Portal de Periódicos da CAPES

The Immunoglobulin Class Switch: Beyond “Accessibility”

1997; Cell Press; Volume: 6; Issue: 3 Linguagem: Inglês

10.1016/s1074-7613(00)80324-6

1097-4180

Clifford M. Snapper, Kenneth B. Marcu, Piotr Zelazowski,

Immune Cell Function and Interaction

Induction of germline constant heavy (CH) gene transcription by cytokines, B cell activators, or both is a key event for targeting the CH gene for subsequent switch rearrangement. An increase in transcriptional activity may confer to the CH locus a state of enhanced accessibility for the binding of additional factors important for mediating the switching event. This regulatory paradigm is known as the "accessibility" model of immunoglobulin class switching. Here we discuss recent studies demonstrating the ability of interleukin-4 (IL-4), IL-5, IL-10, interferon-γ (IFNγ), membrane immunoglobulin–mediated activation, and the transcription factor, p50/NF-κB to mediate distinct and profound changes in immunoglobulin class switching in the absence of corresponding alterations in germline CH RNA expression. These studies suggest novel modes of regulation of the immunoglobulin class switch that may involve levels of accessibility distinct from transcriptional activation or alterations in the recombination machinery itself (or both). The immunoglobulin class switch is a process that is critical for the generation of functional diversity of a humoral immune response (for review see12Coffman R.L Lebman D.A Rothman P.B Adv. Immunol. 1993; 54: 229-270Crossref PubMed Scopus (454) Google Scholar, 26Harriman W Volk H Defranoux N Wabl M Annu. Rev. Immunol. 1993; 11: 361-384Crossref PubMed Scopus (97) Google Scholar, 76Snapper C.M Finkelman F.D Fundamental Immunology, Third Edition (New York: Raven Press), pp. 1993Google Scholar, 42Lorenz M Radbruch A Cytokine Regulation of Humoral Immunity, (Chichester, United Kingdom: John Wiley & Sons), pp. 1996Google Scholar, 79Stavnezer J Adv. Immunol. 1996; 61: 79-146Crossref PubMed Google Scholar). At the cellular level, switching is manifested by the transition from B cells expressing membrane immunoglobulin M (mIgM), IgD, or both to those expressing IgE, IgA, or one of four IgG subclasses. Immunoglobulin classes are encoded by CH genes aligned in tandem, 3′ to a rearranged VDJ gene that encodes for antigen specificity. 5′ to each CH gene, except CHδ, are blocks of tandemly repetitious DNA sequences, termed switch (S) regions. At the molecular level, the predominant mode of immunoglobulin class switching comprises a recombination event that includes looping out and deletion of all CH genes 5′ to the CH gene that is to be expressed (28Iwasato T Shimizu A Honjo T Yamagishi H Cell. 1990; 62: 143-149Abstract Full Text PDF PubMed Scopus (208) Google Scholar, 50Matsuoka M Yoshida K Maeda T Usuda S Sakano H Cell. 1990; 62: 135-142Abstract Full Text PDF PubMed Scopus (204) Google Scholar, 82von-Schwedler U Jack H.M Wabl M Nature. 1990; 345: 452-456Crossref PubMed Scopus (170) Google Scholar) (Figure 1). In a B cell that is initially mIgM+, this recombination event occurs between Sμ and the S region of the downstream CH gene. Sequential switching may also occur through the deletion mechanism (55Mills F.C Thyphronitis G Finkelman F.D Max E.E J. Immunol. 1992; 149: 1075-1085PubMed Google Scholar, 56Mills F.C Mitchell M.P Harindranath N Max E.E J. Immunol. 1995; 155: 3021-3036PubMed Google Scholar, 74Siebenkotten G Esser C Wabl M Radbruch A Eur. J. Immunol. 1992; 22: 1827-1834Crossref PubMed Scopus (87) Google Scholar, 48Mandler R Finkelman F.D Levine A.D Snapper C.M J. Immunol. 1993; 150: 407-418PubMed Google Scholar). Prior to switch rearrangement, populations of activated B cells typically express one or more characteristic RNA species encoded by different CH genes in the germline (80Stavnezer-Nordgren J Sirlin S EMBO J. 1986; 5: 95-102Crossref PubMed Scopus (287) Google Scholar, 85Yancopoulos G.D DePinho R.A Zimmerman K.A Lutzker S.G Rosenberg N Alt F.W EMBO J. 1986; 5: 3259-3266Crossref PubMed Scopus (181) Google Scholar). Germline CH RNAs are spliced products of distinct I exons, located 5′ to every S region, and the immediate 3′ CH gene (Figure 1). Transcription initiates within a promoter 5′ to the I exon, proceeds through the S region, and terminates at the 3′ end of the CH gene. Splicing out of the S region creates the final germline CH RNA (i.e., 5′ IHCH 3′) (45Lutzker S Alt F.W Mol. Cell. Biol. 1988; 8: 1849-1852Crossref PubMed Scopus (118) Google Scholar, 22Gaff C Gerondakis S Int. Immunol. 