CD72-Deficient Mice Reveal Nonredundant Roles of CD72 in B Cell Development and Activation

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Immunity, Vol. 11, 495–506, October, 1999, Copyright 1999 by Cell Press CD72-Deficient Mice Reveal Nonredundant Roles of CD72 in B Cell Development and Activation

Chin Pan,* Nicole Baumgarth,† or to anti-IgM and synergizes with IL-1 in the induction and Jane R. Parnes*‡ of B cell proliferation (Snow et al., 1986; Yakura et al., *Department of Medicine 1986; Laurindo et al., 1987). In addition, mouse anti- Division of Immunology and Rheumatology CD72 causes a modest increase of class II major histo- † Department of Genetics compatibility complex (MHC) gene expression (Polla et Stanford University School of Medicine al., 1988; Subbarao et al., 1988) on splenic B cells. Stanford, California 94305 In other systems, however, inhibitory effects have been observed in response to CD72 cross-linking. Addi- tion of anti-CD72 inhibits the primary antibody response Summary to T-dependent antigens, although it does not affect responses to T-independent antigens (Yakura et al., CD72, a B cell surface protein of the C-type lectin 1981; Subbarao and Mosier, 1984). Anti-CD72 also superfamily, recruits the tyrosine phosphatase SHP-1 strongly inhibits LPS and IL-4 initiated IgG1 induction through its ITIM motif(s). Using CD72-deficient (CD72Ϫ/Ϫ) (Yakura et al., 1988). Furthermore, anti-CD72 mAb treat- mice, we demonstrate that CD72 is a nonredundant ment can partially rescue mouse splenic B cells from regulator of B cell development. In the bone marrow apoptosis induction by surface immunoglobulin (sIg) of CD72Ϫ/Ϫ mice, there was a reduction in the number hyper-cross-linking (Nomura et al., 1996). of mature recirculating B cells and an accumulation Recent signaling studies on CD72 might explain these of pre-B cells. In the periphery of CD72Ϫ/Ϫ mice, there observed negative regulatory effects of anti-CD72. The were fewer mature B-2 cells and more B-1 cells. In cytoplasmic domain of CD72 contains two immunore- addition, CD72 is a negative regulator of B cell activa- ceptor tyrosine-based inhibitory motifs (ITIM). Adachi et tion, as CD72Ϫ/Ϫ B cells were hyperproliferative in re- al. (1998) showed that cross-linking of the BCR en- sponse to various stimuli and showed enhanced kinet- hances both tyrosine phosphorylation of CD72 and as- ics in their intracellular Ca2؉ response following IgM sociation of CD72 with the protein tyrosine phosphatase cross-linking. SHP-1 in the WEHI-231 B cell line. Thus, CD72 might negatively regulate BCR signaling by recruiting SHP-1, Introduction a negative regulator of BCR signaling (Cyster and Good- now, 1995). Furthermore, Wu et al. (1998) demonstrated The development and function of B lymphocytes are that CD72 is an in vivo substrate of SHP-1 in B cells. critically controlled by the strength and quality of signals They showed that preligation of CD72 with anti-CD72 transmitted through the B cell antigen receptors (BCR) mAb attenuates BCR-induced growth arrest/apoptosis and positive and negative signals delivered by additional signals in primary immature and mature B cells or B cell cell surface molecules, often called coreceptors. The lines and demonstrated a strong correlation between properties of the antigen, as well as the nature of core- tyrosine phosphorylation of CD72 and BCR-induced ceptor ligation, partially influenced by the microenviron- growth arrest/apoptosis. Wu et al. (1998) hypothesized ment in which the B cell receives its signals, determine that tyrosine-phosphorylated CD72 transmits signals for the cell fate of the B cell during positive and negative BCR-induced apoptosis and that SHP-1 dephosphory- selection, maturation, and functional activation. lates CD72 and reverses the BCR-mediated apoptotic CD72, a 45 kDa type II transmembrane glycoprotein signal. and a member of the calcium-dependent lectin super- On the other hand, several biochemical studies sup- family, is one such coreceptor (Nakayama et al., 1989). port a positive signaling role for anti-CD72. Inositol 1,4,5- It is expressed as a disulfide-linked homodimer on all triphosphate (IP3) and Ca2ϩ appear to be important in B cells but not on terminally differentiated plasma cells. the signal transduction pathway of CD72. Mouse anti- In some mouse strains, CD72 is also expressed at low CD72 weakly induces phosphatidyl inositol turnover levels on subpopulations of T cells (Robinson et al., (Grupp et al., 1987) and induces an early increase in 2ϩ 1997). [Ca ]i (Subbarao et al., 1988; Pezzutto et al., 1990). The in vivo role of CD72 during B cell development and Protein tyrosine phosphorylation also appears to be inactivation has not been studied. However, accumulating volved in the CD72 signaling pathway. Venkataraman evidence from in vitro studies suggests that CD72 is an et al. (1998b) showed that anti-CD72 induces tyrosine important costimulatory molecule for B cell activation. phosphorylation of a variety of proteins in mouse splenic Engagement of CD72 by monoclonal antibodies (mAb) B cells, including phospholipase C-␥2 and CD19, and can transform a subset of small resting B cells into activates Lyn, Blk, and Btk kinases. In addition, CD72 blast cells and induce proliferation of both resting and ligation induces transient association of CD72 with activated B cells (Subbarao and Mosier, 1982, 1984; CD19 (Venkataraman et al., 1998a), which is a positive Yakura et al., 1986). Engagement of mouse CD72 en- regulator of BCR signaling (Engel et al., 1995; Rickert hances the B cell proliferative response either to IL-4 et al., 1995; Sato et al., 1995, 1996b). CD19 plays a critical role in B-1 cell development and is necessary ‡ To whom correspondence should be addressed (e-mail: jrparnes@ for optimal induction of T cell–independent responses leland.stanford.edu). (Rickert et al., 1995; Sato et al., 1995, 1996b). Immunity 496

Figure 1. Generation of CD72Ϫ/Ϫ Mice (A) CD72 targeting strategy: gene structure of wild-type CD72b allele (top), CD72 targeting construct (middle), and the resultant CD72 mutant allele (bottom). The intron/exon structure of the 14.2 kb CD72 clone from 129 mice is derived from CD72a allele (Ying et al., 1995). Exons encoding 5Ј or 3Ј untranslated sequences are shown as open boxes, and exons encoding coding sequences are shown by closed boxes. The 2.1 kb fragment containing exons I to IV was replaced with the pGK-MC1-neo cassette from pPNT (Tybu- lewicz et al., 1991). (B) Southern blot analysis. To assess the ge- notype of mice (wild-type [W]; heterozygous [H]; homozygous mutants [M]), cDNA corre- sponding to exons V to VIII was used as a probe for Southern analysis of EcoRI- digested mouse genomic DNA. The 8.5 kb EcoRI fragment represents the wild-type CD72 allele, and the 5.5 kb fragment depicts the targeted allele. Positions of EcoRI sites are shown in (A). The presence of CD72-targeted allele was also determined by PCR using primers as indicated in (A). (C) Splenocytes from CD72ϩ/ϩ (bold line), CD72ϩ/Ϫ (thin line), and CD72Ϫ/Ϫ (shaded area) mice were stained for CD72 surface expression with anti-CD72 mAb (K10.6) specific for both CD72a and CD72b.

