160: 2786–2793. http://www.jimmunol.org/ -Myb belongs to the myb family of transcription factors, of mature tonsillar B lymphocytes, with the phenotype CD38ϩ, which include in mammals the v-myb oncogene, its nor- CD39Ϫ, and mostly sIgMϪ and is therefore presumably localized A mal cellular equivalent c-myb, and the two structurally within GC (19, 20). Furthermore, the study of a panel of B cell related B-myb and A-myb genes (1–4). All members of the Myb lines and fresh B leukemia samples representing B cell differen- family have been shown to be involved in the control of prolifer- tiation from pre-B cells to plasma cells showed that A-Myb is ation and/or differentiation of different hemopoietic cells (3, 4). In expressed only in Burkitt’s lymphoma cells, presumably represent- particular, v-Myb and c-Myb block the differentiation of myeloid, ing the neoplastic counterpart of GC cells, and in a fraction of erythroid, and immature multipotent progenitors (5–8). C-Myb is B-CLL (21). These findings have led to the hypothesis that A-Myb by guest on October 4, 2021 required for the proliferation of hemopoietic cell lines of different plays a role in the control of proliferation and/or differentiation of lineages (9), and c-myb knockout animals die in utero due to a mature B cells (22). Very recently, this hypothesis has gained complete block of fetal hemopoiesis (10). B-Myb has been shown much strength from transgenic experiments. Mice expressing ec- to be required for the proliferation of many cell types including topic A-Myb in a wide range of tissues showed specific deregu- fibroblasts and smooth muscle cells, and its expression, in general, lation of the B cell compartment, with hyperplasia of the spleen correlates strictly with proliferation (11–14). The third member of and lymph nodes and accumulation of a mature polyclonal B cell the family, A-Myb, shows the most restricted pattern of expres- ϩ sion: in the adult mouse, it is expressed in some proliferating neu- population with a CD38 phenotype (23). ronal precursor cells, in spermatocytes, breast ductal epithelium, Germinal centers are central to the differentiation and function and germinal center (GC)3 B lymphocytes (15–18). In human he- of B lymphocytes (24–26). Naive mature B lymphocytes migrate mopoietic cells, A-Myb has been detected only in a subpopulation from the bone marrow to peripheral lymphoid organs. Upon Ag encounter in the T cell areas of spleen and lymph nodes, B cells migrate to the follicles where the activated cells will form GC. GC † *Istituto Ricerche Farmacologiche “Mario Negri,” Fondazione Matarelli, Hospital B cells undergo extremely rapid proliferation (estimated doubling Fatebenefratelli e Oftalmico, and ‡Department of Pathological Anatomy, Hospital L. Sacco, Milan, Italy; §Department of Hematology, University College London Medical time, 6–7 h) and somatic mutation of their Ig gene within the dark School, London, United Kingdom; and ¶Department of Experimental Oncology A, zone of GC, which contains mostly large proliferating B cells or Istituto Nazionale dei Tumori, Milan, and ʈDepartment of Experimental Medicine and Pathology, Immunopathology Section, Universita`di Roma La Sapienza, Rome, Italy centroblasts. Further differentiation of centroblasts to centrocytes Received for publication July 28, 1997. Accepted for publication November 18, 1997. is accompanied by migration to the light zone of GC and arrest of The costs of publication of this article were defrayed in part by the payment of page the proliferation. The light zone contains the follicular dendritic charges. This article must therefore be hereby marked advertisement in accordance cells (FDC) and GC helper T cells, which are thought to be im- with 18 U.S.C. Section 1734 solely to indicate this fact. portant in regulating the B cell differentiation processes taking 1 This work was supported in part by the Italian Association for Cancer Research place in this region of GC: apoptosis of B cells with low affinity for (AIRC) and the Istituto Superiore di Sanita`(Rome, Italy, AIDS project). J.G. is a 1997 Fellow of the Angelo and Angela Valenti Foundation, Milan, Italy. N.S.B.T. is sup- the Ag, leading to affinity maturation, and differentiation to mem- ported by the Kay Kendall Leukaemia Fund. G.L. is supported by the Associazione ory cells or to preplasma cells, which will then migrate to the bone Italiana Leucemia (AIL), Milan, Italy. marrow for full differentiation to plasma cells. The latter processes 2 Address correspondence and reprint requests to Dr. Jose´e Golay, Istituto Ricerche Farmacologiche “Mario Negri,” via Eritrea 62, 20157 Milano, Italy. E-mail address: also involve isotype switching for the production of different Ig [email protected] classes. Thus, the very fine compartmentalization of the GC has 3 Abbreviations used in this paper: GC, germinal center; CD40L, CD40 ligand. allowed a precise understanding of the sequence of events that

Copyright © 1998 by The American Association of Immunologists 0022-1767/98/$02.00 The Journal of Immunology 2787

