1. A Model for the γδ Resolvase Synaptic Complex
Gary J Sarkis, Laura Lea Murley, Andres E Leschziner, Martin R Boocock, W.Marshall Stark, Nigel D.F Grindley 
Molecular Cell 
Volume 8, Issue 3, Pages (September 2001)
DOI: /S (01)

2. Figure 1
The γδ res Site
(A). The three binding sites for resolvase dimers are indicated by I, II, and III, with the open arrows representing the 12 bp resolvase recognition sequence; L and R indicate the left and right halves of the three sites. Vertical black arrowheads show the crossover point in site I. The distances between the centers of each binding site and the lengths of the intrasite spacers are shown.
(B) Model for the formation of the γδ resolvase synaptic complex. Left: three resolvase dimers (shaded elipses) bind to each res site, enabling those at sites II and III to make a 2-3′ interaction in cis, forming an effective synaptic surface. Synapsis (i) is initiated by interactions in trans between the resolvase dimers bound at sites II and III of each res. Interwrapping and bending of the res sites, plus interactions between the resolvase dimers at sites I and III, allow the crossover sites to pair (ii) completing assembly of the synaptic structure. DNA breakage and rejoining (iii) forms the two-noded catenated products. Adapted from Stark and Boocock (1995) and modified to incorporate ideas from Murley and Grindley (1998) and this paper
Molecular Cell 2001 8, DOI: ( /S (01) )

3. Figure 2 Structures of the γδ Resolvase
(A) Array of four dimers of the resolvase catalytic domain as seen in the original crystals of this domain (PDB ID 2RSL); an almost identical array is also seen in a different crystal form (PDB ID 1GDR) (Rice and Steitz, 1994a, 1994b; Sanderson et al., 1990). The magenta patches indicate the surface (composed of residues 2, 32, 54, and 56) that forms the 2-3′ interface. The two yellow dimers are held together by this interface, as are the two green dimers. The blue patches indicate the surface associated with residues 99–102, and marks the N terminus of α helix E that is largely responsible for dimer formation. Residues Ser10 are colored dull blue, and Arg119 (at the C terminus of the visible portion of the E helices) are colored red.
(B) A schematic diagram showing the same arrangement. The cylindrical projections represent the E helices at the dimer interface. Note the three orthogonal 2-fold axes through the center of this assembly.
(C) Structure of the resolvase crossover site complex (PDB ID 1GDT) (Yang and Steitz, 1995); the coloring is as in (A). Structures were rendered using GRASP (Nicholls et al., 1993)
Molecular Cell 2001 8, DOI: ( /S (01) )

4. Figure 3
Effect of the Rice and Steitz Model of Imposing a 2-3′ Connectivity on the Juxtaposition of Sites I and III
(A) Model of the resolvase synaptic complex proposed by Rice and Steitz (1994a). res DNA is represented by the yellow and blue “worms” with binding sites I, II, and III as indicated; DNA binding domains are not shown. The tetramer of dimers that binds sites II and III is arranged precisely as in Figure 2A. Note that the dimers that bind site I are connected to those bound to site III by a 2-3 interaction, and the 2-3′ interface (magenta) of these subunit is unused.
(B) Two views of a model showing the consequence of shifting the interaction between the central (site III-bound) and lower (site I-bound) dimers from 2,3 (as in A) to 2-3′. Note that in both “filaments” of resolvase dimers, each dimer is connected to its neighbor in the same filament by the 2-3′ interface
Molecular Cell 2001 8, DOI: ( /S (01) )

5. Figure 4
Alternative Ways to Consider Pairing Filaments of Three 2-3′-Connected Resolvase Dimers
The three dimer filaments shown here were constructed from PDB accession number 2RSL (Rice and Steitz, 1994b). The black line passing through a 2-3′ interface represents the 2-fold axis of the upper pair of resolvase dimers (labeled II and III) in each 3-dimer filament. Pairings of the yellow filament (B) with either (A) or (C) were explored by (i) “sliding” the filaments together along the 2-fold axis and rotating them about the same 2-fold. Docking of filaments (A) and (B) gave, as best fit, the arrangement shown in Figure 3B, whereas filaments (B) and (C) gave the arrangement shown in Figure 5. For further details, see the text
Molecular Cell 2001 8, DOI: ( /S (01) )

6. Figure 5 New Model for the Resolvase Synaptic Complex
DNA binding domains are not shown but would radiate out from the red patches (Arg 119).
(A) Three views of the resolvase core (obtained simply by rotating about the y axis as indicated) showing the close juxtaposition between yellow and green dimers at each position along the filament. We anticipate that minor adjustments to side chain conformations at the synaptic surface and to the relative juxtapositions of protomers across the dimer and 2,3′ interfaces would result in a tighter packing of all the resolvase subunits than is apparent in this model.
(B and C) Alternative wrappings of the two res sites (for further discussion, see the text)
Molecular Cell 2001 8, DOI: ( /S (01) )

7. Figure 6
The G101S, E102Y, M103I Triple Mutant of γδ Resolvase Forms a Stable Synaptic Complex with Two Site I DNAs and a Tetramer of Resolvase
Both panels show native polyacrylamide gel shift assays of resolvase-site I complexes.
(A) Gel shifts with two different sizes of site I DNA fragments (39 bp or 54 bp as indicated). WT = wild-type resolvase; 101/3 = the activated (G101S, E102Y, M103I, E124Q) mutant of resolvase with the additional mutations R68H and E56K to eliminate catalysis and the 2,3′ interaction, respectively.
(B) Gel shifts with two different sizes of activated resolvase. 101/3 is as in (A); +HSV is a resolvase with virtually the same mutations but with wild-type E124+ (instead of E124Q) plus a 35 residue C-terminal HSV tag. The two proteins were mixed either immediately (Mix 0) or 3 days (Mix 3) before adding them to the 39 bp site I substrate
Molecular Cell 2001 8, DOI: ( /S (01) )