Supporting Information

Files in this Data Supplement:

Detailed Procedure for Purification of Ligase IV/XRCC4 (LX) Complex. Sf-9 cells were infected at a density of 2 ´ 10⁶ cells per ml with 50 ml of viral stock per liter of cells to give a multiplicity of infection of 5-10. Cells were harvested after 48-72 hours by centrifugation at 500 ´ g (IEC Centra MP4R centrifuge) for 10 min at 4° C and snap frozen in liquid nitrogen and stored at –80° C. Cell pellets were thawed on ice, and five volumes of lysis buffer (50 mM Tris·HCl, pH 8.0/10 mM 2-mercaptoethanol/1% Nonidet P-40) was added then mixed for 10 min at 4° C. Cell debris was removed by centrifugation at 10,000 ´ g (Biofuge, Heraeus) for 10 min at 4° C. 5ml of Talon resin (Clontech) per liter of cells was centrifuged at 500 ´ g for 3 min at 4° C (IEC Centra MP4R centrifuge), then washed with an equal amount of lysis buffer, before being added to the cleared cell extract. After incubation with mixing at 4⁰C for 2 hours the Talon was centrifuged at 500 ´ g for 3 min at 4° C. The resin was washed with 10 volumes of lysis buffer for 20 min at 4° C, centrifuged as before, then washed with 10 volumes of high salt buffer (20 mM Tris·HCl, pH 8.0)/500 mM NaCl/10 mM 2-mercaptoethanol/0.01% Nonidet P-40/10 mM imidazole) twice and 10 volumes of low salt buffer (20 mM Tris·HCl, pH 8.0/100 mM NaCl/10 mM 2-mercaptoethanol/0.01% Nonidet P-40/10 mM imidazole). The tagged proteins were eluted by adding three volumes of elution buffer (20 mM Tris·HCl, pH 8.0/100 mM NaCl/10 mM 2-mercaptoethanol/50 mM imidazole) incubated by mixing at 4° C for 20 min. The resin was centrifuged at 500 ´ g for 10 min, and the supernatant was collected and concentrated on a Vivaspin column (Sartorius, Goettingen, Germany) with a 30-kDa molecular weight cut-off, centrifuged at 4,000 ´ g (Heraeus) at 4° C to a final protein concentration of 1-10 mg/ml. The LX complex preparations were further purified by using the strong anion-exchange Mono Q PC 1.6 column and the SMART System (Applied Biosystems). Protein purity was assessed by SDS/PAGE, and subsequent protein was stained with Brilliant blue G-colloidal Coomassie (Sigma). Protein aliquots were snap frozen in liquid nitrogen and stored at –80° C in 20 mM Tris·HCl, pH 8.0/10 mM 2-mercaptoethanol/100 mM NaCl.

High stringency of the washing buffer allowed purification to near homogeneity in a single-affinity purification step (both XRCC4 and ligase IV proteins in the LX complex were 6His-tagged). Nevertheless, control experiments have been performed to exclude a possibility that copurifying ligation activity originating from insect cells may be present. The LX preparations were further purified over two more column chromatography steps (Mono S and gel filtration) with the SMART System (Applied Biosystems). Titration of the ligation activity of the purified complex after TALON affinity and Mono Q steps compared with the complex purified through all four chromatography steps did not show any differences in ligation activity.

The possibility of a contaminating ligation activity being present was further excluded by our findings with mutant LX complex purified by identical two-step procedure (1).

In this study, we have made several site-directed mutational changes in DNA ligase IV. In at least two of these changes, the double-stranded ligation activity is reduced concomitantly with the adenylation activity. In several cases, the double-stranded ligation activity is completely abolished. The results also correlate with those obtained after expression of the same mutations in Chinese hamster ovary cells.

In Vitro Double-Stranded Ligation Assay. Substrate preparation. Double-stranded DNA fragments were produced from the Bluescript plasmid (Stratagene) to give a 445-bp substrate with 4-bp overhangs at each end. These cohesive ends are not complementary to limit circularization. Bluescript was digested with the restriction enzymes PstI and AflIII (New England Biolabs), and the 445-bp DNA fragment was purified by electrophoresis on a preparative 0.8% agarose gel, followed by electroelution. Electroeluted substrates were precipitated with Pellet Paint (Novagen) according to the manufacturer’s instructions, then resuspended in 10 mM Tris·HCl, pH 8.5, and stored at –20° C.