takes place in GC, through the study of cell markers and morphol- remove unhybridized and nonspecifically bound probe. Autoradiography ogy, expression of specific genes, and the mutation and recombi- was performed with Kodak NT/B2 emulsion. Exposure times were be- nation of Ig genes in normal and genetically modified animals tween 2 and 3 wk. After developing, the sections were mounted in DPX (Merck, Darmstadt, Germany). Adjacent sections were processed for his- (24–26). In turn, knowledge of the localization of new gene prod- tologic staining or immunohistochemistry. ucts within GC can indicate the potential function of such prod- ucts. Previous experiments using purified tonsillar B cell subpopu- Proliferation and differentiation assays in vitro lations had suggested that the A-Myb transcription factor is Recombinant trimeric human CD40 ligand (CD40L) was obtained from expressed by GC B cells, but its possible subcompartmentalization Immunex (Seattle, WA) and used at 1 ␮g/ml. rIL-2 (Eurocetus, Amster- within this structure had not been investigated (20). Given the dam, Holland) was used at 200 U/ml, rIL-4 (Immunex, Seattle, WA) at 50 important role demonstrated for this transcription factor in B cell ng/ml, rIL-6 (Serono, Roma, Italy) at 5 ng/ml, rIL-10 (Shering-Plough, Milano, Italy) at 20 ng/ml, and rIL-13 (Sanofi, Montpellier, France) at 40 differentiation, our goal was to determine the pattern of A-Myb ng/ml; IFN-␥ was from Roussel-UCLAF (Paris, France) at 200 U/ml and expression in GC in vivo and in models of B cell proliferation and TNF-␣ (BASF Pharma Knoll, Ludwigshafen, Germany) at 10 ng/ml. Ј ␮ differentiation in vitro. We show that A-Myb is specifically ex- F(ab )2 Goat anti-human -chain (Cappel, Organon Teknika Corp., West ␮ pressed in proliferating centroblasts generated in vivo and is not Chester, PA) was used at 25 g/ml. The purified anti-CD19 Ab, HD37, was a kind gift of Dr. G. Molden- induced in B cells activated in vitro. Furthermore, it is down-reg- hauer (Deutsches Krebsforschungszentrum, Heidelberg, Germany) and ulated during differentiation of centroblasts to more mature cells. was used at 1 ␮g/ml for stimulation. The Ab against CD20, 1F5, was a kind Thus, among known B cell-restricted transcription factors, A-Myb gift of Dr. E. A. Clark (University of Washington, Seattle, WA) and used 6 shows the most restricted pattern of expression. at 2 ␮g/ml. Resting or GC B cells were cultured at 1 to 2 ϫ 10 /ml with the indicated stimuli. Double immunofluorescence staining was performed with CD20-FITC and CD38-phycoerythrin Abs (Becton Dickinson, San Downloaded from Materials and Methods Jose´, CA). Tissues and cells Cell sorting Tonsils were obtained from routine tonsillectomies. Axillary and mesen- teric lymph nodes were obtained from specimens of patients with unrelated 7 ϫ 105 cells/ml were stained with 20 ␮g/ml Hoechst 33342 and sorted pathologies. All tissues used for in situ hybridization were formalin fixed with a FACSVantage instrument (Becton Dickinson) equipped witha5W and embedded in paraffin. Total tonsillar B lymphocytes were purified as argon-ion laser emitting 20 mW of UV light (352 nm). Cells were lysed in described previously (20). Briefly, tonsils were minced to a single-cell SDS sample buffer containing protease and phosphatase inhibitors. In some http://www.jimmunol.org/ suspension and centrifuged over a Ficoll-Hypaque gradient (Seromed, Ber- experiments, cells were fixed in 80% EtOH and stained with propidium lin, Germany). T cells were removed by rosetting with aminoethylisothio- iodide before sorting. Both methods gave the same results. A fraction of uronium-treated SRBC and centrifuging through a Ficoll-Hypaque gradi- sorted cells were reanalyzed and found to be at least 95% pure with respect ent. Total B cells at the interface were further separated into different to cell cycle phase. fractions on a discontinuous gradient made of 80, 57, 50, 40, and 30% Percoll (Pharmacia, Uppsala, Sweden). The 80 to 57% Percoll interface Northern and Western blots consisted mostly of resting small B lymphocytes. The 40 to 50% interface RNA was extracted by standard guanidium isothiocyanate and cesium was highly enriched (at least 80%) with GC B cells strongly positive for ␮ ϩ ϩ chloride gradient purification. Total RNA (18–20 g) was run in 1% form- both CD38 and CD20 . B cell purity of both the GC and resting B cell aldehyde-agarose gels and blotted onto GeneScreen Plus membrane (New fractions was 96 to 98%, as judged by staining with an anti-CD40 or England Nuclear, Boston, MA), and hybridization was performed accord- by guest on October 4, 2021 anti-CD19 Ab. ing to the manufacturer’s instructions. The A-Myb-specific antiserum has Daudi and Ramos cell lines were from the American Type Culture been described previously (29) and was used a a 1/1000 dilution. The Collection (Rockville, MD). All cells were grown in RPMI 1640 (Se- anti-cdk2, anti-p27, and anti-actin antisera were from Santa Cruz Biotech- romed), supplemented with 10% FCS (HyClone Laboratories, Logan, UT), nology (Santa Cruz, CA). To control for A-Myb antiserum specificity, the ␮ glutamine (Life Technologies, Paisley, Scotland), and 50 g/ml gentami- A-Myb-negative myeloma cell line IM-9 was transfected with the complete ϫ 5 cin (Life Technologies). Cells were resuspended at 3 10 /ml 24 h before A-myb cDNA cloned in the episomial pCEP4 vector (Invitrogen Leeks, The sorting. Netherlands) or with the empty vector, using electroporation (270 mV, 960 ␮ Plasmid constructions and probes F in a Bio-Rad Gene Pulser; Bio-Rad, Hercules, CA). Transfected cells were selected by culture in 500 ␮g/ml hygromycin B (Boehringer Mann- For A-myb-specific in situ hybridization, a fragment of A-myb spanning heim, Mannheim, Germany). Pools of transfected hygromycin resistant nucleotide (nt) 1190–1550 (2) was cloned in the pGem3 vector (Promega cells were obtained after 3 to 4 wk of culture. Then, 1 to 2 ϫ 106 trans- Corp., Madison, WI). Briefly the XbaI-HindIII 360-bp fragment from plas- fected or untransfected B cells were lysed in SDS loading buffer and run in mid A-myb-PCR was subcloned in the same sites of pGem3. For B-myb in a 10% SDS-polyacrylamide gel. The gels were electroblotted onto nitro- situ hybridization, a 390-bp SacI-PstI fragment (nt 742–1133) was sub- cellulose filters (Schleicher and Schuell, Dassel, Germany) for5hat60V, cloned in the same sites of the pGem4 plasmid. The probes for in situ according to standard procedures. The blots were incubated with the Abs hybridization were labeled with[35S]UTP (Amersham, Little Chalfont, diluted in PBS containing 5% nonfat milk powder and washed in the same Buckinghamshire, U.K.) using the Riboprobe kit (Promega). The A-myb, solution containing 0.5% Nonidet P-40. Detection was performed using the B-myb, and cyclin A probes used for Northern blot hybridization have been enhanced chemiluminescence system (ECL; Amersham). described elsewhere (19, 27) and were labeled with[32P]dCTP (Amersham) with the Megaprime DNA labeling system (Amersham). Results In situ hybridization A-myb expression is restricted to the dark zone of GC In situ hybridization was performed essentially as described by Wilkinson Previous data both in the mouse and in human B cells had indi- (28) with minor modifications. Briefly, myb sense and antisense probes cated that A-Myb is expressed by GC B cells, but the localization were generated by transcription with the T7 and SP6 polymerases (Ribo- of A-myb gene expression in the different GC compartments had probe kit, Promega) in the presence of[35S]CTP (Amersham). The template not been investigated (15, 20). We decided, therefore, to use in situ was then degraded in RNase-free DNase (Pharmacia), and the labeled hybridization to answer this question. A 360-bp fragment of hu- probe was purified through a Sephadex G-50 (Pharmacia) column followed by NH4 acetate precipitation. The probes were dissolved at a working man A-myb cDNA spanning a region not conserved with other myb concentration of 106 cpm/␮l in hybridization mix. Before hybridization, genes was subcloned in the pGem3 vector for generation of anti- the sections were treated with 3 ␮g/ml proteinase K, fixed in 4% parafor- sense and control sense RNA probes using the SP6 and T7 poly- maldehyde in PBS, and finally washed in a solution of 0.25% acetic an- merases, respectively. As a control, a similarly unique 390-bp frag- hydride in 0.2 M triethanolamine. Then, 30 ␮l of the appropriate probe was added to each slide, and the hybridization was conducted overnight at ment of human B-myb cDNA was subcloned in the pGem4 vector. 55°C. The slides were then washed under stringent conditions (65°C, 2ϫ B-myb expression correlates with proliferation in many cell types, SSC, 50% formamide) and treated with 100 ␮g/ml RNase A (Sigma) to including tonsillar B cells, and was therefore expected to stain the 2788 A-myb EXPRESSION IN CENTROBLASTS

