The mechanism of osmotic transfection of avian embryonic erythrocytes: analysis of a system for studying developmental gene expression.

We have undertaken a study of the mechanism of DNA transfer into primary chicken erythrocytes by a method named osmotic transfection. The cells are subjected to controlled osmotic swelling in NH4Cl and then ruptured in a lower osmotic strength solution containing DNA and DEAE-dextran. The osmotic rupture results in transient formation of a single hole in the cell membrane, which is followed within hours by recovery of near normal levels of RNA and protein synthesis. The association of DNA with the cells is much greater for ruptured than for unruptured cells or for cells that have been lysed and resealed before DNA is added. Transient formation of pores in the cell membrane is apparently essential for high rates of macromolecular transfer into the cell. DEAE-dextran increases the amount of DNA associated with the cells, especially after cell rupture. Our understanding of the mechanism has allowed us to extend the application of osmotic transfection to essentially all developmental stages of avian erythroid differentiation. Osmotic transfections were done with plasmids containing the chloramphenicol acetyl transferase (cat) gene placed between the chicken beta-globin promoter and the 3' beta-globin enhancer. The pattern of CAT expression at sequential developmental stages parallels that of the endogenous gene, showing that osmotically transfected cells appear to retain developmental fidelity. The approach provides a convenient, sensitive, and flexible system for the study of transient gene expression as a function of development.

Increase in double-stranded DNA break-related foci in early-stage thymocytes of aged mice.

Cellular and molecular mechanisms involved in aging are notoriously complex. Aging-related immune decline of T lymphocyte function is partly caused by attrition of thymic T cell development, which involves programmed creation and repair of DNA breaks for generating T cell receptors. Aging also leads to significant alterations in the cellular DNA repair ability. We show that higher levels of gamma-phosphorylated H2AX (pH2AX), which marks DNA double-stranded breaks (DSBs), were detectable in early thymocyte subsets of aged as compared to young mice. Also, while only 1-2 foci of nuclear accumulation of pH2AX were detectable in these early thymocytes from young mice, cells from aged mice showed higher numbers of pH2AX foci. In CD4-CD8- double-negative (DN) thymocytes of aged mice, which showed the highest levels of DSBs, there was a modest increase in levels of the DNA repair protein MRE11, but not of either Ku70, another DNA repair protein, or the cell cycle checkpoint protein p53. Thus, immature thymocytes in aged mice show a marked increase in DNA DSBs with only a modest enhancement of repair processes, and the resultant cell cycle block could contribute to aging-related defects of T cell development.

Cutting and closing without recombination in V(D)J joining.

Open and shut junctions are rare (V(D)J joining products in which site-specific recognition, cleavage and re-ligation of joining signals has been uncoupled from recombination. Here, we investigate the relationship of opening and shutting to recombination in two ways. First, we have tested a series of substrates containing one or two joining signals in an in vivo assay. Opening and shutting can be readily observed in substrates that have only one consensus joining signal. Thus, unlike recombination, the majority of open and shut events do not require interactions between two canonical joining signals. Next we examined two-signal substrates to investigate the effect of signal proximity on the frequency of dual open and shut events. These experiments indicate that at least some of the time opening and shutting can be a two-signal transaction. Together these results point to two mechanistically related, but distinct origins for open and shut joining events. In one case, cutting and closing may occur without interaction between two signals. In the other, we suggest that interaction of a canonical signal with 'cryptic' signal-like elements whose sequence is extensively diverged from canonical signals, may bias the V(D)J recombination machinery towards opening and shutting rather than recombination. Open and shut operations could in this way provide a means whereby mistakes in target recognition by the V(D)J recombination machinery produce a non-recombinant outcome, avoiding deleterious chromosomal rearrangements in lymphoid tissues.

Definition of a core region of RAG-2 that is functional in V(D)J recombination.

