key: cord-0252737-8t9nypah
authors: nan
title: Double isotype production by a neoplastic B cell line. II. Allelically excluded production of mu and gamma 1 heavy chains without CH gene rearrangement
date: 1986-08-01
journal: J Exp Med
DOI: nan
sha: 61f4cc62aa5a8b65d63664101a7e4af71a7c1df4
doc_id: 252737
cord_uid: 8t9nypah

In our accompanying paper, we described a switch variant (BCL1.2.58) that expresses membrane and secreted forms of IgM and IgG1. Both IgM and IgG1 share the same idiotype and use the same VDJ rearrangement. Here, a detailed Southern blot analysis of the entire constant region of the Ig heavy chain (Ig CH) locus of parental (BCL1.B1) and variants (BCL1.B2) DNA showed no detectable rearrangement. Similar analysis of the JH-C mu region led to the conclusion that two heavy chain alleles present in the IgM/IgG1-producing variants carried the same VDJ rearrangement but differed in their 3' flanking regions. One chromosome 12 did not carry any Ig CH genes, whereas, the other chromosome 12 carried one copy of CH genes. In BCL1.B1, however, each of the chromosome 12 alleles carried a full copy of CH genes. Karyotypic analysis confirmed the presence of two translocated t(12;16) chromosomes in both BCL1.2.58 and BCL1.B1 cells, with a break 5' to the VH locus at the distal region (12F2) of chromosome 12, and at the proximal region below the centromere (16B3) of chromosome 16. We conclude that double production of IgM and IgG1 in BCL1.B2 is accomplished by transcription of the corresponding CH genes in germline configuration using a single VDJ on the same chromosome 12.

transcript. Because of the induction of avid Fc receptors for IgE after infection of SJA/9 mice with such parasites, there is a strong possibility that the majority of the^-10% sIgE+ cells were not synthesizing IgE, but were absorbing it from the culture medium . This hypothesis is strongly supported by recent data of Katona et al . (19) . Studies by Perlmutter and Gilbert (18) also support the longtranscript model . Using a combination of cell sorting, Southern blotting, and sandwich RNA hybridization, they concluded that sIgG I' splenic cells have long RNA transcripts generated from unrearranged DNA that contain both -y and A sequences . However, as in the case of the report by Yaoita et al., their interpretations depend upon cell homogeneity . For example, they claimed to have successfully sorted 1-3% sIgG1 + cells to 99% purity, yet no data were shown to estimate the level of contaminating sIgM+ cells in their sIgGl' population .

The mechanism of double isotype production could give insights into the genetic regulation of Ig synthesis in differentiating B cells. To analyze such mechanisms requires a constitutively expressing clonal analogue. In our accompanying paper, we described a switch variant derived from the BCL,B, in vitro line in which each cell expresses both surface IgM and surface IgGI and also secretes both isotypes . The IgM and IgGI share the same idiotype and use the same VDJ rearrangement . In this paper, we present evidence, based on the C gene context of several independently isolated IgM/IgGI-producing clones (collectively termed BCL, .B2), that both A and -yl heavy chains are transcribed in germline DNA configuration from a single copy of chromosome 12.

Cell Lines and Subdoning. The derivation of the two BCL, .B, subclones, BCL,2 .62 (IgM producer) and BCL, .2.58 (IgM/IgGI producer), is described in the accompanying paper . BCL,2.54 is a sister subclone of BCL,2 .58 . For DNA analysis, we have employed additional independently derived IgM/IgGI subclones of the parental BCL, .B, line. The subclone BCL, . 13 .92 was I of the 12 BCL, .B2 subclones obtained at a frequency of 1 .35 x 10 -2 (12/888) by depositing one cell per well using a FACS III . The medium used for subdoning is the same as described in the accompanying paper. The subclone BCL, .6 .1 is a subclone of BCL, .2.58, derived in the same way. The subclones BCL,.15.9 and BCL, .15.19 were two of the seven BCL, .B2 subclones derived similarly at a frequency of 5 x 10' (7/1429), except that no WEHI 274 SN was added to the medium . As in the case of BCL,.2.58, all subclones except BCL, .6.1 secrete more IgM than IgGI and preliminary data of surface immunofluorescence indicate that all subclones express sIgM and sIgG. BCL,.6.1 secretes more IgGI than IgM (Y.-W . Chen et al., unpublished data).