in Figure 2, the pattern of expression was similar in tonsil sections, with a clear correspondence between A-myb (Fig. 2C), B-myb (2D), and Ki67 (2B) expression within the GC. Of note, however, was the fact that the layer of proliferating epithelial cells (E) clearly marked by Ki67 also strongly hybridized with the B-myb probe (marked with a double arrow in Fig. 2D) but not with the A-myb (Fig. 2C). This demonstrates the specificity of A-myb ex- pression for the proliferating centroblasts of the dark zone, whereas B-myb is expressed in all proliferating cell populations present, including centroblasts. Also, in spleen sections A-myb was specifically expressed in the dark zone of GC (data not shown). A-myb is cell cycle regulated The association of A-myb expression with the dark zone of GC and therefore with proliferating centroblasts suggested that A-Myb may play a role in proliferation and may be regulated in the cell cycle in these cells (24, 25). To verify this hypothesis, we stained

purified tonsillar B cells with DNA dyes and sorted the B cell Downloaded from populations according to the cell cycle phase. Equivalent numbers of sorted cells were then analyzed for A-myb expression in West- ern blots, as verified by Ponceau red staining of the blot. It was demonstrated that the anti-A-Myb antiserum was specific for A- Myb by Western analysis of the IM-9 cell line transfected with

either the complete A-myb cDNA in an expression vector or with http://www.jimmunol.org/ the empty vector. The antiserum specifically recognized a 90-kDa band in A-myb-transfected and not in vector-transfected cells (Fig. 3A). In sorted tonsillar B cell populations, the 90-kDa A-Myb

protein could be detected only in the S and G2/M phases of the cell cycle but not in G0/G1. To verify that sorting according to cell cycle phase had been successful, the same blot was stained with an

anti-p27 Ab. As expected, p27 is strongly expressed in the G0/G1 cell population and is efficiently down-regulated in S/G2/M phase

cells (Fig. 3A) (32). Furthermore, the blot was stripped and re- by guest on October 4, 2021 probed with anti-Actin Ab, showing that each of the lanes con- tained equivalent amounts of protein (Fig. 3A). A-Myb is also highly expressed in most Burkitt’s lymphoma cell lines, which are probably the transformed equivalent of normal centroblasts (20, 21, 33). We therefore verified whether A-Myb FIGURE 1. In situ hybidization of the lymph node. Serial sections of was cell cycle regulated also in A-Myb-positive BL lines. The 35 human lymph nodes were hybridized with A-myb (C) and B-myb (B) S- exponentially growing lines Ramos and Daudi were sorted on the labeled probes. Adjacent sections were stained with the Ki67 Ab (A) (31). FACS according to cell cycle phase. Again, A-Myb protein was FM, follicular mantle; LZ, light zone; DZ, dark zone. found to be strongly expressed in the S and G2/M phases of the cell cycle but not in G1. That all lanes contained protein and that cell proliferating B cells in the dark zone of GC (3, 4). 35S-labeled separation was adequate was verified by Ponceau red staining of sense and antisense RNA probes were generated with SP6 and T7 the blot (data not shown) and by testing the same blots for either