1990; 2: 1143-1148Crossref PubMed Scopus (17) Google Scholar, 23Gerondakis S Proc. Natl. Acad. Sci. USA. 1990; 87: 1581-1585Crossref PubMed Scopus (81) Google Scholar, 32Lebman D.A Nomura D.Y Coffman R.L Lee F.D Proc. Natl. Acad. Sci. USA. 1990; 87: 3962-3966Crossref PubMed Scopus (120) Google Scholar, 63Radcliffe G Lin Y.-C Julius M Marcu K Stavnezer J Mol. Cell. Biol. 1990; 10: 382-386PubMed Google Scholar, 67Rothman P Chen Y.-Y Lutzker S Li S.C Stewart V Coffman R Alt F.W Mol. Cell. Biol. 1990; 10: 1672-1679Crossref PubMed Google Scholar). Multiple chimeric germline CH transcripts (IμCε, IεCμ, IμCγ4, IγCμ, IγCε, IεCγ, and Iγ4Cα1) have also been demonstrated in IgD+ human B cells stimulated by IL-4 alone, in the apparent absence of either genomic switch recombinations or their circular by-products (21Fujieda S Lin Y.Q Saxon A Zhang K J. Immunol. 1996; 157: 3450-3459PubMed Google Scholar). The authors interpret their findings as evidence for trans-splicing of unrearranged germline immunoglobulin RNA transcripts. Additionally, spliced hybrid IμCγ transcripts can form subsequent to switch recombination through processing of a single RNA transcript containing the juxtaposed Iμ and Cγ regions and accumulate along with the respectively mature Cγ RNA (38Li S.C Rothman P.B Zhang J Chan C Hirsh D Alt F.W Int. Immunol. 1994; 6: 491-497Crossref PubMed Scopus (66) Google Scholar). These data are consistent with the model that simultaneous germline expression of both 5′ and 3′ CH genes (Cμ and Cγ) is involved in targeting their subsequent switch recombination. Germline CH RNAs lack VDJ-encoded sequence and hence cannot, by themselves, direct synthesis of intact immunoglobulin molecules. One group has shown that a germline CH transcript has the potential to encode a truncated Cμ protein in vitro (2Bachl J.C Truck W Wabl M Eur. J. Immunol. 1996; 26: 870-874Crossref PubMed Scopus (17) Google Scholar). However, the potential existence of germline CH RNA–encoded, truncated CH proteins in vivo remains controversial, as most I exons contain translation stop codons in each reading frame (34Lennon G.G Perry R.P Nature. 1985; 318: 475-478Crossref PubMed Scopus (149) Google Scholar). Upon switch rearrangement, the I and S regions are deleted and the relevant CH gene is brought into proximity with the VDJ gene (Figure 1). In this position it encodes a productive VDJ-CH RNA for synthesis of an intact immunoglobulin molecule expressing the same antigen specificity but a new immunoglobulin class. The immunoglobulin class switch is influenced in both a positive and negative manner by a number of cytokines and B cell activators. The mechanism for this appears to lie in part in the ability of cytokines and activators to regulate transcription selectively, which initiates upstream of each of the I exons. This effect is manifested by changes in the steady-state levels of the different germline CH RNA after activation. The selectivity of this process lies in the presence of unique sequences 5′ to each I exon for binding of a series of regulatory proteins whose expression within the B cell may be influenced by the activation conditions. However, the function of the I exon itself is unknown. Based upon precedents in other cellular systems, it has been proposed that the events that initiate germline CH transcription also confer a level of accessibility to the CH locus for the binding of additional regulatory elements that participate in switch recombination (i.e., the accessibility model of immunoglobulin class switching; 80Stavnezer-Nordgren J Sirlin S EMBO J. 1986; 5: 95-102Crossref PubMed Scopus (287) Google Scholar, 85Yancopoulos G.D DePinho R.A Zimmerman K.A Lutzker S.G Rosenberg N Alt F.W EMBO J. 1986; 5: 3259-3266Crossref PubMed Scopus (181) Google Scholar). Analyses of switch-recombinase substrates, either chromosomally integrated or in an episomal state, have engendered further support for the accessibility model by demonstrating a requirement for substrates to be transcriptionally active in order to undergo switch rearrangement (60Ott D.E Alt F.W Marcu K.B EMBO J. 1987; 6: 577-584Crossref PubMed Scopus (41) Google Scholar, 59Ott D.E Marcu K.B Int. Immunol. 1989; 1: 582-591Crossref PubMed Scopus (23) Google Scholar, 36Leung H Maizels N Proc. Natl. Acad. Sci. USA. 1992; 89: 4154-4158Crossref PubMed Scopus (64) Google Scholar, 37Leung H Maizels N Mol. Cell. Biol. 1994; 14: 1450-1458Crossref PubMed Scopus (37) Google Scholar, 3Ballantyne J Ozsvath L Bondarchuk K Marcu K.B Curr. Topics Microbiol. Immunol. 1995; 194: 439-448PubMed Google Scholar, 15Daniels G.A Lieber M.R Nucl. Acids Res. 1995; 23: 5006-5011Crossref PubMed Scopus (177) Google Scholar). Further, this transcriptional requirement is DNA-strand specific (64Reaban M.E Griffin J.A Nature. 