DNM-JNK1 or JNK1 siRNA inhibits UVA-induced phosphorylation of H2AX
(A) JB6 stable transfectants (CMV-neo, DNM-ERK2, DNM-p38β or DNM-JNK1) were exposed to UVA and histones extracted at the indicated time after UV. Cells not exposed to UVA served as negative controls (−). Bottom panels represent control experiments to verify that each dominant negative mutant specifically inhibited its respective targeted kinase activity. (B) JNK2^−/− MEFs transfected with mock siRNA or siRNA JNK1were treated with UVA (80 kJ/m²). H2AX phosphorylation at Ser139 (γH2AX) and histone H2A total protein level were determined. Right panel shows the effectiveness of siRNA JNK1 to suppress JNK level.

H2AX co-localizes and interacts with JNK in vivo following UVA exposure
(A) JB6 cells were exposed to UVA (80 kJ/m²), fixed with paraformaldehyde, stained for pJNK (green), H2AX (red) or γH2AX (red) and observed by immunofluorescence microscopy. Localization of pJNK1 and total H2AX are indicated (left panels) with (bottom) or without (top) exposure to UVA and localization of pJNK1 and γH2AX are indicated (right panels) with (bottom) or without (top) exposure to UVA.
(B) The endogenous H2AX-pJNK complex was immunoprecipitated from UVA-treated or untreated cells with a JNK or pJNK antibody and H2AX was detected by immunoblotting with an H2AX antibody (top upper, middle panels). Two blots (UVA-treated or untreated cells) were probed with a JNK antibody to monitor the amount of JNK precipitated in each reaction (bottom upper, middle panels). Inputs are representative of whole cell lysates. The positive control (upper and middle panels) consists of histones extracted by acid as described in “Experimental Procedures”. The H2AX-immunoprecipitated chromatin proteins were probed with JNK, pJNK, H2AX and γH2AX antibodies (lower panels). Precipitation with normal IgG served as a negative control.
(C) The in vivo interaction of H2AX and JNK1 or JNK1 deletion mutants was assessed by mammalian two-hybrid assay. Luciferase activity indicates the fold-increase in relative luminescence units normalized to the negative control (value for cells transfected with only pGL-Luc and pACT-H2AX = 1.0). Data are represented as means ± S.D.

H2AX phosphorylation is required for cells to undergo apoptosis
(A) H2AX-wt and H2AX^−/− MEFs were exposed to UVA and harvested at the indicated time after UVA. One group of each cell type was prepared for the DNA fragmentation ladder assay (upper panels) and another group of each type was divided into two parts. One part was prepared for determination of sub-G1 fraction by flow cytometry (middle panels) and the other was used to detect γH2AX or total H2A (lower panels).
(B) H2AX-wt and H2AX^−/− MEFs were exposed to the indicated dose of UVA and cell survival was evaluated at 24 h after UVA (left panel) or at the indicated day (right panel) by MTS assay as described in “Experimental Procedures”. Cells not exposed to UVA served as control.
(C) H2AX-wt MEFs were treated with SP600125 (5 μM) 1 h before exposure to UVA and harvested at the indicated time after UVA. One group of cells was prepared for the DNA fragmentation ladder assay (upper panel). Another group of cells was divided into two parts. One part was prepared for determination of sub-G1 fraction by flow cytometry (middle panel) and the other part was used for detection of γH2AX and total H2A (lower panels).
(D) H2AX-wt MEFs were treated with the caspase-3 inhibitor, Z-VAD (40 μM), 1 h before exposure to UVA. Harvested cells were used for the DNA ladder assay (upper panel) or to detect γH2AX, caspase-3 (full length and active fragments), pJNK and JNK, p-c-Jun (Ser63) and c-Jun. Detection of total H2A and β-actin was used to confirm equal protein loading.
(E) As for (A), cells following UVA exposure were lysed at the indicated time to detect caspase-3 (full length and active fragments). β-actin was used to confirm equal protein loading. For all experiments, cells not exposed to UVA served as negative controls (−).
(F) H2AX^−/− MEFs 48 h after transfection with pcDNA4 (vector), pcDNA4/H2AX-wt, or pcDNA4/H2AX-139A and H2AX-wt and H2AX^−/− MEFs without transfection were treated with UVA. Harvested cells were used for the DNA ladder assay (upper panel) and western analysis to detect γH2AX and H2AX or H2AX-139A from the anti-his-immunoprecipitated complex.

UVA, UVB or UVC induces phosphorylation of H2AX
(A) JB6 cells were exposed to (A) UVA, (B) UVB, or (C) UVC and histones were extracted as described in “Experimental Procedures” after the incubation time indicated (left panels) or after 60 min following indicated doses of UVA (right panels). Cells not exposed to UV served as negative controls (−). For all experiments, histones were resolved by 15% SDS-PAGE followed by western analysis with antibodies against γH2AX or total H2A.

