Mutation Research, 234 (1990) 97-106
Elsevier MUTENV 08753
Immunodetection of DNA-protein crosslinks by slot blotting * Charles A. Miller III and Max Costa Department of Environmental Medicine. New York Unwersity School of Medicine, New York, N Y 10016 (U.S.A.)
(Received 14 August 1989) (Revision received 26 October 1989) (Accepted 2 November1989)
Kevwords: Chromium; cis-Diamminedichloroplatinum(ll); Formaldehyde; DNA-protein crosslinks; DNA damage; Slot blotting
Summary Ultraviolet light, formaldehyde, c i s - d i a m m i n e d i c h l o r o p l a t i n u m ( I I ) , chromate (Cr6+), or chromium chloride (Cr 3+) under the appropriate conditions caused the formation of D N A - p r o t e i n crosslinks in intact Chinese hamster ovary (CHO) cells or in cell nuclei. The D N A - p r o t e i n crosslinks were isolated, applied to nitrocellulose filters, and reacted with antibodies to nuclear proteins. An antiserum to a 97-kD nuclear protein detected p97-DNA complexes in C H O nuclei and cell cultures treated with UV light, cis-Pt and formaldehyde. Exposure to Cr 3+ induced p97-DNA crosslinks only in isolated nuclei, while ~chromate (Cr 6+) treatment resulted in significant crosslink formation only in intact cells. Analysis of western blots with the p97 antiserum indicated that crosslinks induced by formaldehyde or ultraviolet light required DNAase I digestion of D N A for migration of the p97 complexes into the gel. In contrast, the 97-kD antigen from the metal-induced crosslinks was released from D N A and resolved in the gel when 2-mercaptoethanol was included in the electrophoresis sample buffer. Assay of slot blots with an antihistone monoclonal antibody indicated that formaldehyde, but not cis-Pt or chromate, crosslinked histones to the DNA. These results illustrate the utility of immuno-slot blots in detecting and characterizing D N A - p r o t e i n complexes induced by diverse chemical and physical agents.
Most mutagens and carcinogens are thought to induce genetic changes by interacting with DNA. The biologically significant lesions responsible for the critical alterations that result in cancers are unknown. One lesion induced by several genotoxic
* This work was supported by grants ES 04895 and ES 04715 from the National Institute of Environmental Health Sciences and by grant CA 43070 from the National Cancer Institute. Correspondence: Dr. Max Costa, Department of Environmental Medicine, New York University School of Medicine, Long Meadow Road, Tuxedo, NY 10987 (U.S.A.).
agents of environmental and occupational significance is the D N A - p r o t e i n crosslink (reviewed by Oleinick et al., 1987). D N A - p r o t e i n crosslinks are thought to be important lesions because they might impede the replicative, transcriptional, or repair enzymes of the cell. The repair rate of D N A - p r o tein crosslinks is relatively slow or absent with respect to other genetic damage (Sugiyama et al., 1986, Cosma and Marchok, 1988). Consequently, D N A - p r o t e i n crosslinks may be present during D N A replication and may have the potential to induce permanent genetic changes. Numerous studies with ultraviolet light (UV), formaldehyde (HCHO), c i s - d i a m m i n e d i c h l o r o -
0165-1161/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
98 platinum(II) (cis-Pt), and chromium compounds have revealed that these carcinogens induce D N A - p r o t e i n crosslinks (Banjar et al., 1983, Cosma and Marchok, 1988, Cupo and Wetterhahn, 1985, Wedrychowski et al., 1985). These agents are quite different and consequently each should form D N A - p r o t e i n crosslinks by a distinctly different mechanism. UV light of 254 nm forms radicals in the ring structures of D N A bases, which then react with adjacent proteins to form a covalent bond (Peak and Peak, 1986). H C H O forms covalent bonds between the amine groups of proteins and D N A bases (Singer and Grunberger, 1983). Cr 3+ and the dechlorinated form of cis-Pt have an affinity for histidyl, methionyl and cysteinyl groups of proteins and the nitrogens of D N A bases (Howe-Grant and Lippard, 1980; Nieboer and Jusys, 1988). Cr 3+ has an additional affinity for oxygens of phosphoryl and carboxylic groups. Metal compounds containing chromium and platinum preferentially crosslink many of the same non-histone proteins to DNA, in contrast, H C H O actively forms hist o n e - D N A crosslinks (Wedrychowski et al., 1985; Solomon et al., 1988). UV light crosslinks both non-histone proteins and histones to D N A (Bouliakis, 1986; Angelov et al., 1988). Conditions of reduced p H enhance the ability of UV light to crosslink a number of non-nuclear proteins to D N A in vitro (Czichos et al., 1989). The present experiments describe the use of an alternative technique to alkaline elution for the detection and identification of proteins that are crosslinked to DNA. UV light, cis-Pt, H C H O and two chromium compounds were tested for their ability to cause stable D N A - p r o t e i n crosslinks following exposure of intact C H O cells or isolated nuclei. The formation of D N A - p r o t e i n crosslinks was assessed by measuring the reactivity of an anti-histone monoclonal antibody and an antiserum to a 97-kD nuclear protein with purified D N A - p r o t e i n crosslinks. All 5 agents crosslinked p97 to DNA, however, the two Cr compounds crosslinked p97 to D N A under specific conditions. Cr 3+ formed p97-DNA crosslinks only in isolated nuclei, while chromate strongly induced crosslinks in intact living cells. However, the chemical stability and bonding of p97 to D N A differed among the 5 agents. H C H O produced immunologically
detectable h i s t o n e - D N A crosslinks, whereas the crosslinks induced by metals contained no detectable immunoreactivity with the anti-histone antibody.