Taken together, a number of functional and biochemi- Impaired B Cell Development in CD72Ϫ/Ϫ Mice cal studies in vitro indicate that ligation of CD72 by its B cell development in the bone marrow of adult 8- to mAb leads to multiple effects upon B cell signaling. In 12-week-old CD72Ϫ/Ϫ mice and their littermate controls order to begin to unravel the mechanisms of these in was analyzed by flow cytometry as described (Hardy et vitro effects of anti-CD72 mAb and particularly to eluci- al., 1991). Total numbers of bone marrow cells recovered date the function of CD72 during B cell development from one femur and one tibia of CD72Ϫ/Ϫ and wild-type and activation in vivo, we generated CD72Ϫ/Ϫ mice by mice were comparable. B220loCD43ϩ B cells, containing targeted gene inactivation. Here, we demonstrate that the earliest stages of B cell development (pre-pro-B, CD72 plays a nonredundant role in B cell development fraction A; pro-B, fraction B; and late pro-B, fraction C, in vivo and negatively regulates responsiveness of B cells. Figure 2) were present at normal numbers in the mutant mice. However, the later stages of B cell development Results were altered. Although the numbers of B220ϩCD43Ϫ B cells from mutant and wild-type mice were comparable, Generation of CD72Ϫ/Ϫ Mice there was a significant decrease in the number of mature A 14.2 kb CD72 genomic clone was isolated from a 129/ IgMϩIgDϩ B cells (fraction F, Figure 2) and an increase SV/EV mouse genomic library (Stratagene) with probes in the number of IgMϪIgDϪ pre-B cells (fraction D, Figure derived from the CD72 cDNA sequence. The targeting 2). The number of IgMϩIgDϪ immature B cells (fraction vector was constructed by removing a 2.1 kb DNA frag- E, Figure 2) was minimally affected in CD72Ϫ/Ϫ mice. ment including the CD72 minimal promoter and se- The decreased ratios of immature to pre-B cells and quence encoding the cytoplasmic and transmembrane mature to immature B cells suggest that CD72 may be domains to ensure disruption of CD72 protein expres- important for the efficient transitions from pre-B to im- sion (Figure 1A). Transfection of the embryonic stem mature B cells and from immature to mature B cells (ES) cell line E14.1 (kindly provided by Dr. Richard Mur- during development. ray, DNAX Research Institute) resulted in two clones carrying one copy of the homologously recombined Alterations in Peripheral B Cell Pools CD72 mutant allele as demonstrated by Southern blot of CD72Ϫ/Ϫ Mice and PCR analysis of genomic DNA from ES cell clones. Total numbers of cells from spleen, peripheral lymph Standard blastocyst injection techniques were used to nodes, and Peyer’s patches of 8- to 12-week-old CD72Ϫ/Ϫ generate chimeric mice from CD72 mutant ES cells. and wild-type mice were comparable. However, the total Germline transmission of the CD72 targeted allele was number of splenic B cells (B220ϩ) in adult CD72Ϫ/Ϫ mice assessed by Southern blot analysis (Figure 1B). Cell was decreased by 24 Ϯ 11% compared to wild-type surface staining of spleen cells confirmed the lack of controls (Figure 3A). The mutant mice also had de- CD72 expression in homozygous mutants and reduced creased numbers of B cells in lymph nodes (32 Ϯ 12% expression of CD72 in heterozygous mutants compared decrease) and Peyer’s patches (20 Ϯ 10% decrease) to wild-type controls (Figure 1C). (Figure 3A). Multiple B Cell Defects in CD72-Deficient Mice 497

Increased Numbers of IgMϪIgDϪ B Cells in Peyer’s Patches and Lymph Nodes of CD72Ϫ/Ϫ Mice Analysis of Peyer’s patches and mesenteric lymph nodes from adult CD72Ϫ/Ϫ mice revealed an increase in IgMϪIgDϪ B cells in these tissues compared to wild-type controls (Figures 3D and 3E). A comparable frequency of these IgMϪIgDϪ B cells in both types of mice stained highly with peanut agglutinin (Figures 3D and 3E) and expressed high levels of Fas antigen (data not shown), therefore suggesting that CD72Ϫ/Ϫ mice have increased numbers of B cells with the phenotypic characteristics of germinal center cells in Peyer’s patches and lymph nodes (Rose et al., 1980; Butcher et al., 1982; Smith et al., 1995). Consistent with this, there was an increase in the number (but not percentage) of IgA-expressing cells in both lymph nodes and Peyer’s patches (Figures 3E and 3D). This increase in the number of IgMϪIgDϪ cells, including germinal center B cells and isotype switched B cells in CD72Ϫ/Ϫ mice, suggests that CD72 regulates the levels of B cell activation in vivo. Consis- tent with this, Percoll gradient purification of wild-type and mutant B cells showed that mutant mice had ap- proximately half the number of resting, high buoyant density splenic B cells (60%–70% Percoll gradient interface) and about twice the number of activated, low buoyant density B cells (50%–60% Percoll gradient interface), further indicating that the absence of CD72 expression increases the number of activated B cells in vivo.

Delayed B Cell Development and Maturation Figure 2. Impaired Early B Cell Development in the Bone Marrow in Young CD72Ϫ/Ϫ Mice of CD72Ϫ/Ϫ Mice To further study the role of CD72 during B cell matura- B cell development in the bone marrow was analyzed by eight- tion in the periphery, we analyzed B cell develop- color FACS according to Hardy et al. (1991). Shown at the top are ment in young mice. Consistent with the data from adult representative 5% probability contour plots of total bone marrow mice, 3- to 5-week-old CD72Ϫ/Ϫ mice showed a delay cells from adult CD72 wild-type (ϩ/ϩ) and mutant (Ϫ/Ϫ) mice. Com- parable numbers of bone marrow cells were recovered from one in B cell maturation. Significantly increased numbers of hi lo femur and one tibia of wild-type and CD72Ϫ/Ϫ mice. B220ϩCD43ϩ immature IgM B220 and reduced numbers of mature pro-B cells and B220ϩCD43Ϫ cells including pre-B, immature, and IgMintB220hi cells were found in spleens of young mature B cells were gated for further analysis of surface expression CD72Ϫ/Ϫ mice (Figure 4A). Flow cytometric analyses of BP-1 versus HSA and IgD versus IgM, respectively. Gates are confirmed that the majority of IgMhiB220lo splenic cells indicated by boxes, and percent of gated cells of total cells within were immature B cells, since they were IgDϪ/lo, CD23Ϫ, the plot is indicated. CD22lo, and HSAhi (data not shown). The IgMintB220hi splenic cells were IgDhi, CD22hi, HSAlo and were therefore CD72Ϫ/Ϫ mice had a 38 Ϯ 11% reduction in IgMintCD23ϩ mature B cells (data not shown). Three- to five-week-old mature long-lived follicular B cells (Figures 3A and CD72Ϫ/Ϫ mice also had significantly decreased numbers 3B) in the spleen compared to wild-type control mice. (53 Ϯ 15%) of mature CD23ϩ follicular B cells in the In contrast, CD72Ϫ/Ϫ mice had normal or increased spleen (Figure 4A) and decreased numbers of B cells in numbers of other B cell populations in the spleen. In lymph nodes (data not shown). particular, they had increased numbers of B-1 cells In summary, the absence of CD72 led to a delay rather (B220ϩCD23ϪIgDϪ/loCD43ϩ, Figure 3A), some of which than a complete block in B cell maturation, since the are CD5ϩ B-1a cells (Figure 3B), and a modest increase deficiency in mature follicular B cells in adult CD72Ϫ/Ϫ in the number of B cells with the phenotypic features of mice was not as severe as in young CD72Ϫ/Ϫ mice. It marginal zone cells (IgMhiIgDϪ/loCD21hi) (data not shown). took longer for CD72Ϫ/Ϫ newborn mice to reach a steady- The number of immature B cells (IgMhiIgDϪ/loHSAhiCD43Ϫ state level of mature B cell accumulation in the spleen CD21lo) was similar to that in control mice (data not and to accumulate adult numbers of B cells in lymph shown). nodes (data not shown). Consistent with the findings in the spleen, CD72Ϫ/Ϫ mice had increased numbers of B-1 cells (Figure 3C) Adult CD72Ϫ/Ϫ Mice Have a Slower Rate of Generating and reduced numbers of conventional B-2 cells (CD23ϩ, Mature Recirculating B Cells data not shown) in the peritoneal cavity. In summary, Although adult mice are capable of generating 2 ϫ 107 CD72Ϫ/Ϫ mice had a reduction in mature, long-lived con- new B cells every day (Osmond, 1991), only about 10%– ventional B-2 cells in the periphery and an expansion 20% of them are recruited into the mature recirculating of B-1 cells. peripheral B cell pool (Allman et al., 1993). To determine Immunity 498

