- Email: [email protected]
Positive and negative regulation of transcription by a cleavage product of Ada protein
J. Mol. Biol. (1990) 216, 261-273
Positive and Negative Regulation of Transcription by a Cleavage Product of Ada Protein Hiroshi Akimaru, Kunihiko Sakumi, Tomonori Yoshikai, Motoaki Anai~ and Mutsuo Sekiguchi:~ Department of Biochemistry, Faculty of Medicine K y u s h u University 60, F u k u o k a 812, J a p a n (Received 2 March 1990; accepted 16 J u l y 1990) The 39,000 M r Ada protein of Escherichia coli that carries two distinct methyltransferase activities and activity to promote transcription of the ada and the alkA genes is cleaved by a cellular proteinase. As a result, the 20,000 and the 19,000 M r proteins are formed, which are derived from the N-terminal and the C-terminal halves of the protein, respectively. To elucidate the molecular mechanism of transcriptional control by Ads protein, the N-terminal 20,000 Mr protein was overproduced by manipulating the cloned ada gene. The protein possessed an activity to transfer a methyl group from the methylphosphotriester of the alkylated DNA to its own molecule and retained the potential to promote transcription of the alkA gene. The methylated form of the 20,000 Mr proteins binds to the proper allcA regulatory sequence, as does the intact Ada protein, and facilitates further binding of RNA polymerase to the promoter, thus forming an active transcription initiation complex. The non-methylated 20,000 M r protein was incapable of binding itself or supporting RNA polymerase binding to the alkA promoter. When the 20,000 M r protein was produced under the control of the lac promoter in E. coli and then exposed to a methylating agent, a considerable amount of 3-methyladenine-DNA gtycosylase II, the product of the alkA gene, was formed. Thus, the results obtained in in vitro experiments were confirmed by the events observed in vivo. The methylated 20,000 M r protein also binds to the ada promoter; however, it does not facilitate further binding of RNA polymerase to the promoter nor does it promote ada transcription in vitro. These findings indicate that the N-terminal half of Ada protein is mainly responsible for recognition of and binding to alkA and the ada regulatory sequence. The methylated 20,000 M r protein occupies the same region of the ada promoter to which the intact Ada protein would bind, thereby suggesting that it acts as a repressor for expression of the ada gene. The ada transcription promoted by the Ada protein was greatly inhibited by the methylated, but not the non-methylated, form of the 20,000 M r protein. In an in vivo system, formation of the 20,000 M r protein leads to inhibition of transcription from the ada promoter. We suggest that termination of the adaptive response may come about by proteolytic cleavage of the Ada protein, the result being a loss of the activator as well as formation of the repressor for ada transcription.
1. I n t r o d u c t i o n
Treatment of Escherichia coli cells with low doses of simple aikylating agents, such as MNU§, results Visiting scientist from the Department of Medical Technology, School of Health Sciences, Kyushu University, Japan. :~Author to whom all correspondence should be addressed. § Abbreviations used: MNU, N-methyl-N-nitrosourea; MNNG, N-methyl-N'-nitro-N-nitrosoguanidine; MMS, methyl methanesulibnate; IPTG, isopropylfl-D-thiogalactopyranoside; X-gal, 5-bromo-4-chloro-3indolyl-fl-D-galactopyranoside; kb, 103 base-pairs; bp, base-pair(s). 261 0022-2836/90/220261-13 $03.00/0
in induction of formation of a series of enzymes that repair alkylation lesions in DNA (Karran et al., 1982; Evensen & Seeberg, 1982; Demple et al., are 0 -methylguanine-DNA 1982). Included methyltransferase and 3-methyladenine-DNA glycosylase II, coded by the ada (Teo et al., 1984; Nakabeppu et al., 1985) and alkA genes (Nakabeppu et al., 1984a,b; Clarke et al., 1984), respectively. In this process Ada protein, the product of the ada gene, plays a central role; it acts as a transcriptional regulator for its own gene, ada, and other g e n e s belonging to the ada regulon (for a review, see Sekiguchi & Nakabeppu, 1987; Lindahl et al., 1988). © 1990 Academic Press Limited