Materials and methodology Formaldehyde and potassium chromate were from Fisher Scientific and J.T. Baker respectively. Ultraviolet light with a wavelength of 254 nm was generated from a General Electric lamp (No. G15T8) that produced a fluence of 2.8 j / m 2 / s e c at a distance of 12.5 cm. Chromium chloride hexahydrate and cis-Pt were obtained from Sigma Chemical Company. The routine culture of Chinese hamster ovary cells (CHO) has been described in earlier studies (Miller and Costa, 1988, 1989). Cells were cultured in alpha minimal essential medium and 10% fetal bovine serum (Hazelton) containing 100 U / m l penicillin and 100 # g / m l streptomycin (Gibco). Treatment of C H O cells with cis-Pt, CrC13 . 6H20, or K2CrO 4 was for 24 h in complete medium. H C H O treatments were for 1.5 h in medium without serum. Ultraviolet irradiation of cells was performed in 60-mm tissue culture dishes (Costar) containing 0.5 ml phosphate-buffered saline. For in vitro exposures, nuclei were prepared by hypotonic lysis in 0.5% nonidet-P 40 detergent (Miller and Costa, 1989) and were exposed on ice in 0.34 M sucrose containing 10 mM T r i s - H C l , 5 mM MgC12, and 1 mM phenylmethylsulfonylfluoride (PMSF), p H = 7.5, at a concentration of approx. 5 × 107 nuclei per ml. Nuclei were treated for 24 h with 0.01-1 mM metal compound in the sucrose solution (vide supra). Isolated nuclei were exposed for 1.5 h at concentrations up to 13 m M H C H O . Nuclei were irradiated with UV light exposures up t o 4 k J / m 2 in 60-mm petri dishes containing 1.5 × 108 nuclei suspended in 1 ml of the sucrose solution. Cells utilized for cytotoxicity assays were harvested after exposure to crosslinking agents and seeded at 100 cells/dish for colony formation. Dye exclusion assays (0.2% trypan blue dye in buffered saline) were carried out in an aliquot of these cells. Cells for D N A - p r o t e i n crosslink isolation were washed in phosphate-buffered saline and scraped
99 from 150-mm dishes using a rubber policeman. The nuclei were isolated by hypotonic lysis in nonidet-P 40 detergent, purified through a discontinuous sucrose gradient, and lysed in 35 ml of a 2% SDS solution containing 1 m M PMSF buffered with 10 m M Tris-HC1, p H = 7.5. Samples treated with cis-Pt contained 1 m M thiourea to prevent monofunctional cis-Pt adducts from forming crosslinks during all fractionation procedures. The nuclear lysates were rocked for 4 h on a platform shaker and then homogenized using a tight-fitting teflon tipped homogenizer. The material was sedimented in a SW-27 rotor (Beckman) for 16 h at 1 0 0 0 0 0 × g and the resulting pellet was resuspended in 5 M urea with 1 m M PMSF at ice-bath temperature. The crosslinks were shaken for 4 h on ice and homogenized again as described above. The SDS concentration was adjusted to 2% and the complexes were centrifuged again for 16 h at 100000 X g. The D N A pellet was suspended in a solution of 10 m M Tris-HC1 p H = 7.5 containing 1 m M PMSF, disrupted by sonication, precipitated with cold acetone, and then resuspended a second time by sonication in 10 m M Tris-HC1 ( p H = 7.5) with 1 m M PMSF. The concentration of the crosslinked material was determined by measuring absorbance at 260 nm. Equivalent units (0.05 A260 units/lane) were separated on 1% agarose gels, stained with ethidium bromide, and the fluorescence was compared to ensure the D N A concentration of the D N A - p r o t e i n crosslink was correctly estimated (Miller and Costa, 1988). Crosslinks were isolated from nuclei exposed in vitro by a similar procedure. The production and characterization of the rabbit antiserum to chromate-induced D N A - p r o t e i n crosslinks has been previously described in detail (Miller and Costa, 1989). Analysis with this antiserum, control rabbit serum, anti-histone monoclonal antibody (Chemicon, E1 Segundo, CA), and a non-specific monoclonal antibody (to SV-40 T antigen) was performed by slot blotting (Hawkes et al., 1982). 1-25 /~g of D N A containing the crosslinked proteins was applied to a nitrocellulose filter using a Schleicher and Schuell Minifold II apparatus (Keene, NH). Supernatants from the nonidet-P 40 nuclei isolation procedure (described above) were used as the source of cytoplasmic proteins and nuclear proteins came from nuclei