Figure 3. Reduced Numbers of B Cells and Altered Distribution of B Cell Subpopulations in the Periphery of CD72Ϫ/Ϫ Mice (A) Average absolute numbers of B cells in the spleen (n ϭ 20), mesenteric lymph nodes (MLN) (n ϭ 16), Peyer’s patches (PP) (n ϭ 8), and numbers of splenic CD23ϩ follicular B cells (FO) (n ϭ 20) and B-1 cells (n ϭ 12) of 8- to 12-week-old wild-type and CD72Ϫ/Ϫ mice. (B) Representative 5% probability contour profiles of spleen cells of wild-type and CD72Ϫ/Ϫ mice. Comparable numbers of splenocytes were recovered from wild-type and mutant mice. Note the decreased frequency of follicular B cells and increased frequency of B-1 cells in the spleen of mutant mice. (C) Representative 5% probability contour profiles of lymphocytes in the peritoneal cavity of wild-type and CD72Ϫ/Ϫ mice. Note the increased frequency of B-1 cells in the peritoneum of mutant mice. (D and E) Increased numbers of IgA-expressing and PNAhi B cells in Peyer’s patches and mesenteric lymph nodes of CD72Ϫ/Ϫ mice. B cells (B220ϩ) from Peyer’s patches and mesenteric (mes.) lymph nodes of wild-type and CD72Ϫ/Ϫ mice were analyzed for expression of IgM and IgD. IgMϪIgDϪ B cells were further analyzed for their expression of IgA and binding to PNA. Note the increased frequency of the IgMϪ/IgDϪ fraction and the increased numbers of IgA-expressing and PNAhi B cells in Peyer’s patches and mesenteric lymph nodes of CD72Ϫ/Ϫ mice. Shown are 5% contour plots, and percentages of gated cells of total cells within the plot are indicated. Multiple B Cell Defects in CD72-Deficient Mice 499

CD23 expression (Figure 4B). The reduction of CD23ϩ B cells in CD72Ϫ/Ϫ mice was mainly in the IgMint fraction, which is the more mature population. The CD23ϩIgMhi fraction, which is less mature, was not affected in CD72Ϫ/Ϫ mice. Furthermore, reduced numbers of B220ϩ cells were found in lymph nodes of CD72Ϫ/Ϫ mice (Figure 4B) at day 13 after irradiation. We therefore conclude that CD72 expression is important for normal B cell development in both developing and adult mice.

Overexpression of BCL-2 Does Not Completely Change the B Cell Developmental Defect in CD72Ϫ/Ϫ Mice Bcl-2 is an important regulator of B cell apoptosis/sur- vival during development. To test whether enforced overexpression of this inhibitor of apoptosis negates the effects on B cell development introduced by the deletion of CD72, CD72Ϫ/Ϫ mice were crossed with E␮-BCL-2- 22 transgenic mice, which express BCL-2 at all stages of B cell development (Strasser et al., 1991). Overexpression of BCL-2 led to equivalent accumulation of B cells in the periphery of 4- to 6-week-old CD72Ϫ/Ϫ and control mice and a partial correction of the B cell maturation defect observed in CD72Ϫ/Ϫ mice. However, mice lacking CD72 and overexpressing BCL-2 still had a reduction in the frequency of mature B cells (CD23ϩHSAlo, or CD23ϩIgMint/hi,orIgDϩIgMint/hi) and a higher frequency of immature B cells (CD23ϪHSAhi or IgDϪ) in the spleen (Figure 5A). In the bone marrow, overexpression of BCL-2 in CD72Ϫ/Ϫ mice led to a less severe reduction in IgDϩ cells, yet these mice still had a higher frequency of immature B cells (IgMϩIgDϪ) and a slightly lower frequency of mature B cells, including the more mature IgMintIgDϩ population and the less mature IgMhiIgDϩ population, than CD72-expressing, BCL-2 transgenic controls (Figure 5B). Therefore, the B cell developmental defect in CD72Ϫ/Ϫ mice is likely in part a result of increased apoptosis blockable by enforced expression of Bcl-2, but the maturation deficiency is not Ϫ/Ϫ Figure 4. Delayed B Cell Maturation in Young and Adult CD72 solely a result of increased apoptosis of B cells as they Mice mature. (A) Representative 5% probability contour profiles of spleen cells from wild-type (ϩ/ϩ) and CD72Ϫ/Ϫ (Ϫ/Ϫ) mice of 4 weeks of age. Ϫ/Ϫ (B) Delayed generation of mature B cells in sublethally irradiated Hyperproliferation of CD72 B Cells In Vitro CD72Ϫ/Ϫ mice. Representative 5% probability contour profiles of Next, we determined whether the loss of CD72 affects splenic B cells and cervical lymph node cells from wild-type and the ability of B cells to respond with proliferation to CD72Ϫ/Ϫ adult mice 13 and 15 days after sublethal dosage of irradia- various stimuli. In particular, we tested the effects of ϩ tion are shown. Plots of CD23 versus IgM of B220 splenic B cells are lack of CD72 costimulation on the response to activation shown. Percentages of CD23hiIgMint mature B cells and CD23hiIgMhi via the B cell receptor by IgM cross-linking and on the transitional B cells out of total splenic B cells are indicated. Plots of B220 versus CD3⑀ of cervical lymph node cells from day 13 response to polyclonal stimulation by the mitogen lipo- reconstituted mice are shown, and percentages of B220ϩ B cells polysaccharide (LPS) (Figure 6A). To minimize subpopu- and CD3ϩ T cells out of total live lymph node cells are indicated. lation differences, T cell–depleted splenic B cells from adult CD72Ϫ/Ϫ and wild-type mice were purified through Percoll gradient centrifugation, and resting B cells re- whether the alterations in the B cell compartments of covered from the interface of 60%–70% Percoll gradient adult CD72Ϫ/Ϫ mice resulted from maturation defects were used for proliferation assays; these cells were present only during ontogeny or whether such defects mostly mature CD23ϩIgDϩ cells from both wild-type and extended into adulthood, we compared B cell replen- CD72Ϫ/Ϫ mice (Figure 6A). These purified resting splenic ishment of adult wild-type and CD72Ϫ/Ϫ mice at 13 and B cells from CD72Ϫ/Ϫ mice showed increased prolifera- 15 days after sublethal irradiation. tive responses to low levels of IgM cross-linking in the Following irradiation, CD72Ϫ/Ϫ and control mice had presence or absence of IL-4 and to stimulation with LPS similar numbers of total B cells in the spleen. However, compared to wild-type controls. Increased proliferation more of the B cells in the mutant mice were of the of B cells from CD72Ϫ/Ϫ mice was most prominent at immature phenotype, as demonstrated by their lack of suboptimal concentrations of F(abЈ)2 anti-IgM and LPS Immunity 500

6A, mutant B cells proliferated more upon BCR-Fc␥RIIB co-cross-linking than wild-type controls. However, both types of B cells showed a strong reduction in proliferation compared to stimulation with F(abЈ)2 anti-IgM. Therefore, CD72 expression is not necessary for the Fc␥RIIB-mediated block in proliferation.

Altered In Vitro Ca2؉ Responses by CD72Ϫ/Ϫ B Cells SHP-1 mutant B cells have an exaggerated intracellular Ca2ϩ response to antigen cross-linking (Cyster and Goodnow, 1995). Given that CD72 associates with SHP-1 (Adachi et al., 1998; Wu et al., 1998), the ability of wild-type and mutant splenic B cells to increase their levels of intracellular free Ca2ϩ after IgM cross-linking was examined. As shown in Figure 6B, Ca2ϩ modulation in B cells from CD72Ϫ/Ϫ mice was distinctively altered in 2ϩ that the increase in [Ca ]i following surface IgM cross- linking was more rapid in the mutant B cells. This phenotype was observed over a range of F(abЈ)2 anti-IgM concentrations used for stimulation. These data indicate that CD72 negatively regulates the proximal Ca2ϩ response to BCR signaling.

