Rag-1 mutations associated with B-cell-negative scid dissociate the nicking and transesterification steps of V(D)J recombination

Abstract

Some patients with B-cell-negative severe combined immune deficiency (SCID) carry mutations in RAG-1 or RAG-2 that impair V(D)J recombination. Two recessive RAG-1 mutations responsible for B-cell-negative SCID, R621H and E719K, impair V(D)J recombination without affecting formation of single-site recombination signal sequence complexes, specific DNA contacts, or perturbation of DNA structure at the heptamer-coding junction. The E719K mutation impairs DNA cleavage by the RAG complex, with a greater effect on nicking than on transesterification; a conservative glutamine substitution exhibits a similar effect. When cysteine is substituted for E719, RAG-1 activity is enhanced in Mn(2+) but remains impaired in Mg(2+), suggesting an interaction between this residue and an essential metal ion. The R621H mutation partially impairs nicking, with little effect on transesterification. The residual nicking activity of the R621H mutant is reduced at least 10-fold upon a change from pH 7.0 to pH 8.4. Site-specific nicking is severely impaired by an alanine substitution at R621 but is spared by substitution with lysine. These observations are consistent with involvement of a positively charged residue at position 621 in the nicking step of the RAG-mediated cleavage reaction. Our data provide a mechanistic explanation for one form of hereditary SCID. Moreover, while RAG-1 is directly involved in catalysis of both nicking and transesterification, our observations indicate that these two steps have distinct catalytic requirements.

FIG. 1

Impairment of V(D)J recombination and DNA cleavage in vivo by RAG-1 mutations associated with B-cell-negative SCID. (A) Diagram of the RAG-1 fusion protein. The RAG-1 core (amino acids 384 to 1008), MBP, polyhistidine tag (H), and c-Myc epitope (M) are indicated. Positions of SCID-associated point mutations are indicated. (B) Wild-type or mutant RAG-1 fusion protein was coexpressed with the RAG-2 core in 293 cells and signal joint formation was quantitated by using the extrachromosomal V(D)J recombination substrate pJH200. Percent recombination was calculated as described by Hesse et al. (17); values represent the means of at least three independent experiments. Lanes: 1, wild-type RAG-1; 2, RAG-1(R621H) and RAG-1(E719K); 3, RAG-1(R621H) and RAG-1(Y935Stop); 4, RAG-1(E719K) and RAG-1(Y935Stop); 5, RAG-1(R621H); 6, RAG-1(E719K); 7, RAG-1(Y935Stop). All assays included wild-type RAG-2 core. (C) Signal ends (upper panel) were assayed by ligation-mediated PCR. Products were detected by staining with ethidium bromide. Total pJH200 was assayed by PCR amplification of a backbone sequence (lower panel), as described in Materials and Methods. Lanes are numbered as for panel B.

FIG. 2

Impairment of RSS cleavage in vitro by SCID-associated RAG-1 mutations. (A) Wild-type or mutant RAG-1 core fusion proteins, diagrammed in Fig. 1A, were coexpressed with an MBP-tagged RAG-2 core in 293 cells and purified by affinity chromatography as described elsewhere (29, 45). Equal volumes of purified protein (25 μl) were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and detected by silver staining. Lane 1, wild-type RAG-1; lane 2, wild-type RAG-1 and RAG-2; lane 3, RAG-1(R621H) and RAG-2; lane 4, RAG-1(E719K) and RAG-2. Positions of RAG-1 and RAG-2 fusion proteins are indicated by arrows. (B) In vitro cleavage assay. Cleavage of ³²P-labeled 12-spacer (lanes 1 to 5, 11 to 15, and 21 to 25) or ³²P-labeled 23-spacer (lanes 6 to 10, 16 to 20, and 26 to 30) substrate (0.02 pmol in a 10-μl reaction volume) was assayed in the presence of Mg2+ as described elsewhere (19). Lanes 2 to 5 and 7 to 10, wild type RAG-1 and RAG-2; lanes 12 to 15 and 17 to 20, RAG-1(R621H) and RAG-2; lanes 22 to 25 and 27 to 30, RAG-1(E719K) and RAG-2; lanes 1, 6, 11, 16, 21, and 26, and RAG added. Lanes 4, 5, 14, 15, 24, and 25, addition of equimolar unlabeled 23-spacer substrate (u). Lanes 3, 5, 8, 10, 13, 15, 18, 20, 23, 25, 28, and 30, addition of HMG-1 to 8μg/ml. Positions of nicked and hairpin products are indicated by arrows at left.