1990; 348: 342-344Crossref PubMed Scopus (190) Google Scholar, 15Daniels G.A Lieber M.R Nucl. Acids Res. 1995; 23: 5006-5011Crossref PubMed Scopus (177) Google Scholar, 16Daniels G.S Lieber M.R Proc. Natl. Acad. Sci. USA. 1995; 92: 5625-5629Crossref PubMed Scopus (71) Google Scholar). Thus, a cellular switch-recombinase activity appears to operate on appropriate switch sequences even when they are present outside of their natural chromosomal contexts. The S sequence–specific-recombinase activity acting on chromosomally integrated switch substrates also appears to be B-cell specific (60Ott D.E Alt F.W Marcu K.B EMBO J. 1987; 6: 577-584Crossref PubMed Scopus (41) Google Scholar, 59Ott D.E Marcu K.B Int. Immunol. 1989; 1: 582-591Crossref PubMed Scopus (23) Google Scholar, 36Leung H Maizels N Proc. Natl. Acad. Sci. USA. 1992; 89: 4154-4158Crossref PubMed Scopus (64) Google Scholar, 37Leung H Maizels N Mol. Cell. Biol. 1994; 14: 1450-1458Crossref PubMed Scopus (37) Google Scholar, 3Ballantyne J Ozsvath L Bondarchuk K Marcu K.B Curr. Topics Microbiol. Immunol. 1995; 194: 439-448PubMed Google Scholar, 4Ballantyne J Henry D.L Marcu K.B Int. Immunol., in press. 1997; Google Scholar, 15Daniels G.A Lieber M.R Nucl. Acids Res. 1995; 23: 5006-5011Crossref PubMed Scopus (177) Google Scholar). One study demonstrated that Sγ and Sα substrate sequences could recombine with Sμ in the same pre-B cell line, implying that a common recombinase may exist for all S sequences (35Lepse C.L Kumar R Ganea D DNA Cell Biol. 1994; 13: 1151-1161Crossref PubMed Scopus (22) Google Scholar). In contrast, replicating extrachromosomal switch substrates (unlike their natural genomic counterparts or integrated, proviral substrates; 60Ott D.E Alt F.W Marcu K.B EMBO J. 1987; 6: 577-584Crossref PubMed Scopus (41) Google Scholar, 59Ott D.E Marcu K.B Int. Immunol. 1989; 1: 582-591Crossref PubMed Scopus (23) Google Scholar, 3Ballantyne J Ozsvath L Bondarchuk K Marcu K.B Curr. Topics Microbiol. Immunol. 1995; 194: 439-448PubMed Google Scholar, 4Ballantyne J Henry D.L Marcu K.B Int. Immunol., in press. 1997; Google Scholar) undergo significant recombination in inappropriate cell types and can also produce non-S sequence recombination products (36Leung H Maizels N Proc. Natl. Acad. Sci. USA. 1992; 89: 4154-4158Crossref PubMed Scopus (64) Google Scholar, 37Leung H Maizels N Mol. Cell. Biol. 1994; 14: 1450-1458Crossref PubMed Scopus (37) Google Scholar, 15Daniels G.A Lieber M.R Nucl. Acids Res. 1995; 23: 5006-5011Crossref PubMed Scopus (177) Google Scholar), implying that constraints of substrate specificity are more relaxed on such replicating episomal, recombination targets. Recent work with single copy, chromosomally integrated retroviral substrates indicates that recombinase activity is not only B-cell specific, but B cell–stage specific (restricted to cell lines representing late stage pre-B and mature B cells; 4Ballantyne J Henry D.L Marcu K.B Int. Immunol., in press. 1997; Google Scholar). Furthermore, the recombinase activity was found to act in a stochastic fashion upon constitutively accessible (i.e., transcribed) S region targets, in that switch recombinations were found to be chance events likely occurring with similar probability from one cell generation to the next (4Ballantyne J Henry D.L Marcu K.B Int. Immunol., in press. 1997; Google Scholar). Whether the switch-recombinase activity itself is subject to regulation by various B cell stimuli remains an open question. Two recent reports employing homologous gene targeting in mice have now shown that I regions strongly target class switching to specific S regions beyond their ability simply to induce transcription (7Bottaro A Lansford R Xu L Zhang J Rothman P Alt F.W EMBO J. 1994; 13: 665-674Crossref PubMed Scopus (185) Google Scholar, 43Lorenz M Jung S Radbruch A Science. 1995; 267: 1825-1828Crossref PubMed Scopus (207) Google Scholar). In one of these studies it was proposed that the spliced IHCH RNA may also be required (43Lorenz M Jung S Radbruch A Science. 1995; 267: 1825-1828Crossref PubMed Scopus (207) Google Scholar). 