Tumour dormancy and cell signalling--III: Role of hypercrosslinking of IgM and CD40 on the induction of cell cycle arrest and apoptosis in B lymphoma cells

Polyclonal anti-IgM antibodies were more effective than monoclonal antibodies in inducing dormancy in SCID mice bearing a murine B lymphoma (BCL1). Under saturating conditions, both polyclonal and monoclonal anti-Ig antibodies induced cell cycle arrest (CCA) in both BCL1 cells and human B lymphoma cells (Daudi) but polyclonal antibodies were far more effective at inducing apoptosis. A mixture of several monoclonal antibodies specific for noncrossreactive epitopes on C mu mimicked the effects of a polyclonal anti-mu. Hypercrosslinking mIgM by a polyclonal antibody against the primary monoclonal anti-mu markedly increased apoptosis and CCA. Hence, the extent of crosslinking of IgM and the resultant singnalling may be a major factor in inducing and maintaining dormancy and in determining whether lymphoma cells respond by apoptosis or CCA. In contrast to mIgM, another B cell receptor, CD40, which induces CCA when crosslinked did not induce apoptosis after hypercrosslinking. The results are consistent with the hypothesis that aspects of the CCA and apoptotic pathways are independent. When anti-CD40 was added with anti-mu to Daudi cells, the proportion of cells undergoing apoptosis was increased.

Tumor dormancy and cell signaling. II. Antibody as an agonist in inducing dormancy of a B cell lymphoma in SCID mice

Tumor dormancy can be induced in a murine B cell lymphoma (BCL1) by immunizing BALB/c mice with the tumor immunoglobulin (Ig) before tumor cell challenge. In this report, we have investigated the immunological and cellular mechanisms underlying the induction of dormancy. BCL1 tumor cells were injected into SCID mice passively immunized with antibody against different epitopes on IgM or IgD with or without idiotype (Id)-immune T lymphocytes. Results indicate that antibody to IgM is sufficient to induce a state of dormancy. Antibodies against other cell surface molecules including IgD and CD44 (Pgp1) had no effect on tumor growth. Id-immune T cells by themselves also had no effect on tumor growth in SCID mice. However, simultaneous transfer of anti-Id and Id-immune T cells enhanced both the induction and duration of the dormant state. In vitro studies indicated that antibody to IgM induced apoptosis within several hours and cell cycle arrest by 24 h. Hyper cross-linking increased apoptosis. The Fc gamma RII receptor played little or no role in the negative signaling. Antibodies that did not negatively signal in vitro did not induce dormancy in vivo. The results suggest that anti-IgM plays a decisive role in inducing tumor dormancy to BCL1 by acting as an agonist of IgM-mediated signal transduction pathways.

Lyn tyrosine kinase signals cell cycle arrest but not apoptosis in B-lineage lymphoma cells

Signal transduction initiated by binding of antibodies to cell surface molecules can have an important impact on the growth of tumor cells. The malignant behavior of the murine lymphoma BCL1 can be suppressed and the neoplastic cells can be induced to enter a dormant state by in vivo ligation of membrane immunoglobulin. Anti-CD19 antibodies can prolong the survival of SCID mice challenged with the human Burkitt lymphoma cell line, Daudi. Here, we show that cross-linking of membrane immunoglobulin on both murine BCL1 and human Daudi cells initiates a cascade of signals leading to the induction of both apoptosis and cell cycle arrest in vitro. Using antisense oligonucleotides, we demonstrate that the immunoglobulin-associated Lyn tyrosine kinase is required for anti-immunoglobulin-mediated cell cycle arrest but is not required for the signal leading to apoptosis. These results define a branch point in the cytosolic signaling pathways mediating cell cycle arrest and apoptosis. In Daudi cells, Lyn is also critical for cell cycle arrest induced by anti-CD19 signaling. Thus, the Lyn tyrosine kinase may be an important mediator of cell cycle arrest in neoplastic B lymphocytes and, perhaps, other cell types.

Induction of B cell tumor dormancy by anti-idiotypic antibodies

Long-term dormancy of murine B-cell lymphomas can be experimentally induced by immunizing the host with the idiotype expressed on the tumor. Interaction of the cells with anti-idiotype antibodies is sufficient to induce and maintain the dormant state. The growth of lymphoma cells interacting with anti-idiotype antibodies is arrested and they undergo dramatic changes in their morphology, cell-cycle status and oncogene expression. Regrowth of a tumor after long-term dormancy results from the emergence of a tumor cell variant that no longer responds to the antibodies with growth inhibition. These data demonstrate the feasibility of reversing a malignant phenotype of cells by specific growth arrest signals and suggest new approaches for therapeutic intervention in cancer.

