An immunoradiometric quantitative assay of Escherichia coli recA protein

An immunoradiometric quantitative assay of Escherichia coli recA protein

MGmoires originaux BIOCH1MIE, 1982, 64, 239-246. An immunoradiometrie quantitative assay of Escherichia coli recA protein. Claude PAOLETTI * , Berna...

657KB Sizes 6 Downloads 35 Views

MGmoires originaux

BIOCH1MIE, 1982, 64, 239-246.

An immunoradiometrie quantitative assay of Escherichia coli recA protein. Claude PAOLETTI * <>, Bernard SALLES ** and Paolo GIACOMONI *** (Refu le 26-12-81, acceptd apr~s rdvision le 4-3-82). * Department of Biochemistry, Stanford University School of Medicine, Stanford, California 94305. On leave from Laboratoire de Biochimie, Institut Gustave Roussy, rue Camille Desmoulins, 94800 Villejuif, France.

** Laboratoire de Pharmacologie et de Toxicologie Fondamentales du CNRS, 205, Route de Narbonne, 31400 Toulouse, France. *** Laboratoire associd du C N R S n ° 147, Unitd I N S E R M n ° 140, lnstitut Gustave Roussy, Rue Camille Desmoulins, 94800 Ville]uif, France.

R~sum~.

Summary.

Un damage radioimmunologique de la protdine recA chez Escherichia coli est ddcrit ; la limite de ddtection est de 0,1 ng de protdine recA ; il donne une rdponse lindaire pour des quantit#s de protdine recA de 0,1 gt 7 ng. II peut ~tre directement employ6 pour des extraits d'Escherichia coli obtenus aprds sonication.

A two-site immunoradiometric assay of Escherichia coli recA protein is described ; its sensitivity allows the detection of 0.1 ng of recA protein ; it yields a linear response for amounts of recA protein in the 0.1-7 ng range. It can be directly applied to extracts of Escherichia coli obtained by sonication.

Des extraits de Salmonella typhimurium donnent une rdaction croisde avec les anticorps anti-recA prot~ine d'Escherichia coli et contiennent donc des ddterminants antig~niques communs avec la prot~ine recA d'Escherichia coll.

Salmonella typhimufium extracts contain some cross-reacting material which share common antigenic determinants with Escherichia coli recA protein but differ from it.

Mots-cl~s : fonetions SOS / protA~ine recA.

Key-words : SOS functions / recA protein.

Introduction.

The recA gene controls homologous recombination [1] and the so called SOS functions in Escherichia coli (see [2] and,,[3] for reviews). The expression of its product, a 37,842 molecular weight protein, is repressed by the product of the lexA gene. RecA protein can be activated by unidentified effector(s) when DNA is damaged or its synthesis arrested [4]. The activated recA protein is capable of removing the lexA protein repressor through a <>To whom all correspondence should be addressed.

proteolytic cleavage [5] ; it therefore stimulates the transcription of its own gene and increases its cellular concentration. This regulatory model was proposed mainly on genetic arguments (reviewed in ref 2) and is now biochemically confirmed [6]. It raises the question of correlating the amount of intracellular recA protein with the level of expression of the functions controled by the recA-lexA genes. No direct quantitative determination of recA protein is available in spite of some preliminary work [7]. A new two19

240

C. Paoletti and coll.

site immunoradiometric assay (IRMA) for quantification of recA protein is described in this paper. Its basis has been outlined previously [8-101. It takes advantage of the tight attachment of rabbit anti-recA protein immunoglobulin G (IgG) to polystyrene tubes under proper conditions. This antibody coated tubes bind recA protein quantitatively. The tubes are then incubated with antirecA protein IgG labelled with radioactive iodine

I

level. Mutations arise from the coordinated expression of multiple functions, a major one being under the control of the umuC gene in E. coli whereas changes of recA protein content seem to depend only on a reduced number of events according to the regulatory model commonly accepted (damages to DNA, activation of recA protein into a protease, lexA repressor inactivating cleavage and transcription and expression of recA gene). It is therefore expected that these events will be made more easily amenable to an in vivo analysis, in connection with the type of D N A damages, through the use of our new immunometrie method, than with the conventional Ames'test.

tll

Material and Methods. Preparation o1 recA protein (E. coli).