7Bottaro A Lansford R Xu L Zhang J Rothman P Alt F.W EMBO J. 1994; 13: 665-674Crossref PubMed Scopus (185) Google Scholar have shown that when the natural lipopolysaccharide (LPS)/IL-4- inducible Iε promoter and exon is replaced by an LPS-inducible Eμ enhancer/VH promoter, switching to IgE is inhibited, in cis, by 10- to 100-fold in spite of substantial transcription through the Sε region of the targeted allele. Therefore, optimal switching requires an intact I region or I region promoter in cis (or both) beyond their contributions to S region transcription. Lorenz, Jung, and Radbruch went on to report that the Iγ1 exon sequences encompassing the splice donor signal were necessary and sufficient for normal levels of switching to IgG1, implying that the spliced germline transcript or the spliceosome complex was in some way required for subsequent switch recombination. However, in this latter study the mutant γ1 locus that retained the I exon splice donor possessed much higher transcriptional activity than the switching-defective, mutant γ1 locus which lacked the same splice donor signal. Therefore, a more direct role of the processed germline transcript or its splicing intermediate in subsequent switch-recombination remains controversial. It has been suggested that transcription of S sequences prior to their recombination may produce an RNA:DNA triplex switch recombination substrate. Reaban and Griffin first demonstrated that the supercoiled state of Sα polypurine repeats is lost upon their transcription, but that the configuration of these repeats remains stable by virtue of DNA strand–specific RNA: DNA hybrid formation. It was suggested that such stabilized, triplex strands of RNA:DNA conformers would exist just prior to switch recombination (64Reaban M.E Griffin J.A Nature. 1990; 348: 342-344Crossref PubMed Scopus (190) Google Scholar). In vitro transcription of Sγ sequences has also recently been shown to produce strand-specific RNA:DNA hybrid structures, providing additional support for the idea of an RNA:DNA intermediate culminating in switch recombination (16Daniels G.S Lieber M.R Proc. Natl. Acad. Sci. USA. 1995; 92: 5625-5629Crossref PubMed Scopus (71) Google Scholar). Another group has independently hypothesized that such S sequence–specific RNA:DNA hybrids could indirectly enhance the overall recombinagenic activity of S regions by specifically facilitating the formation of nuclease-sensitive, short stem loop structures among the self-complementary S region tandem repeats (1Baar J Pennell N.M Schulman M.J J. Immunol. 1996; 157: 3430-3435PubMed Google Scholar). Over and above control of germline CH RNA expression by a promoter 5′ to the I exon, IgH locus transcriptional control in more general terms is orchestrated through the actions of an intronic enhancer (Eμ), located between JH and CHμ, and a developmentally regulated enhancer complex located 3′ of Cα (3′E) and consisting of four enhancers: α3′E, 3′E-HS1,2, 3′E-HS3, and 3′E-HS4 (17Dariavach P Williams G.T Campbell K Pettersson S Neuberger M.S Eur. J. Immunol. 1991; 21: 1499-1504Crossref PubMed Scopus (130) Google Scholar, 39Lieberson R Giannini S.L Birshtein B.K Eckhardt L.A Nucl. Acids Res. 1991; 19: 933-937Crossref PubMed Scopus (104) Google Scholar, 51Matthias P Baltimore D Mol. Cell. Biol. 1993; 13: 1547-1553Crossref PubMed Scopus (90) Google Scholar, 46Madisen L Groudine M Genes Dev. 1994; 8: 2212-2226Crossref PubMed Scopus (223) Google Scholar, 53Michaelson J.S Giannini S.L Birshstein B.K Nucl. Acids Res. 1995; 23: 975-981Crossref PubMed Scopus (92) Google Scholar). Eμ is implicated, through its early transcriptional activation of DJ-CHμ sequences, in making Sμ accessible for subsequent recombination with downstream S regions (25Gu H Zou Y.-R Rajewsky K Cell. 1993; 73: 1155-1164Abstract Full Text PDF PubMed Scopus (795) Google Scholar). One study employing episomal S sequence substrates observed that Eμ was even capable of stimulating switch recombination above its effects on transcriptional activity (37Leung H Maizels N Mol. Cell. Biol. 1994; 14: 1450-1458Crossref PubMed Scopus (37) Google Scholar). The evidence for a role of the 3′E region in germline CH transcription comes from homologous gene targeting in mice. Here, replacement of 3′E-HS1,2 by a heterologous neomycin gene expression cassette strongly repressed most germline CH RNAs (13Cogne M Lansford R Boaro A Zhang J Gorman J Young F Hwei-Ling C Alt F.W Cell. 1994; 77: 1-20Google Scholar). As anticipated from the accessibility model, this was associated with a marked reduction in switch rearrangements of the corresponding CH genes. However, the foreign neo expression cassette may have contributed to the latter effects by disturbing, perhaps through enhancer competition, the function of the other enhancers of the 3′E region or other yet to be defined regulatory elements. A study demonstrating that the two most distal enhancers of the 3′E may impart additional properties to the more proximal enhancers is consistent with the notion that the 3′E acts as a locus control region (46Madisen L Groudine M Genes Dev. 1994; 8: 2212-2226Crossref PubMed Scopus (223) Google Scholar). The 3′E region may selectively regulate transcription of germline CH RNAs, but the nature of the cross-talk between individual I exon–associated promoters and specific 3′E enhancers remains to be delineated. An extensive literature lends strong support for the accessibility model of immunoglobulin class switching. This is evidenced by numerous observations that switching to a particular immunoglobulin class is always preceded by the appearance of germline CH RNA corresponding to that class (12Coffman R.L Lebman D.A Rothman P.B Adv. Immunol. 