FIG. 3

Effects of the R621H and E719K mutations on nicking and transesterification. (A) Kinetic analysis of RSS nicking by RAG-1(R621H) and RAG-1(E719K) in Mg²⁺. Lanes 1 to 7, wild-type RAG-1 and RAG-2; lanes 8 to 14, RAG-1(E719K) and RAG-2; lanes 15 to 21, RAG-1(R621H) and RAG-2. Assays were carried out in Mg²⁺ as described in Materials and Methods by using the mutant 12-spacer substrate C17A, which undergoes nicking in the absence of transesterification (29). Samples were withdrawn at times indicated. The position of nicked product is indicated by the shaded arrow at left. (B) Kinetic analysis of hairpin formation by RAG-1(R621H) and RAG-1(E719K) in Mg²⁺. Lanes 1 to 6, wild-type RAG-1 and RAG-2; lanes 7 to 12, RAG-2(E719K) and RAG-1; lanes 13 to 18, RAG-1(R621H) and RAG-2. Assays were carried out in Mg²⁺ using a prenicked substrate as described in Materials and Methods. Samples were withdrawn at times indicated. The positions of prenicked substrate (shaded arrow) and hairpin product (filled arrow) are indicated at left. (C) Kinetic analysis of hairpin formation by RAG-1(R621H) and RAG-1(E719K) in Mn²⁺. Lanes 1 to 8, wild-type RAG-1 and RAG-2; lanes 9 to 16, RAG-1(R621H) and RAG-2; lanes 17 to 24, RAG-1(E719K) and RAG-2. Assays were carried out in Mn²⁺ using a prenicked substrate as described in Materials and Methods. Samples were withdrawn at times indicated. The positions of prenicked substrate (shaded arrow) and hairpin product (filled arrow) are indicated at left. (D and F) The yield of nicked products in panel A and hairpin products in panel B was quantitated by phosphorimager and plotted as a function of time. Filled squares, wild-type RAG-1; filled diamonds, RAG-1(E719K); dotted squares, RAG-1(R621H). (E and G) The kinetics of nicking (E) or hairpin formation (G) by RAG-1(R621H) and RAG-1(E719K) were assayed for panels D and F except that reactions were performed in the presence of Mn²⁺. Products were quantitated by phosphorimager and plotted as a function of time. Symbols are as defined for panel D.