Cancer dormancy: isolation and characterization of dormant lymphoma cells

"Tumor dormancy" is an operational term used to describe a prolonged quiescent state in which tumor cells are present, but tumor progression is not clinically apparent. Although clinical examples of tumor dormancy abound, little is known regarding the mechanisms underlying this state. Here we utilize an antibody-induced dormancy model of an aggressive murine B-cell lymphoma (BCL1) and show that the induction of the dormant state is accompanied by dramatic changes in tumor cell morphology and cell cycle status. These data indicate the feasibility of altering the malignant phenotype of transformed cells by specific signals originating at the cell surface, and they suggest new opportunities for therapeutic intervention in cancer.

The tetrameric structure of NF-mu NR provides a mechanism for cooperative binding to the immunoglobulin heavy chain mu enhancer

We have analyzed the structure and DNA-binding characteristics of the IgH enhancer-binding regulatory protein NF-mu NR. We estimate that there are at least 5000 molecules of NF-mu NR protein/nucleus of non-B cells, based on the yield of active protein following purification. Purified NF-mu NR exists as a tetramer of 40-kDa subunits, even in the absence of its DNA-binding substrate. The protein complex binds to four binding sites flanking the immunoglobulin heavy chain mu enhancer core. Separately, individual sites bind to the tetrameric complex with affinities varying over a 100-fold range. However, when a low affinity site and a high affinity site are present on the same DNA molecule, both are occupied at the same NF-mu NR concentration; thus, there is a strong cooperative interaction between binding sites. The tetrameric structure provides a mechanism for binding cooperativity in which initial binding is mediated through a high affinity site on the DNA molecule followed by the engagement of the low affinity site juxtaposed to adjacent protein subunits. The presence of multiple binding sites flanking the mu enhancer core may reflect the influence of NF-mu NR binding on enhancer three-dimensional structure, transcription factor binding, and/or nuclear matrix interactions for cell type-specific enhancer function.

Loss of p53 expression in Myc-induced B lineage tumors

Tumors are formed following the accumulation of several genetic changes in genes which normally function to regulate cell growth. As yet it is unclear why multiple mutations are required, which type of alterations can collaborate with each other, and if collaboration is cell-type specific. In our myc transgenic mouse model system both point mutations and loss of mRNA expression for the p53 tumor suppressor gene have been found in the myc-induced B-lineage tumors arising spontaneously in these mice. This demonstrates the collaboration between these two growth control genes in cellular transformation. The observation that alterations in the expression of p53 is a common phenomenon in tumors formed in myc transgenic mice as well as a variety of different types of human tumors suggests that inactivation of the p53 growth control pathway may be required for transformation, and that alterations in p53 itself might be the most efficient way to achieve this inactivation. An analysis of the molecular mechanism for p53 alterations has implications for what kind of factors, both environmental and physiological, can influence tumor formation. The identification of collaboration groups has implications for the process of tumor formation, growth regulation, and will some day be important for the diagnosis of cancer, the prognosis of the individual and the design of specific therapeutic agents for treatment.

Anti-IgM antibodies down modulate mu-enhancer activity and OTF2 levels in LPS-stimulated mouse splenic B-cells

Stimulation of small, resting, splenic B cells with bacterial lipopolysaccharide (LPS) induces proliferation, differentiation to plasma cell formation, and the expression of immunoglobulin heavy chain (IgH). When this is combined with agents which crosslink surface Ig, differentiation and the induction of surface immunoglobulin are suppressed even though proliferation proceeds. We find that anti-mu antibodies suppresses Ig gene expression of transfected mu constructs, even if either the membrane or secretory segments have been deleted. We examined the effects of anti-mu treatment on the IgH enhancer (IgHE) attached to a heterologous test gene (CAT). Indeed the IgH enhancer alone was subject to anti-mu suppression, while the SV40 enhancer was insensitive. To determine what was responsible for suppression of enhancer function by anti-mu we examined nuclear extracts from stimulated splenic B cells for the presence of sequence-specific DNA binding activities to various sites within the enhancer. We found two specific differences--an induction in mu E5 binding activity, and a reduction in octamer transcription factor 2 (OTF2) binding activity, after anti-mu treatment. Analysis of these cells by in situ immunofluorescence with anti-OTF2 antibodies suggests that the nuclear localization of OTF2 in anti-mu treated cells may change, as well as its absolute level.