FIo. I . -

Diagram oJ the two-site hnmtmoradiomelrfc assay o] recA protein. Symbols : ~ reeA protein (0.I to 10ng) ;......~ unlabelled recA protein lgG (8 to 12ng) ;tl~....~ a'-'~I labelled anti-recA protein IgG (l to 100ng).

so that other determinants of the bound antigen can in turn bind the iodinated antibody (see fig. 1). This assay is used here for measuring the content of recA protein in Escherichia coll. It has been applied to follow the changes of recA protein concentration in bacteria after U V irradiation (B. Salles & C. Paoletti, in preparation) or exposure to some anticancer drugs (B. Salles & C. Lesca ; P. Giacomoni & C. Paoletti, en preparation). Preliminary results aiming at the extention of this assay to Salmonella typhimurium are also described. This bacterial species has been used by Ames et al. [11] for detecting mutagens, and correlatively cancerogens, which are assumed to trigger mutagenic events from damages occuring on D N A . It would be interesting to compare two of the effects of standard mutagens: the first one remote from the initial damages of DNA, i.e. the mutation expression , the second, much closer to these damages, i.e. the variation of recA protein BIOCHIMIE, 1982, 64, n ° 4.

It was a generous gift from Drs. K. McEntee and G. Weinstock. It was prepared according to the method described by Weinstock et al. [12] and subsequently modified by Cox et al. [13]. It was more than 95 per cent (first method) or 99 per cent (second method) pure, as determined by visual inspection of polyacrylamide electrophoresis gels. RecA protein solutions in buffer A (see below) are stable for months at 2°C. Preparation of antibodies against recA protein (E. coli). 600 Fxgof recA protein were suspended in one milliliter of Freund incomplete bacto adjuvant (Difco Lab.). The emulsion was injected intramuscularly into a New Zealand rabbit. A booster injection of an identical emulsion was performed three weeks after the first one. The blood was collected one week later and the serum diluted (1/5) in 0.15 M NaC1. The proteins which had precipitated in the presence of ammonium sulfate (30 per cent w/v) after three hours stirring at 2°C were spun down, redissolved (serum volume × 2) in 20 mM sodium phosphate buffer, pit 7.15, and dialyzed against this buffer. The solution was centrifuged (10,000 rpm at 4oc for I0 minutes); 8 ml of the supernatant were fractionated by chromatography on a mixed column, the bottom of which consisted of 4 ml of diethylaminoethyl cellulose (Whatman, DE52) and the top of which consisted o,f 6 ml carboxymethyl cellulose (Whatman, CM52), equilibrated with 20 mM sodium phosphate buffer, pH 7.15. After loading the proteins, the column was washed with 20 mM sodium phosphate pH 7.15. Immunoglobulins G (IgG) are nol retained by such a column and are eluted in a peak containing about 25 per cent of the total protein load. Protein concentration was determined according to the Lowry technique. It was tmually in the 1-3 mg/ml range. Saturation experiments such as the one described in figure 2 established that about 10 per cent of IgG are anti-recA protein antibodies.

E . c o l i recA protein.

1~5I labeling o] anti-recA protein - - IgG. We used the method first described by Miles et al. [8] and subsequently modified by Mirault et al. [9]. Iodo-Gen (1.3,4,6-tetrachloro-3a, 6ct-diphenyl glycoluril from Pierce Chemicals) was dissolved in CHCI~ to form a 0.02 per cent w / w solution. 20 ~I of this solution were evaporated in a glass tube. 100 ~.g of IgG proteins (the peak described in the preceding section) were added in 180 M of 0.5 M potassium phosphate buffer, pH 8. The iodination reaction was started by the addilion of 2 mCi of Na125I in 20 M (New England Nuclear, specific activity 17 Ci/mg, 100 mCi per ml). The iodination reaction was run for 5 minutes at 0°C with constant stirring and was stopped by addition of 50 ~1 of 10 mM tyrosine in 1 M potassium phosphate buffer, pH 8, and 50 Ixl of 0.1 M NaI in water. The resulting soIution was run on a 3 mI Sephadex G25 column equilibrated with 40 mM sodium phosphate, pH 7.5, 0.15 NaC1, 1 m g / m l bovine serum albumin. Elution rate was 1 ml per minute. 150 ~1 fractions were collected and 1 Ixl was assayed for total and acid insoluble radioactivity. The iodinated proteins containing anti-recA protein IgG were recovered in the first peak (about 50 to 70 per cent of the total ~25I). The specific activity of the labelled antibodies ranged from 0.5 x 106 to 2 × 106 cpm/Ixg. On the average, one molecule of IgG out of eight bound one atom of 125I.