Radioactive labeling. Two micrograms of the 445-bp substrate was incubated with 4 units of shrimp alkaline phosphatase (SAP, Roche) and an appropriate amount of 10´ SAP buffer (Roche) at 37° C for 30 min. The enzyme was then heat-inactivated at 65° C for 15 min. The substrate was then end-labeled with ³²P by incubation with 4 m l (10 units/m l) of T4 polynucleotide kinase (PNK; Roche), an appropriate amount of 10´ PNK buffer (Roche) and 2 m l of [g -³²P]ATP (7,000 Ci/mmol, end-labeling grade; ICN) at 30° C for 2 h. The reaction was stopped by the addition of 2 m l of 0.5 M EDTA, and the enzyme was heat-killed at 65° C for 15 min. Three m l of 10% sodium dodecyl sulphate (SDS) and 5 m l of DNA loading buffer (30% glycerol/0.1% xylene cyanol/0.1% bromophenol blue) was added, and the sample was gel-purified on an 10% neutral PAG in 1´ TBE . The sample was then electroeluted and precipitated with Pellet Paint (Novagen) according to the manufacturer’s instructions. The substrate was resuspended in 100 m l of 10 mM Tris·HCl, pH 8.5, and stored at –20° C.

Ligation assay. The radioactively labeled substrate was thawed on ice, then diluted to give 40 fmol of DNA per reaction. A standard 30-m l reaction contained 50 mM triethanolamine (pH 7.5), 2 mM Mg(OAc)₂, 2 mM DTT, 50 μg/ml BSA, 12% PEG, 1.75 pmol of LX, and 40 fmol of the 445-bp DNA substrate. Reactions were incubated at 37° C for 3 h.

Reactions were stopped by the addition of 1.5 m l of 10% SDS and 3.5 m l of proteinase K at 10mg/ml (Sigma) and incubated at 37° C for 20 min. The reactions were extracted twice with phenol/chloroform/isoamyl alcohol (25:24:1) then precipitated with Pellet Paint (Novagen) according to the manufacturer’s instructions. The DNA pellet was resuspended in 15 m l of 10 mM Tris·HCl, pH 8.5, and 3 m l of 6´ DNA loading buffer was added. The reactions were heated to 65° C for 5 min, then cooled rapidly to 4° C on ice. The reactions were electrophoresed on a 0.8% agarose gel at 150 V for 1.5 h. The gel was vacuum-dried (Bio-Rad) at 50° C for 3 h onto DE 81 anion exchange paper (Whatman) and visualized by using a STORM PhosphorImager (Molecular Dynamics).

Notes on Efficiency of Double-Stranded Ligation by LX. With respect to the high efficiency of our ligation reaction, the following comments based on the analysis of published data can be provided.

In our initial experiments with recombinant baculovirus-produced LX complex, we found that the efficiency of double-stranded ligation is extremely dependent on the reaction conditions, where even small changes in the reaction buffer composition and other conditions result in substantial inhibition of the double-stranded ligation by LX complex.

The most critical parameters found are the following (approximately in order of importance).

Overall salt concentration. In our hands, the presence of as little as 50 mM KCl in the reaction buffer significantly (≈50%) reduces the ligation efficiency.

Presence of 12% PEG (MW 8000). Three- to 5-fold stimulation of ligation is observed in the presence of 12% PEG.

Size of the DNA substrate. Of all the substrates tested, the reaction worked the best with substrates between 400-500 bp (53 bp, 157 bp, 1.56 kbp, and 2.96 kbp tested at the equimolar concentrations of DNA ends). Only ≈30% ligation efficiency has been observed with plasmid-sized substrates compared with the 445-bp substrate.

Presence of ATP. We observed 15–25% reduction in ligation efficiency in the presence of 5 mM ATP.

Temperature and time of incubation. We found 37° C to be the optimum temperature for our recombinant LX complex (4° C,16° C, room temperature, 30° C, and 37° C were tested).