Humoral Immune Responses in CD72Ϫ/Ϫ Mice Serum immunoglobulin (Ig) levels in nonimmunized mice were determined to assess any functional alterations the loss of CD72 might have on this in vivo steady-state B cell function. At 2 months of age, CD72Ϫ/Ϫ mice had a small but statistically significant increase in the level of serum IgM compared to wild-type controls, whereas the levels of all other isotypes were comparable (Figure 7A). To test whether B cells from CD72Ϫ/Ϫ mice could mount normal humoral responses to T cell–dependent and type 2 T cell–independent antigens, CD72Ϫ/Ϫ mice and wild-type littermates were immunized with 2,4,6- trinitrophenol-conjugated keyhole limpet hemocyanin (TNP-KLH) or TNP-Ficoll. At 7 and 14 days after immunization, sera from these mice were assayed by TNP- specific and isotype-specific ELISA to determine the levels of antigen-specific IgM, IgG1, IgG2a, IgG2b, and Figure 5. Overexpression of BCL-2 Does Not Completely Rescue IgG3. Mice were rechallenged with TNP-KLH at day 21 the B Cell Maturation Defect in CD72Ϫ/Ϫ Mice and bled at day 28 and day 35 to determine their ability Representative 5% probability contour profiles of spleen and bone to mount a secondary antibody response. Despite the marrow B cells from 5-week-old wild-type (ϩ/ϩ) and mutant (Ϫ/Ϫ) observed alterations in their B cell compartment, CD72Ϫ/Ϫ mice expressing E␮-BCL-2-22 transgene. (A) CD23 versus HSA, mice showed normal T-independent responses except ϩ CD23 versus IgM, and IgD versus IgM contour plots of B220 splenic for a slightly stronger IgM response to TNP-Ficoll at day B cells are shown. Percentages of CD23ϩHSAlo mature B cells and 7 (Figure 7B). Overall, the T cell–dependent primary and CD23ϪHSAhi immature B cells; mature CD23ϩIgMint or IgDϩIgMint Ϫ/Ϫ cells, transitional CD23ϩIgMhi or IgDϩIgMhi cells, and immature secondary responses to TNP-KLH in CD72 mice were CD23ϪIgMhi or IgDϪIgMhi B cells out of total B220ϩ cells are indi- also similar to those seen in wild-type controls (Figures cated. (B) IgD versus IgM contour plots of B220ϩ bone marrow B 7C–7F). Therefore, CD72 expression does not appear cells are shown. Percentages of immature IgMϩIgDϪ B cells, transi- to be necessary for the induction of normal humoral ϩ hi ϩ int tional IgD IgM B cells, and mature IgD IgM B cells out of total immune responses to T cell–dependent and T cell–inde- bone marrow B cells are indicated. pendent antigens or for isotype switching. However, CD72 plays some role in regulating IgM responses as and gradually disappeared with increasing concentra- CD72Ϫ/Ϫ mice had a slightly higher level of serum IgM tions of stimuli (data not shown). Thus, the data suggest and a stronger IgM T-independent response. that CD72 is a negative regulator of activation-induced B cell proliferation. To test whether CD72 is required for Discussion negative signaling through Fc receptors, we analyzed B cell proliferation in response to BCR-Fc␥RIIB co-cross- In this study, we demonstrate that lack of CD72, a B linking using whole anti-IgM Abs. As shown in Figure cell surface protein that recruits and is a substrate for Multiple B Cell Defects in CD72-Deficient Mice 501

Figure 6. Hyperproliferation and Enhanced Kinetics of Intracellular Ca2ϩ Response of Splenic B Cells from CD72Ϫ/Ϫ Mice (A) Hyperproliferation of splenic B cells from CD72Ϫ/Ϫ mice. T-depleted splenic B cells were purified through Percoll gradient centrifugation, and IgD versus CD23 contour plots of resting B cells recovered from the 60%–70% interface and more activated B cells recovered from the 50%–60% interface are shown. Proliferative response of Percoll gradient purified resting splenic B cells (60%–70% interface) from wild-type (ϩ/ϩ) and mutant (Ϫ/Ϫ) mice to various stimuli is shown. All assays were performed in triplicate, and standard deviations are shown. The results shown are representative of six individual experiments. The open boxes represent the wild-type mice, and the closed boxes represent the mutant mice. 1 or 5 IgM, 1 or 5 ␮g/ml F(abЈ)2 goat anti-mouse polyclonal antibodies; 0.1 or 1 LPS, 0.1 or 1 ␮g/ml LPS; 2 or 10 whole, 2 or 10 ␮g/ml goat anti-mouse polyclonal whole antibodies. (B) Intracellular Ca2ϩ response of splenic B cells from wild-type and CD72Ϫ/Ϫ mice upon IgM cross-linking. Splenocytes from wild-type and CD72Ϫ/Ϫ mice stimulated with 5, 10, and 15 ␮g/ml of F(abЈ)2 goat anti-mouse polyclonal 2ϩ 2ϩ antibodies were examined by flow cytometry for [Ca ]i modulation. Indo-1 violet/blue ratio correlates with [Ca ]i level, and the mean indo-1 ratio was calculated for every 10 sec (sec) fraction. Representative plots from six experiments of indo-1 violet/blue ratio over time are shown. The time of addition of stimuli is indicated by the arrow. the tyrosine phosphatase SHP-1, affects the develop- velopment but rather affects the number or the kinetics ment and the functional activation of B cells in vivo. The of the cells selected into the mature B cell pool. Consis- most prominent effects of CD72 deficiency on B cell tent with the effect of CD72 deletion on immature B development were seen at the transition from the imma- cells, we found that normal B cells at this stage express ture to mature B cell stage. Decreased numbers of ma- the highest level of CD72 compared to all other B cell ture B cells were found in the bone marrow and periphery maturation stages (data not shown). of CD72Ϫ/Ϫ mice. In both young and adult mice, there Lack of CD72 expression also affected the pre-B cell was ongoing delayed B cell maturation. Our data there- compartment. This compartment (Hardy’s fraction D, fore provide direct evidence that CD72 plays an impor- Figure 2) was increased in size, and the ratio of immature tant role in the efficiency of this developmental transi- B cells to pre-B cells was reduced, suggesting that CD72 tion, possibly by negatively regulating BCR signaling. might be important for the efficient transition from the However, mature B cells are found in CD72Ϫ/Ϫ mice, pre-B cell to the immature B cell stage. However, imma- albeit at reduced numbers, demonstrating that the lack ture B cells were present at relatively normal rather than of CD72 signaling does not completely block B cell de- reduced numbers in the bone marrow of CD72Ϫ/Ϫ mice; Immunity 502

this could be the result of a combination of a relatively decreased proportion of immature B cells from inefficient transition from the pre-B to immature B cell stage and the opposite effect from inefficient transition from the immature to mature B cell stage. Alternatively, the accumulation of pre-B cells could also be a result of increased or abnormal proliferation of pre-B cells, as increased proliferation following BCR-stimulation was seen in mature, CD72Ϫ/Ϫ B cells. It is conceivable that CD72 could negatively regulate pre-BCR signaling. Immature B cells express very low levels of Bcl-2 (Carsetti et al., 1995), rendering these cells exquisitely sensitive to apoptosis induction. Data from transgenic mouse models indicate that negative selection occurs at this stage of B cell development (Goodnow et al., 1995), facilitated in part by the lack of Bcl-2 expression. Excessive signaling through the BCR in mature B cells could also lead to their increased apoptosis as evi- denced by CD22Ϫ/Ϫ mice, in which the reduction in mature recirculating B cells was secondary to their increased turnover (Otipoby et al., 1996). The decrease in the number of mature B cells in CD72Ϫ/Ϫ mice was partially corrected by overexpression of Bcl-2, indicating that the deficiency of mature B cells may result in part from increased apoptosis of mature B cells or immature B cells on their way to maturation. In support of this interpretation, BrdU incorporation studies also showed a slight increase in B cell turnover in the spleen of CD72Ϫ/Ϫ mice (data not shown). However, the B cell maturation defect was not fully reversed by overexpression of Bcl-2, suggesting that CD72 plays an additional role in B cell maturation beyond regulation of cell death of immature and/or mature B cells. In contrast to the reduced numbers of mature conventional B-2 cells found in CD72Ϫ/Ϫ mice, the mutant mice had increased numbers of B-1 cells in the spleen and peritoneal cavity. These data are reminiscent of findings in a number of other gene-targeted mice that lack molecules important in BCR signaling with divergent conse- quences for B-1 and B-2 cell development. Our data therefore provide further evidence that B-1 and B-2 cell development is differently affected by signaling through the BCR and its coreceptors and demonstrate that CD72 is one such coreceptor. Mice deficient in molecules that negatively regulate BCR signaling, including SHP-1 (Sid- man et al., 1986) and CD22 (O’Keefe et al., 1996; Sato et al., 1996a), show an expansion of B-1 cells, whereas mice deficient in several other molecules known to posi- tively regulate BCR signaling, including the nonreceptor tyrosine kinase Btk (Khan et al., 1995), the B cell surface molecule CD19 (Engel et al., 1995; Rickert et al., 1995),