preparations that were disrupted by 10 sec of sonication. After binding of the crosslinked material, the filter was blocked in a solution of 2% Carnation non-fat dry milk, 25 m M Tris-HC1, 1% NaCI, 0.05% tween-20, and 0.05% N a N 3, p H = 7.5, for at least 2 h at room temperature. Rabbit antiserum was added to the same blocking solution at dilutions of 1/200 (0.25 ml s e r u m : 5 0 nil buffer) while monoclonal antibodies were added at 1/500 dilutions. Antibody binding proceeded for at least 3 h and then unbound antibodies were removed with 4 washes of 5 min each in blocking buffer without the non-fat dry milk. A 1/2000 dilution of peroxidase conjugated goat anti-mouse immunoglobulin or anti-rabbit immunoglobulin antiserum (Boehringer Mannheim Biochemical) was added in blocking buffer and reacted as described above. The secondary antiserum was washed out and the murine antibody binding was visualized by a peroxidase reaction following addition of 60 ml PBS solution containing 4-chloronaphthol (0.5 m g / m l final concentration) and hydrogen peroxide (0.01%). Color development was complete after approx. 15 min of agitation on a platform shaker. Detection of rabbit antibody binding to filters was performed by the peroxidase reaction or by addition of 105 cpm 125I protein A (New England Nuclear) in blocking buffer as described above. The unbound protein A was removed by washing and the filter was air-dried and exposed to film for autoradiography. To quantitate film exposure, the autoradiographs were scanned with an LKB laser densitometer. Detection of p 9 7 - D N A interactions after western blotting has been previously described (Miller and Costa, 1989). In selected instances, purified crosslink material was digested with 1/~g D N A a s e I (Boehringer Mannheim Biochemical) per 100/xg D N A for 1 h. The samples were precipitated in cold acetone and the precipitates were resuspended in SDS electrophoresis sample buffer containing 2-mercaptoethanol (2%). The samples were electrophoretically separated in a 12.5% SDS gel, and transferred to nitrocellulose (Burnette, 1981). Antibody-antigen reactions were detected using the same methods described for slot blots. The nitrocellulose sheet was blocked, incubated with antiserum, reacted with 125I-protein A, dried, and exposed to X-ray film.
+ Crosslink Serum
T h e effects of H C H O , cis-Pt, U V light, a n d c h r o m a t e exposure on c o l o n y f o r m a t i o n is shown in Fig. 1. A 5 J / m 2 e x p o s u r e to U V light p r o d u c e d a 50% r e d u c t i o n in c o l o n y f o r m a t i o n (Fig. 1). cis-Pt was the most toxic chemical tested, reducing c o l o n y f o r m a t i o n to 50% at a t r e a t m e n t of 0.5/~M, while c h r o m a t e a n d f o r m a l d e h y d e r e d u c e d c o l o n y f o r m a t i o n b y 50% at 20 a n d 1450/~M respectively (Fig. 1). C h r o m i u m chloride h e x a h y d r a t e d i d n o t alter colony f o r m a t i o n after 24 h i n c u b a t i o n of C H O cells at c o n c e n t r a t i o n s up to 1 m M ( d a t a not shown). Each agent tested at any of the c o n c e n t r a tions in Fig. 1 d i d not affect cell m e m b r a n e integrity i m m e d i a t e l y after t r e a t m e n t as m e a s u r e d b y t r y p a n blue exclusion ( d a t a not shown). Shown in Fig. 2 is an a u t o r a d i o g r a p h d e m o n strating the differential reactivity of the a n t i s e r u m to c h r o m a t e - i n d u c e d D N A - p r o t e i n crosslinks with either c y t o p l a s m i c or nuclear p r o t e i n fractions. T h e a n t i s e r u m reacted m o r e strongly with nuclear p r o t e i n s than with those in the cytoplasm. C o m p a r i s o n b y d e n s i t o m e t r i c s c a n n i n g i n d i c a t e d the nuclear p r o t e i n fraction was approx. 25-fold richer in reactivity as c o m p a r e d to the cytosolic p r o t e i n
0.6 1.2 1.8 2.4 Formaldehyde (rnM)
15 30 45 C h r e m a f e (uM)
=~ 75+ ~o 50~
o ~ 0
- 0.1 ~g Cytoplasm
- 1.0 I~g Cytoplasm
- 10 #g Cytoplasm
- 0.05 I~g Nucleus
- 0.5 Izg Nucleus
- 5.0 Izg Nucleus
Fig. 2. Enrichment of antiserum reactivity in the nuclear protein fraction. The nonidet-P 40 soluble cytoplasmic proteins and proteins from a lysate of sedimented nuclei were quantitated and applied to nitrocellulose sheets. Antibody reaction with the proteins was followed by binding of ]251-protein A, which was visualized by autoradiography. Duplicate samples were reacted with rabbit sera containing either antibodies to p97 (left panel) or control immunoglobulins (right panel).
oo , ~ 25~
+ Normal Serum
~ , 6 12 18 24 0 2 5 4 n m UVL ( k J / / m 2)
..... 300 cis-Pf
Fig. ]. Inhibition of colony formation by crosslJnking agents.