Southern Hybridization Analysis . Procedures for genomic DNA isolation and blotting were described in the preceding paper. All washings were done in 3x SSC plus 0 .1 % SDS, three times at room temperature, followed by washing in O.lx SSC, 0 .1% SDS at 42°C for 1 h, with the exception of the pe-(see probes) probed filter, which was washed at 65°C. Restriction endonucleases were purchased from Boehringer Mannheim Biochemicals (Indianapolis, IN), New England Biolabs (Beverly, MA), and Bethesda Research Laboratories (Gaithersburg, MD) and used according to the manufacturer's instructions . Electrophoresis was carried out as described in the accompanying paper (in 1 % agarose at 40 V for 18 h).

Probes. Genomic or cDNA probes were prepared by gel isolation (20) of restriction fragments from previously characterized recombinant clones, except for pyl and pe where the whole plasmids were used. To insure that no repetitive sequences were present in the probes, particularly in fragments derived from introns, DNAs were digested with various enzymes and subjected to Southern analysis with "P-labeled mouse genomic DNA as a probe . Only nonhybridizing fragments were employed for further experiments .

Relative positions of the probes are shown in Figs . 2, A-R and 6, S-U . Ali DNAs were labeled by nick translation (21) to a specific activity of^-10 8 cpm/tag . Probe A : A 700-bp Pvu II/Pst I fragment of pu3741 (cDNA clone) (22) Karyotyping. Cells from the BCL,B, and BCL, .2 .58 lines were grown to exponential phase and cells were harvested for karyotypic G-banding analysis (32, 33) .

Densitometry . For band quantification, autoradiographs were scanned with a densitometer (model GS 300, Hoefer Instruments), and densities were calculated using a Bio-Rad model 33E2A integrator . C., s and C,2, (9 .2, 6.6, and 6 .4 kbp, respectively) in liver (lane 2), BCLI IgM producers (lanes 4 and 5) and BCL, .B 2 (lanes 3, 6-8) . Likewise, with the more upstream probes L, K and J (Fig . 4,b-d) , the BCLI and BCL I .B2 patterns are identical to those of the liver. Although no rearrangement occurred between y3 and y,, it still remained possible that expression of y l might be affected by a recombination between b and y3 . For example, a deletion within this region would appreciably reduce the length of a putative J r-Ca-Cy transcription unit . As summarized in Fig. 2 , we detected no rearrangements in BCL I .B, and BCLI .B 2 relative to liver within thiŝ -55 kbp area with the exception of only 7 .9 kbp of repetitive DNA (between fragment 4 and 7) that was difficult to confirm . Similarly, in the BCL I .B 2 cells, downstream genes from Cy 1 to Ca remain in germline context. Taken together, these results strongly argue that DNA rearrangement of CH genes is not responsible for expression of y, in BCL, . 1 and 4) , C,s (probe V) (lanes 3 and 6), and a mixture of probes S and V (lanes 2 and 5) . In Fig. 5c, lanes 2, 3, 5, and 6 , the 17 kbp and 7.5 kbp fragments were C,2n gene crosshybridizations with probe W. The 6.2 kbp fragment is the Bgl I fragment containing VDJ sequences located on the H' allele (see Fig. 6 ) . kb markers are shown. (Fig . 3) . Both lines contain two copies of the t(12 ;16) chromosome . But a single, apparently identical variable region rearrangement is observed when Eco RI blots of BCL I .2.58, BCL I .B I and BCL I .2.62 DNA are probed with a JH region fragment (probe S and Fig . 6 of accompanying paper) . In addition, a single VDJ H2-containing band was obtained in Xba I, Hind III, Bam HI and Bgl I digests of all the BCL I .B 2 subclones probed with JH (accompanying paper and data not shown) . These data suggested that each BCL I .B 2 allele possessed an identically rearranged VDJ . However, Kpn I digests revealed a rearrangement of one of the two BCL I .B2 alleles with respect to BCL I .B I . As shown in Fig . 5a 4-6, 8) , suggesting that all BCL I .B 2 lines, although independently cloned, were derived from a common parental variant . When a similar Kpn I blot was probed with Cu ( Fig . 5 b) , only a single hybridizing band, corresponding to the 17 .8-kbp fragment from the unrearranged allele of BCL1 .B 1 , was detected in BCL, .B 2. To map the rearrangement more precisely, we performed similar Kpn I blots with probes (T and U) that extend directly from the Eco RI site in the JH-Cp intron . In all cases, only the 17 .8-kbp fragment hybridized (data not shown) . Since there is only a single Eco RI band in J3J,,4 blots, the rearrangement had to occur just 3' (<50 bp) to the Eco RI site in the J,,-Cp intron . Otherwise, the 12 .5-kbp band would have been detected by probe T or U in BCL, .B 2.