RNA polymerases and used to hybridize serial sections of forma- Cdk2 (which is more strongly expressed in S phase than in G1 lin-fixed human lymphoid tissues, including tonsils, lymph nodes, phase) or p27 expression (Fig. 3A). Furthermore, the purity of the and spleen. Adjacent sections were also stained with hematoxylin different cell cycle phases was regularly determined on the FACS or with the Ki67 Ab to identify the dark zone of GC (30, 31). and found to be Ͼ82% for all phases. One representative example Representative results obtained with a lymph node and a tonsil are of such cell cycle analysis on the Ramos cell line, before and after shown in Figures 1 and 2, respectively. In the lymph node section sorting, is shown in Figure 3B. We conclude that A-Myb expres- of Figure 1, the compartmentalization of the GC is most evident, sion is restricted to the S and G2/M phases of the cell cycle in both with clear staining of dark zone B cells with the Ki67 Ag (Fig. 1A). normal and BL cell lines. As expected, B-myb positivity correlated with expression of the Ki67 Ag and therefore with proliferating cells in the dark zone of A-myb expression is not induced following in vitro-stimulated GC (Fig. 1B). The follicular mantle, T cell area, and light zone of proliferation of resting tonsillar B cells GC were mostly negative for B-myb expression. The pattern of Resting tonsillar B cells can be induced to proliferate in vitro by a A-myb expression was very similar to that of B-myb and correlated number of agents, including crude polyclonal B cell activators closely with Ki67 expression. Small differences in A-myb, B-myb, such as SAC (formalin-killed Staphylococcus aureus Cowan I and Ki67 positivity probably reflect the fact that different probes strain bacteria), drugs such as PMA and Ca2ϩ ionophores, Abs were used on adjacent sections. Sections from the same tissues against cell surface molecules, and presumably more physiologic were also hybridized with the sense probes as negative controls signals such as different combinations of anti-␮, trimeric CD40L, and did not show any specific signal (data not shown). As shown and a variety of cytokines (34, 35). We had previously observed The Journal of Immunology 2789 Downloaded from http://www.jimmunol.org/

FIGURE 2. In situ hybridization of the tonsil. Serial tonsil sections were hybridized with A-myb (C)orB-myb (D) probes or stained with hematoxylin- eosin (A) or with the Ki67 Ab (B). GC, germinal center; T, T cell area; E, epithelial layer. by guest on October 4, 2021 that the induction of B cell proliferation in vitro by nonphysiologic sponding roughly to the levels of thymidine uptake or S/G2/M ϩ signals (SAC, PMA calcium ionophore, etc.) was not accom- phase induction. Indeed the percentage of cells in the S/G2/M panied by the induction of A-Myb (20). This was somewhat puz- phases of the the cell cycle obtained with the most effective stim- zling in view of the finding described above that A-Myb is ex- uli, such as anti-␮ ϩ CD40L ϩ IL-2 ϩ IL-10, corresponded ap- pressed by the highly proliferating centroblasts and is cell cycle proximately to that of purified GC B cells (Table I). However, the regulated in these cells. We therefore set out to investigate whether levels of A-myb expression were clearly very different in the two more physiologic means of inducing proliferation of resting B populations (Fig. 4, lane 1 compared wiht lanes 3–15). Most stim- lymphocytes in vitro could induce A-myb expression. As shown in ulation regimens were also tested for A-Myb protein expression in Figure 4 and Table I, we used a wide combination of signals in- Western blots. Again, A-Myb could be detected in purified tonsil- cluding anti-␮ or anti-CD20 Abs with trimeric CD40L and/or dif- lar GC B cells but not following stimulation of resting B cells (data ferent cytokines, i.e., IL-2, IL-4, IL-6, IL-10, IL-13, TNF-␣, and not shown). IFN-␥, or an anti-CD19 Ab. The cytokines were chosen among We conclude that A-Myb is expressed in centroblasts generated those known to induce either growth or differentiation of B cells or in vivo but not in tonsillar resting B cells induced to proliferate in both. TNF-␣ was chosen because it is known to be required for GC vitro. The latter, however, do not have a full centroblast pheno- formation in vivo. IFN-␣ as well as anti-CD19 were chosen be- type, in agreement with several previous studies (36). Although cause they have been reported to induce a partial GC phenotype to IFN-␥ did lead to doubling of CD38 levels as reported (36), other resting B cells in vitro (36). Proliferation was measured by thy- markers of centroblasts were not obtained during our culture con- midine incorporation and in some cases cell cycle analysis. All ditions (in particular, all activated B cell populations showed high combinations used stimulated strong proliferation of resting B CD39 expression, unlike GC B cells, which are CD39Ϫ; data not cells, the combination of anti-CD20 and IFN-␥ being the only shown). relatively weak stimulus (Table I). RNA was extracted from rest- ing B cells at the beginning of culture and 3 days after stimulation A-myb is rapidly down-regulated during in vitro-induced of proliferation and analyzed for A-myb expression in Northern differentiation of GC B cells blots. As shown in Figure 4, whereas GC B cells express high CD38ϩCD20ϩ GC B cells can be induced to proliferate and dif- levels of A-myb RNA, resting B cells induced to proliferate did not ferentiate in vitro toward either memory or plasma cells, a process express detectable levels of A-myb. To verify that other cell cycle- that can be followed by a down-regulation in CD38 or CD20, regulated genes characteristic of the S/G2/M phases of the cell respectively (37–39). Given the restricted expression of A-myb in cycle could be induced in the same conditions, the blots were centroblasts in vivo, we investigated whether differentiation of rehybridized with a cyclin A cDNA probe and, in some cases, also these cells in vitro along both of the differentiation pathways was with B-myb. As shown in Figure 4, both B-myb and cyclin A were accompanied by down-regulation of A-myb expression. Purified efficiently induced by all of the stimuli employed to levels corre- GC B cells were cultured in the presence of CD40L, IL-2, and 2790 A-myb EXPRESSION IN CENTROBLASTS