262
H. A k i m a r u eta;l:
When Ada protein accepts a methyl group- from m e t h y l p h o s p h o t r i e s t e r of alkylated D N A to the residue Cys69 of its own molecule, the protein acquires the potential to bind to the ada p r o m o t e r to enhance transcription of the gene (Sakumi & Sekiguchi, 1989). Once Ada protein is overproduced in this manner, it further p r o m o t e s transcription of the alkA and other genes. Ada protein is cleaved to smaller polypeptides b y a cellular proteinase(s) in vitro (Teo et al., 1984; N a k a b e p p u et al., 1985). On the basis of the N-terminal amino acid sequence of the products and the nucleotide sequence of the ada gene, the m a j o r cleavage site was found to be between L y s l 7 8 and Glnl79 of Ada protein (Demple et al., 1985). The two cleavage products possessed distinct methyltransferase activities, b u t neither p r o m o t e d transcription of the ada gene. Thus, proteolysis of the Ada protein m a y be related to cessation of the a d a p t i v e response. I t was found, however, t h a t the m e t h y l a t e d N-terminal half of protein, b u t not the m e t h y l a t e d C-terminal half, is capable of p r o m o t i n g transcription of the a l k A gene (Yoshikai et al., 1988), t h e r e b y implying t h a t the potential transcription-promoting a c t i v i t y resides on the N-terminal half of the protein. To elucidate the molecular mechanism involved in transcriptional regulation b y Ada protein, we investigated potentials of the N-terminal half of the Ada protein, as a transcriptional regulator. We used footprinting techniques to examine the potential of the intact Ada and 20,000 M r proteins, with or without methylation, to form a transcription initiation complex. Effects of the 20,000 M r protein on expression of the a l k A and ada gene both in vitro and in vivo were also investigated. D a t a obtained support the view t h a t the 20,000 M r protein acts as a repressor for ada transcription while acting as an a c t i v a t o r for alkA transcription.
2. Materials and Methods (a) Enzymes and chemicals Restriction enzymes, bacteriophage T4 polynucleotide kinase, calf intestine alkaline phosphatase, T4 DNA ligase, and the Klenow fragment of DNA polymerase I were purchased from Takara Shuzo Co., Ltd (Kyoto). E. coli RNA polymerase holoenzyme, bovine serum albumin, poly(dA), poly(dT) and calf thymus DNA were obtained from Pharmacia. Heparin, rifampicin, IPTG and X-gal were products from Wako Pure Chemical Industries, L t d (Osaka). Ribosomal RNA was from Boehringer-Mannheim Bioehemieals. [a-a2p]UTP 3 [a- 3 2 p ]d CT P , [~-3 2 P]ATP and [methyl-H]MNU were' purchased from New England Nuclear and Amersham. MNNG and MMS were obtained from Nakarai Chemicals Co., Ltd (Tokyo). (b) Bacteria, and plasmids E. coli strain YN180 (A(ada-alkB) : : kan), constructed by Takano et al. (1988), was used for overproduction of the 20,000 M r protein. Strain WB373 (Barnes et al., 1983) was used for site-directed mutagenesis. Plasmid pYN3089
(Takano et al., 1988) was introduced into appropriate bacterial strains. To produce a truncated ads gene for the N-terminal half of the Ada protein, Ml3 single-stranded DNA carrying the entire ada gene was annealed with a synthetic primer containing the termination codon (TGA) and the EcoRV site, and the complementary strand was synthesized by the DNA polymerase I reaction. After transfection of E. coil WB373, mutant phage were selected and sequenced using the dideoxy method (Sanger et al., 1980). The cloned DNA carrying the termination codon within the ada gene was digested with SalI and EcoRV, and the 263 bp fragment containing the termination