The products of the RAG-1 and RAG-2 genes cooperate to allow V(D)J recombination in lymphoid and non-lymphoid cells. As one step toward understanding the role of RAG-2, we have constructed mutated RAG-2 genes and examined their ability to support recombination of plasmid substrates in a fibroblast cell line. The mutations define essential and dispensable parts of the RAG-2 gene. Mutations in the N-terminal part eliminate almost all activity. In the central region of the protein, some but not all local alterations still allow recombination. On the other hand, proteins with large deletions from the C-terminal end, including one truncated by 25%, still retain activity, even though this part of the protein is highly conserved between species. Similar results were obtained with substrates that retain either a signal joint or a coding joint, or perform an inversion. Thus all basic features of V(D)J joining are retained in a RAG-2 protein with only the first 75% of the sequence.

DNA binding of Xrcc4 protein is associated with V(D)J recombination but not with stimulation of DNA ligase IV activity.

Mammalian cells are protected from the effects of DNA double-strand breaks by end-joining repair. Cells lacking the Xrcc4 protein are hypersensitive to agents that induce DNA double-strand breaks, and are unable to complete V(D)J recombination. The residual repair of broken DNA ends in XRCC4-deficient cells requires short sequence homologies, thus possibly implicating Xrcc4 in end alignment. We show that Xrcc4 binds DNA, and prefers DNA with nicks or broken ends. Xrcc4 also binds to DNA ligase IV and enhances its joining activity. This stimulatory effect is shown to occur at the adenylation of the enzyme. DNA binding of Xrcc4 is correlated with its complementation of the V(D)J recombination defects in XRCC4-deficient cells, but is not required for stimulation of DNA ligase IV. Thus, the ability of Xrcc4 to bind to DNA suggests functions independent of DNA ligase IV.

Identification and radiochemical purification of the recA protein of Escherichia coli K-12.

The product of the recA gene of E. coli has been identified by labeling proteins synthesized in UV-treated cells after infection with specialized transducing phages carrying the recA gene. Following infection of UV-treated cells by lambda precA, which carries the recA+ gene, a major protein with a molecular weight of 43,000 is detected on polyacrylamide gels containing sodium dodecyl sulfate. This protein is also made after infection of suppressing hosts by lambda precA99, which carries an amber recA- mutation, but is not synthesized after infection of nonsuppressing hosts by this transducing phage. A spontaneous recatrevertant of lambda preca99 induces synthesis of this protein after infection of a nonsuppressing host. The product of the recA gene is a soluble protein found in a complex with a molecular weight of approximately 150,000 after mild detergent lysis of cells.

Role of cyclic adenosine 3',5'-monophosphate in the in vivo expression of the galactose operon of Escherichia coli.

Studies of levels of galactokinase in Escherichia coli with mutations affecting synthesis of, or response to, cyclic adenosine 3',5'-monophosphate show that this nucleotide does not play a major role in expression of the galactose operon, causing at most a twofold stimulation. The discrepancy between our in vivo results and the marked stimulation by cyclic adenosine 3',5'-monophosphate in in vitro systems indicates that current cell-free systems lack a factor which allows efficient expression of the galactose operon even in the absence of cyclic adenosine 3',5'-monophosphate or of the binding protein for this nucleotide.

Novel strand exchanges in V(D)J recombination.

We describe novel products of V(D)J recombination in which signal sequences become joined to coding elements, in contrast to the standard reaction whose products are junctions of two signal sequences or two coding elements. In this variant reaction, the recombination machinery evidently recognizes signal sequences and introduces strand breaks at the normal positions, but then connects the elements in unusual combinations. The lack of fixed directionality indicates that recombination sites are not uniquely aligned when strand exchange occurs. The discovery of these variant junctions suggests a model for the evolution of the antigen receptor loci.

Developmental stage specificity of the lymphoid V(D)J recombination activity.