FIGURE 3. Cell cycle analysis of A-Myb ex- pression. A, To verify the specificity of the A- Myb antiserum, lysates from the IM-9 myeloma cells transfected with an A-Myb-expressing or empty vector were analyzed in Western blots. Tonsillar B cells or the exponentially growing Daudi and Ramos BL lines were stained with Hoechst 33342 and sorted for cell cycle phase. Equivalent numbers of cells (1–1.5 ϫ 105) were loaded in each lane of a 10% SDS-polyacryl- amide gel. Blotted proteins were analyzed for A- Downloaded from Myb (90 kDa), Cdk2 (34 kDa), p27 (27 kDa), and Actin (43 kDa) expression by Western blot- ting. The results are representative of at least two independent experiments for each cell line. B, Both the Hoechst-stained unfractionated cells (starting population) and the fractions sorted ac- cording to cell cycle phase (sorted fractions) http://www.jimmunol.org/ were analyzed on the FACS. One representative result obtained with the Ramos cell line is shown. The % purity of each fraction is indicated. by guest on October 4, 2021

IL-10for 3 days, washed, and replated in medium containing IL-2 led to partial differentiation of the GC B cells toward memory and IL-10 in either the presence or absence of CD40L for an ad- (CD20ϩCD38Ϫ) (MC) or plasma-like cells (CD38ϩCD20Ϫ) (PC), ditional 4 days. As reported by others (37), these culture conditions respectively (Fig. 5A). Plasma-like cells increased from 1 to 27%

Table I. Proliferative response of resting B cells to different stimuli

TdR (SD) ϫ 3 Stimulus ( 10 cpm) % S/G2/M Expt. 1 GC B cells ND 24.8 Medium 0.7 (0.3) 4.0 Anti-␮ ϩ CD40L 26.7 (3.9) 22.7 Anti-␮ ϩ CD40L ϩ IL-10 26.9 (0.2) 21.6 Anti-␮ ϩ CD40L ϩ IL-10 ϩ IL-2 24.7 (1.2) 23.6 Anti-␮ ϩ IFN-␥ 17.4 (0.5) 21.6

Expt. 2 Medium 0.2 (0.1) ND Anti-␮ ϩ CD40L 65.9 (2.3) ND Anti-␮ ϩ CD40L ϩ IFN-␥ 54.4 (0.8) ND Anti-␮ ϩ CD40L ϩ TNF-␣ 66.0 (5.7) ND Anti-␮ ϩ CD40L ϩ IL-13 73.6 (3.9) ND CD20 ϩ IFN-␥ 24.5 (2.6) ND CD20 ϩ CD40L ϩ IFN-␥ 67.5 (2.6) ND FIGURE 4. A-myb expression in GC B cells and resting B cells induced to proliferate in vitro. Dense resting B cells were induced to proliferate Expt. 3 Anti-␮ ϩ CD40L 54.8 (12.9) ND with the indicated stimuli. RNA was extracted from the GC-enriched frac- Anti-␮ ϩ CD40L ϩ IL-4 54.9 (4.0) ND tion (lane 1) or from the resting B cells at the beginning of culture (lane Anti-␮ ϩ CD40L ϩ IL-6 78.2 (4.6) ND ␮ ϩ ϩ 2) or after 66 h (lanes 3–15) and analyzed by Northern blotting. The same Anti- CD40L IL-2 70.8 (7.1) ND ␮ ϩ ϩ blots were hybridized with A-myb, B-myb, cyclin A, and ␤-actin. Anti- CD40L CD19 66.2 (2.2) ND The Journal of Immunology 2791

centroblasts is rapidly down-regulated both in vitro and in vivo during differentiation to centrocytes and then to memory or plasma cells.

Discussion In this report, we demonstrate by in situ hybridization of human tonsils and lymph nodes that the A-Myb transcription factor is expressed during a narrow window of B cell differentiation taking place in the dark zone of GC. Furthermore, in both normal cen- troblasts and Burkitt’s lymphoma cell lines, which may represent the transformed equivalent of centroblasts, A-Myb is expressed

only during the S and G2/M phases of the cell cycle and not in G0/G1. Perhaps surprisingly, however, A-Myb is not up-regulated during in vitro stimulation of purified resting tonsillar B cells in- duced to proliferate with a variety of physiologic stimuli such as anti-␮, CD40L, IL-2, IL-10, IL-4, IL-6, IL-13, TNF-␣, IFN-␥, anti-CD20, and anti-CD19 Abs in different combinations. Thus,

the simple induction of proliferation in vitro is not sufficient for Downloaded from A-Myb expression but may require the acquisition of a true cen- troblast phenotype. In vitro differentiation of GC B cells with CD40L, IL-2, and IL-10, however, led to the rapid disappearance of A-myb mRNA levels, in accordance with in situ hybridization data showing that centrocytes are negative for A-myb expression.

We conclude that A-Myb is a transcription factor that is specifi- http://www.jimmunol.org/ cally induced in centroblasts during the in vivo GC reaction. We had previously investigated A-myb expression in different subpopulations of human tonsillar B cells using sorting by FACS and PCR analysis (20). These data had already suggested that A- Myb is expressed in GC B cells. Furthermore, studies of A-myb expression in mice by in situ hybridization had confirmed a GC localization for this gene (15). In neither case, however, had it been possible to analyze the localization of A-myb within the different by guest on October 4, 2021 FIGURE 5. Differentiation of GC B cells in vitro. The GC-enriched GC subcompartments. In this report, we have demonstrated that fraction of tonsillar B cells was cultured with the indicated stimuli for a A-myb is specifically expressed in the dark zone of GC and there- total of 7 days. At the indicated times, a fraction of cells was either double fore in centroblasts, which show an extremely high rate of prolif- stained for CD20 and CD38 (A) or used for RNA extraction and Northern eration (with a cell cycle of about 7 h) and undergo somatic mu- blot analysis for A-myb, cyclin A, and ␤-actin expression (B). tations of their Ig (24, 25, 40). Further differentiation of centroblasts to centrocytes, which move to the light zone of GC, leads to the elimination of cells with low affinity for the Ag after 7 days of culture and memory cells from 20 to 36% (Fig. 5A). through apoptosis, resulting in affinity maturation (24, 25, 41). In addition, an early and transient peak of proliferation was ob- Also, the processes of isotype switching and further differentiation served at 24 h, as determined by thymidine uptake and cell count to memory and plasma cells is thought to take place in the light (data not shown). We then investigated the time course of A-myb zone of GC, which in addition to centrocytes contains FDC and expression during the same culture period. As shown in Figure 5B, some T cells (24, 41). Thus expression of A-Myb in centroblasts A-myb expression was rapidly down-regulated, having decreased suggests that this transcription factor may be associated with the at least threefold after 24 h and disappearing completely by 66 h. process of somatic mutation or with the particularly high rate of Differentiation toward either memory or plasma cells did not lead proliferation of these cells. to later induction of A-myb. Expression of cyclin A, on the other A role for A-Myb in centroblast proliferation is also supported hand, reflected the proliferative activity of the cells during the in by its regulation during the cell cycle. Also, in bovine smooth vitro culture (Fig. 5B). We conclude that A-myb expression in muscle cells, A-myb has been shown to be induced during the G1/S phase transition (42). Furthermore, these data are reminiscent of the specific A-myb expression only in proliferating populations of cells in the mouse (immature spermatocytes and neuronal and breast ductal epithelium) (15, 18). However, simple proliferation of mature human B cells, induced either by polyclonal B cell ac- tivators (20) or by more physiologic signal as reported here, is not sufficient to induce A-myb expression, even though the percentage