codon was isolated and joined to a SalI-SmaI fragment derived from pYN3089. This plasmid, in which the truncated gene can be transcribed under the control of the lac promoter, was termed pHA511. To investigate effects of the 20,000 M r protein on alkA and ada transcription in vivo, plasmids pMF511 and pCM511 were constructed. Plasmid pMF511 was constructed by insertion of a 0"8 kb P v u I I - P v u I I fragment, derived from pHA511 and containing the lac promoter and the truncated gene coding for the 20,000 M r protein, into the EcoRI site of plasmid pMF3 (Manis & Kline, 1977). The same fragment was inserted into the EcoRV site of plasmid pKT410, which contains the chloramphenicol acetyltransferase gene with the ada promoter (Takano et al., 1988). The resulting plasmid was termed pCM511. (c) Site-directed mutagenesis Oligonucleotide-directed in vitro mutagenesis was performed using Amersham's system. Oligonucleotides were chemically synthesized by an Applied Biosystem DNA synthesizer 381A. (d) Purification of 20,000 M, protein The 20,000 M r protein was purified from YNI80 cells harboring plasmid pHA511, by a modification of the procedure for purification of Ada protein (Nakabeppu el at., 1985). Transformed cells were cultured in 1 l of PYG broth (1% (w/v) polypeptone, 0"5% (w/v) yeast extract, 0"01% (w/v) MgSOa. TH20, 0"25% (v/v) glycerol, 0"l M-potassium phosphate (pH 7'0)) containing 50 pg/ml each of ampieillin and kanamycin, at 300C to A66o = 0"3. After addition of IPTG (a final concentration, 5 x l0 -4 M) the culture was incubated at 37 °C for 2 h. The cells were harvested, washed and then suspended in 25 ml of buffer A (20mM-Tris'HCl (pHS'5), 1 mM-EDTA, l mM-2mercaptoethanol, 5% glycerol) and disrupted in an ultrasonic disintegrator. The lysate was centrifuged at 12,000 g for 30 min, and the supernatant was taken as the crude extract (fraction I). After treatment with 0"5 % Polymin-P to remove nucleic acids, the sample was precipitated by addition of (NH4)2SO 4 to 52 % saturation. The precipitate was dissolved in 15 ml of buffer A (fraction II). Fraction I I was applied to a phosphoeellulose P-11 column, and eluted with a linear gradient from 0"15 M to 0"5 M-NaCI in buffer A. Fractions showing a peak of the activity were pooled and concentrated (fraction III), and applied to a Superose-12 column. The eluted fractions were collected and concentrated by Centricon 10 (fraction
IV). (e) Transcription in vitro In vitro transcription experiments were performed as described by Sakumi & Sekiguchi (1989). Purification of
Transcriptional Regulation by Ada Cleavage Product Ada protein and preparation of methylated DNA were performed as described by Takano et al. (1988). DNA templates for in vitro transcription were a 265bp MluI-AccI fragment containing the alkA promoter, a 176bp EcoRI-HindIII fragment containing the ada promoter and a 205bp EcoRI-EcoRI fragment containing the laeUV5 promoter, prepared from pYN3072, pYN3066 and pYN3077, respectively (Nakabeppu & Sekiguchi, 1986). The concentration of Ada protein was determined spectrophotometrically {Nakabeppu & Sekiguchi, 1986). To quantify the 20,000 Mr protein, dye-strained SDS/polyacrylamide gels were scanned, using a chromatoscanner (Shimadzu CS-930). The reaction mixture (35 ~l) contained 50 mMTris'HCI (pH 7"8), 3 mM-magnesium acetate, 0"1 mMEDTA, 0"l mM-dithiothrcitol, 50 mM-NaCl, 25 ~g nuclease-free bovine serum albumin/ml, 3pmol RNA polymerase, 0"3 pmol each of template DNA fragments, methylated or non-methylated 20,000 Mr or Ada protein. Amounts of the 20,000 -SIr and Ada protein are given in the Figure legends. After incubation at 37 °C for 30 rain, 15pl of a pre-warmed solution containing 50pM[~-32P]UTP (2/~Ci/reaetion), 160 pM each of ATP, CTP and GTP, and 200pg heparin]ml were added. The mixture was incubated at 37°C for l0 rain and the reaction was terminated by adding 50~1 of stop solution containing 40 mM-EDTA, 100 #g E. coli ribosomal