Imrnunoradiornetric Assay (IRMA).

241

part of the polystyrene .tubes with radioactive solution must be avoided. Control tubes without antigen or without cold IgG or with extracts of the bacterial strain 159 recA A21 (lacking endogenous recA protein) were always run in duplicate.

Preparation of bacterial extracts. Wild type E. Coli AB 1157 strain and its mutant 159 recAh21, deleted in the control and structural part of recA gene, were gifts from Dr. K. McEntee. Dr. G. Weinstock also provided K12-W3110 E. Coli and D 2 1 S. typhimurium (wild type). Throughout this work, the cells were grown in L broth (Bacto-tryptone ." 10 g ; bactoyeast : 5 g ; NaC1 : 10 g per liter) at 37°C. Aliquots were withdrawn (1 to 5 ml) and centrifuged. The bacterial pellet was resuspended in IRMA-A and IRMA-B (for total protein content determination) ; the I R M A - A suspension was chilled in ice and sonJcated (30 seconds × 3 ; 50 watts (Branson Sonic Power, Danbury)) ; the sonicate was either used as such (5 ~1 to 100 ~I in assay) or properly diluted out or frozen at - - 7 0 ° C . Cell disruption was monitored under microscope. Portions of the sonicated extracts (5 ~I to 100 p,1) were assayed for total protein concentration and for recA protein. Results were usually expressed in ng of recA protein per ~xg of bacterial proteins. The actual concentration of bacterial cells cannot be measured precisely after treatment by U V or D N A damaging chemicals due to the filamentation. A typical experiment is described in the legend of figure 4.

(a) BuHers: Buffer A was made of 0.01 M sodium barbital - - 0.5 M NaC1 - - 0.003 M NaNa (0.02 per cent) - - 0.1 per cent (w/v) bovine serum albumin - - 0.1 per cent (v/v) calf serum - - pH 7.30. Buffer B was the same as buffer A but deprived of BSA and calf senma. (b) Preparation o] recA protein antibody coated tubes (IRMA tubes). 10 to 15 Ixg of IgG in 0.5 ml of 0.2 M NaHCO, were incubated 24 hours at room temperature in sterile polystyretie tissue-cu~lture tubes (Falcon n ° 2025, 12 × 75 ram, clear without cap). The IgG solution was then decmated, 0.6 ml of I R M A - A buffer was added and the tubes ~vere stored at - - 2 0 ° C . The same solution of IgG in 0.2 M NaHCO8 could be used to e coat >> at least four sets of tubes with IgG. Before use, the tubes were thawed, decanted and rinsed with 0.6 ml of I R M A - A buffer. All decanting was by aspiration. Each tube bound 0.8 per cent of total IgG, as measured with radioactive antibodies. Therefore, each I R M A tube was coated with roughly 8-12 ng of anti-recA protein IgG. (c) Assay. Varying amounts of antigen (recA protein) or various biological extracts in 0.5 ml of I R M A - A were added to the tubes and incubated 4 hours at 37°C. After decanting and rinsing with 0.6 rrd of IRMA-A buffer, 125I~labelled antibodies (2 × 105 to 5 × 105 cpm in 0.5 ml) were added and incubated overnight at room temperature. After decanting and rinsing twice with 0.6 ml of IRMA-A, the amount of bound l~I-labelled IgG was measured in a LK-Minigamma counter. Wetting the upper

BIOCH1MIE, 1982, 64, n ° 4.

Results. A ) SELECTION OF I R M A ASSAY.