gens. Levels of TNP-specific IgM, IgG1, and IgG3 in CD72Ϫ/Ϫ mice and wild-type littermates (n ϭ 10) at 7 and 14 days after immunization with TNP-Ficoll are shown. (C–F) Humoral responses to T cell- dependent antigens. Levels of TNP-specific IgM, IgG1, IgG2a, IgG2b, and IgG3 in CD72Ϫ/Ϫ mice and wild-type littermates at 7 and 14 days after primary and secondary immunizations with TNP-KLH (n ϭ 12) are shown. Statistical analyses were performed using Stat- Ϫ/Ϫ Figure 7. Humoral Immune Responses in CD72 Mice view 4.5 software (Abacus Concepts). P values were calculated (A) Serum immunoglobulin (Ig) levels in 8-week-old nonimmunized using Unpaired t-test, and statistically significant differences (p Ͻ wild-type (ϩ/ϩ) and mutant (Ϫ/Ϫ) mice on 100% 129 background 0.05) between responses of wild-type and CD72Ϫ/Ϫ mice were (n ϭ 10). (B) Humoral responses to type 2 T cell–independent anti- marked by an asterisk. Multiple B Cell Defects in CD72-Deficient Mice 503

the C3d complement receptor CD21 (Ahearn et al., and CD72Ϫ/Ϫ mice were masked by the use of a strong 1996), and the signaling molecule Ig-␣ (Torres et al., adjuvant or by high doses of antigen. 1996), all have various degrees of reduction in the num- At the molecular level, the negative regulatory function ber of B-1 cells (reviewed in Tedder et al., 1997). CD72 of CD72 could be exerted through the two ITIMs in its gene-targeted mice fall within the first group of mice, cytoplasmic tail. Upon phosphorylation of its ITIM mo- therefore suggesting that CD72 negatively regulates tif(s), CD72 recruits (Adachi et al., 1998; Wu et al., 1998) BCR signaling in vivo. and serves as a substrate for the tyrosine phosphatase Our functional studies on B cells from CD72Ϫ/Ϫ mice SHP-1 (Wu et al., 1998). The first ITIM sequence (5ITY- strongly support such a conclusion. These B cells showed ADL10) has been shown to be critical for SHP-1 associa- heightened, antigen-dose-dependent responses to in tion, while the second (37LTYENV42) is capable of in vitro vitro stimulation. In particular, B cells from CD72Ϫ/Ϫ mice binding with Grb2 (Wu et al., 1998). The potential associ- proliferated at increased levels to various stimuli, includ- ation of CD72 with Grb2 in vivo provides another mecha- ing BCR cross-linking and polyclonal activation through nism by which CD72 might regulate downstream ef- LPS, suggesting that CD72 might negatively regulate a fector proteins, since Grb2 binds to Sos, a guanine common step in the pathway leading to B cell prolifera- nucleotide exchange factor for Ras (Egan et al., 1993; tion. The proproliferative effects seen with anti-CD72 Li et al., 1993). mAb treatment may be explained by mAb blocking of The phenotype of CD72Ϫ/Ϫ mice is similar to but more negative regulatory signals. Notably, the Fc␥RIIB-medi- subtle than that of mev mice, which express a catalyti- ated inhibitory pathway was still functional in CD72Ϫ/Ϫ cally impaired form of SHP-1. Both types of mice have mice, as CD72Ϫ/Ϫ B cells, like control B cells, were an expansion of B-1 cells and a reduction of B-2 cells strongly inhibited in their proliferative response to stimu- (Sidman et al., 1986). Their B cells both mount an in- lation with whole anti-IgM Abs, which bind to both IgM creased calcium response to BCR cross-linking (Cyster and the Fc␥RIIB. Also consistent with a negative regula- and Goodnow, 1995). Ϫ/Ϫ Ϫ/Ϫ tory role for CD72, Ca2ϩ flux, a very early signaling event CD72 mice are also similar to CD22 mice after BCR cross-linking, was more rapidly induced in (O’Keefe et al., 1996; Otipoby et al., 1996; Sato et al., CD72Ϫ/Ϫ mice following IgM cross-linking. 1996a; Nitschke et al., 1997). CD22, a B cell coreceptor, The difference in proliferation between purified resting also contains ITIMs in its cytoplasmic tail and has been splenic B cells from CD72Ϫ/Ϫ and wild-type mice was shown to associate with SHP-1 (Campbell and Klinman, mainly a result of loss of expression of CD72, since the 1995; Doody et al., 1995; Law et al., 1996). Both CD72- Percoll gradient-purified CD72Ϫ/Ϫ and CD72ϩ/ϩ resting and CD22-deficient mice have a great reduction in ma- B cells (60%–70% interface) used for these assays were ture, recirculating B cells in the bone marrow, and their B phenotypically very similar (Figure 6A). In contrast, the cells are hyperproliferative to LPS stimulation. However, altered calcium response of CD72Ϫ/Ϫ splenic B cells CD22 plays a larger role in regulating the intracellular could result, at least in part, from the altered compo- calcium response of B cells to BCR stimulation, as Ϫ/Ϫ sition of B cell subpopulations, since the intracellular CD22 B cells have a much greater calcium flux upon calcium assay measured the calcium response of total low-level BCR cross-linking, while CD72 plays a more splenic B cells. Calcium mobilization occurs more important role in early B cell development. CD22 is a quickly and continues for a longer time in marginal zone negative regulator of antigen receptor signaling, and its (MZ) B cells than in follicular B cells and newly formed increased expression at the mature B cell stage may B cells (Oliver et al., 1997), and the CD72Ϫ/Ϫ mice had serve to raise the antigen concentration threshold required for B cell triggering. CD72 is also a negative a slightly higher percentage of MZ cells and a lower regulator of BCR signaling; however, its reduced expres- percentage of follicular B cells. Furthermore, the in- sion at the mature B cell stage may be critical for the creases in the steady-state levels of serum IgM and the regulation of mature B cell activation. Moreover, CD22 increase in the IgM-response to TNP-Ficoll might also and CD72 may be functionally redundant in some as- be explained, at least in part, by the increased numbers pects of in vivo function, as the phenotype of mice defi- of B-1 cells and MZ cells found in CD72Ϫ/Ϫ mice, since cient in either one of these proteins was not as severe B-1 cells have been shown to be largely responsible for as mev mice. Experiments are in progress to study the basal serum Igs and to be critical for T cell–independent potential redundancy in CD72Ϫ/Ϫ CD22Ϫ/Ϫ double mu- type 2 responses (Kantor and Herzenberg, 1993). In ad- tant mice. dition, MZ B cells have been proposed to have an impor- Finally, the phenotype of CD72Ϫ/Ϫ mice closely resem- tant role in the response to T cell–independent type 2 bles that of the tyrosine kinase LynϪ/Ϫ mice (Hibbs et antigens (Oliver et al., 1997). al., 1995; Nishizumi et al., 1995; Wang et al., 1996). Like Our findings of heightened B cell responsiveness in CD72Ϫ/Ϫ mice, LynϪ/Ϫ mice have greatly reduced num- vitro are in apparent contrast to the normal levels of bers of mature recirculating B cells in the bone marrow serum Ig and the normal in vivo responses to immuni- and in the periphery. Like young CD72Ϫ/Ϫ mice, young zation in CD72Ϫ/Ϫ mice. However, CD72Ϫ/Ϫ mice had LynϪ/Ϫ mice also show delayed B cell maturation and increased numbers of germinal center and isotype- Lyn mutant B cells are hyperresponsive to IgM cross- switched B cells in Peyer’s patches and lymph nodes, linking and LPS stimulation (Wang et al., 1996). The suggesting heightened steady-state B cell activation in similarity of these mice suggests that Lyn might be in vivo. Furthermore, we noted increased numbers of acti- the signaling pathway of CD72. Consistent with this, vated and decreased numbers of resting splenic B cells CD72 is phosphorylated by Lyn tyrosine kinase in vitro in CD72Ϫ/Ϫ mice. It is possible that differences in in vivo (Adachi et al., 1998). Lyn mutant mice produce autoanti- responses to antigen immunization between wild-type bodies and develop glomerulonephritis, suggestive of Immunity 504