CHO cells were treated with the amounts of agents shown. After treatment, the cells were detached from dishes using trypsin and seeded at low density for colony formation. The colony-forming efficiency of untreated cells (approx. 80% formed colonies) served as a standard for comparing the proliferative capacity of treated cells. The means and standard errors are shown in the figures.
fraction. T h e right c o l u m n of Fig. 2 is a d u p l i c a t e b l o t that was r e a c t e d with p r e i m m u n e r a b b i t serum. R e a c t i v i t y was only observed when the i m m u n e s e r u m was used, i n d i c a t i n g the d e t e c t i o n signal was d u e to specific a n t i b o d y - a n t i g e n interactions. Shown in Fig. 3 is an a u t o r a d i o g r a p h of a slot b l o t of crosslink m a t e r i a l from either untreated nuclei or f r o m nuclei treated with 1 m M CrC13. T h e slots were l o a d e d with 1 - 2 5 /~g of crosslink s a m p l e as m e a s u r e d b y A 260- This figure shows the a n t i s e r u m reactivity is p r o p o r t i o n a l to the a m o u n t of c r o s s l i n k e d m a t e r i a l l o a d e d on the filter. This assay was very sensitive in distinguishing b e t w e e n c o n t r o l a n d treated material. A s little as 1 /tg of C r 3 + - i n d u c e d crosslinks could be detected. D N A a s e I t r e a t m e n t d i d not e l i m i n a t e the i m m u n o r e a c t i v i t y (fourth row of Fig. 3), whereas p r o t e i n a s e K digestion of samples p r i o r to filter a p p l i c a t i o n resulted in a c o m p l e t e loss of antis e r u m reactivity (fifth row, Fig. 3), i n d i c a t i n g the a n t i s e r u m recognizes a crosslinked p r o t e i n anti-
3 mM HCHO - 1 t.tg D N A
13 mM HCHO
- 5lag D N A
2 kJ/m 2 UVL
- 25~tg D N A
4 kJ/m 2 UVL
- 251xg D N A + D N a s e I
lO0!aM cis Pt - 251xg D N A + Proteinase K
2001aM cis Pt Fig. 3. Detection of Cr ~*-induced crosslinks formed in isolated nuclei. Purified nuclei were treated for 24 h with 1 mM Cr 3+ and DNA-protein crosslinks were isolated as described in the methods. The antibody detection methodology was the same as in Fig. 2. Various amounts of DNA were applied to the filters and nuclease or proteinase digestion of samples was performed prior to filter binding to determine the antigenic reactivity of the antiserum.
gen. Collectively, Figs. 2 a n d 3 clearly show that the a n t i s e r u m to c h r o m a t e - i n d u c e d D N A - p r o t e i n crosslinks recognizes a nuclear p r o t e i n a n t i g e n that becomes complexed to D N A b y Cr 3+ exposure. The a u t o r a d i o g r a p h shown in Fig. 4 illustrates the c o n c e n t r a t i o n d e p e n d e n t effect of Cr 3÷ add i t i o n on p 9 7 - D N A crosslink f o r m a t i o n in isolated nuclei. The differences were statistically sig-
- Contml -200~M -400~M -800~M -IO00~M Fig. 4. DNA-protein crosslinking by increasing Cr ~* exposures in isolated nuclei. 20 /xg of DNA containing crosslinked proteins was applied in triplicate to a nitrocellulose filter using the treatment concentrations indicated. The binding of antibodies to p97 was detected by 12sI protein A and autoradiography. Densitometric scans of the autoradiograph were performed.
7501xM cis Pt Fig. 5. Antiserum detection of p97-DNA crosslinks rormed m isolated nuclei by UV light, formaldehyde, and cis-Pt. Purified nuclei were crosslinked with the various agents by conditions specified in the methods. DNA-protein crosslinks were isolated, applied to nitrocellulose, and assayed as in Fig. 3.