of S/G2/M phase cells obtained after 3 days of culture with many combinations of stimuli was similar to that of purified GC B cells that expressed high levels of A-myb. It is known, however, and has been verified in our culture conditions, that in vitro stimulation of FIGURE 6. Pattern of expression of B cell-specific transcription factors resting B cells does not to lead to the formation of true centroblast- during differentiation. The stages of B cell differentiation during which the like cells (36). Anti-␮ together with IFNs induce CD38 and CD95 indicated transcription factors are expressed are represented schematically. but not carboxypeptidase-M, a marker of in vivo centroblasts; nor 2792 A-myb EXPRESSION IN CENTROBLASTS do they down-regulate CD44, which is normally absent from cen- References troblasts (36). We found that whereas anti-␮ ϩ IFN-␥ was indeed 1. Klempnauer, K. H., G. Ramsay, J. M. Bishop, M. G. Moscovici, C. Moscovici, the most effective stimulus to induce CD38 on resting B cells, none J. P. McGrath, and A. D. Levinson. 1983. The product of the retroviral trans- of the culture conditions used led to a down-regulation of CD39 forming gene v-myb is a truncated version of the protein encoded by the cellular (data not shown). We suggest, therefore, that A-Myb is a marker oncogene c-myb. Cell 33:345. 2. Nomura, N., M. Takahashi, M. Matsui, S. Ishii, T. Date, S. Sasamoto, and of proliferating centroblasts produced in vivo and that culture con- R. Ishizaki. 1988. Isolation of human cDNA clones of myb-related genes, A-myb ditions that do not induce a true centroblast phenotype cannot up- and B-myb. Nucleic Acids Res. 16:11075. 3. Introna, M., M. Luchetti, M. Castellano, M. Arsura, and J. Golay. 1994. The myb regulate A-Myb expression. Similarly, in vitro differentiation of oncogene family of transcription factors: potent regulators of hematopoietic cell GC B cells led to a rapid disappearance of A-myb expression, in proliferation and differentiation. Semin. Cancer Biol. 5:113. accordance with the in situ data showing that centrocytes do not 4. Golay, J., L. Basilico, L. Loffarelli, S. Songia, V. Broccoli, and M. Introna. 1996. Regulation of haematopoietic cell proliferation and differentiation by the myb express this gene. Thus, A-Myb is the B cell-restricted transcrip- oncogene family of transcription factors. Int. J. Clin. Lab. Res. 26:24. tion factor with the narrowest pattern of expression during B cell 5. Beug, H., A. Leutz, P. Kahan, and T. Graf. 1984. Ts mutants of E26 leukemia differentiation. As shown in Figure 6, Bcl-6 (43) is most similar to virus allow transformed myeloblasts, but not erythroblasts or fibroblasts, to dif- ferentiate at the nonpermissive temperature. Cell 39:579. A-Myb, but is expressed in both centroblasts and centrocytes. 6. Selvakumaran, M., D. A. Liebermann, and B. Hoffman Liebermann. 1992. De- Blimp-1 (44) is present only during the latest stages of B cell regulated c-myb disrupts interleukin-6- or leukemia inhibitory factor-induced my- differentiation, whereas BSAP (pax-5) (45) and the coactivator eloid differentiation prior to c-myc: role in leukemogenesis. Mol. Cell Biol. 12:2 493. OCA-B (OBF-1/Bob-1) (46) show a more widespread expression 7. McClinton, D., J. Stafford, L. Brents, T. P. Bender, and W. M. Kuehl. 1990. pattern in B cells. Differentiation of mouse erythroleukemia cells is blocked by late up-regulation of ac-myb transgene. Mol. Cell Biol. 10:705. Downloaded from It is worth noting that A-Myb is cell cycle regulated in both 8. Frampton, J., K. McNagny, M. Sieweke, A. Philip, G. Smith, and T. Graf. 1995. Burkitt’s lymphoma lines and normal B cells. Thus, the transfor- V-myb DNA binding is required to block thrombocytic differentiation of myb- mation event in Burkitt’s lymphoma has not apparently altered this ets-transformed multipotent haematopoietic progenitors. EMBO J. 14:2866. 9. Gewirtz, A. M., and B. Calabretta. 1988. A c-myb antisense oligodeoxynucleotide regulation of A-myb expression, despite the localization of the A- inhibits normal human hematopoiesis in vitro. Science 242:1303. myb genome relatively close to the translocated c-myc gene (3). 10. Mucenski, M. L., K. McLain, A. B. Kier, S. H. Swerdlow, C. M. Schreiner, It is interesting to note that all members of the Myb family in T. A. Miller, D. W. Pietryga, W. J. J. Scott, and S. S. Potter. 1991. A functional c-myb gene is required for normal murine fetal hepatic hematopoiesis. Cell. 65: mammals have been shown to play a direct role in the control of 677. http://www.jimmunol.org/ either proliferation or differentiation or both. C-myb knockout 11. Arsura, M., M. Introna, F. Passerini, A. Mantovani, and J. Golay. 1992. B-myb mice die in utero due to the complete loss of fetal hemopoiesis antisense oligonucleotides inhibit proliferation of human hematopoietic cell lines. Blood 79:2708. (10). C-myb is cell cycle regulated in mature hemopoietic cells and 12. Reiss, K., S. Travali, B. Calabretta, and R. Baserga. 1991. Growth regulated is required for their proliferation (9, 47). Furthermore, c-myb is a expression of B-myb in fibroblasts and hematopoietic cells. J. Cell. Physiol. crucial regulator of hemopoietic cell differentiation (6, 7). B-myb is 148:338. 