systemic lupus erythematosus (SLE) (Hibbs et al., 1995; APC, and anti-IgD PE were purchased from Southern Biotechnology Nishizumi et al., 1995). We have not seen evidence of Associates. All other antibodies were purchased from PharMingen. autoimmune diseases in CD72Ϫ/Ϫ mice to date. Samples were analyzed on a modified Vantage (Beckton Dickinson) equipped with an Argon laser (488 nm) and a dye laser (595 nm). Ϫ/Ϫ In summary, by generating CD72 mice, we have For the analysis of bone marrow samples, eight-color FACS analy- shown that CD72 plays a nonredundant role in B cell sis was performed using the following antibodies: anti-CD3 Cascade development and function. Loss of CD72 leads to a Blue (CB) (mAb 2C11); anti-CD4 CB (GK.1.5); anti-CD8 CB (53-6.7); reduction in mature conventional B-2 cells and an anti-macrophage antigen F4/80 CB; anti-BP-1 FITC (6C3); anti-CD43 expansion of B-1 cells similar to that seen in mice defi- phycoerythrin (PE) (S7); anti-B220 Cy5-PE; anti-IgD Cy7-PE (1126); cient in other proteins that negatively regulate BCR sig- anti-CD24 TR (30. F1); anti–stem cell antigen AA4.1 APC; and anti- IgM Cy7-APC (331). The antibodies were purified from hybridoma naling. On mature B cells, the accentuated signaling supernatant and conjugated as outlined (Hardy, 1986; Roederer et resulting from CD72-deficiency lowers the threshold for al., 1996). For eight-color analysis, a custom-built triple laser FACS B cell activation. was used as described previously (Roederer et al., 1997). Data analysis was performed using the FlowJo software (Treestar). Experimental Procedures Cell Proliferation Assays Cloning of the CD72 Gene and Construction Splenocytes from 8- to 12-week-old mice were depleted of T cells of the CD72 Targeting Vector by incubating with an antibody cocktail of Thy1.2 (3HO12), CD4 A 14.2 kb CD72 genomic clone from the 129 mouse strain was (GK1.5), and CD8 (3.155) monoclonal antibodies, followed by T cell isolated by screening a 129/SV/EV genomic library (Stratagene) with lysis with 3–4 week rabbit complement (Pel-Freez). Resting B cells two probes derived from the CD72 cDNA sequence. Based on the were separated from macrophages, dead cells, and debris by cen- sequence information of the 8.5 kb CD72 gene from C57L mice trifugation in a discontinuous Percoll gradient of 50%, 60% and 5 (Ying et al., 1995), restriction mapping analysis was performed on 70% and collected from the 60%–70% interface. 2 ϫ 10 purified B the 14.2 kb clone. This genomic clone was shown to contain 7.1 kb cells were plated in each well in 96-well flat-bottom plates in com- of CD72 5Ј flanking sequence and 7.1 kb of sequence containing plete RPMI medium containing 5% FCS and ␤-mercaptoethanol Ϫ5 all the nine exons and eight introns of CD72 (Figure 1A). The targeting (5 ϫ 10 M). B cells were stimulated with indicated concentrations strategy was to disrupt the minimal promoter and the cytoplasmic of affinity-purified F(abЈ)2 goat anti-mouse IgM antisera (Jackson and transmembrane domains of CD72. To construct the CD72 Immune Research Laboratories), with or without 10 ng/ml IL-4, lipo- knockout allele, the 2.1 kb fragment containing exons I, II, III (encod- polysaccharide (Sigma), or affinity-purified whole Ig goat anti-mouse ing the cytoplasmic tail), and exon IV (encoding the transmembrane IgM (Jackson Immune Research Laboratories). One microcurie of 3 domain) sequences of CD72 was replaced with the pGK-MC1-neo [ H]-thymidine was added for the last 6 hr of a 48 hr culture. Cells 3 cassette. Therefore, the neo cassette was flanked with 6.9 kb CD72 were harvested, and [ H]-thymidine uptake was measured. 5Ј homologous sequence and 5.3 kb 3Ј homologous sequence. The pGK-tk cassette was then inserted upstream of the 6.9 kb CD72 5Ј Immunizations and Serum Antibody Assays Ϫ/Ϫ flanking sequence. The linearized targeting plasmid DNA was used Wild-type and CD72 mice of 8 weeks of age with 100% 129 to transfect the E14.1 ES cells by electroporation. After double- genetic background were prebled and immunized intraperitoneally 28 selection with G418 and gancyclovir, 150 colonies were screened with either TNP-KLH (100 ␮g) (TNP -KLH, Biosearch Technologies), for homologous recombination of the CD72 targeted allele by South- with RIBI adjuvant (RIBI ImmunoChem Research), or with TNP-Ficoll ern blot analysis and PCR using one primer derived from the neo (20 ␮g) (Biosearch Technologies) in PBS. RIBI adjuvant was used sequence and the other primer derived from CD72 sequence down- according to manufacturer’s instructions. Mice were bled at day 7 stream of the 5.3 kb CD72 3Ј homologous sequence within the and day 14 after primary immunization with TNP-KLH and TNP- targeting plasmid. Two clones with homologous recombination were Ficoll. Mice were rechallenged with TNP-KLH with RIBI adjuvant at identified. day 21 and bled at day 28 and day 35 to assess the secondary response. Sera were collected, and the anti-TNP antibodies were determined by TNP- and isotype-specific ELISA. In brief, ELISA Generation, Screening, and Breeding of CD72-Deficient Mice plates were coated with TNP-BSA (10 ␮g/ml) (TNP31-BSA, Biosearch 10–15 ES cells from the two CD72 knockout clones were injected into Technologies) and blocked with BSA (1%) and Tween-20 (0.05%). 3.5-day blastocysts harvested from C57BL/6 mice. The blastocysts Sera were serially diluted in triplicate and detected with horseradish were transferred to pseudopregnant C57BL/6 ϫ CBA F1 foster peroxidase-conjugated rabbit anti-mouse isotype-specific antisera mothers. Male chimera were crossed with C57BL/6 mice, and germ- (Southern Biotechnology Associates). line transmission was scored by agouti coat color. Germline transmission of CD72-targeted allele was further assayed by Southern Calcium Mobilization Assay blot and PCR analysis. Mice with the CD72 null allele were either 2ϩ To monitor changes in [Ca ]i, splenocytes were resuspended at Ϫ/Ϫ bred with 129/SV/EV mice to establish a CD72 line with 100% 107 per ml in RPMI 1640 medium containing 5% FCS (fetal calf 129 background or backcrossed to C57BL/6 up to eight generations serum) and 10 mM HEPES, loaded with 1 ␮M indo-1/AM ester (Mo- Ϫ/Ϫ to obtain a CD72 line with C57BL/6 background. lecular Probes) for 45 min at 37ЊC, and stained with anti-FITC-CD3 (145-2C11, PharMingen) for the last 5 min at 37ЊC. Cells were washed Flow Cytometric Analysis and resuspended at 2 ϫ 106 per ml in the same RPMI medium. The Single-cell suspensions were prepared from mouse spleen, cervical, ratio of indo-1 violet/blue of CD3Ϫ B cells was analyzed by flow or mesenteric lymph nodes, and Peyer’s patches. Bone marrow cytometry using FACSTAR (Beckton Dickinson). Data were collected cells were obtained from one tibia and one femur. Peritoneal cavity for 45 sec to establish baseline violet/blue ratios; cells were then cells were collected by flushing the peritoneal cavity with PBS ϩ stimulated with anti-IgM F(abЈ)2, and data were collected for a total 2% FCS (fetal calf serum). One million cells of the various tissues of 5 min. tested were stained with the following antibodies: anti-CD3⑀ (145- 2C11), anti-CD4 (RM4-5), anti-CD8␣ (53-6.7), anti-CD5 (53-7.3), anti- Sublethal Irradiation of Mice CD43 (S7), anti-B220 (RA3-6B2), anti-CD19 (6D5), anti-CD21 (7G6), Wild-type and mutant mice on a C57BL/6 background were sub- anti-CD22 (Cy34.1), anti-CD23 (B3B4), anti-CD24 (HSA, M1-69), jected to 500 rad whole body irradiation. Self-reconstitution of pe- F(abЈ)2 anti-IgM (polyclonal goat anti-mouse), and anti-IgD (11-26) ripheral B cells was analyzed 13 and 15 days after irradiation. conjugated with FITC (fluorescein), R-PE (R-phycoerythrin), TR (Texas red), APC (allophycocyanin), or biotin. Streptavidin-TR or Acknowledgments streptavidin-PerCP were used as second-step reagents for biotinyl- ated antibodies. F(abЈ)2 anti-IgM-FITC or TR were purchased from The authors thank Richard Murray (DNAX) for kindly providing E14.1 Jackson Immune Research Laboratories, and anti-CD19 FITC, B220 ES cells and primary mouse embryonic fibroblasts and for help Multiple B Cell Defects in CD72-Deficient Mice 505