nificant at the two higher Cr 3+ c o n c e n t r a t i o n s when a b s o r b a n c e i n t e g r a t i o n s from d e n s i t o m e t r i c scans were c o m p a r e d b y S t u d e n t ' s t-test ( p < 0.05). The next experiments were p e r f o r m e d to ascertain the ability of the p97 a n t i s e r u m to react with crosslinks that were formed in isolated nuclei by various agents differing in their m e c h a n i s m of crosslink formation. T r e a t m e n t of nuclei with form a l d e h y d e p r o d u c e d p 9 7 - D N A crosslinks that were significantly greater ( p < 0.05) than the control sample (Fig. 5). Similarly, ultraviolet light a n d cis-Pt p r o d u c e d p 9 7 - D N A crosslinks that were d e p e n d e n t u p o n the t r e a t m e n t c o n c e n t r a t i o n . The presence of p97 in the crosslinked material shown in Fig. 5 was c o n f i r m e d by western transfer a n d reaction with p97 a n t i s e r u m (Fig. 6). The autoradiograph shown in Fig. 6 also c o m p a r e s the conditions I required to detect p 9 7 - D N A - p r o t e i n crosslinks ",',in the D N A - p r o t e i n crosslinks i n d u c e d by cis-Ptl U V light a n d H C H O . The p 9 7 - D N A crosslinks i n d u c e d by cis-Pt were released from the D N A a n d resolved b y gel electrophoresis w i t h o u t D N A a s e I treatment. The instability of cis-Pt crosslinks was a t t r i b u t a b l e to the presence of the reducing agent 2 - m e r c a p t o e t h a n o l d u r i n g electrophoresis. The p r o t e i n c o m p o n e n t of D N A - p r o tein crosslinks i n d u c e d by Cr c o m p o u n d s b e h a v e d in a similar m a n n e r as those of cis-Pt (Miller a n d
+ DNase I 13-
> (D T
I 116 66 45 31
Molecular Weight x 10 -3
21 Fig. 6. Comparison of the p97 D N A interactions induced by different crosslinking agents using western blotting and antiserum detection. The purified crosslink material of the slot blot shown in Fig. 5 was run in a 12.5% SDS gel under reducing conditions. Prior to electrophoresis and blotting, some samples were digested with D N A a s e I as indicated, p97 was detected with antiserum and 1251-protein A followed by autoradiography.
Costa, 1989). In contrast to metal-induced crosslinks, p97 crosslinked to D N A by H C H O and UV light was not released from the D N A during electrophoresis (left portion of Fig. 6). The right portion of Fig. 6 shows the release of p97 from H C H O - and UV light-induced crosslinks following treatment of the crosslink with DNAase. This release allows detection of p97 with the antibody as it migrated according to its molecular weight in the gel. In the left portion, no immune reactivity was resolved in the gel using crosslinks formed by UV and H C H O , since p97 was not free to enter the gel. H C H O , cis-Pt, chromium chloride hexahydrate and potassium chromate treatments were given to live C H O cells and compared to p97-DNA crosslinks formed in isolated nuclei. The crosslinks were isolated from cells, bound to nitrocellulose filters and tested for the presence of p 9 7 - D N A protein complexes (Fig. 7). Different amounts of D N A were added to the filter and comparisons
were made by scanning densitometry. All treatments with crosslinking agents (except for Cr 3+, not shown) produced significant increases in p97 crosslinking in living cells ( p < 0.05). Antibody binding was increased when more material was added to the filter and the control samples were weakly positive when a high amount of D N A was applied (20 /,g). The weak control reactivity may represent a limitation in the ability to remove all the loosely bound p97 from D N A or it may be due to small amounts of p97 that are covalently linked to the D N A in untreated cells (also seen in the control of Fig. 4). Fig. 7 also illustrates the reactivity of the antiserum for the p97 antigen was lost when samples were digested with proteinase K but not with DNAase. Additional studies with the antiserum indicated that p97 formed crosslinks with D N A when C H O cell monolayers were irradiated with UV light (Cosma et al., 1989). Chromate induced significant crosslinking of p97 in living cells at concentrations of 50 /,M or
104 chowski et al., 1986a,b; Chiu et al., 1986; Banjar et al., 1983). Actin and a few other crosslinked nuclear proteins were detected using 2-D gels and silver staining after treatments with concentrations of chromate that inhibit colony formation by approx. 60% (Miller and Costa, 1988). The mechanisms for the causes of cytotoxicity produced by these agents are unknown, however, the data presented here suggest that D N A - p r o t e i n crosslink formation is independent of the inhibition of colony formation. A possible mechanism for the cytotoxicity of platinum compounds was attributed to its inhibition of essential m R N A transcripts and this may be applicable to the other agents described here as well (Sorenson and Eastman, 1988). A significant finding of this report is the immunological detection of the in vitro formation of p 9 7 - D N A crosslinks by Cr 3 +. The reaction rate of Cr 3÷ was relatively slow, but once a product formed, it was very stable in the presence of ionic detergents and high salt conditions. These results are consistent with the kinetically inert properties of trivalent chromium. Treatment of living cells with Cr 3+ did not induce crosslinks. Treatment of isolated nuclei with 1 mM chromate (Cr 6+) for 24 h was only weakly effective in increasing the amount of protein associated with