13. Lin, D., M. Fiscella, P. M. O’Connor, J. Jackman, M. Chen, L. L. Luo, A. Sala, induced in the S/G2/M phases of the cell cycle and is required for S. Travali, S. Appella, and W. E. Mercer. 1994. Constitutive expression of B-myb the proliferation of many cell types (11, 13). More recent evidence bypass p53-induced Waf1/Cip1-mediated G1 arrest. Proc. Natl. Acad. Sci. USA 91:10079. also suggests a role for B-myb in differentiation. Finally, A-myb 14. Arsura, M., M. M. Luchetti, E. Erba, J. Golay, A. Rambaldi, and M. Introna. by guest on October 4, 2021 expression is restricted to a very few cells and during a restricted 1994. Dissociation between p93B-myb and p75c-myb expression during the pro- stage of their differentiation: it is expressed in immature prolifer- liferation and differentiation of human myeloid cell lines. Blood 83:1778. 15. Trauth, K., B. Mutschler, N. A. Jenkins, D. J. Gilbert, N. G. Copeland, and ating spermatocytes, mammary gland ductal epithelial cells, and K. H. Klempnauer. 1994. Mouse A-myb encodes a transactivator and is expressed some neuronal cells (15, 18). It has been demonstrated recently, in mitotically active cells of the developing CNS, adult testis and B-lymphocytes. using A-myb knockout and transgenic animals, that A-myb plays an EMBO J. 13:5994. 16. Mettus, R. V., J. Litvin, A. Wali, A. Toscani, K. Latham, K. Hatton, and important role in these restricted cell types (18, 23). The former E. P. Reddy. 1994. Murine A-myb: evidence for differential splicing and tissue- show defects in spermatocytes and breast duct development (18). specific expression. Oncogene 9:3077. Transgenic animals expressing ectopic A-myb in most tissues show 17. Latham, K. E., J. Litvin, J. M. Orth, B. Patel, R. Mettus, E. P. Reddy. 1996. Temporal pattern of A-myb and B-myb gene expression during testis develop- a restricted effect on B cells. These animals show hyperplasia of ment. Oncogene 13:1161. the follicles of peripheral lymphoid organs with an accumulation 18. Toscani, A., R. V. Mettus, R. Coupland, H. Simpkins, J. Litvin, J. Orth, ϩ K. S. Hatton, and E. P. Reddy. 1997. Arrest of spermatogenesis and defective of CD38 B cells in the follicles (23). These data demonstrate that breast development in mice lacking A-myb. Nature 386:713. A-Myb regulates the proliferation and/or differentiation of GC B 19. Golay, J., A. Capucci, M. Arsura, M. Castellano, V. Rizzo, and M. Introna. 1991. cells. Thus, the present report further defines the role of A-Myb as Expression of c-myb and B-myb, but not A-myb, correlates with proliferation in human hematopoietic cells. Blood 77:149. being specific for centroblasts. 20. Golay, J., E. Erba, S. Bernasconi, G. Peri, and M. Introna. 1994. The A-myb gene A further understanding of A-Myb function in centroblasts is preferentially expressed in tonsillar CD38ϩ, CD39Ϫ, sIgMϪ B lymphocytes will require the definition of its molecular targets in the context and Burkitt’s lymphoma cell lines. J. Immunol. 153:543. 21. Golay, J., M. Luppi, S. Songia, C. Palvarini, L. Lombardi, A. Aiello, D. Delia, of B lymphocytes. Recently, we and others have shown that the K. Lam, D. H. Crawford, A. Biondi, T. Barbui, A. Rambaldi, and M. Introna. c-myc promoter is up-regulated by A-Myb (48). Furthermore, 1996. Expression of A-myb, but not c-myb and B-myb, is restricted to Burkitt’s ϩ we have recently shown that A-Myb may be active as a tran- lymphoma, sIg -B acute lymphoblastic leukemia and a subset of chronic lym- phocytic leukemias. Blood 87:1900. scriptional factor in B cells only and not in other hemopoietic 22. Golay, J., V. Facchinetti, G.-G. Ying, and M. Introna. 1997. The A-myb tran- cells, probably through interaction with a B cell-specific cofac- scription factor in neoplastic and normal B cells. Leuk. Lymphoma 26:271. tor (48). Thus, the definiton of the target genes, cofactors, and 23. DeRocco, S. E., R. Iozzo, X. P. Ma, R. Schwarting, D. Peterson, and B. Calabretta. 1997. Ectopic expression of A-myb in transgenic mice causes fol- specific signals that induce A-Myb in vivo should prove crucial licular hyperplasia and enhanced B lymphocyte proliferation. Proc. Natl. Acad. elements for the understanding of A-Myb function during the Sci. USA 94:3240. GC reaction in response to Ags and in B cell pathologies such 24. MacLennan, I. C. M. 1994. Germinal centers. Annu. Rev. Immunol. 12:117. 25. Kelsoe, G. 1996. Life and death in germinal centers (redux). Immunity 4:107. as Burkitt’s lymphoma. 26. Caligaris-Cappio, F., and M. Ferrarini. 1996. B cells and their fate in health and disease. Immunol. Today 17:206. 27. Songia, S., A. Mortellaro, S. Taverna, C. Fornasiero, E. Scheiber, E. Erba, F. Colotta, A. Mantovani, A. Isetta, and J. Golay. 1997. Characterization ofthe Acknowledgments new immunosuppressive drug undecylprodigiosin in human lymphocytes: reti- noblastoma protein, cyclin-dependent kinase 2, and cyclin-dependent kinase 4 as We thank Felice and Vincenzo de Ceglie for the graphics. molecular targets. J. Immunol. 158:3987. The Journal of Immunology 2793

