Abstract
Introduction. V(D)J recombination is a form of site-specific DNA rearrangement that is responsible for assembling antigen receptor genes during lymphocyte development. This process is initiated when two lymphoid cell specific proteins, RAG-1 and RAG-2 (Recombination Activating Gene), bring two gene coding segments into close proximity (synapsis) and introduce a DNA double-strand break (DSB) at the end of each coding segment. A recombination signal sequence (RSS) adjoining each gene coding segment serves as the binding site for RAG proteins, and directs the location of DNA cleavage. Each RSS contains a conserved heptamer and nonamer sequence, separated by either 12 or 23 nucleotides (12-RSS and 23-RSS, respectively). Generally, synapsis and cleavage only occur between two RSSs whose spacer lengths differ (the 12/23 rule). While RAG proteins are sufficient to mediate RSS binding and cleavage in vitro, these activities are promoted in the presence of the DNA binding and bending factors HMGB1 or HMGB2 (High Mobility Group Box 1 and 2).|Specific aims. This study examines the biochemistry of V(D)J recombination in three ways: (1) characterizing the first gain-of-function E649A RAG-1 mutant, and examining the global influence of the E649 region on RAG activity and 12/23 restriction (Chapters III and IV); (2) examining the interaction of RAG proteins with other factors in the RAG recombinase complex, such as HMGB1 and the significance of the separate domains in RAG-mediated cleavage under two divalent metals (Chapter V); and (3) designing a protein purification method that yields concentrated mammalian-expressed core RAG-1/2 proteins.|Background information, results, and conclusions for each specific aim are summarized below:|1) Specific aim #1. Previous mutational analysis of the RAG proteins has identified various catalysis-deficient and separation-of-function mutants. However, mutations that enhance RAG activity have not yet been reported. The purpose of this study was to characterize a novel gain-of-function E649A RAG-1 mutant that exhibits increased cleavage activity due to enhanced catalysis of transesterification. This gain-of-function mutant was characterized based on the catalytic activity of the RAG proteins in vitro (binding, cleavage, and other DNA strand transfer reactions) and in vivo (cleavage and joining). Wild-type (WT) and mutant RAG-1 possess comparable RSS binding and nicking activity, but the E649A mutant exhibits a 10- to 25-fold increase in overall cleavage in vitro. Other DNA strand transfer reactions, such as transposition, disintegration, and hybrid joint formation are similarly affected. Interestingly, WT and mutant RAG-1 support similar levels of V(D)J recombination when assayed in cell culture using a plasmid substrate containing a complementary (12/23) pair of RSSs. However, on substrates containing a 12/12 RSS pair, mutant RAG-1 supports higher levels of cleavage and recombination compared to WT (2.9-fold increase), in violation of the 12/23 rule. The E649A mutant is also able to cleave mismatched (12/12 and 23/23) and unpaired (12 only or 23 only) at higher levels. Since the E649 region of murine RAG-1 is conserved among many species, upstream and downstream residues were mutated. In vitro and in vivo experiments demonstrated that these mutants could not reproduce the gain-of-function phenotype of the E649A RAG-1 mutant, interestingly, substitution of four amino acids upstream of E649 with four alanines exhibited a more stringent requirement for synapsis. In contrast to E649A RAG-1, which can cleave without partner RSS, this mutant required a nick, HMGB1, and partner RSS for cleavage. In conclusion, our study suggests that this novel E649A RAG-1 mutation impairs the sensing of 12/23-regulated synapsis by RAG proteins; this sensing is normally required to promote cleavage at both RSSs. Thus, the mutation "tricks" the RAG complex into falsely sensing synapsis, causing the complex to cleave a bound RSS as if it were integrated into a synaptic complex. Mutations elsewhere within the E649 region do not phenocopy cleavage hyperactivity, but produce a mutant with more stringent requirement for synapsis that needs both partner RSSs and a nick for cleavage in in vitro experiments.|2) Specific aim #2. HMGB1 contains two DNA binding domains. Two previous studies (Bergeron et at., 2005, and Dai et al., 2005) reached different conclusions regarding whether one or both of these domains is sufficient to stimulate RAG-mediated binding and cleavage of naked DNA in vitro. I tested whether the choice of divalent metal ion and the concentration of HMGB1 used in the cleavage reaction contribute to this apparent discrepancy. In this study, I used electromobility shift assays (EMSA) and in vitro cleavage assays to demonstrate that single HMG-box domains of HMGB1 stimulate RAG-mediated RSS cleavage in a concentration-dependent manner. I found that single HMG-box domains support RAG-mediated cleavage in the presence of Mn2+, but not Mg2+. Interestingly, the inability of a single HMG-box domain to stimulate cleavage in the presence of Mg2+ is overcome by the addition of partner RSS to promote synapsis. Furthermore, I show that mutant forms of HMGB1 that fail to stimulate cleavage in the presence of Mg2+ can be largely rescued by replacing Mg2+ with Mn2+. The conflicting data published by Bergeron et al. and Dai et al. is explained by the choice of divalent metal ion and the abundance of HMGB1 in the cleavage reaction. These data also show that synapsis bypasses the need for both HMG-boxes, raising the possibility that synaptic complex assembly in vitro is associated with conformational changes that alter the interactions of RAGs and HMGB1 with DNA.|3) Specific aim #3. Only two reports exist detailing the crystal structure of RAG proteins; in both studies, a 100-residue stretch was described in non-core regions of RAG-1 and RAG-2 (Bellon et al., 1997 and Matthews et al., 2007, respectively). Both of these reports relied on singly expressed RAG proteins in a bacterial system to increase the yield of expressed RAG proteins. We wondered whether RAG-1 and RAG-2 when coexpressed in mammalian cells can promote production of a compact structure to allow for x-ray diffraction crystal formation. I therefore optimized purification and concentration conditions necessary before proteins can be submitted for crystallization. Here I describe two protein concentration methods: sucrose bed and centrifugal filtration devices. Using this protocol, I obtained mammalian-expressed RAG proteins at the desired concentration of at least 10 mg/ml. Our preliminary crystallization screen yielded precipitate formation in 48% of samples. In conclusion, I have developed an effective method to concentrate RAG proteins. This protocol can be used to study less abundant proteins, such as full- length RAG-1 and RAG-2.|Conclusions. I characterized the first gain-of-function E649A RAG-1 mutant. This mutation enhances the transesterification step of the cleavage phase, and cleaves independently of synapsis both in vitro and in vivo. The mutant supports rearrangement of a 12/12 extrachromosomal substrate 2.9-fold higher than wild type (WT) RAG-112 in vivo. Additionally, the E649A region mutations do not phenocopy the original E649A phenotype, but provide a new RAG-1 mutant with more stringent requirement for synapsis. I resolved two conflicting reports (Bergeron et at., 2004 and Dai et at.,2004), by demonstrating that single HMG-box domains are not sufficient to stimulate cleavage of RSS substrates in the presence of Mg2+ under any concentration, while single HMG-box domains are sufficient to stimulate cleavage in the presence of Mn2+. Removal of the acidic tail and the basic linker lead to an overall increase of HMGB1 activity. I designed RAG protein purification and concentration strategies that allowed us to obtain concentrated RAG-1/2 proteins at any desired concentration. I now have a set of conditions that promote precipitation and a set that inhibits precipitation. As a result, an effective technique has been developed to facilitate the concentration of other rare proteins, such as full-length RAG-1 and RAG-2 that are difficult to obtain in high yield in an active form.