1993; 54: 229-270Crossref PubMed Scopus (454) Google Scholar, 26Harriman W Volk H Defranoux N Wabl M Annu. Rev. Immunol. 1993; 11: 361-384Crossref PubMed Scopus (97) Google Scholar, 76Snapper C.M Finkelman F.D Fundamental Immunology, Third Edition (New York: Raven Press), pp. 1993Google Scholar, 42Lorenz M Radbruch A Cytokine Regulation of Humoral Immunity, (Chichester, United Kingdom: John Wiley & Sons), pp. 1996Google Scholar, 79Stavnezer J Adv. Immunol. 1996; 61: 79-146Crossref PubMed Google Scholar). Further, several studies have demonstrated a direct correlation between cytokine-mediated induction of germline CH RNA and enhanced transcriptional rate at the relevant CH locus (68Rothman P Li S.C Gorham B Glimcher L Alt F Boothby M Mol. Cell. Biol. 1991; 11: 5551-5561Crossref PubMed Google Scholar, 72Shockett P Stavnezer J J. Immunol. 1991; 147: 4374-4383PubMed Google Scholar, 43Lorenz M Jung S Radbruch A Science. 1995; 267: 1825-1828Crossref PubMed Scopus (207) Google Scholar), although a single report challenged this notion (33Lebman D.A Park. M.J Hansen-Bundy S Pandya A Int. Immunol. 1994; 6: 113-119Crossref PubMed Scopus (15) Google Scholar). Although germline CH RNA can be induced in a resting B cell, such cells will not undergo switch rearrangement unless they have also entered the cell cycle (69Severinson E Bergstedt-Lindqvist S van der Loo W Fernandez C Immunol. Rev. 1982; 67: 73-85Crossref PubMed Scopus (31) Google Scholar, 30Kenter A.L Watson J.V J. Immunol. Meth. 1987; 97: 111-117Crossref PubMed Scopus (23) Google Scholar, 44Lundgren M Strom L Bergquist L.-O Skog S Heiden T Stavnezer J Severinson E Eur. J. Immunol. 1995; 25: 2042-2051Crossref PubMed Scopus (47) Google Scholar). Several reports have suggested a direct role for DNA synthesis as an element or component of the switch rearrangement process (20Dunnick W Wilson M Stavnezer J Mol. Cell. Biol. 1989; 9: 1850-1856Crossref PubMed Scopus (73) Google Scholar, 19Dunnick W Stavnezer J Mol. Cell. Biol. 1990; 10: 397-400PubMed Google Scholar) and even germline CH RNA expression (44Lundgren M Strom L Bergquist L.-O Skog S Heiden T Stavnezer J Severinson E Eur. J. Immunol. 1995; 25: 2042-2051Crossref PubMed Scopus (47) Google Scholar), but the molecular mechanisms remain to be elucidated. Although DNA synthesis and germline CH gene transcription appear to be necessary for switch rearrangement to occur, a large body of evidence, utilizing normal and transformed murine B cells, now strongly suggests that these two processes are not sufficient. Indeed, a series of regulated events in cycling B cells has been observed that leads to striking changes in switch rearrangement without corresponding changes in germline CH RNA expression. These observations, discussed in detail below, imply that in addition to germline CH transcription, additional events under the control of a number of cytokines, B cell activators, and transcription factors (summarized in) or other targeting factors may also be central to controlling the immunoglobulin class switch. The identification of these regulated components represents a key challenge for investigators in the field. In most of the studies discussed below, three methodologies were utilized to assess immunoglobulin class switching: flow cytometric analysis of the percentage of B cells expressing various mIg isotypes consequent to in vitro activation; digestion circularization–polymerase chain reaction (DC–PCR) for quantitating specific switch rearrangement events at the DNA level (10Chu C.C Paul W.E Max E.E Proc. Natl. Acad. Sci. USA. 1992; 89: 6978-6982Crossref PubMed Scopus (82) Google Scholar); and semi-quantitative reverse transcription–PCR for comparative assessment of germline CH RNA expression. IL-4 is a switch factor for the IgG1 (27Isakson P.C Pure E Vitetta E.S Krammer P.H J. Exp. Med. 1982; 155: 734-748Crossref PubMed Scopus (338) Google Scholar, 69Severinson E Bergstedt-Lindqvist S van der Loo W Fernandez C Immunol. Rev. 1982; 67: 73-85Crossref PubMed Scopus (31) Google Scholar) and IgE isotypes (11Coffman R.L Carty J J. Immunol. 