R,NA/m[. The RNA products were collected by precipitation with ethanol and analyzed by denaturing 10% polyacrylamide gel electrophoresis. In the case of competition experiments, l pmol of the methylated Ada protein was added to a mixture (35 pl) containing 0-3pmol of DNA fragment with the ada promoter, 0 to 7 pmol methylated or non-methylated 20,000 Mr protein and 2 pmo] RNA polymerase. After incubation at 37°C for 30 min, run-off transcription was performed. The transcripts were analyzed by denaturing 10~/o polyacrylamide gel electrophoresis, followed by autoradiography, and quantified using a chromatoscanner. (f) DNase 1 protection assay (footprintinff ) Plasmid pYN3072 containing the alkA promoter region was digested with CfrlOI. For the 5' end-labeling, the DNA fragments were treated with calf intestinal alkaline phosphatase and then with phage T4 polynucleotide kinase and [T-a2P]ATP (:>3000 Ci/mmol). After digestion with AcclII, the 276 bp fragment was purified using 80/o polyacrylamide gel etectrophoresis and recovered with DE81 paper. The 3' end-labeling was performed by filling the (::frl0I site of the DNA fragment with the Klenow fragment and [a-a2P]dCTP. Preparation of the probe for the ada promoter and DNase I protection assay (footprinting) were performed as described by Sakumi & Sekiguchi (1989). (g) Adaptive treatment and in vivo ffene expression To observe in vivo expression of the alkA gene, E. coti YN180 (A(ada-alkB) : : kan) cells harboring plasmid pMF511 were incubated in 100 ml of M9 salt medium (pH 7"4) (Takano et al., 1988) containing ampicillin and kanamycin (50/~g/ml each) at 37°C to A66o = 0"3. After addition of 0"5 mM-IPTG, the cells were cultured at 37°C for 2 h, collected, washed with M9 salt (pH 6-0), and then suspended in 100 ml of M9 salt (pH 6"0) with or without MNNG. For adaptation, the cells were treated with 5 #g MNNG/ml for 10min at 37°C. The adapted and
263
unadapted cells were collected by centrifugation, washed with M9 salt (pH 7-4), suspended in 100 m l o f M9 medium (pH 7"4) and incubated at 37°C. Portions of the cultures were taken at various times and the cells collected and sonicated in 1 ml of buffer A. The iysate was centrifuged at 12,000 g for 30 min and the supernatant was taken as the crude extract. Using 400 pg protein of the crude extract, methyltransferase and 3-methyladenine-DNA glycosylase II activities were determined as described (Nakabeppu et al., 1984a). In vivo expression of the ada gene was investigated as follows. E. coli ABlI57 (ada +) dells harboring plasmid pCM511 or pKT410 were incubated at 37°C to A660 -~ 0"3 in 100 ml of M9 medium (pH 7"4) containing 50 ]~g ampicillin/ml. After addition of 0"005% MMS, the cultures were incubated at 37 °C for 1 h (adaptive treatment). The cells were collected by centrifugation, washed with M9 salt (pH 7-4), and suspended in 100 ml of M9 medium (pH 7"4) containing 1 mM-IPTG, 0-005% MMS. The cells were taken at various times, collected and sonicated in l ml of buffer A. After centrifugation at 12,000g for 30 min, the supernatant was taken as a crude extract. Using 50/~g protein of the crude extract, chloramphenicol acetyltransferase activity was determined as described by Shaw (1975). 3. Results (a) Preparation and characterization of the 20,000 M, protein The N-terminal half of Ada protein (a 20,000 M, protein) has been generated b y cleavage of Ada protein with a cellular proteinase, as used in other studies (Yoshikai et al., 1988). Since the proteolytic cleavage site is between Lys178 and Gln179 of the protein (Demple et al., 1985), we introduced a termination codon (TGA) into the site corresponding to the 179th residue of Ada protein (Fig. 1). To confirm t h a t the plasmid constructed, pHA511, carries the desired sequence, a b o u t a 500 bp region coding for the 20,000 M, protein was sequenced. Cells harboring pHA511 were incubated a t 37°C in medium containing 0"5 mM-IPTG and a large a m