D N A but p97 was not detected in these samples (data not shown). The minimal effects of chromate (Cr 6÷) in isolated nuclei was probably due to the generation of Cr 3+ by reduction reactions with the thiol groups of proteins or the ribonucleotides contained in the cell nucleus (Goodgame et al., 1982). Generally, hexavalent chromate is thought to be potent in inducing genetic damage, mutagenesis and carcinogenesis in living cells (Nieboer and Shaw, 1988). In the studies presented here, chromate (Cr 6÷) treatment of cells and Cr 3+ treatment of nuclei both resulted in the same genetic lesion: p97-DNA crosslinks. Evidence of Cr 3÷ genotoxicity when it gains access to the D N A a n d / o r chromatin is apparent from other studies. A prokaryotic transfection system was used to demonstrate that bacteriophage D N A having Cr 3+ adducts was mutated when taken up by the host bacterium (Snow and Xu, 1989). Cr 3+ compounds were also shown to induce D N A - p r o t e i n crosslinks and sister-chromatid exchanges when cells
phagocytized particulate Cr 3+ compounds (Wedrychowski et al., 1986b; Z. Elias et al., 1983). Actin, which is the major protein crosslinked to D N A in intact cells by either chromate or cis-Pt treatment, can also be crosslinked in vitro to D N A by Cr 3+ (C. Miller, M. Cohen and M. Costa, unpublished observation). The acidic p97 protein represents a second example of a nuclear protein that is crosslinked to D N A by Cr 3+. Thus, it is highly probable that Cr 3+ is responsible for some of the D N A - p r o t e i n crosslinks once chromate, the parent compound, is taken up and reduced within the cell. This is consistent with a previously described uptake and reduction model for chromium cytotoxicity (Connett and Wetterhahn, 1983). A number of studies have indicated that distinct groups of proteins are crosslinked to D N A by chemical and physical agents. Metals preferentially crosslink non-histone proteins to DNA, formaldehyde crosslinks histories to DNA, and UV light crosslinks both histone and non-histone proteins (Wedrychowski et al., 1985, 1986a,b; Solomon et al., 1988; Angelov et al., 1988; Bouliakis, 1986). The formation of p97-DNA crosslinks by all the agents tested in this report was surprising, considering the agent specific protein crosslinking reported in these other studies. One likely explanation for p 9 7 - D N A crosslinking by the various agents is that p97 may contain multiple domains of differing amino acid composition. An amine-rich center oriented appropriately toward the D N A could be a region for H C H O crosslink formation, while methionyl, cysteinyl and histidyl groups of another domain could serve as sites for metal-induced crosslinking. Ultraviolet light may crosslink proteins through either of these domains, provided that p97 is in an appropriate orientation and distance with respect to the DNA. The nature of the interactions mediating the crosslinks was addressed in Fig. 6. UV light and H C H O induced p97-DNA crosslinks that were covalent and therefore should remain associated as large molecular weight complexes that could not enter a protein separatory gel unless DNAase I digestion of the complexes was performed prior to electrophoresis. The crosslinks produced by chromium and cis-Pt were mediated by coordinant
- Control - 50p, M c i s - P t - 200gM - 13mM
- Control - 50gM cis-Pt - 2001,tM C h r o m a t e - 13mM
- Control + DNase
- 50pM cis-Pt + DNase - 2001xM Chromate - 13mM
- C o n t r o l + Prot. K - 50gM cis-Pt + Prot. K - 2001,tM C h r o m a t e - 13mM
+ Prot. K Prot. K
Fig. 7. Detection of p97 crosslinks formed in cultured cells using p97 antiserum. CHO cells were treated with metals or formaldehyde and crosslinked material was isolated as described in the methods. Variable amounts of DNA were bound to the nitrocellulose sheets (in triplicate) and antigen detection was achieved with antiserum and lzSI-protein A. In some instances the complexes were digested with DNAase 1 or proteinase K prior to their application to the filter to assess the nature of the antigenic reactivity. Densitometric scanning was used to quantitate X-ray film exposure.
greater (data not shown). Exposure of CHO cells to 1 mM Cr 3+ for 24 h caused no detectable D N A - p r o t e i n crosslinking (data not shown). The reactivity of an anti-histone monoclonal antibody with D N A - p r o t e i n crosslink material isolated from cultured cells is shown in Fig. 8. HCHO, cis-Pt and chromate were compared for their ability to induce histone-DNA crosslinks in living cells. The only reactivity detected by the immunoperoxidase reaction was with the highest H C H O treatment (13 mM). These results indicate that formaldehyde induces histone-DNA crosslinks, but chromate and cis-Pt do not crosslink detectable amounts of histones. The antiserum to p97 appears more sensitive than the anti-histone monoclonal antibody in the detection of crosslinks using slot blotting, even though histones are abundant targets for crosslinking. The reason for the greater reactivity of the antiserum could be due to differences in antibody-antigen affinity, the polyclonal nature of the antiserum, a n d / o r the differences in signal visualization with 125I protein A autoradiography (p97 detection) compared to color detection by immunoperoxidase reaction (histone detection). Another possible explanation for the greater immunological detection of p97 may be that it is intrinsically better than histone as a substrate for the crosslinking agents.