H2AX is phosphorylated by JNK
(A) The H2AX protein was used as a substrate for active protein kinases including p38α, ERK1, JNK1 and JNK2. Reactive products were subjected to SDS–PAGE and western blot to detect H2AX phosphorylation (γH2AX). Coommassie blue staining shows the level of H2AX and kinases in each reaction.
(B) H2AX and a GST-c-Jun fusion protein were used as substrates for JNK, which was immunoprecipitated from JB6 cells treated or not treated with UVA (80 kJ/m²). The normal IgG immune complex served as a negative control. The reactive products were subjected to SDS–PAGE and western blot analysis to detect γH2AX and H2AX (upper 2 panels). Phosphorylation of GST-c-Jun (Ser63) served as positive control (middle panel). Bottom panel shows the amount of JNK precipitated in each reaction.
(C) Equal molar amounts of H2AX or GST-c-Jun (left panels) or GST-H2AX or GST-H2AX-139A (right panels) were used as substrates for active JNK1 or JNK2. Reactive products were resolved by SDS–PAGE followed by autoradiography or Coommassie blue staining.
(D) Five peptides were designed for in vitro kinase assays with active JNK1. Reaction samples were subjected to SDS-PAGE and autoradiography to detect phosphorylation (Lane 1, control without peptide).
(E) HEK 293 cells were co-transfected with pcDNA3.1/JNK2 and pcDNA4/H2AX or pcDNA4/H2AX-139A and 48 h later treated with UVA (80 kJ/m²). The expression level of JNK2 was detected with anti-V5. His-H2AX or his-H2AX-139A was immunoprecipitated with anti-his and the phosphorylation of the expressed H2AX or H2AX-139A was detected with an antibody against γH2AX.

Phosphorylation of H2AX and caspase-3 activation are suppressed by DNM-JNK1 JB6 stable transfectants (CMV-neo, DNM-JNK1) were exposed to UVA. One group of each stable cell line was used to extract histones at the indicated time following UV. Another group of each stable cell line was harvested at the indicated time following UV for detection of caspase-3 (full length and active fragments), β-actin and measurement of activation of JNK by in vitro kinase assay with c-Jun as substrate. Phosphorylated c-Jun was detected after SDS-PAGE by autoradiography (upper right, left panels). Cells not exposed to UVA served as negative controls (−).

H2AX phosphorylation results in enhanced DNA cleavage by CAD
(A) H2AX-wt and H2AX^−/− MEFs were exposed to UVA (40 kJ/m²) or treated with a caspase-3 inhibitor, Z-VAD, 1 h before UVA exposure. Then the cytosolic or S-100 fractions were extracted and incubated with naked DNA (plasmid pcDNA3) as described in “Experimental Procedures.” The reaction samples were separated by 1.8% agarose gel electrophoresis (left panel). The S-100 (80 μg) fractions isolated from UVA (40 kJ/m²)-treated H2AX^−/− MEFs were incubated with DNA (8 μg) and H2AX protein (0.5 μg) that had or had not been preincubated at 30 °C for 30 min with an active JNK2 to promote H2AX phosphorylation. The reaction samples were separated by 1.8% agarose gel electrophoresis (right panel). (B) The S-100 (80 μg) fractions isolated from UVA (40 kJ/m²)-stimulated H2AX^−/− MEFs were incubated with DNA (8 μg) and different doses of nonphosphorylated H2AX. Each reaction sample was divided into two parts. One part was separated by 1.8% agarose gel electrophoresis for DNA degradation analysis (upper panel) and another part was used to monitor H2AX protein level (lower panel).
(C) The S-100 fractions (80 μg) isolated from UVA (40 kJ/m²)-exposed H2AX^−/− MEFs were incubated with DNA (8 μg) and different doses of H2AX, which had or had not been preincubated at 30 °C for 30 min with an active JNK2 to induce phosphorylation. The reaction samples were separated by 1.8% agarose gel electrophoresis. The western blot (lower panel) indicates the level of γH2AX.
(D) H2AX was phosphorylated first by JNK2 and then mixed with DNA (8 μg) and CAD immunoprecipitated from H2AX-wt or H2AX^−/− MEFs exposed or not exposed to UVA. The reaction samples were separated by 1.8% agarose gel electrophoresis to visualize DNA degradation. The western blot indicates that CAD was precipitated from H2AX-wt or H2AX^−/− MEFs (lower panel).
(E) H2AX was incubated with PBS (lane 1, input), Ni-NTA agarose (lane 2), or Ni-NTA agarose-DFF40 (lane 3) for protein binding. Then the binding was analyzed by western blot and Coommassie blue staining.
(F) A model of apoptosis regulated by cooperation between the JNK/H2AX pathway and caspase-3/CAD (DFF40) pathway is proposed. When cells are exposed to UVA, JNK is activated and translocated into the nucleus where it phosphorylates H2AX. At the same time, UV-activated JNK can also stimulate caspase-3 activation through the release of cytochrome c from the mitochondria. Phosphorylated nuclear H2AX regulates DNA fragmentation mediated by CAD (DFF40) during apoptosis.