with the ES cell culture technique. We thank Julie Christiansen and immune system of Lyn-deficient mice, culminating in autoimmune Andreea Salinas (Stanford University) for performing blastocyst in- disease. Cell 83, 301–311. jection and Xiuling Xu for her assistance in knockout mice screening. Kantor, A.B., and Herzenberg, L.A. (1993). Origin of murine B cell The research was supported by National Institutes of Health (NIH) lineages. Annu. Rev. Immunol. 11, 501–538. grant CA68675 (to J. R. P.). C. P. was supported in part by NIH Khan, W.N., Alt, F.W., Gerstein, R.M., Malynn, B.A., Larsson, I., Rath- postdoctoral fellowships AI-07290 and AI-09462 and N. B. was sup- bun, G., Davidson, L., Mu¨ ller, S., Kantor, A.B., Herzenberg, L.A., et ported by NIH grant AI-34762 to L. A. Herzenberg. al. (1995). Defective B cell development and function in Btk-deficient mice. Immunity 3, 283–299. Received June 15, 1999; revised August 22, 1999. Laurindo, I., Finkelman, F.D., Mizuguchi, J., and Mond, J.J. (1987). B-cell activation stimulated by monoclonal anti-Lyb2.1 antibody is mediated through a receptor distinct from the BSF-1 receptor. Cell. References Immunol. 108, 473–482. Law, C.L., Sidorenko, S.P., Chandran, K.A., Zhao, Z., Shen, S.H., Adachi, T., Flaswinkel, H., Yakura, H., Reth, M., and Tsubata, T. Fischer, E.H., and Clark, E.A. (1996). CD22 associates with protein (1998). The B cell surface protein CD72 recruits the tyrosine phos- tyrosine phosphatase 1C, Syk, and phospholipase C-gamma(1) phatase SHP-1 upon tyrosine phosphorylation. J. Immunol. 160, upon B cell activation. J. Exp. Med. 183, 547–560. 4662–4665. Li, N., Batzer, A., Daly, R., Yajnik, V., Skolnik, E., Chardin, P., Bar- Ahearn, J.M., Fischer, M.B., Croix, D., Goerg, S., Ma, M., Xia, J., Sagi, D., Margolis, B., and Schlessinger, J. (1993). Guanine-nucleo- Zhou, X., Howard, R.G., Rothstein, T.L., and Carroll, M.C. (1996). tide-releasing factor hSos1 binds to Grb2 and links receptor tyrosine Disruption of the Cr2 locus results in a reduction in B-1a cells and kinases to Ras signaling. Nature 363, 85–88. in an impaired B cell response to T-dependent antigen. Immunity Nakayama, E., von Hoegen, I., and Parnes, J.R. (1989). Sequence 4, 251–262. of the Lyb-2 B-cell differentiation antigen defines a gene superfamily Allman, D.M., Ferguson, S.E., Lentz, V.M., and Cancro, M.P. (1993). of receptors with inverted membrane orientation. Proc. Natl. Acad. hi Peripheral B cell maturation. II. Heat-stable antigen splenic B cells Sci. USA 86, 1352–1356. are an immature developmental intermediate in the production of Nishizumi, H., Taniuchi, I., Yamanashi, Y., Kitamura, D., Ilic, D., Mori, long-lived marrow-derived B cells. J. Immunol. 151, 4431–4444. S., Watanabe, T., and Yamamoto, T. (1995). Impaired proliferation Butcher, E.C., Rouse, R.V., Coffman, R.L., Nottenburg, C.N., Hardy, of peripheral B cells and indication of autoimmune disease in lyn- R.R., and Weissman, I.L. (1982). Surface phenotype of Peyer’s patch deficient mice. Immunity 3, 549–560. germinal center cells: implications for the role of germinal centers Nitschke, L., Carsetti, R., Ocker, B., Ko¨ hler, G., and Lamers, M.C. in B cell differentiation. J. Immunol. 129, 2698–2707. (1997). CD22 is a negative regulator of B-cell receptor signaling. Campbell, M.A., and Klinman, N.R. (1995). Phosphotyrosine-depen- Curr. Biol. 7, 133–143. dent association between CD22 and protein tyrosine phosphatase Nomura, T., Han, H., Howard, M.C., Yagita, H., Yakura, H., Honjo, 1C. Scand. J. Immunol. 41, 563–569. T., and Tsubata, T. (1996). Antigen receptor-mediated B cell death Carsetti, R., Ko¨ hler, G., and Lamers, M.C. (1995). Transitional B cells is blocked by signaling via CD72 or treatment with dextran sulfate are the target of negative selection in the B cell compartment. J. and is defective in autoimmunity-prone mice. Int. Immunol. 8, Exp. Med. 181, 2129–2140. 867–875. Cyster, J.G., and Goodnow, C.C. (1995). Protein tyrosine phospha- O’Keefe, T.L., Williams, G.T., Davies, S.L., and Neuberger, M.S. tase 1C negatively regulates antigen receptor signaling in B lympho- (1996). Hyperresponsive B cells in CD22-deficient mice. Science cytes and determines thresholds for negative selection. Immunity 274, 798–801. 2, 13–24. Oliver, A.M., Martin, F., Gartland, G.L., Carter, R.H., and Kearney, Doody, G.M., Justement, L.B., Delibrias, C.C., Matthews, R.J., Lin, J.F. (1997). Marginal zone B cells exhibit unique activation, prolifera- J., Thomas, M.L., and Fearon, D.T. (1995). A role in B cell activation tive and immunoglobulin secretory responses. Eur. J. Immunol. 27, for CD22 and the protein tyrosine phosphatase SHP. Science 269, 2366–2374. 242–244. Osmond, D.G. (1991). Proliferation kinetics and the lifespan of B Egan, S.E., Giddings, B.W., Brooks, M.W., Buday, L., Sizeland, A.M., cells in central and peripheral lymphoid organs. Curr. Opin. Immunol. and Weinberg, R.A. (1993). Association of Sos Ras exchange protein 3, 179–185. with Grb2 is implicated in tyrosine kinase signal transduction and Otipoby, K.L., Andersson, K.B., Draves, K.E., Klaus, S.J., Farr, A.G., transformation. Nature 363, 83–85. Kerner, J.D., Perlmutter, R.M., Law, C.L., and Clark, E.A. (1996). Engel, P., Zhou, L.J., Ord, D.C., Sato, S., Koller, B., and Tedder, CD22 regulates thymus-independent responses and the lifespan of T.F. (1995). Abnormal B lymphocyte development, activation, and B cells. Nature 384, 634–637. differentiation in mice that lack or overexpress the CD19 signal Pezzutto, A., Dorken, B., Valentine, M.A., Shu, G.L., and Clark, E.A. transduction molecule. Immunity 3, 39–50. (1990). Activation and proliferation of human B-cells: effects of mAb Goodnow, C.C., Cyster, J.G., Hartley, S.B., Bell, S.E., Cooke, M.P., in the B-cell panel of the fourth international workshop. In Leukocyte Healy, J.I., Akkaraju, S., Rathmell, J.C., Pogue, S.L., and Shokat, Typing IV, W. Knapp et al., eds. (Oxford: Oxford University Press), K.P. (1995). Self-tolerance checkpoints in B lymphocyte develop- pp. 178–182. ment. Adv. Immunol. 59, 279–368. Polla, B.S., Ohara, J., Paul, W.E., Nabavi, N., Myer, A., Liou, H.C., Grupp, S.A., Harmony, J.A., Baluyut, A.R., and Subbarao, B. (1987). Shen, F.W., and Gillis (1988). Differential induction of class II gene Early events in B-cell activation: anti-Lyb2, but not BSF-1, induces expression in murine pre-B-cell lines by B-cell stimulatory factor-1 a phosphatidylinositol response in murine B cells. Cell. Immunol. and by antibodies to B-cell surface antigens. J. Mol. Cell. Immunol. 110, 131–139. 3, 363–373. Hardy, R.R. (1986). Purification and coupling of fluorescent proteins Rickert, R.C., Rajewsky, K., and Roes, J. (1995). Impairment of T-cell- for use in flow cytometry. In Handbook of Experimental Immunol- dependent B-cell responses and B-1 cell development in CD19- ogy., D.M. Weir, ed. (Oxford: Blackwell Scientific Publications), pp. deficient mice. Nature 376, 352–355. 31–37. Robinson, W.H., Landolfi, M.M., and Parnes, J.R. (1997). Allele-spe- Hardy, R.R., Carmack, C.E., Shinton, S.A., Kemp, J.D., and Haya- cific expression of the mouse B-cell surface protein CD72 on T cells. kawa, K. (1991). Resolution and characterization of pro-B and pre- Immunogenetics 45, 195–200. pro-B cell stages in normal mouse bone marrow. J. Exp. Med. 173, Roederer, M., Kantor, A.B., Parks, D.R., and Herzenberg, L.A. (1996). 1213–1225. Cy7PE and Cy7APC: bright new probes for immunofluorescence. Hibbs, M.L., Tarlinton, D.M., Armes, J., Grail, D., Hodgson, G., Mag- Cytometry 24, 191–197. litto, R., Stacker, S.A., and Dunn, A.R. (1995). Multiple defects in the Roederer, M., De Rosa, S., Gerstein, R., Anderson, M., Bigos, M., Immunity 506