Control 50t.tM cis-Pt
Previous reports from this laboratory have investigated the induction and repair of chromateinduced D N A - p r o t e i n crosslinks and characterized the protein component of the crosslink (Sugiyama et al., 1986; Miller and Costa, 1988). Alkaline elution studies indicate strand breaks and D N A - p r o t e i n crosslinks form at chromate treatments that do not alter the colony-forming ability of CHO cells, although it is not possible to study the crosslinked proteins or the nature of their chemical association with the DNA using these methods (Sugiyama et al., 1986). The crosslinked proteins induced by the exposure of cells to metals, radiation, and other chemical agents have been characterized using biochemical and immunological methods, however, highly lethal treatment conditions were required for the detection of specific proteins in crosslinked material (Wedry-
50~tM Chromate 100~tM Chromate 200~tM Chromate 0.6mM HCHO 3 mM HCHO 13mM HCHO Fig. 8. Detection of histone-DNA crosslinks formed in cultured cells treated with cis-Pt, formaldehyde or chromate. Purified crosslink material generated by the treatment concentrations listed in the figure was evaluated as in Fig. 7 using an anti-histone monoclonal antibody. Antibody binding was amplified with a secondary peroxidase-conjugated goat antimouse IgG serum and visualized by chromogenic reaction.
105 c o v a l e n t b o n d s that were susceptible to d i s r u p t i o n by thiourea or 2 - m e r c a p t o e t h a n o l (Miller and Costa, 1989; W e d r y c h o w s k i et al., 1986a; H o w e G r a n t and Lippard, 1980). C o n s e q u e n t l y , p97 was released f r o m the D N A a n d m i g r a t e d a c c o r d i n g to its m o l e c u l a r weight in a separatory gel in the presence of 2 - m e r c a p t o e t h a n o l , which was disruptive to m e t a l - i n d u c e d crosslinks (Fig. 6). T h e results of this study indicated that detection of crosslinked proteins using a n t i b o d i e s to nuc le a r proteins m a y be a useful m e t h o d ; however, greater sensitivity is n e e d e d for the assessm e n t of crosslinks in samples from h u m a n s expose d to chemical agents. W e are currently trying to i m p r o v e the d e t e c t i o n of crosslinked proteins by p r e p a r i n g m o n o c l o n a l antibodies to nuclear p r o t e i n s which have greater specificity and lower b a c k g r o u n d reactivity than the p97 a n t i s e r u m and the anti-histone m o n o c l o n a l a n t i b o d y used in this study (see Fig. 7 n o t i n g the c o n tr o l background). T h e d e v e l o p m e n t of i m m u n o l o g i c a l reagents of high affinity to the m a j o r proteins that are crosslinked by different agents should e n h a n c e the d e t e c t i o n of D N A - p r o t e i n crosslinks. T h e u l t i m a t e ai m of these e x p e r i m e n t s is to u n d e r s t a n d the biological c o n s e q u e n c e s of D N A p r o t e i n crosslink f o r m a t i o n in the cell. Currently, little is k n o w n a b o u t this lesion o th e r than the fact that it is p r o d u c e d by several c a r c i n o g e n i c agents of e n v i r o n m e n t a l an d o c c u p a t i o n a l significance (Oleinick et al., 1987). A study d e m o n s t r a t i n g that a D N A p o l y m e r a s e c o m p l e x c o u ld not replicate o v e r a D N A - p r o t e i n c o m p l e x that b l o c k e d its procession p r o v i d e s insight into the p o t e n t i a l l y deleterious effects of D N A - p r o t e i n crosslinks (Bedinger et al., 1983). T h e p o t e n ti a l m u t a g e n i c i t y of D N A - p r o t e i n crosslinks m a y be particularly applicable to the p o p u l a r c o n c e p t of suppressor gene deletions as a part of the c a r c i n o g e n i c process (Bouck and Benton, 1989).
References Angelov, D., V. Stefanovsky, S. Dimitriov, V. Russanova, E. Keskinova and I. Pashev (1988) Protein-DNA crosslinking in reconstituted nucleohistone, nuclei, and whole cells by picosecond UV laser irradiation, Nucl. Acids Res., 16, 4525-4538. Banjar, Z.M., L. Hnilica, R. Briggs, J. Stein and G. Stein (1983) Crosslinking of chromosomal proteins to DNA in