STANDARD

CONDITION

FOR

1. O p t i m a l a m o u n t s of 125I antibodies. I n o r d e r to d e t e r m i n e the o p t i m a l c o n c e n t r a t i o n of r a d i o a c t i v e a n t i b o d i e s to b e u s e d in t h e assay, w e m e a s u r e d the f r a c t i o n a n d t h e a m o u n t of r a d i o a c t i v i t y b o u n d to p o l y s t y r e n e t u b e s v e r s u s t h e c o n c e n t r a t i o n of r a d i o a c t i v e a n t i b o d i e s . T h i s e x p e r i m e n t was p e r f o r m e d w i t h two sets of t u b e s p r e c o a t e d w i t h a s a t u r a t i n g a m o u n t of u n l a b e l l e d a n t i - r e c A p r o t e i n I g G . 11 ng a n d 1 1 0 0 n g of r e c A p r o t e i n p e r t u b e w e r e b o u n d to e a c h set, r e s p e c tively. T h e results o f s u c h an e x p e r i m e n t are s h o w n in figure 2 a a n d 2 b. T h e v a l u e s o n the a b s c i s s a e a n d o n the o r d i n a t e s are c a l c u l a t e d f r o m t h e t o t a l r a d i o a c t i v i t y i n p u t a n d r e t e n t i o n , assum i n g t h a t the I g G m o l e c u l e s are h o m o g e n e o u s l y l a b e l l e d , t h a t 10 p e r c e n t of total m o l e c u l e s a r e a n t i b o d i e s against r e c A p r o t e i n , a n d t h a t e a c h antir e c A p r o t e i n I g G m o l e c u l e has the s a m e p r o b a -

242

C. Paoletti and coil. 2. Linearity and sensitivity of the assay.

bility to interact with its antigenic counterpart. For an exces of recA protein (1100 ng per tube) up to 100 per cent of specific antibodies could be bound at low concentrations of labelled IgG ; at high concentrations, 75 per cent was still bound. For amounts of antigen of practical use in this assay, i.e. those allowing a linear dose-response curve (fig. 3), here 11 ng per tube, an increase of the displayed 125I antibody concentration allowed an increase of the amount of measurable radioactivity for the same amount of antigen. However, this advantage is compensated by two drawbacks : a larger consumption of radioactive antibodies, the limiting material in this assay, and an increase of the background level due to nonspecific binding of labelled IgG. Consequently we selected a standard value of 500 ng per ml of total IgG, i.e. 50 ng per ml of specific anti-recA protein IgG. g

.

i

i

i

i

i

i

)(

It appeared from the experiment in figure 2 b that the use of a hundred-fold greater amount of recA protein led to only a five-fold greater retention of labelled antibodies. It was therefore necessary to carefully study the response of the assay versus the amount of added recA protein. Figure 3 shows the variations Of .the radioactivity retained on the IRMA tubes as a function of the amount of recA protein added to the tubes. The response is linear between 0.1 and 7 ng of recA protein (fig. 3 b). The slope of the linear curve depends on the origin of the antibodies and the aging of l~zI labelled IgG. The whole set of our control experiments indicated that 1 ng of specific IgG is bound by amounts of recA protein ranging between 0.36 and 2.08 ng ~data not shown). Above 7 ng of recA protein, the curve levels off (fig. 3 b) ; for amounts of recA protein higher than 300 ng, the I R M A tubes are saturated (fig. 3 a) ; the stoichiometry of these reactions is discussed below. Below 0.1 ng of antigen, variations of the background might disturb the measurements. The unspecific binding of labelled antibodies on IRMA tubes in the absence of recA protein was less than 0.05 per cent of the total radioactivity, which was usually around 5 × 105 cpm.

t 1 1 1 1 1

b

u

.2

/o i~

l 0

I 20

l 40

I 60

Anti-recA

I I I 80 100 120 14~

protein

IgG

¢.7

~

o

. . . . . 40

6(I

80

concentration

1~

120 14(I

(ng/ml)