Hybridization of the Kpn I digests to probe M had shown that the 12 .5-kbp fragment did not contain C,., (Fig . 2, fragment 18) . However, there are a number of Kpn I sites between C,.3 and C,, in germline DNA (25) , and there is a slight possibility of fragment comigration in the analysis of Fig. 4 . Therefore, we digested BCL, .B, and BCL 1 .B 2 DNAs with Bgl I, an enzyme that generates a fragment that spans the entire C,3-S,, germline region (Fig . 2, fragment 20a) , and cuts at a single site between JH and C, -480 by downstream of the Eco RI site of the rearrangement junction (P . W. Tucker, unpublished results) . As seen in Fig. 5c , hybridization either separately or jointly with JH (probe S) and C, 3 (probe V) gave different sized bands (30 and 32 kbp, respectively) . These data prove thatJH and C,., are not linked on either allele of the IgM/IgGI producing variants .

Simultaneous Expression of u and yl in BCL I.B2 is Allelically Excluded . The above results, taken with the analyses of the C genes, suggest that the observed rearrangement on one allele of BCL, .B 2 is unusual. It deletes, at a minimum, part of J-Cy intron, all Cy and, by direct inference of the Ckt-C6 linkage (Fig .  3) , Cb . However, the germline patterns (Figs. 2 and 4 ) and the analysis of Fig.  5 c show unequivocally that this rearrangement is not accomplished by recombination into downstream C sequences. Therefore, the 3' side of the rearrangement in BCL I .B 2 appears to be unrelated to Ig CH DNA, yet it is not large enough to be detected cytogenetically.

A proposed map of the two IgH alleles in BCL 1 .B 2 is shown in Fig. 6 . The model predicts that downstream C genes on the productive chromosome (i .e ., on the H+ allele) exist in single copy in BCL I .B 2 but in double copy in BCL1 .B1, BCL, .2 .62, or liver. On the other hand, J genes should exist in double copy in parental, variant, and germline DNA. To test this hypothesis, we hybridized Eco RI digests (Fig . 7a ) of the appropriate DNAs to a mixture of J,, (probe S) and C,-3 (probe V), and Hind III digests (Fig . 7b) to a mixture of J,, (probe S) and C,, (probe M) . The hybridization intensities were quantified by densitometry and the ratio ofJN/C, 3 (4 kbp:18 kbp, Fig. 7 a) and J /C,1 (2 .55 kbp:23 kbp, Fig. 7 b) were determined . As shown in Table I, expressed from a single rearranged VDJ segment on the same chromosome 12 . Therefore, we conclude that simultaneous production of IgM/IgGI in BCLI .B2, as in the case of IgM and IgD in BCL I .B l , is controlled exclusively at the RNA level.

Our findings provide the first example of allelically excluded, double isotype expression in a neoplastic B cell clone . A similar conclusion was drawn to explain dual u and ti2b production in a derivative of a 1A-producing Abelson murine leukemia-transformed cell line (40) . Contrary to the initial report, it is now clear (41, 42) that -y2b synthesis in this line is accompanied by deletion of sequences between JH and C,,26 , and thus cannot be explained by differential RNA processing. Our data also support the observations of Yaoita et al . (17) and Perlmutter and Gilbert (18) , who used purified populations of normal lymphocytes. The caveats associated with these studies (19) IgGI . Furthermore, the clonal BCL, .B2 lines have provided the opportunity for a more extensive analysis of the DNA context. In addition to establishing germline configuration of sequences anticipated to undergo rearrangement during switch recombination (e .g., S,, C, and S,,), we have also shown that the remainder of the C locus is unrearranged on the productive allele . This observation argues against the hypothesis that CH genes might be translocated, via nonclassical switch sequences and deletion of Cb, to a position downstream to C,, possibly replacing b with another isotype (7, 11) .