1986; 136: 949-954PubMed Google Scholar), and this property is associated with IL-4 induction of germline CHγ1 and CHε RNA (81Stavnezer J Radcliffe G Lin Y.-C Nietupski J Berggren L Sitia R Severinson E Proc. Natl. Acad. Sci. USA. 1988; 85: 7704-7708Crossref PubMed Scopus (243) Google Scholar, 66Rothman P Lutzker S Cook W Coffman R Alt F.W J. Exp. Med. 1988; 168: 2385-2389Crossref PubMed Scopus (174) Google Scholar). A role for IL-4 in switching to IgA was recently assessed using a B cell lymphoma model. CH12F3, a subclone of the CH12.LX B cell lymphoma which constitutively expresses germline CHα RNA, is induced to switch at a high frequency (∼50%) in response to the combination of CD40 ligand, IL-4, and transforming growth factor β (TGFβ) (58Nakamura M Kondo S Sugai M Nazarea M Imamura S Honjo T Int. Immunol. 1996; 8: 193-201Crossref PubMed Scopus (151) Google Scholar). IL-4 alone enhances switching to IgA (up to 12%) in CH12F3 cells without altering their steady-state levels of germline CHα RNA. In contrast, CD40 ligand or TGFβ alone stimulated an increase in the levels of germline CHα RNA, and this was associated in each case with an increase in switching to IgA (up to 12%). However, an additive effect on switching to IgA by CD40 ligand plus TGFβ (22%) is observed in the absence of an associated increase in germline CHα RNA levels above that seen with either agent acting alone. This latter finding suggested the possibility that, in addition to IL-4, CD40 ligand, TGFβ, or both may enhance Sμ–Sα rearrangement events independently of alterations in germline CHα transcription. IL-4 also augments switching to IgA in normal murine B cells activated in vitro with LPS or CD40 ligand plus IL-5, anti-Ig-dex, and TGFβ (52McIntyre T.M Kehry M.R Snapper C.M J. Immunol. 1995; 154: 3156-3161PubMed Google Scholar). Dextran-conjugated anti-immunoglobulin antibodies (anti-Ig-dex) stimulate resting B cells to synthesize DNA (9Brunswick M Finkelman F.D Highet P.F Inman J.K Dintzis H.M Mond J.J J. Immunol. 1988; 140: 3364-3372PubMed Google Scholar). In consideration of the accessibility model it was predicted that induction of germline CHγ1 transcription in response to the IL-4 IgG1 switch factor would be sufficient to trigger Sμ–Sγ1 rearrangement in B cells induced to enter the cell cycle and synthesize DNA upon anti-Ig-dex activation. However, despite IL-4 induction of germline CHγ1 RNA in anti-Ig-dex-activated B cells, no corresponding increase in the percentage of mIgG1+ cells or Sμ–Sγ1 rearrangement events was observed (49Mandler R Chu C.C Paul W.E Max E.E Snapper C.M J. Exp. Med. 1993; 178: 1577-1586Crossref PubMed Scopus (39) Google Scholar). Surprisingly, further addition of IL-5, a cytokine initially described for its ability to promote B cell differentiation to immunoglobulin secretion, strongly induced the appearance of mIgG1+ cells and Sμ–Sγ1 rearrangements. IL-5 mediated this effect without altering the levels of germline CHγ1 RNA and with only a modest enhancement in DNA synthesis. In the absence of IL-4-mediated targeting of the CHγ1 gene, IL-5 by itself fails to induce switching to IgG1 in anti-Ig-dex-activated cells. This suggested that IL-5 provides a critical signal that allows switching to IgG1 to occur in a cycling B cell with a transcriptionally active germline CHγ1 gene. Since IL-4 is required for germline CHγ1 RNA expression, another independent role(s) for this cytokine in promoting Sμ–Sγ1 rearrangement cannot be assessed in this system. IL-5 has also been found to promote switching to IgE, as well as IgG1, without altering the corresponding levels of germline CHε and CHγ1 RNA in anti-immunoglobulin-activated normal murine B lymphoblasts cultured in the presence of LPS (62Purkerson J.M Isakson P.C J. Exp. Med. 1992; 175: 973-982Crossref PubMed Scopus (93) Google Scholar). IL-10 strongly inhibits the generation of mIgA+ cells and Sμ–Sα rearrangements in normal murine B cells in response to combined activation with LPS, IL-4, IL-5, anti-Ig-dex, and TGFβ (73Shparago N Zelazowski P Jin L McIntyre T.M Stuber E Peçanha L.M.T Kehry M.R Mond J.J Max E.E Snapper C.M Int. Immunol. 1996; 8: 781-790Crossref PubMed Scopus (41) Google Scholar). This effect of IL-10 occurs in the absence of significant alterations in germline CHα RNA expression or substantial changes in DNA synthesis. In contrast, IL-10 augments the generation of mIgG3+ cells and Sμ–Sγ3 rearrangements in response to activation with LPS alone (73Shparago N Zelazowski P Jin L McIntyre T.M Stuber E Peçanha L.M.T Kehry M.R Mond J.J Max E.E Snapper C.M Int. Immunol. 