B) I R M A ASSAY OF recA PROTEIN IN BACTERIA

FIG. 2. (a and b). - - Binding of increasing amounts o/ anti-recA protein lgG to a constant amount o] recA protein antigen. This antigen was previously immobilized o n recA protein I g G coated polystyrene tubes. (@) l l n g , (©) l l 0 0 n g of recA protein antigen. Specific activity of antibodies : 0.27 l~Ci per p,g, i.e. 479 c p m per ng. F o r details see text. 200

tO0

i

---

I

1. Reconstitution experiments. Reconstitution experiments were carried out using extracts of strain 159 recA ± 21 deprived of endogenous recA protein to which were added

300

400

500

I

1

I

c.,*° O

I

"o

20

g

/

aa

O

/



~5

B

10

/

/

S

I

*l 0

.ff 00.1 1

Amount

BIOCHIM1E, 1982, 64, n ° 4.

of recA

protein

/ I 2 (ng)

I 3

I 4

I 5

I 10

FIG. 3. - - Binding of recA prorein IgG as a Junction of the amount of recA protein in the assay. A) Saturation curve : in each assay, 51ng of labelled recA protein I g G (specific activity: 0.29 jxCi p e r ~,g, i.e. 515 c p m per ng). B) Standard curve (initial part o6 the saturation curve) : in each assay, 54 ng o f labelled recA protein I g G (specific activity 0.24 ~Ci per jxg, i.e. 427 c p m p e r ng).

E. coli recA protein. k n o w n amounts of pure E. coU recA protein. A t r e c A protein concentrations of 7.3, 73.5 or 367.5 ng per ml, the recovery of recA protein was complete within --- 20 per cent. W h e n the bacterial protein concentration in the assay was kept under 1 m g / m l , no proteolysis could be observed. M o r e generally, recA protein appears to be remarkably stable : the content of bacterial extracts in r e c A protein was not modified after one m o n t h at 4 ° C in I R M A - B buffer.

2. Unspecific binding. Control experiments carried out on an extract of strain 159 recA ~ 21 showed that a m a j o r part of the b a c k g r o u n d ranging between 0.05 and 0.2 per cent of the total radioactivity arise f r o m material present in this extract, cross-reacting with r e c A protein. This material might arise f r o m proteins contaminating r e c A protein samples (less than 5 per cent) which were used for preparation of recA protein antibodies.

W h e n different concentrations of extracts f r o m non-induced or induced bacteria were a d d e d ins|

tead of purified r e c A protein, the variation of retained radioactivity versus the amount of bacterial extract was linear. Therefore, no variation in the antibody binding activity is to be ascribed to endogenous bacterial effectors or inhibitors (fig. 4). C) CROSS-REACTION BETWEEN E. coli r e c A PROTEIN AND S. typhimurium PROTEINS. T w o experiments were undertaken. 1. The first one was aimed at establishing whether some proteins of S. typhimurium share c o m m o n antigenic determinants with E. coli recA protein. As depicted in figure 5, this is the case because proteins in an S. typhimurium extract were able to bind insolubilized E. coli recA protein antibodies and subsequently to interact with labelled E. coli 6 v

.a

3. L a c k of interfering components.

I

243

$

o

I

.c §

lO

E.co antigen ( n g ) /

0

I

IO0

!

2OO

I

3OO

I

4OO

$,typhi antigen larbitrary

o c

g E <~ o 5

Volume

10

of bacterial

15

extract~jJI)

FIG. 4. - - Relationship between the volume o] bacterial extract in assay and the amount ol bound recA protein antibodies. 1 ~g per ml o~ mitomycin C was added to E. coli AB 1157 strain. The treatment of E. coli by this well known inducer of recA protein was aimed at increasing the concentration of the protein and allowing better detection of it after dilution. After 60 minutes of growth with mitomycin C, 10 ml were removed and immediately added to 5 ml o.f froze,n L broth. The suspension was sonicated as described in Material and Methods and processed to a 3 ml bacterial extract which was assayed undiluted. (0) or after 1 to 10 dilution (©) in IRMA B. Background was 16 cpm. a2~I antibodies had a specific activity of 1070 cpm per ug ; 226800 epm in 0.5 ml were added to IRMA tubes having reacted with increasing volumes of undiluted or diluted bacterial extraots (total volume = 0.5 ml). When no extract was added (0.5 ml of IRMA-B buffer) cotmting was 336 cpm (0.15 per cent of added radioactivity).

BIOCHIMIE, 1982, 64, n ° 4.