o u n t of the 20,000 M r protein was produced. The crude e x t r a c t was treated with P o l y m i n - P to remove nucleic acids, after which column chromatog r a p h y on phosphocellulose was performed. The protein was eluted from the column with a gradient of 0"15M to 0-5M-NaCI and t h e n subjected to Superose-12 gel filtration. Purification was followed by SDS/polyacrylamide gel electrophoresis (Fig. 2) as well as assaying for methyltransferase activity. Table 1 summarizes the result of a typical purification. The enzyme was purified a b o u t 150-fold a t the final step (fraction IV). SDS/polyacrylamide gel electrophoresis revealed t h a t this fraction contained a polypeptide with a molecular mass of 20,000, as the major component. About 0.17 mg of the homogeneous protein was obtained from 3"4 g of cells, wet weight. Among the two methyltransferase activities associated with the intact 39,000 M r Ada protein, the 20,000 M r protein would carry o n l y one of the activities, one to transfer a m e t h y l group from methylphosphotriester of m e t h y l a t e d DNA. To
H. A kimaru et al.
264
M~orproteasedeavage
(a)
1 NH2-Met
--
-- --
Met -ThrATG
(b)
1 N H 2 - Met - - - - - -
178~17g
Ala- Lys-GIn-Phe-Arg
ACG GCT AAA
CAA
TIC
s~e
-His-Gly-Gly
--
--
--
354
Arg - C O O H
CGT CAC GGT GGC
178 Met - T h r - Ala - Lys - COOH ATG
ACG
GCT AAA ~
TAT CGA TAC GGT GGC
t
EcoRV
Figure 1. Construction of a truncated gene coding for the N-terminal half of the Ada protein. To introduce a termination eodon (TGA) next to the codon for the 178th lysine residue, which is located at the major proteolytic cleavage site of Ada protein, a 30-mer oligonucleotide was synthesized, annealed with M13 single-strand DNA carrying the o~a ceding sequence, and the complementary strand synthesized. As a result, the modified gene carried the EeoRV site near the termination codon. (a) The 39,000 _Mr Ada protein and a part of its coding sequence. (b) The oligonucleotide primer used for production of a truncated ada gene, coding for the N-terminal half of the Ada protein. confirm this, we incubated the two types of proteins with [3H]MNU-treated DNA or [aH]MNU-treated poly(dA) annealed with poly(dT), the former carrying 06-methylguanine and methylphosphotriester while the latter only methyiphosphotriester as the major methyl donors. When the reaction products
M
!
D
Trr
Nr (x m"s) 94 ~
67l=
43
30
20
were analyzed by SDS/polyacrylamide gel electrophoresis followed by fluorography, the result shown in Figure 3 was obtained. In all cases, a band corresponding to each protein was detected, but the relative ratios of transfer of methyl groups to the 20,000 Mr and to the intact Ada protein differed. Densitometric scanning revealed that the ratio with methylated poly(dA)-poly(dT) was approximately 1 while the value with methylated DNA was 0-5, as deduced from the supposition that the 20,000 M r protein carries only methylphosphotriester-DNA methyltransferase activity. Further quantitative determinations of methyl transfer to the two types of proteins were made, using methylated DNA as the substrate. From the result shown in Figure 4, it is evident that the 20,000 M r protein accepts the methyl group to the extent of about half of that accepted by the intact Ada protein. (b) Promotion of alkA transcription by the 20,000 M, protein in vitro
"q
1 4 . 4 "~
Figure 2. SDS/polyacrylamide gel electrophoresis of the 20,000 Mr protein. Samples were run in 12-5% polyacrylamide gels containing SDS at 25 mA for 4 h and stained with Coomassie brilliant blue. Each lane contained the following: lanes I and II, 40/~g protein of fraction I and II; lanes III and IV, l0 pg protein of fraction III and IV; lane M, molecular mass markers: phosphorylase b (94,000Mr), bovine serum albumin (67,000Mr), ovalbumin (43,000Mr), carbonic anhydrase (30,000Mr), soybean trypsin inhibitor (20,000 Mr), and ce-lactalbumin (14,400 Mr). An arrow indicates the 20,000 M, protein.