1996; 8: 781-790Crossref PubMed Scopus (41) Google Scholar). The inductive effect of IL-10 on LPS-mediated switching to IgG3 is also not associated with significant alterations in germline CHγ3 RNA, although IL-10 inhibits LPS-mediated DNA synthesis. Thus, IL-10 selectively regulates immunoglobulin class switching in murine B cells in a manner that appears to be independent of changes in germline CH transcription. IL-10 also induces switching to IgG1 and IgG3 in CD40-activated naive human B cells (8Briere F Servet D.C Bridon J.M Saint R.J Banchereau J J. Exp. Med. 1994; 179: 757-762Crossref PubMed Scopus (290) Google Scholar, 47Malisan F Briere F Bridon J.M Harindranath N Mills F.C Max E.E Banchereau J Martinez V.H J. Exp. Med. 1996; 183: 937-947Crossref PubMed Scopus (139) Google Scholar), but the mechanism underlying this selectivity has not yet been reported. However, activation of human B cells with anti-CD40 antibody alone induces multiple germline CH RNAs, including IgG1 and IgG3 (29Jumper M.D Splawski J.B Lipsky P.E Meek K J. Immunol. 1994; 152: 438-445PubMed Google Scholar). This suggests that, as in the mouse, IL-10 may promote switching in human B cells in a manner that is independent of alterations in germline CH RNA expression. Alternatively, IL-10 may up-regulate the levels of germline CHγ1 and CHγ3 RNA above some threshold that is critical for switching to occur. IFNγ promotes switching to IgG2a in LPS- and anti-Ig-dex-activated B cells (75Snapper C.M Paul W.E Science. 1987; 236: 944-947Crossref PubMed Scopus (1564) Google Scholar, 77Snapper C.M McIntyre T.M Mandler R Peçanha L.M.T Finkelman F.D Lees A Mond J.J J. Exp. Med. 1992; 175: 1367-1371Crossref PubMed Scopus (218) Google Scholar). These effects of IFNγ are associated with a corresponding induction of germline CHγ2a RNA (70Severinson E Fernandez C Stavnezer J Eur. J. Immunol. 1990; 20: 1079-1084Crossref PubMed Scopus (112) Google Scholar, 14Collins J.T Dunnick W.A Int. Immunol. 1993; 5: 885-891Crossref PubMed Scopus (77) Google Scholar). IFNγ also induces switching to IgG3 in anti-Ig-dex-activated B cells and, likewise, this induction correlates with an increase in the levels of germline CHγ3 RNA (77Snapper C.M McIntyre T.M Mandler R Peçanha L.M.T Finkelman F.D Lees A Mond J.J J. Exp. Med. 1992; 175: 1367-1371Crossref PubMed Scopus (218) Google Scholar). In contrast, IFNγ inhibits LPS-mediated switching to IgG3 and IgG2b (75Snapper C.M Paul W.E Science. 1987; 236: 944-947Crossref PubMed Scopus (1564) Google Scholar). Whereas inhibition of switching to IgG2b is associated with a reduction in the levels of germline CHγ2b RNA, as anticipated by the accessibility model, IFNγ suppression of LPS-induced switching to IgG3 is not associated with a corresponding reduction in the levels of germline CHγ3 RNA (70Severinson E Fernandez C Stavnezer J Eur. J. Immunol. 1990; 20: 1079-1084Crossref PubMed Scopus (112) Google Scholar, 86Zelazowski P Collins J.T Dunnick W Snapper C.M J. Immunol. 1995; 154: 1223-1231PubMed Google Scholar). The ability of IFNγ to suppress switching to IgG3 in LPS-activated cells could either result from a direct inhibitory effect of IFNγ on Sμ–Sγ3 recombination or perhaps reflect sequential switching from IgM to IgG3 to IgG2a. Multivalent mIg cross-linking, using anti-Ig-dex, by itself induces a modest increase in the levels of germline CH RNA specific for IgG3, IgG1, and IgG2b in resting B cells (86Zelazowski P Collins J.T Dunnick W Snapper C.M J. Immunol. 1995; 154: 1223-1231PubMed Google Scholar). This is associated with the induction of small amounts of secreted IgG3 and IgG1, although not IgG2b, by anti-Ig-dex-activated B cells in the presence of IL-5. The additional capacity of anti-Ig-dex to inhibit switching to IgE in B cells activated with LPS plus IL-4 is associated with a marked reduction in the levels of germline CHε RNA, consistent with the accessibility model (61Peçanha L.M.T Yamaguchi H Lees A Noelle R.J Mond J.J Snapper C.M J. Immunol. 1993; 150: 2160-2168PubMed Google Scholar, 86Zelazowski P Collins J.T Dunnick W Snapper C.M J. Immunol. 1995; 154: 1223-1231PubMed Google Scholar). However, the 50% inhibition of LPS plus IL-4–mediated switching to IgG1 by anti-Ig-dex is associated with a greater than 7-fold increase in the levels of germline CHγ1 RNA (86Zelazowski P Collins J.T Dunnick W Snapper C.M J. Immunol. 1995; 154: 1223-1231PubMed Google Scholar). Hence, mIg cross-linking can selectively regulate immunoglobulin class switching in a manner that appears to be both dependent and independent of changes in CH gene transcription. The IgH locus contains multiple binding sites for the NF-κB/Rel family of transcription factors. In particular, p50-binding sites have been identified in Iγ3 (24Gerondakis S Gaff C Goodman D.J Grumont R.J Immunogenetics. 