I

5OO

I ¢¢..J soo- 75o

unitl

FIG. 5. - - Binding o] S. typhimurium recA-like protein on IRMA tubes. (0) Standard curve ; (©) D21 S. typhimurium extract (protein concentration : 2 mg/ml, reeA-like protein crossreacting proteins : 2.23 ~l*g/ml) ; (A) 159 recA h 21 extract (protein concentration : 4.2 mg/ml). One arbitray unit = 1 V.1 of extract. 125I recA protein IgG had a specific activity of 545 cpm per ~ag. Background level : 20 cpm. When no extract was added (0.5 ml of IRMA-B buffer) counting was 274 cpm (0.08 per cent of total radioactivity).

recA protein antibodies in solution. Extracts of E. coli deleted in the recA gene yielded at saturation a binding which was only one third of that of S. typhimurium extracts. It is therefore unlikely that the E. coli recA-like antigen(s) of S. typhimurium might interact non-specifically. Moreover, there is a linear relationship within the lowest range of concentrations between the volume o f e x t r a c t added in the assay and the a m o u n t of b o u n d 125I antibodies. These observations implied that an immunoradiometric assay

244

C. PaoIetti and coll.

might also be established for measuring the amount of recA-like protein in S. typhimurium. However, it will not be possible to standardize such an assay as long as the E. coli recA-like protein has not been purified from S. typhimurium. 2. However, the previous results suggested that most of the E. coli recA protein binding sites of E. coli antibodies were unable to be recognized by the S. typhimurium proteins. This hypothesis was directly checked in a second experiment (fig. 6). When E. coli antibodies fixed on I R M A tubes were first nearly saturated by S. typhimurium proteins, they were subsequently still able to bind E. coli recA protein (curve B). The extend of this binding is about twice as high when no presaturation of I R M A tubes by S. typhimurium proteins is carried out. It is therefore likely that some S. typhimurium protein(s) inhibit(s) the recA protein antigen-antibody interactions. Such an inhibition was also suggested by a complementary experiment (curve A) in which S. typhimurium proteins where added to I R M A after about 3 / 4 saturation by E. coli recA protein. In this case, the S. typhimurium proteins appeared to mask some of the antigenic sites of the 0

Amount of E.coli recA protein ~ n g ) 100 300 500 700

16

~

A

fixed E. coli recA protein offered to 1~5I antibodies. Any quantitative interpretation of these results would be presently unwarranted as long as an S. typhimurium protein analogous to E. coli recA protein has not been purified and characterized. This work is progressing in our laboratory (A. Pierr6 & C. Paoletti). D) Content of recA protein in wild type E. coli strain. The number of molecules of recA protein per bacterium varied depends on the growth stage of cell population. Values around 2,000 molecules per bacterium were obtained at cell population density of around l 0 s ceils per ml. A t higher densities, when cultures entered into the stationary phase, the abundance of recA protein was slightly diminishing; at cell concentration higher than 109 ceils per ml, 800-900 molecules per cell have been found. Values under 1 000 molecules per cell were also found at cell concentrations below 107 cells per ml. In seven different experiments, E. coli A B l 1 5 7 grown at a cell population density ranging between 3 × 10 r and 7 × 107 cells per ml was found to contain an average value of 800 molecules per cell. In induced cells (figure 4), recA protein concentration were found to be around, 100,000 molecules per cell after one hour of exposure to mitomycine C. It represents 4-5 per cent of the total cell proteins. Similar values were previously published [14].

u

~

B

E o 0 tO0

5 0[0

Discussion.

Amount of S typhi recA protein (arbitcacy unlt~"

FIG. 6. - - Competitive binding of E. coli recA protein and S. typhimuritma recA-like protein on IRMA tubes partially presaturated by one of these antigens. A - - Partial saturation by E. coli recA protein (96 rig) incubated at 37°C for 6 hours in 0.5 ml IRMA-A buffer ; they were then washed by IRMA-A buffer. They were divided in two sets • one set received increasing amounts of recA protein from 0.4ng to 770ng in 0.5 ml of IRMA-A buffer (37°C, 4 hours) (Q) , the other set received increasing volumes of a S. typhimurium extract from 0.17 ~1 to 100 ~1 (*). B - - Partial saturation by S. typhimurium recA like cross-reacting proteins. Same conditions as in A. The value of 96 ng of S. typhimurium recA-like protein was determined according to experiments described in figure 5. Subsequent additions of E. coli recA protein (O) or S. typhimurium extract (*) under the same conditions as in A. 1~5I IgG = 1566 cpm per ng. For other details, see Material and Methods. BIOCHIMIE, 1982, 64, n ° 4.