Although we have provided strong evidence for use of a single VDJ, it resides on a chromosome 12 that has undergone an unusual translocation, distal to the Ig locus, with chromosome 16 . At the microscopic level, the t(12;16) appears to be identically duplicated in both the BCL, .BI and BCL, .B2 cell lines. However, our blotting data define the nonproductive t(12;16) allele (H -in Fig. 6 ) by virtue of its loss of Ig-related sequences 3' to the Eco RI site within the JH-CA intron . These findings have important implications in considering both the derivation of the in vivo BCL, leukemia, its subsequent in vitro adaptation, and the eventual mechanism for allelic exclusion in BCLI .B2. The in vivo BCLI leukemia cells appear to have a stable karyotype, with the exception of chromosome 12 . Two early reports (34, 35) (35) found no evidence for normal copies ofchromosome 12, but identified three translocated alleles. We observed a single copy of the same t (12 ;16) in the in vivo line (Y .-W . Chen and G. V. Dev, unpublished results) that is duplicated in the in vitro BCL, .B, and BCL, .B2 lines, both of which have a modal chromosome number of --60 . We suggest that the precursor to the original leukemic cell carried a normal chromosome 12 as nonproductive and retained it during early passages, at which time the productive chromosome 12 underwent translocation(s) . Adaptation to cell culture resulted in a duplication of most of the karyotype, including t(12 ;16), and may have required a second more subtle event (e.g., somatic mutation), which was undetectable at the cytogenetic or DNA rearrangement level, to effect allelic exclusion in BCL, .B, . Alternatively, both of the t(12;16) alleles remain active in BCL, . If the former is true, then the third event, the rearrangement 3' to VDJ on the Hallele of BCL, .B2, is not required for allelic exclusion, but may instead reflect some trans-acting regulatory requirement for switching the productive allele from ,u/b to ,u/-y 1 . The significance of this DNA rearrangement is underscored by the fact that all independent BCL, .B2 variants maintain it.

Aside from its unusual karyotype, do the BCI., .B2 variants have a counterpart amongst normal B cells? Based on its surface Ig phenotype, its low level of constitutive IgM secretion (reviewed in reference 43), and its heavy chain gene transcription profile (44) , the parental BCL, .B, line appears to represent an immature B cell . It has been employed as a model for numerous functional studies by virtue of its ability to undergo further differentiation when cultured with LPS (45), anti-Ig (46), or T cell-derived lymphokines (47) . The fact that all BCL, .B2 variants isolated for IgG secretion secrete both IgM and IgGI, as well as express them on their surface, may reflect their derivation from a single precursor. It may also denote a basic difference between these cells and the nonsecreting memory cell populations analyzed by others (17, 18) at the molecular level. Regardless of whether these differences in mode of expression are real or are a consequence of the transformed state, the BCL, .B 2 phenotype shows that high expression of the secreted and the membrane form of y mRNA can be generated from C genes in germline configuration.