28. Wilkinson, D. G. 1992. The theory and practice of in situ hybridization. In In Situ 39. Choe, J., H.-S. Kim, X. Zhang, R. J. Armitage, and Y. S. Choi. 1996. Cellular and Hybridization: A Practical Appoach. D. Rickwood and B. D. Hames, eds. IRL molecular factors that regulate the differentiation and apoptosis of germinal cen- press, Oxford, pp. 1–13. ter B cells. J. Immunol. 157:1006. 29. Golay, J., L. Loffarelli, M. Luppi, M. Castellano, and M. Introna. 1994. The 40. Ku¨ppers, R., M. Zhao, M.-L. Hansman, and K. Rajewsky. 1993. Tracing devel- human A-myb protein is a strong activator of transcription. Oncogene 9:2469. opment in human germinal centers by molecular analysis of single cells picked 30. Hardie, D. L., G. D. Johnson, M. Kahn, and I. C. M. MacLennan. 1993. Quan- from histological sections. Embo J. 12:4955. titative analysis of molecules which distinguish functional compartments within germinal centers. Eur. J. Immunol. 23:997. 41. Liu, Y. J., F. Malisan, O. de Bouteiller, C. Guret, S. Lebecque, J. Banchereau, 31. Gerdes, J., U. Schwab, H. Lemke, and H. Stein. 1983. Production of a mouse F. C. Mills, E. E. Max, and H. Martinez-Valdez. 1996. Within germinal centers, monoclonal antibody reactive with human nuclear antigen associated with cell isotype switching of immunoglobulin genes occurs after the onset of somatic proliferation. Int. J. Cancer 31:13. mutation. Immunity 4:241. 32. Pater, M., and I. Herskowitz. 1994. Joining the complex: cyclin-dependent kinase 42. Marhamati, D. J., R. E. Bellas, M. Arsura, K. E. Kypreos, and G. E. Sonenshein. inhibitory proteins and the cell cycle. Cell 79:181. 1997. A-myb is expressed in bovine vascular smooth muscle cells during the late 33. Gregory, C. D., T. Tursz, C. F. Edwards, C. Tetaud, M. Talbot, M. Caillou, G1/S phase transition and cooperates with c-myc to mediate progression to S A. B. Rickinson, and M. Lipinski. 1987. Identification of normal B cells with a phase. Mol. Biol. Cell 17:2448. Burkitt’s lymphoma (BL)-like phenotype. J. Immunol. 139:313. 43. Cattoretti, G., C.-C. Chang, K. Cechova, J. Zhang, B. H. Ye, B. Falini, 34. Clark, L. B., T. M. Foy, and R. J. Noelle. 1996. CD40 and its ligand. Adv. D. C. Louie, K. Offit, R. S. K. Chaganti, and R. Dalla-Favera. 1995. BCL-6 Immunol. 63:43. protein is expressed in germinal center B cells. Blood 86:45. 35. Cambier, J. C., L. B. Justement, K. Newell, Z. Z. Chen, L. K. Harris, V. M. Sandoval, M. J. Kemsz, and J. T. Ransom. 1987. Transmembrane signals 44. Turner, C. A., Jr., D. H. Mack, and M. M. Davis. 1994. Blimp-1, a novel zinc and intracellular “second messengers” in the regulation of quiescent B lympho- finger-containing protein that can drive the maturation of B lymphocytes into cytes activation. Immunol. Rev. 95:37. immunoglobulin-secreting cells. Cell 77:297. 36. Galibert, L., N. Burdin, B. de Saint-Vis, P. Garrone, C. Van Kooten, 45. Neurath, M. F., E. R. Stu¨ber, and W. Stober. 1995. BSAP: a key regulator of B J. Banchereau, and F. Rousset. 1996. CD40 and B cell antigen receptor dual cell development and differentiation. Immunol. Today 16:564. triggering of resting B lymphocytes turns on a partial germinal center phenotype. 46. Gstaiger, M., L. Knoepfel, O. Georgiev, W. Schaffner, and C. M. Hovens. 1995. J. Exp. Med. 183:77. A B-cell coactivator of octamer-binding transcription factors. Nature 373:360. Downloaded from 37. Arpin, C., J. Dechanet, C. Van Kooten, P. Merville, G. Grouard, F. Briere, J. Banchereau, and Y.-J. Liu. 1995. Generation of memory B cells and plasma 47. Thompson, C. B., P. B. Challoner, P. E. Neiman, and M. Groudine. 1986. Ex- cells in vitro. Science 268:720. pression of the c-myb proto-oncogene during cellular proliferation. Nature 319: 38. Malisan, F., F. Briere, J.-M. Bridon, N. Harindranah, F. C. Mills, E. E. Max, 374. J. Banchereau, and H. Martinez-Valdez. 1996. Interleukin-10 induces immuno- 48. Ying, G.-G., M. Arsura, M. Introna, and J. Golay. 1997. The DNA binding globulin G isotype switch recombination in human CD40-activated naive B lym- domain of the A-myb transcription factor is responsible for its B cell specific phocytes. J. Exp. Med. 183:937. activity and binds to a B cell 110 kD nuclear protein. J. Biol. Chem. 272:24921. http://www.jimmunol.org/ by guest on October 4, 2021