Any amount of E. coli recA protein larger than 0.1 ng can be measured b y the immunoradiometric assay described in this article. The titration curve is linear in the 0.1-7 ng range. Under our experimental conditions, the background fluctuated between 200 and 400 cpm, so that for recA protein amount higher than 2 ng the experimental error was less than 10 per cent. A stoichiometric molecular binding of E. coli recA protein (MW = 37 842) [15] and its specific I g G (MW _~ 155 000) would allow 1 ng of recA protein to interact with 4 ng of anti-recA protein

IgG. We observed the some discrepancies from this expected ratio: (a) the IgG immobilized on

E. coli recA protein. the polystyrene tubes (8-12 rig) bound at saturation more recA protein (40 rig) than expected ( _~ 3 ng). (b) in contrast, the amounts of labeled antibodies which could interact with recA protein fixed upon the IRMA tubes in the linear part of the curve were slightly less than expected (0.5-2.5 ng instead of 4 ng per ng of antigen). These discrepancies cannot be interpreted since the actual stoichiometry of the antigen-antibody interaction is not known under the experimental conditions of this assay, i.e. binding on a polystyrene surface. In any case, any quantitative treatment of the data would be complicated by the fact that E. coli recA protein is capable of forming aggregates [12] although the low protein and high salt concentrations in this assay do not favor such a process. It should be noted that an aggregation process of recA protein - - if any - - does not seem to depend heavily on the environmental conditions since a pure recA protein solution dissolved in recA A 21 strain extracts, at identical concentrations, yield identical results. This immunoradiometric assay can be directly applied to one step extracts of E. coli obtained through sortication whithout any further treatment. Control experiments have shown that a more elaborate procedure (Brij-EDTA-lysozyme lysis) yielded the same values of recA protein concentration. Although various detergents have been reported to modify antigen-antibody interactions [16], we found that several detergents do not interfere in tiffs assay at concentrations below 0.1 per cent. The recA protein content ranged from 600 to 2 000 molecules per bacterium, following the phase of growth. Some other unknown parameters might come into play to explain these variations. Values of the same order of magnitude have also been reported [17, 18] after fractionation of E. coli proteins on O'Farrell two dimensional isoelectric focusing and electrophoretic polyacrylamide gel [19] and characterization of recA protein on these gels. These values varied from 650 (rich medium) to 1 800 (minimum acetate medium) molecules per bacterial genome at a population density of 1.5 × 108 cells per ml. RecA protein molecules might have different enzymatic functions in vivo. This assay does not discriminate between them. Consequently, the concentration of total recA protein does not always express the functions related to this protein; for instance, the concentration of non activated recA protein in the bacteria is not a critical factor in itself for controling its overall activity; this concentration can be increased by a nearly one hunBIOCHIM1E, 1982, 64, n ° 4.

245

dred-fold factor in the 9, lysogenie strain KM1862pLCL1842 with no correlative enhancement of the susceptibility to X induction [20]. In addition, some ~ lysogenic bacterial mutants such as infA are unable to produce phages in spite of induction of reeA protein after DNA damage [21, 22]. On the other hand, ), induction can proceed without full induction of recA protein synthesis [23, 24]. Cross experiments described in this work establish that S. typhimurium contains at least one protein species which carries common antigenic determinants with E. coli recA protein ; however, this protein is not identical to the E. coli protein in spite of some homology iaa the genotype of these two bacterial strains. Our results indicate than an immunoradiometric assay such as the one described in this paper might be set up as a practical convenient tool for detecting DNA damage and supplementing the information provided by the Ames' test.