We performed an in vitro transcription experiment to measure the transcription-promoting activity of the 20,000 M r protein. A DNA fragment containing the alkA promoter region, together with a fragment containing the lacUV5 promoter as a control, were incubated with E. coli RNA polymerase holoenzyme and the purified preparations of Ada and 20,000 M r proteins. To methylate the proteins, MNU-treated DNA was included in the reaction mixture. Transcription was initiated by adding a mixture of ATP, GTP, CTP and [a-32P]UTP, and the transcription products were analyzed by polyacrylamide gel electrophoresis followed by autoradiography. The result of the in vitro transcription experiment is shown in Figure 5. In agreement with previous data (Nakabeppu & Sekiguehi, 1986), both nonmethylated and methylated forms of Ada protein had an almost equal potential to promote transcription of the alkA gene, yet with the 20,000 Mr protein
Transcriptional Regulation by Ada Cleavage Product
265
Table 1
Purification of the 20,000 M, protein from an extract of E. coli YN180 harboring pHA511 Fraction I II Ill IV
Crude extract Polymin-P Phosphocellulose Superose-I2
Total activity (units)
Total protein (mg)
Specific activity (units/rag)
Purification (fold)
Recovery (%)
14,000 13,900 7670 2380
426 305 2-12 0"48
32"8 45-4 3620 4950
l'O0 1-38 110 151
100 98"6 54.8 17"0
One unit is defined as the activity required to transfer 1 pmol methyl group//Jg protein.
different results were obtained. The m e t h y l a t e d form of the 20,000 Mr protein was as active as the intact Ada protein in promoting transcription of the alkA gene, but the transcription-promoting activity of the non-methylated form of 20,000 M r protein remained at a low level. To investigate the interactions of the 20,000 M r protein with the alkA promoter, we used the footprinting technique. A CfrlOI-AccIII fragment M
I
2
3
4
M r ( x l O "s) 94 67
containing the alkA p r o m o t e r sequence, radioactively labeled in either the lower (transcribed) or the upper (non-transcribed) strand, was incubated with various amounts of the non-methylated or methylated form of the 20,000 M r protein. After a brief exposure to DNase I, the products were analyzed by polyacrylamide gel electrophoresis followed b y autoradiography (Fig. 6). No detectable protection of both upper and lower strands of the alkA promoter region was observed with the non-methylated form of the 20,000 M r protein. On the other hand, the m e t h y l a t e d forms of both intact 39,000 and 20,000 M r proteins protected DNA regions corresponding to positions - 5 2 to - 4 9 and - 4 5 to - 2 9 of the upper strand
4;5
39,000
9
b8 x C
~ 9 1 ' - - 2 0 t 000
m
14"4
i
Figure 3. Methyltransferase activities of the 39,000 Ada and 20,000 Mr proteins. Purified preparations of the 2 types of proteins were incubated with [3H]MNU-treated calf thymus DNA (5"4 × 10a disints/min) or [3H]MNU-treated poly(dA) (7 x 104 disints/min) annealed with poly(dT), at 37°C for 15 rain in 50~1 of MT buffer (70 mM-Hepes- KOH (pH 7"8), 1 mM-dithiothreitol, 5 mM-EDTA). The reaction was terminated by heating at 90°C for 3 min after addition of 20/~l of 250 mM-Tris" HCI (pH 6"8), 20% (v/v) 2-mercaptoethanol, 9"2% (w/v) SDS, 40% (v/v) glycerol. The samples were applied to 15% polyacrylamide/SDS gels. After electrophoresis, the gels were fixed in 30% methanol, 10% acetic acid for 15 min, and [3H]methyl-accepted proteins were detected by fluorography. Lane 1, 20,000M r protein incubated with methylated DNA; lane 2, 39,000 Mr Ada protein with methylated DNA; lane 3, 20,000 M~ protein with methylated poly(dA)'poly(dT); lane 4, 39,000 Mr Ada protein with methylated poly(dA)'poly(dT); M, molecular mass markers as described in the legend to Fig. 2.
o
O
0
8
I 16
I
I
I
24 32 40 Protein (pmol)
I
!