Stovel, R., Nozaki, T., Parks, D., and Herzenberg, L.A. (1997). 8-color, Yakura, H., Kawabata, I., Ashida, T., and Katagiri, M. (1988). Differen- 10-parameter flow cytometry to elucidation of complex leukoctye tial regulation by Ly-5 and Lyb-2 of IgG production induced by heterogeneity. Cytometry 29, 328–329. lipopolysaccharide and B cell stimulatory factor-1 (IL-4). J. Immunol. Rose, M.L., Birbeck, M.S.C., Wallis, V.J., Forrester, J.A., and Davies, 141, 875–880. A.J.S. (1980). Peanut lectin binding properties of germinal centres Ying, H., Nakayama, E., Robinson, W.H., and Parnes, J.R. (1995). of mouse lymphoid tissue. Nature 284, 364–366. Structure of the mouse CD72 (Lyb-2) gene and its alternatively Sato, S., Steeber, D.A., and Tedder, T.F. (1995). The CD19 signal spliced transcripts. J. Immunol. 154, 2743–2752. transduction molecule is a response regulator of B-lymphocyte differentiation. Proc. Natl. Acad. Sci. USA 92, 11558–11562. Sato, S., Miller, A.S., Inaoki, M., Bock, C.B., Jansen, P.J., Tang, M.L., and Tedder, T.F. (1996a). CD22 is both a positive and negative regulator of B lymphocyte antigen receptor signal transduction: altered signaling in CD22-deficient mice. Immunity 5, 551–562. Sato, S., Ono, N., Steeber, D.A., Pisetsky, D.S., and Tedder, T.F. (1996b). CD19 regulates B lymphocyte signaling thresholds critical for the development of B-1 lineage cells and autoimmunity. J. Immu- nol. 157, 4371–4378. Sidman, C.L., Shultz, L.D., Hardy, R.R., Hayakawa, K., and Herzen- berg, L.A. (1986). Production of immunoglobulin isotypes by Ly-1ϩ B cells in viable motheaten and normal mice. Science 232, 1423–1425. Smith, K.G.C., Nossal, G.J.V., and Tarlinton, D.M. (1995). FAS is highly expressed in the germinal center but is not required for regulation of the B-cell response to antigen. Proc. Natl. Acad. Sci. USA 92, 11628–11632. Snow, E.C., Mond, J.J., and Subbarao, B. (1986). Enhancement by monoclonal anti-Lyb-2 antibody of antigen-specific B lymphocyte expansion stimulated by TNP-Ficoll and T lymphocyte-derived fac- tors. J. Immunol. 137, 1793–1796. Strasser, A., Whittingham, S., Vaux, D.L., Bath, M.L., Adams, J.M., Cory, S., and Harris, A.W. (1991). Enforced BCL2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease. Proc. Natl. Acad. Sci. USA 88, 8661–8665. Subbarao, B., and Mosier, D.E. (1982). Lyb antigens and their role in B lymphocyte activation. Immunol. Rev. 69, 81–97. Subbarao, B., and Mosier, D.E. (1984). Activation of B lymphocytes by monovalent anti-Lyb-2 antibodies. J. Exp. Med. 159, 1796–1801. Subbarao, B., Morris, J., and Baluyut, A.R. (1988). Properties of anti-Lyb-2-mediated B-cell activation and the relationship between Lyb-2 molecules and receptors for B-cell stimulatory factor-1 on murine B lymphocytes. Cell. Immunol. 112, 329–342. Tedder, T.F., Inaoki, M., and Sato, S. (1997). The CD19-CD21 complex regulates signal transduction thresholds governing humoral immunity and autoimmunity. Immunity 6, 107–118. Torres, R.M., Flaswinkel, H., Reth, M., and Rajewsky, K. (1996). Aberrant B cell development and immune response in mice with a compromised BCR complex. Science 272, 1804–1808. Tybulewicz, V.L., Crawford, C.E., Jackson, P.K., Bronson, R.T., and Mulligan, R.C. (1991). Neonatal lethality and lymphopeniain mice with a homozygous disruption of the c-abl proto-oncogene. Cell 65, 1153–1163. Venkataraman, C., Lu, P.J., Buhl, A.M., Chen, C.S., Cambier, J.C., and Bondada, S. (1998a). CD72-mediated B cell activation involves recruitment of CD19 and activation of phosphatidylinositol 3-kinase. Eur. J. Immunol. 28, 3003–3016. Venkataraman, C., Muthusamy, N., Muthukkumar, S., and Bondada, S. (1998b). Activation of lyn, blk, and btk but not syk in CD72- stimulated B lymphocytes. J. Immunol. 160, 3322–3329. Wang, J., Koizumi, T., and Watanabe, T. (1996). Altered antigen receptor signaling and impaired Fas-mediated apoptosis of B cells in Lyn-deficient mice. J. Exp. Med. 184, 831–838. Wu, Y., Nadler, M.J., Brennan, L.A., Gish, G.D., Timms, J.F., Fusaki, N., Jongstra-Bilen, J., Tada, N., Pawson, T., Wither, J., et al. (1998). The B-cell transmembrane protein CD72 binds to and is an in vivo substrate of the protein tyrosine phosphatase SHP-1. Curr. Biol. 8, 1009–1017. Yakura, H., Shen, F.W., Kaemmer, M., and Boyse, E.A. (1981). Lyb-2 system of mouse B cells. Evidence for a role in the generation of antibody-forming cells. J. Exp. Med. 153, 129–135. Yakura, H., Kawabata, I., Ashida, T., Shen, F.W., and Katagiri, M. (1986). A role for Lyb-2 in B cell activation mediated by a B cell stimulatory factor. J. Immunol. 137, 1475–1481.