Acknowledgements. The authors are grateful to Prof. G. Stark for practical guidance in establishing the tmmunoradiometric assay, to Prof. I. R. Lehman, Dr. G. Weinstock and Dr K. Mc Entee for many valuable discussions and to B. Swyryd for technical help. Useful comments have been made by Drs M. Defais and G. Villani on our manuscript. Part of this work was carried out during the participation of one of us (C.P.) in an exchange visitor program sponsored by NIH under a French-American agreement for cancer research (INSERM). Generous support to this work was also provided by rAssociation Fran£aise pour [e D(veloppement de la Recherche sur le Cancer (Ville]uif, M. J. Crozemarie). One of us (B. S.) & Assistant to the Facultd de Pharmacie, Universitd Paul Sabatier, Toulouse).

REFERENCES. 1. Clark, A. J. & Margtdies, A. D. (1965) Proc. Natl. Acad. Sci. USA, 53, 451-459. 2. Witkin, E. M. (1976) Bacteriol. rev., 49, 869-907. 3. Hanawa'lt, P. C., Cooper, P. J., Ganesan, A. K. Smith, C. A. (1979) Ann. Rev. Biochem., 48, 783-836. 4. Roberts, J. W., Roberts, C. W. & Craig, N. L. (1978) Proc. Natl. Acad. Sci. USA, 75, 4714-4718. 5. Little, J. W., Edmiston, S. H., Pacelli, L. Z. & Moun~, D. W. (1980) Proc. Natl. Acad. Sci. USA, 77, 3225-3229.

246

C. Paoletti and coll.

6. Little, J. W., Mount, D. W. ~ Yanisch-Perron, C. R. (1981) Proc. Natl. Acad. Sci. USA, 78, 4199-4203. 7. Moreau, P. L., Fanica, M. & Devoret, R. (1980) Carcinogenesis, 5, 837-848. 8. Miles, L. E. M., Bieber, C. P., Eng, L. F. & Lipschitz, D. A. (1974) Anal. Biochem., 61, 209-217. 9. Mirault, M. E., Reed, S. I. & Stark, G. R. (1974) Cold Spring Harbor Symp. Quant. Biol., XXXIX, 295-303. 10. Broome, S. ~ Gi~lbert, W. (1978) Proc. Natl. Acad. Sei. USA, 75, 2746-2749. 11. Mc Cann, J., Choi, E., Yamasaki, E. & Ames, B. N. (1975) Proe. Natl. Aead. Sci. USA, 72, 5135-5139. 12. Wein~tock, G. M., Mc Entee, K. & Lehman, I. R. (1979) Proc. Natl. Acad. Sci. USA.. 76, 126-130. 13. Cox, M. M., Mc Entee, K. ,~ Lehman, I. R. (1981) J. Biol. Chem., 256, 4676-4678. 14. Gudas, L. J. & Pardee, A. B. (1976) 1. Mol. Biol., 101, 459-477.

BIOCH1M1E, 1982, 64, n ° 4.

15. Sancar, A., S~aehelek, C., Konisberg, W. ~ Rupp, D. W. (1980) Proc. Natl. Aead. ScL USA, 77, 2611-2615. 16. Crttmpton, M. J. & Parkhouse, R. M. E. (1972) FEBS Lett., 22, 210-212. 17. Pedersen, S., Bloch, P. L., Reeh, S. & Neidhardt, F. C. (1978) Cell, 14, 179-190. 18. Bloch, P. L., Phillips, T. A. & Neihardt, F. C. (1980) J. Bact., 141, 1409-1420. 19. O'Farrel, P. H. (1975) J. Biol. Chem., 250, 4007-4021. 20. Mc Entee, K. (1978) in <> (Hanawalt, P. C., Friec~berg, E. C. & Fox, C. F. eds) 349-360, New-York Acad. Press. 21. Emmerson, P. T. & West, S. C. (1977) Mol. Gen. Genet.. 155. 77-85. 22. Bai'lone, A., Blan¢o, M. & Devoret, R. (1975) Mol. Gen. Genet., 136, 291-307. 23. Moreau, P. L., Fanica, M. & Devoret, R. (1980) Bioehimie, ~ , 687-694. 24. Baluch, J., Sussman, R. & Resnick, J. (1980) Mol. Gen. Genet., 178, 317-323.