We have examined the level of immunoglobulin gene V(D)J recombination activity in a number of cell lines derived from lymphoid or nonlymphoid lineages. The assay we employed uses extrachromosomal DNA as substrate and thereby avoids difficulties associated with the use of chromosomally integrated substrates. The recombination activity decreases during B-lymphoid development. It is highest at the earliest stages of committed B-cell differentiation and then falls progressively, reaching undetectable levels at the mature B-cell stage. The activity is also present in multipotential progenitors of myeloid cells and in pre-T cells but not mature T cells. No activity was found in several nonhematopoietic cell lines. Recombination was seen only among substrate molecules which had replicated in the eukaryotic cells. Several possible interpretations of this result are discussed.

Extrachromosomal DNA substrates in pre-B cells undergo inversion or deletion at immunoglobulin V-(D)-J joining signals.

Sequences encoding immunoglobulin variable domains are known to be assembled from variable (V), diversity (D), and joining (J) segments by site-specific recombination. We present a sensitive and rapid assay for V-(D)-J recombination that uses plasmid DNA transiently introduced into transformed pre-B cells, and demonstrates that the recombination is independent of any unique chromosomal context. Sequences sufficient to constitute recombination sites are contained within the 84 and 42 bp flanking, respectively, the murine J kappa 1 and V kappa L8 segments, which include the known heptamer-nonamer V-(D)-J joining signals. Deletion and inversion occur at comparable frequencies. Thus, V-(D)-J recombination may be relatively insensitive to the topological arrangement of sites, and events at the two novel junctions produced by the reaction may be coupled.

Lymphoid V(D)J recombination: nucleotide insertion at signal joints as well as coding joints.

The coding regions of antigen receptor genes assembled by variable-diversity-joining region [V(D)J] recombination are known in many cases to have undergone deletions of several nucleotides and also to contain insertions of noncoded nucleotides at the recombined junction (the coding joint). By using extrachromosomal recombination substrates to transfect lymphoid cell lines, we show that the signal joint (the fusion of the corresponding recognition signal sequences) can also contain insertions; however, nucleotide loss from the signals is very rare. The frequency of nucleotide addition varies among pre-B-cell lines in a manner proportional to their content of terminal deoxynucleotidyltransferase. We also find frequent nucleotide additions (and deletions) at coding joints, but in this case there is no strong correlation with the level of terminal deoxynucleotidyltransferase activity. Inserts at both signal and coding joints are rich in G + C, consistent with the base utilization preference of this enzyme.

Regulated gene expression in transfected primary chicken erythrocytes.

We describe a method for studying transient gene expression in primary avian erythroid cells that involves controlled osmotic shock, followed by DNA transfection using DEAE-dextran. Cells treated in this way reproducibly express high levels of chloramphenicol acetyltransferase (CAT) when transfected with a plasmid having the cat gene coupled to an appropriate viral promoter. An observed correlation between levels of CAT expression and extent of hemoglobin release during controlled shock makes it possible to choose optimum conditions for expression in erythroid cells at various stages of embryonic development. Using these techniques, we have investigated the effect on CAT expression of fusing to the cat gene various portions of the chicken adult beta-globin (beta A) gene. We show that in 9-day or 12-day embryonic erythrocytes, the promoter activity of the 5' flanking region of the beta A gene (in the absence of any viral promoters) is strongly stimulated by a downstream sequence, located in the region 110-588 base pairs on the 3' side of the poly(A) signal, that acts as an enhancer. Its activity is reduced in 5-day embryonic cells and absent in primary chicken fibroblasts and mouse L cells, suggesting that this transient expression system will be useful in studying developmentally regulated globin gene expression.

Sequence homology between two membrane transport ATPases, the Kdp-ATPase of Escherichia coli and the Ca2+-ATPase of sarcoplasmic reticulum.

We have determined the DNA sequences of the genes encoding the three structural proteins of the Kdp-ATPase, an ATP-driven potassium transport system of Escherichia coli. Regions of the predicted amino acid sequence of KdpB, the phosphorylated protein of the system, are homologous to regions of the Ca2+-ATPase of rabbit sarcoplasmic reticulum. The phosphorylated aspartate residue of the latter is within a region of homology.

V(D)J recombination: a functional definition of the joining signals.