A developmental-specific factor binds to suppressor sites flanking the immunoglobulin heavy-chain enhancer

We identified a novel nuclear protein, NF-mu NR, that binds to multiple sites flanking the immunoglobulin heavy-chain enhancer. The expression of NF-mu NR shows a unique developmental pattern; the activity is present in all cells representing early stages of B-cell development, but is absent from more mature cells that express a high level of immunoglobulin heavy chains. NF-mu NR also is present in most cell lines outside of the B-cell lineage (e.g., T cells, macrophages, and fibroblasts). The binding sites for NF-mu NR correlate very well with cis-acting negative regulatory elements of the heavy-chain enhancer defined previously. Indeed, when the segments bound by NF-mu NR are deleted from the enhancer, it is now found to function as a positive transcription element in T cells and macrophages. Taken together, these results suggest that NF-mu NR may function as a negative regulator of enhancer function. The observation that the segments bound by NF-mu NR correspond to the segments bound to the nuclear matrix suggests an intriguing model not only of how enhancers might function but also of how negative regulation might occur.

UmuD mutagenesis protein of Escherichia coli: overproduction, purification, and cleavage by RecA

The mutation rate of Escherichia coli increases approximately 100-fold after treatment with replication-inhibiting agents such as UV light. This enhanced mutation rate requires the action of the UmuD and UmuC proteins, which are induced as part of the SOS response to DNA damage. To initiate a biochemical characterization of the role of these proteins, we have developed a plasmid system that gives efficient expression of the umuD and umuC genes. The umuD and umuC genes were placed under the control of a regulated phage lambda PL promoter and a synthetic ribosome-binding site, and the distance to the UmuD start was adjusted to maximize gene expression. Starting from this overproduction system, we have purified the UmuD protein and studied its interaction with RecA. The SOS response is turned on by the capacity of RecA protein to mediate cleavage of the LexA repressor for SOS-controlled operons. Others have shown that UmuD exhibits sequence homology to LexA around the cleavage site, suggesting a possible cleavage reaction for UmuD. We show that RecA mediates cleavage of UmuD, probably at this site. As with LexA, UmuD also undergoes a self-cleavage reaction. We infer that RecA-mediated cleavage of UmuD is another role for RecA in SOS mutagenesis, probably activating UmuD for its mutagenic function.

Capacity of RecA protein to bind preferentially to UV lesions and inhibit the editing subunit (epsilon) of DNA polymerase III: a possible mechanism for SOS-induced targeted mutagenesis

The RecA protein of Escherichia coli is required for SOS-induced mutagenesis in addition to its recombinational and regulatory roles. Most SOS-induced mutations probably occur during replication across a DNA lesion (targeted mutagenesis). We have suggested previously that RecA might participate in targeted mutagenesis by binding preferentially to the site of the DNA damage (e.g., pyrimidine dimer) because of its partially unwound character; DNA polymerase III (polIII) will then encounter RecA-coated DNA at the lesion and might replicate across the damaged site with reduced fidelity. In this report, we analyze at a biochemical level two major predictions of this model. With respect to lesion recognition, we show that purified RecA protein binds more efficiently to UV-irradiated double-stranded DNA than to nonirradiated DNA, as judged by filter-binding and gel electrophoresis assays. With respect to replication fidelity, Fersht and Knill-Jones [Fersht, A. R. & Knill-Jones, J. W. (1983) J. Mol. Biol. 165, 669-682] have found that RecA inhibits the 3'----5' exonuclease (editing function) of polIII holoenzyme. We extend this observation by demonstrating that RecA inhibits the exonuclease of the purified editing subunit of polIII, epsilon protein. Thus, we suggest that the activities of RecA required for targeted mutagenesis are lesion-recognition, followed by localized inhibition of the editing capacity of the epsilon subunit of polIII holoenzme. In this proposed mechanism, one activation signal for RecA for mutagenesis is the lesion itself. Because UV-irradiated, double-stranded DNA efficiently activates RecA for cleavage of the LexA repressor, the lesion itself may also often serve as an activation signal for induction of SOS-controlled genes.

A separate editing exonuclease for DNA replication: the epsilon subunit of Escherichia coli DNA polymerase III holoenzyme

DNA polymerase III (polIII) holoenzyme of Escherichia coli has 3'----5' exonuclease ("editing") activity in addition to its polymerase activity, a property shared by other prokaryotic DNA polymerases. The polymerization activity is carried by the large alpha subunit, the product of the dnaE gene. Mutations affecting the fidelity of DNA replication in vivo and the activity of 3'----5' exonuclease assayed in vitro are found in the dnaQ gene, which specifies the epsilon subunit. To determine whether epsilon carries the 3'----5' exonuclease activity, we have used an overproduction protocol to purify epsilon separately from the other subunits of polIII holoenzyme. We find that epsilon has 3'----5' exonuclease activity indistinguishable from that of polIII core, the subassembly of polIII holoenzyme consisting of the alpha, epsilon, and theta subunits. We conclude that the editing and polymerization activities of polIII holoenzyme reside on distinct subunits, in contrast to DNA polymerase I of E. coli and DNA polymerase of phage T4. This functional separation may provide for regulation of exonucleolytic editing independently of polymerization, allowing cellular control of replication fidelity.