Different conditions and different enzyme-to-substrate ratios alter the time kinetics of ligation reactions. Under our reaction conditions (at a given enzyme to substrate ratio), the reaction works well, plateau at 2 h with no further increase in ligation product formation being observed at longer incubation times (3-h and overnight incubations were tested).

Although the above parameters were systematically tested only individually, it seems reasonable to expect that combinations of them would have an additive effect on the efficiency of double-stranded ligation by LX, which is, indeed, suggested by several experiments we have performed, in which more than one parameter from the list above has been changed simultaneously.

Comparison of Our Reaction Conditions with the Previously Published Data. Our reaction conditions. The reaction mixture was 50 mM triethanolamine, pH 7.5/ 2mM Mg(OAc)₂/2mM DTT/50 μg/ml BSA/12% PEG/1.75 pmol of LX/40 fmol of 445-bp DNA substrate (44´ molar excess of LX to DNA, 22´ molar excess of LX to DNA ends). Incubation was for 3 h at 37° C.

Reddy et al. (2). The reaction mixture was 25 mM Tris·HCl, pH 7.5/50 m g/ml BSA/45 mM KCl/105 mM NaCl/0.05% Triton X 100/0.1 mM EDTA/2mM DTT/5 mM MgCl₂/0.1 mM ATP/10% PEG -754-bp Taq1 fragment from pUC18/1 nM DNA/1nM LX. Incubation was for only 10 min at 30° C. The probable main causes of low efficiency of ligation reaction include the high salt concentration, short incubation time, and 1:1 enzyme/substrate molar ratio.

Huang and Dynan (3). The reaction mixture was 50 mM triethanolamine, pH 7.5/10 mM Tris·HCl, pH 7.9/65 mM KOac/0.25 mM EDTA/0.5mM DTT/10% glycerol/1.0 mM Mg(OAc)₂/100 ng/m l BSA/1 mM ATP, with linearized pBluescript (2.96 kb) at 0.5 ng/m l in 20 m l (10 ng total) and 100 ng of LX. Incubation was for 30 min at 37° C. The probable main causes of low efficiency of ligation reaction include the high salt concentration, the substrate size, and the absence of crowding agent in the reactions.

Chen et al. (4). The reaction mixture was 60 mM Tris·HCl, pH 8.0/10 mM MgCl₂/5 mM DTT/1 mM ATP/50 μg/ml BSA, with 2.9-kb and 400-bp EcoR1 fragments: 4 pmol of substrate and 0.2 pmol of LX for the 2.9-kb substrate, and 20 fmol of DNA and 500 fmol of LX for the 400-bp substrate. Incubation was at 25° C for 2 h for the 2.9-kb substrate and 1 h for the 400-bp substrate The probable main causes of low efficiency of ligation reaction for the plasmid-sized substrate include the substrate size and the absence of crowding agent in the reactions.

Interestingly, for the 400-bp long substrate (which is close to our 445-bp) and where the enzyme-to-substrate ratios are closer to those used by us (20 fmol DNA and 500 fmol LX) ≈50% of the substrate is converted into circular substrate. Distribution of the products in favour of a circular monomer is likely to be due to the fact that the low concentration of DNA in a relatively large reaction volume (60-μl reactions) favours intramolecular ligation. As the authors point out, the addition of DNA–protein kinase dramatically changes this distribution in favor of intermolecular ligation but does not change the efficiency of the reaction.

The overall efficiency of this reaction is not dissimilar to ours, and it is likely that an addition of PEG would further increase the amount of product and change the distribution to a higher proportion of intermolecular ligation products.

1. Girard, P. M., Kysela, B., Harer, C., Doherty A. J. & Jeggo, P. A. (2004) Hum. Mol. Genet. 13, 1–8.

2. Ding, Q., Reddy, Y. V., Wang, W., Woods, T., Douglas, P., Ramsden, D. A., Lees-Miller, S. P. & Meek, K. (2003) Mol. Cell. Biol. 23, 5836–5848.

3. Huang, J. & Dynan, W. S. (2002) Nucleic Acids Res. 30, 667–674.

4. Chen, L., Trujillo, K., Sung, P., Tomkinson, A. E. (2000) J. Biol. Chem. 275, 26196–26205.