Finally, with regard to mechanisms, the only plausible explanation for the present data is that y and yl mRNA are expressed with a common VDJ gene by alternative RNA processing . This could be accomplished by a discontinuous mechanism in which the RNA polymerase translocates from the template to transcribe discrete segments of DNA, as proposed for trypanosome variable antigen (48) and certain viral genes (49, 50) . Alternatively, there could be a continuous transcription mechanism (i .e ., long transcript model), such as that apparently used by the drosophila bithorax locus (51) , which has been invoked by others (17, 18) to account for their results in normal B cells. In the latter scenario, the y 1 chains could then be translated from mRNA derived by processing a primary transcript that includes sequences from both Cp and Cy 1 genes. The t mRNA may be derived either from an identical transcript of -120 kb or from one that terminates in the intervening sequence between CU and CS . Based on nuclear transcription studies in BCL, .B, cells (44) and other IgM/IgD double producers (52, 53) , we favor the latter alternative . This would require differential regulation of transcriptional termination, perhaps at two points . First, the majority of the RNA polymerases would unload 3' to Cu, since in BCL, .B 2 cells, A RNA abundance is greater than yl RNA abundance (accompanying report and data not shown) . Second, a small proportion of the polymerases would read through the 1 termination region and exit the template downstream of the CH gene to be expressed (Cyl in BCL, .B 2). A choice in the second polymerase unloading event may dictate (and simplify) not only the subsequent RNA splicing decision, but may also bear on the acquisition of isotype commitment ; i.e ., once a cell expresses an isotype other than IgM and IgD, it is then committed and restricted to secretion of that isotype upon stimulation (54) . The best clonal example of this model is the 1 .29 B cell lymphoma, which undergoes spontaneous and/or inducible switch recombination from IgM to IgA (55) . The Ca gene is preferentially open, as defined by hypomethylation, in the IgM-synthesizing cells that are committed to switch to this isotype (56) . Similarly, in the transition of BCL1 to BCL, .B2 , Cyl appears to be exclusively selected . Perhaps transcription through, and equally important, termination beyond a C gene (C,, in BCL, .B 2) is critical in inducing an active chromatin structure for subsequent expression of that same C gene . The resulting prediction is that BCL, .B 2 cells, on receiving the appropriate signal, would undergo exclusive switch recombination to C,,, . This cell line, therefore, provides us with the opportunity of testing the mechanisms of double isotype production and its consequences .

In our accompanying paper, we described a switch variant (BCL I .2 .58) that expresses membrane and secreted forms of IgM and IgGI . Both IgM and IgGI share the same idiotype and use the same VDJ rearrangement . Here, a detailed Southern blot analysis of the entire constant region of the Ig heavy chain (Ig C ) locus of parental (BCL I .B I ) and variants (BCL I .B2 ) DNA showed no detectable rearrangement . Similar analysis of the JH-C/ region led to the conclusion that two heavy chain alleles present in the IgM/IgGI-producing variants carried the same VDJ rearrangement but differed in their 3' flanking regions . One chromosome 12 did not carry any Ig C F, genes, whereas, the other chromosome 12 carried one copy of C genes. In BCL I .B I , however, each of the chromosome 12 alleles carried a full copy of C genes . Karyotypic analysis confirmed the presence of two translocated t(12 ;16) chromosomes in both BCL I .2 .58 and BCL I .B I cells, with a break 5' to the V locus at the distal region (12F2) of chromosome 12, and at the proximal region below the centromere (16113) of chromosome 16. We conclude that double production of IgM and IgGI in BCL I .B 2 is accomplished by transcription of the corresponding C genes in germline configuration using a single VDJ on the same chromosome 12.

We thank Ms . M .-M . Liu, Ms . C. Das, Mr . J . Rybak and Ms . C. Glover-Humphries for excellent technical assistance and Ms . M . Gardner for secretarial assistance . We thank Dr. P. Howard-Peebles for assistance with the cytogenetics and Dr. D. Yuan for assistance with desitometry . We are grateful to Dr. T. Honjo for providing phages bearing Cy intronic sequences .

Immunoglobulin heavy-chain constant-region genes

Organization of immunoglobulin heavy chain genes and allelic deletion model

New views of the immunoglobulin heavy chain switch

Immunoglobulin class switching

Unusual sequences in the murine immunoglobulin u-6 heavy-chain region

Illegitimate recombination generates a class switch from CA to C6 in an IgD-secreting plasmacytoma

Simultaneous expression of immunoglobulin A and 6 heavy chains by a cloned B-cell lymphoma : a single copy of the V gene is shared by two adjacent C genes

Mouse immunoglobulin D : mRNA and genomic DNA sequences

Expression of IgD may use both DNA rearrangement and RNA splicing mechanisms

The role of DNA rearrangement and alternative RNA processing in the expression of immunoglobulin delta genes

Molecula r biology of IgD

Synthesis of multiple immunoglobulin classes by single lymphocytes . Cold Spring Harbor Symp