FIG. 4

(A to C) RAG-1(R621H), RAG-1(E719K), and RAG-1(E719C) retain the ability to form RAG-2-dependent RSS complexes. Binding to a ³²P-labeled 12-spacer substrate and EMSAs were carried out as described in Materials and Methods. (A) EMSAs of wild-type RAG-1 and RAG-1(R621H). Lane 1, wild-type RAG-1 alone; lane 2, RAG-2 alone; lane 3, wild-type RAG-1 and RAG-2, expressed separately and combined; lane 4; wild-type RAG-1 and RAG-2, coexpressed and copurified; lane 5, RAG-1(R621H) and RAG-2, coexpressed and copurified. Positions of the M1 and M1/2 complexes are indicated by arrows at left. (B) Wild-type or mutant MBP–RAG-1 fusion proteins, coexpressed and copurified with MBP–RAG-2, were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and detected by silver staining. Increasing amounts of purified complexes containing wild-type RAG-1 (lanes 1 to 3), RAG-1(E719K) (lanes 4 to 6), and RAG-1(E719C) (lanes 7 to 9) were loaded. Positions of RAG-1 and RAG-2 are indicated at left. (C) EMSAs of wild-type RAG-1 (lanes 3, 6, 9, and 12), RAG-1(E719C) (lanes 1, 4, 7, and 10), and RAG-1(E719K) (lanes 2, 5, 8, and 11) or in the absence of added protein (lane 13). Equal amounts of protein, copurified with RAG-2 and quantitated for panel B, were added. Reactions were carried out in the presence of Mn²⁺ (lanes 1 to 6), Mg²⁺ (lanes 7 to 9), or Ca²⁺ (lanes 10 to 13) at 4°C for 30 min (lanes 1 to 3) or at 37°C for 20 min (lanes 4 to 13). (D and E) RAG-1(R621H) and RAG-1(E719K) yield modification interference patterns identical to that of wild-type RAG-1. (D) DMS modification interference. Cleavage products from free or bound DNA, radiolabeled on the top strand (left panel) or bottom strand (right panel), were fractionated by gel electrophoresis and detected with a phosphorimager. The RSS heptamer (7) and nonamer (9) are indicated by vertical bars. The arrowheads mark positions of strongest interference. Lanes 1 and 9, G-specific sequencing tracts; lanes 2 and 10, T-specific sequencing tracts; lanes 3, 5, 7, 11, 13, and 15, products from free DNA; lanes 4, 6, 8, 12, 14, and 16, products from bound fractions. Lanes 3, 4, 11, and 12, wild-type RAG-1 and RAG-2; lanes 5, 6, 13, and 14, RAG-1(R621H) and RAG-2; lanes 7, 8, 15, and 16, RAG-1(E719K) and RAG-2. (E) KMnO₄ modification interference. Cleavage products from free or bound DNA, radiolabeled on the top strand (left panel) or bottom strand (right panel), were fractionated by gel electrophoresis and detected by phosphorimager analysis. The heptamer (7) and nonamer (9) are indicated by vertical bars. Closed and open arrowheads mark positions at which modification is underrepresented or overrepresented, respectively, in the bound fraction. Lanes 1 and 8, G-specific sequencing tracts; lanes 2, 4, 6, 9, 11, and 13, products from free DNA; lanes 3, 5, 7, 10, 12, and 14, products from bound fractions. Lanes 2, 3, 9, and 10, wild-type RAG-1 and RAG-2; lanes 4, 5, 11, and 12, RAG-1(R621H) and RAG-2; lanes 6, 7, 13, and 14, RAG-1(E719K) and RAG-2.

FIG. 5

Impairment of V(D)J recombination and DNA cleavage in vivo by diverse substitutions at E719 of RAG-1. (A) Wild-type or mutant RAG-1 fusion protein was coexpressed with RAG-2 core in 293 cells, and signal joint formation was quantitated using the extrachromosomal V(D)J recombination substrate pJH200. Percent recombination was calculated; values represent the means of two independent experiments. Lanes: 1, wild-type RAG-1; 2, RAG-1(E719K); 3, RAG-1(E719C); 4, RAG-1(E719A); 5, RAG-1(E719Q). (B) Signal ends were assayed by ligation-mediated PCR (upper panel); total pJH200 was assayed by PCR amplification of a backbone sequence (lower panel), as described in Materials and Methods. Lanes 1 and 2, wild-type RAG-1; lanes 3 and 4, RAG-1(E719K); lanes 5 and 6, RAG-1(E719C); lanes 7 and 8, RAG-1(E719A); lanes 9 and 10, RAG-1(E719Q). All transfections included wild-type RAG-2 core.