HeLa cells by UV, gamma radiation and some antitumor drugs, Biochem. Biophys. Res. Commun., 114, 767-773. Bedinger, P., M. Hochstrasser, C.V. Jongneel and B. Alberts (1983) Properties of the T4 bacteriophage replication apparatus: the T4 dda DNA helicase is required to pass a bound RNA polymerase molecule, Cell, 34, 115-123. Bouck, N.P., and B.K. Benjamin (1989) Loss of cancer suppressors, a driving force in carcinogenesis, Chem. Res. Toxicol., 2, 1-1i. Bouliakis, T. (1986) Protein-protein and protein-DNA interactions in calf thymus nuclear matrix using crosslinking by ultraviolet light, Biochem. Cell. Biol., 64, 474-484. Burnette, W.N. (1981) Western blotting: Electrophoretic transfer of proteins from sodium dodecyl sulfate-polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A, Anal. Biochem., 112, 195-203. Chiu, S., L. Friedman, N. Skokany, L. Xue and N.L. Oleinick (1986) Nuclear matrix proteins are crosslinked to transcriptionally active gene sequences by ionizing radiation, Radiation Res., 107, 24-38. Connett, P.H., and K. Wetterhahn (1983) Metabolism of the carcinogen chromate by cellular constituents, Struct. Bond., 54, 93-124. Cosma, G.N., R. Jamasbi and A. Marchok (1988) Growth inhibition and DNA damage induced by benzo[a]pyrene and formaldehyde in primary cultures of rat tracheal epithelial cells, Mutation Res., 201,161-168. Cosma, G.N., C.A. Miller III and M. Costa (1989) Detection of DNA-protein crosslinks in vitro by an immunological assay, Toxicol. In Vitro, in press. Cupo, D.Y., and K.E. Wetterhahn (1985) Modification of chromium-induced DNA damage by glutathione and cytochromes P-450 in chicken embryo hepatocytes, Proc. Natl. Acad. Sci. (U.S.A.), 82, 6755-6759. Czichos, J., M. Kohler, B. Reckman and M. Renz (1989) Protein-DNA conjugates produced by UV irradiation and their use as probes for hybridization, Nucl. Acids Res., 17, 1563-1572. Elias, Z., O. Schneider, F. Aubry, M.C. Daniere and O. Poirot (1983) Sister chromatid exchanges in Chinese hamster V79 cells treated with trivalent chromium compounds chromic chloride and chromic oxide, Carcinogenesis, 4, 605-611. Goodgame, D.M.L., P.B. Hayman and D.E. Hathaway (1982) Carcinogenic chromium(Vl) forms chromium(V) with ribonucleotides but not with deoxyribonucleotides, Polyhedron, 1,497-499. Hawkes, R., E. Niday and J. Gordon (1982) A dot-immunobinding assay for monoclonal and other antibodies, Anal. Biochem., 119, 142-147. Howe-Grant, M.E., and S.J. Lippard (1980) Aqueous platinum(ll) chemistry; binding to biological molecules, in: H. Sigel (Ed.), Metals in Biological Systems, Vol. 11, Marcel Dekker, New York, pp. 63-125. Miller III, C.A., and M. Costa (1988) Characterization of DNA-protein complexes induced by the carcinogen chromate, Mol. Carcinogen., 1, 125-133. Miller III, C.A., and M. Costa (1989) Immunological detection
106 of DNA-protein complexes induced by chromate, Carcinogenesis, 10, 667-672. Nieboer, E., and A. Jusys (1988) Biologic chemistry of chromium, in: E. Nieboer and J.O. Nriagu (Eds.), Chromium in the Natural and Human Environments, Advances in Science and Technology, Vol. 20, Wiley, New York, pp. 21-79. Nieboer, E., and S.L. Shaw (1988) Mutagenic and other genotoxic effects of chromium compounds, in: E. Nieboer and J.O. Nriagu (Eds.), Chromium in the Natural and Human Environments, Advances in Science and Technology, Vol. 20, Wiley, New York, pp. 399-441. Oleinick, N.L., S. Chiu, N. Ramakrishnan and L. Xue (1987) The formation and significance of DNA-protein crosslinks in mammalian cells, Br. J. Cancer, 55, 135-140. Peak, M.L., and J.G. Peak (1986) DNA-to-protein crosslinks and backbone breaks caused by far- and near-ultraviolet, and visible light radiations in mammalian cells, in: M.G. Simic, L. Grossman and A. Upton (Eds.), Mechanisms of DNA Damage and Repair: Implications for Carcinogenesis and Risk Assessment, Plenum, New York, pp. 193-202. Singer, B., and D. Grunberger (1983) Reactions of directly acting agents with nucleic acids, in: B. Singer and D. Grunberger (Eds.) Molecular Biology of Mutagens and Carcinogens, Plenum, New York, pp. 53-55. Snow, E.T., and L. Xu (1989) Effects of chromium(Ill) on
DNA replication in vitro, Biol. Trace Metals Res., 21, 61-71. Solomon, M.J., P. Larsen and A. Varshavsky (1988) Mapping protein-DNA interactions in vivo with formaldehyde: evidence that histone H4 is retained on a highly transcribed gene, Cell, 53, 937-947. Sorenson, C.M., and A. Eastman (1988) Mechanism of cis-diamminedichloroplatinum(lI)-induced cytotoxicity: role of G2 arrest and DNA double strand breaks, Cancer Res., 48, 4484-4488. Sugiyama, M., S.R. Patierno, O. Cantoni and M. Costa (1986) Characterization of DNA lesions induced by CaCrO 4 in synchronous and asynchronous cultured mammalian cells, Mol. Pharmacol., 29, 606-613. Wedrychowski, A., W.N. Schmidt and L. Hnilica (1986a) The in vivo crosslinking of proteins and DNA by heavy metals, J. Biol. Chem., 261, 3370-3376. Wedrychowski, A., W.N. Schmidt and L.S. Hnilica (1986b) DNA-protein crosslinking by heavy metals in Novikoff hepatoma, Arch. Biochem. Biophys., 251, 397-402. Wedrychowski, A., W.S. Ward, W.N. Schmidt and L. Hnilica (1985) Chromium-induced crosslinking to nuclear proteins and DNA, J. Biol. Chem., 260, 7150-7155. Communicated by M.D. Shelby