Stimulatio n of human B lymphocytes by antibodies to IgM and IgG : Functional evidence for the expression of IgG on B lymphocyte surface membranes

Th e relationship between surface immunoglobulin isotype and function of murine B lymphocytes . III . Expression of a single predominant isotype on primed and unprimed cells

B cell differentiation induced by lipopolysaccharide . IV . Development of immunoglobulin class restriction in precursors of IgG-synthesizing

Differentiation of B lineage cells from livers of neonatal mice : generation of immunoglobulin isotype diversity in vitro

Expression of lymphocyte surface IgE does not require switch recombination

Antibodies of the secondary response can be expressed without switch recombination in normal B cells

B cells that simultaneously express surface IgM and IgE in Nippostrongylus brasiliensis-infected SJA/9 mice do not provide evidence for isotype switching without gene deletion

Sequencing end-labeled DNA with base-specific chemical cleavages

Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA Polymerase I

Nuclea r transcripts of mouse heavy chain immunoglobulin genes contain only the expressed class of c-region sequences

Mouse immunoglobulin D : Construction and characterization of a cloned S chain cDNA

Mappin g of the heavy chain genes for mouse IgM and IgD

Organization of the constant-region gene family of the mouse immunoglobulin heavy chain

Structura l analysis of the murine IgG3 constant region gene

Cloning and complete nucleotide sequence of mouse immunoglobulin yl chain gene

Structure of the constant and 3' untranslated regions of the murine y2b heavy chain messenger RNA

Cloning of mouse immunoglobulin e gene and its location within the heavy chain gene cluster

Mous e IgA heavy chain gene sequence : Implications for evolution of immunoglobulin hinge exons

J genes for immunoglobulin heavy chains in mouse

Techniques for Chromosomal Analysis

A rapid banding technique for human chromosomes

Karyotypic differentiation in a spontaneous mouse B cell leukemia

Cytogenetic analysis of the spontaneous murine B cell leukemia (BCI )

Chromosoma l location of the structural gene cluster encoding murine immunoglobulin heavy chains

Chromosoma l location of structure genes encoding murine immunoglobulin a light chains : genetics of murine X light chains

Th e c-myc oncogene is translocated to the involved chromosome 12 in mouse plasmacytomas

A system of nomenclature for band patterns of mouse chromosomes

Immunoglobulin heavy chain class switching and lipopolysaccharide-dependent heavy chain expression in derivatives of an Abelson leukemia virus-transformed cell line

Immunoglobuli n heavy chain class switching in a pre-B cell line is accompanied by DNA rearrangement

Molecula r basis of heavy chain class switching and switch region deletion in an Abelson virus-transformed cell line

Th e activation of murine B cells : The role of surface immunoglobulins

Regulatio n of expression immunoglobulin M and D in murine B cells

An in vitro line of the B cell tumor BCL, can be activated by LPS to secrete IgM

Acquisition of cell surface IgD after in vitro culture of neoplastic B cells from the murine tumor BCL

Lymphokineinduced IgM secretion by clones of neoplastic B cells

Molecula r biology of trypanosome antigenic variation

Domain s in viroids : Evidence of intermolecular RNA rearrangements and their contribution to viroid evolution

Coronaviru s mRNA synthesis involves fusion of non-contiguous sequences

Th e molecular organization of the antennapedia locus of drosophila

Mode of regulation of immunoglobulin j-and 6-chain expression varies during B-lymphocyte maturation

Transcriptional regulation of the IA-6 heavy chain locus in normal murine B lymphocytes

Suppressio n of immunoglobulin G synthesis as a result of antibody-mediated suppression of immunoglobulin M synthesis in chickens

Induction of immunoglobulin isotype switching in cultured 1 .29 B lymphoma cells ; Characterization of the accompanying rearrangements of heavy chain genes

Immunoglobulin in heavy chain switching in cultured 1 .29 murine B lymphoma cells : Commitment to an IgA or IgE switch

Received for publication 27 February 1986 and in revised form 6 May 1986.Note added in proof. Results of sandwich hybridization and nuclear run-on transcription are consistent with the long transcript model for dual II. and yl expression in BCL I .B2 .