FIG. 6

Substitution of RAG-1 E719 by cysteine, but not by glutamine or lysine, confers preferential cleavage activity in Mn²⁺. (A) Kinetic analysis of RSS cleavage by purified wild-type RAG-1 or RAG-1(E719C) and wild-type RAG-2. Assays were carried out against a 12-spacer substrate in 6 mM Mg²⁺ (left panel) or 1 mM Mn²⁺ (right panel) as described in Materials and Methods. Lanes 1 to 6 and 13 to 18, wild-type RAG-1 and RAG-2; lanes 7 to 12 and 19 to 24, RAG-1(E719C) and RAG-2. Samples were withdrawn at times indicated. Positions of nicked and hairpin products are indicated at right. (B) The total yield of nicked and hairpin products shown in panel A, as quantitated by phosphorimager, is plotted as a function of time. Open squares, wild-type RAG-1; filled diamonds, RAG-1(E719C); left panel, cleavage activity in Mg²⁺; right panel, cleavage activity in Mn²⁺. (C) RAG-1(E719K) and RAG-1(E719Q) do not exhibit altered metal ion specificity. Kinetic analysis of RSS cleavage was carried out by purified proteins against a 12-spacer substrate in 6 mM Mg²⁺ or 1 mM Mn²⁺ as in panel A. Lanes 1 to 6 and 19 to 24, wild-type RAG-1 and RAG-2; lanes 7 to 12 and 25 to 30, RAG-1(E719K) and RAG-2; lanes 13 to 18 and 31 to 36, RAG-1(E719Q) and RAG-2. Positions of nicked and hairpin products are indicated at right. (D) The total yield of nicked and hairpin products in panel C, as quantitated by phosphorimager, is plotted as a function of time. Open squares, wild-type RAG-1; filled diamonds, RAG-1(E719K); filled squares, RAG-1(E719Q); left panel, cleavage activity in Mg²⁺; right panel, cleavage in Mn²⁺.

FIG. 7

Impairment of nicking but not transesterification by a nonconservative substitution at R621 of RAG-1. (A) Kinetic analysis of RSS nicking by RAG-1(R621K) and RAG-1(R621A). Lanes 1 to 9, wild-type RAG-1 and RAG-2; lanes 10 to 18, RAG-1(R621K) and RAG-2; lanes 19 to 27, RAG-1(R621A) and RAG-2. Assays were carried out in Mn²⁺ as described in Materials and Methods by using the mutant 12-spacer substrate C17A, which undergoes nicking in the absence of transesterification (29). Samples were withdrawn at times indicated above. The position of nicked product is indicated by the arrow at left. (B) The yield of nicked products in panel A was quantitated by phosphorimager and plotted as a function of time. Filled squares, wild-type RAG-1; filled diamonds, RAG-1(R621K); dotted squares, RAG-1(R621A). (C) Kinetic analysis of hairpin formation by RAG-1(R621K) and RAG-1(R621A). Lanes 1 to 8, wild-type RAG-1 and RAG-2; lanes 9 to 16, RAG-1(R621K) and RAG-2; lanes 17 to 24, RAG-1(R621A) and RAG-2. Assays were carried out in Mn²⁺ using a prenicked substrate as described in Materials and Methods. Samples were withdrawn at times indicated above. The positions of hairpin product are indicated at left. (D) The yield of hairpin products in panel C was quantitated by phosphorimager and plotted as a function of time. Symbols are as defined for panel B.

FIG. 8

Comparison of nicking by RAG-1(R621H) and RAG-1(R621A) at pH 7.0 and at pH 8.4. (A) Assay at pH 7.0. Assays were carried out in Mn²⁺ under standard conditions (pH 7.0) using the mutant 12-spacer substrate C17A. Samples were withdrawn at various times and products were fractionated by gel electrophoresis. The yield of nicked products was quantitated by phosphorimager and plotted as a function of time. Filled squares, wild-type RAG-1; filled diamonds, RAG-1(R621H); dotted squares, RAG-1(R621A). (B) As in panel A, except that reactions were carried out at pH 8.4.