Two conserved DNA sequences serve as joining signals in the assembly of immunoglobulins and T-cell receptors from V-, (D)-, and J-coding segments during lymphoid differentiation. We have examined V(D)J recombination as a function of joining signal sequence. Plasmid substrates with mutations in one or both of the heptamer-spacer-nonamer sequences were tested for recombination in a pre-B-cell line active in V(D)J recombination. No signal variant recombines more efficiently than the consensus forms of the joining signals. We find the heptamer sequence to be the most important; specifically, the three bases closest to the recombination crossover site are critical. The nonamer is not as rigidly defined, and it is not important to maintain the five consecutive As that distinguish the consensus nonamer sequence. Both types of signals display very similar sequence requirements and have in common an intolerance for changes in spacer length greater than 1 bp. Although the two signal types share sequence motifs, we find no evidence of a role in recombination for homology between the signals, suggesting that they serve primarily as protein recognition and binding sites.

Expression and V(D)J recombination activity of mutated RAG-1 proteins.

The products of the RAG-1 and RAG-2 genes are essential for the recombination of the DNA encoding the antigen receptors of the developing immune system. Little is known of the specific role these genes play. We have explored the sequences encoding mouse RAG-1 by deleting large parts of the gene and by introducing local sequence changes. We find that a RAG-1 gene with 40% of the coding region deleted still retains its recombination function. In addition, a series of small deletions within the strongly conserved remaining 60% of the coding region was tested. Nine out of ten of these prove unable to provide RAG-1 activity, but one is quite active. Certain peptide sequences were also specifically targeted for mutagenesis. The RAG-1 protein generated from this expression system is transported to the nucleus and is degraded with a 15 minute half-life. The fate of the proteins made by the deletion mutants were also assessed. Transport of RAG-1 protein to the nucleus was found even with the most extensive deletions studied. The functionality of the deleted proteins is discussed with relation to an alignment of RAG-1 sequences from five animal species.

RAG-1 mutations that affect the target specificity of V(D)j recombination: a possible direct role of RAG-1 in site recognition.

The RAG-1 protein plays an essential role in V(D)j recombination, but its exact function has not yet been defined. Here we report that a particular mutation in RAG-1 affects recombination by altering the specificity of target sequence usage. Recombination mediated by wild-type RAG-1 is tolerant of a wide range of coding sequences adjacent to the recombination signal. With the mutant RAG-1, recombination is much more demanding; efficient recombination is only found when particular dinucleotides are adjacent to the signal sequence heptamer. The mutant is also more sensitive than wild-type RAG-1 to certain alterations within the signal sequence. We suggest that the RAG-1 protein may interact physically with the target DNA at the coding-signal sequence border.

The bacterial Kdp K(+)-ATPase and its relation to other transport ATPases, such as the Na+/K(+)- and Ca2(+)-ATPases in higher organisms.

The Kdp system is a three-subunit member of the E1-E2 family of transport ATPases. There is sequence homology of the 72 kDa KdpB protein, the largest subunit of Kdp, with the other members of this family. The predicted structure of the 21 kDa KdpC subunit resembles that of the beta subunit of the Na+,K(+)-ATPase, suggesting that these subunits may have a similar function. The 59 kDa KdpA subunit has no known homologue; it is very hydrophobic and is predicted to cross the membrane 10-12 times. Genetic studies implicate this subunit in the binding of K+. As the binding site must be close to the beginning of the transmembrane channel, we suggest that KdpA also forms most or all of the latter. KdpA may have evolved from a K+/H+ antiporter that was recruited by the KdpB precursor to achieve the high affinity and specificity for K+, and the activation of transport by low turgor pressure characteristic of Kdp. Turgor pressure controls the expression of Kdp. This action is dependent on the 70 kDa KdpD and 23 kDa KdpE proteins. We are in the process of sequencing these genes. KdpE is homologous to the smaller protein of other members of a family of pairs of regulatory proteins implicated in control of a variety of bacterial processes such as porin synthesis, phosphate regulon expression, nitrogen metabolism, chemotaxis and nodule formation.