1991; 34: 392-400Crossref PubMed Scopus (27) Google Scholar), Iγ1 (40Lin S.-C Stavnezer J Mol. Cell. Biol. 1996; 16: 4591-4603Crossref PubMed Google Scholar), and Iε (18Delphin S Stavnezer J J. Exp. Med. 1995; 181: 181-192Crossref PubMed Scopus (187) Google Scholar), as well as in 3′E-HS1,2 and 3′E-HS4 (54Michaelson J.S Sing M Snapper C.M Sha W.C Baltimore D Birshtein B.K J. Immunol. 1996; 156: 2828-2839PubMed Google Scholar). In addition, p50-binding sites have been identified in Sγ3, Sγ1, and Sγ2b in proximity to sites of recombination within the switch region (83Wuerffel R.A Nathan A.T Kenter A.L Mol. Cell. Biol. 1990; 10: 1714-1718Crossref PubMed Scopus (45) Google Scholar, 84Wuerffel R Jamieson C.E Morgan L Merkulov G.V Sen R Kenter A.L J. Exp. Med. 1992; 176: 339-349Crossref PubMed Scopus (51) Google Scholar, 31Kenter A.L Wuerffel R Sen R Jamieson C.E Merkulov G.V J. Immunol. 1993; 151: 4718-4731PubMed Google Scholar). In this regard, B cells from mice made genetically deficient in p50/NF-κB (p50−/−) demonstrate selective defects in germline CH RNA expression and immunoglobulin class switching in vitro relative to wild-type controls (78Snapper C.M Zelazowski P Rosas F.R Kehry M.R Tian M Baltimore D Sha W.C J. Immunol. 1996; 156: 183-191PubMed Google Scholar). In p50−/− B cells, defective expression of CHγ3 and CHε RNA correlate with a corresponding inability to switch to IgG3 and IgE, as anticipated by the accessibility model. Likewise, normal expression of germline CHγ1 RNA by p50−/− B cells is associated with near normal levels of switching to IgG1. However, p50−/− B cells demonstrate a substantial defect in switching to IgA despite expressing normal levels of germline CHα RNA. These data suggest a role for p50/NF-κB in regulating immunoglobulin class switching that is both dependent and independent of alterations in germline CH transcription. The nuclear components responsible for effecting the switch recombination event are unknown. Potential candidates may include a variety of regulatory proteins that participate in DNA replication, repair, and/or recombination and whose expression could in theory be regulated. In this regard, a recent report compared the capacity of B cell precursors from RAG-2-deficient and SCID mice to undergo Sμ–Sε recombination in response to anti-CD40 monoclonal antibody plus IL-4 (65Rolink A Melchers F Andersson J Immunity. 1996; 5: 319-330Abstract Full Text Full Text PDF PubMed Scopus (371) Google Scholar). The RAG genes (RAG-1 and RAG-2) are key mediators of VDJ recombination (57Mombaerts P Iacomini J Johnson R.S Herrup K Tonegawa S Papaioannou V.E Cell. 1992; 68: 869-877Abstract Full Text PDF PubMed Scopus (2216) Google Scholar, 71Shinkai Y Rathbun G Lam K.P Oltz E.M Stewart V Mendelsohn M Charron J Datta M Young F Stall A.M Alt F.W Cell. 1992; 68: 855-867Abstract Full Text PDF PubMed Scopus (2093) Google Scholar). SCID mice lack the Ku70/Ku86-associated DNA-dependent serine/threonine protein kinase (DNA-PK; p350), which has also been implicated in VDJ recombination (6Bosma G.C Custer R.P Bosma M.J Nature. 1983; 301: 527-530Crossref PubMed Scopus (1744) Google Scholar, 5Blunt T Finnie N.J Taccioli G.E Smith G.C.M Demengeot J Gottlieb T.M Mizuta R Varghese A.J Alt F.W Jeggo P.A Jackson S.P Cell. 1995; 80: 813-823Abstract Full Text PDF PubMed Scopus (758) Google Scholar). B cell precursors from both the RAG-2 and SCID mice expressed germline CHε RNA in the presence of anti-CD40 monoclonal antibody plus IL-4 (65Rolink A Melchers F Andersson J Immunity. 1996; 5: 319-330Abstract Full Text Full Text PDF PubMed Scopus (371) Google Scholar). However, RAG-2-deficient, but not SCID, B cell precursors undergo Sμ–Sε recombination under these conditions. This suggests that DNA-PK, though not RAG-2, is a key mediator of switch rearrangement. If the level of expression of the functional Ku complex in B cells can be regulated during activation, this would represent one potential pathway for regulating Sμ–Sx recombination independent of alterations in germline CH transcription. Overall, the nature of the switch-recombinase activity and its targeting factors still remains largely unknown. That at least some of these parameters may be regulated within the B cell is suggested by the capacity of various cytokines, B cell activators, and transcription factors to effect changes in switch recombination without altering the expression of the corresponding germline CH genes. These model systems should help to elucidate the nature of these factors.