48
56
64
F i g u r e 4. Transfer of m e t h y l groups to the 20,000 and
39,000 M, Ada proteins. Various amounts of purified preparations of the 20,000 and 39,000 Mr Ada proteins were incubated with [3H]MNU-treated calf thymus DNA (5"4x 104 disints/min) at 37°C for 15 rain in 50gl of MT buffer. The reaction mixture was heated at 90°C for 15 min after addition of 100/~l of 1 mg bovine serum albumin/ml and 400 ~1 of 0"8 M-trichloroacetic acid, and the proteins were precipitated. The pellet was rinsed with 500 gl of 5O/o (w/v) trichloroacetic acid and dissolved in 200 gl of 0"1 M-NaOH. The radioactivity was determined in a liquid scintillation counter. ( 0 ) 20,000 M, protein; (O) 39,000 Mr intact Ada protein.
266
H. A kimaru et ~1.
Non-methylated 39 20 IO'~--~l~(pmol)
39 '0
Methytoted 20 I 3 6' '0 I 3
6' (pmol)
<] -{r i ],'
<~] /oc
:
~] ..
:
.
:.::
4 (o)
.i-! 14
. • ,j
o/kA
(b)
Figure 5. Promotion of transcription of the alkA gene in "vitro by the 20,000 M~ protein. For the reaction, 3 pmol of RNA polymerase holoenzyme and various amounts of non-methylated or methylated forms of the 39,000 Ada and 20,000 M~ proteins were incubated at 37°C for 15 min in the mixture described by Nakabeppu & Sekiguchi (1986). The labeled products were analyzed by denaturing 10°/o polyacrylamide gel electrophoresis, followed by autoradiography. The filled and open triangles indicate the 23-nucleotide alkA and the 63-nucleotide lac transcripts, respectively. Amounts (pmol) of the non-methylated or methylated 39,000 (39) Ada and 20,000 (20) proteins added are shown above each column. (a) Non-methylated 39,000 Ada and 20,000 M~ proteins. (b) Methylated 39,000 Ada and 20,000 M r proteins. The number above each lane indicates the amount of protein used (pmol).
a
b
39
20
c
d
Me .'59
Me 20
A•
-I-I- I+l+J
+1,
toe
35 [
1
+11
+I.
lop
-10l =
-21"
-21"
t5 r
-41 '
-41"
-51'
-51"
-61"
-61"
-35 i ~
(e) Fig. 6.
Transcriptional Regulation by Ada Cleavage Product
267
o
b
c
d
59
20
Me 59
Me 2 0
c°l l
['l-I-I+l+l
TA""
iI
+1
+1
-11
-11
-21
-21 -31
(b) Figure 6. Binding of the 20,000 M, protein and RNA polymerase to the alkA promoter. (a) Upper strand. (b) Lower strand. 0"3 pmol of a 276 bp alkA promoter DNA fragment, labeled at (a) the 5' end or (b) the 3' end, was incubated with various amounts of the non-methylated or methylated forms of 39,000 Ada or 20,000 Mr proteins at 22°C for 60 rain, and treated with 20 ng of DNase I for 30 s. The samples were analyzed by denaturing 8% polyacrylamide gel electrophoresis. Positioned numbers at the right sides of the gel patterns indicate base numbers from the transcription initiation site (Nakabeppu & Sekiguchi, 1986). The - l0 and - 3 5 boxes are indicated by open boxes, DNA sequences protected by the proteins are shown by filled boxes, and RNA polymerase-binding sequences by hatched boxes. DNase I-hypersensitive sites are indicated by arrows. G'A and T" C show the sequences determined by the Maxam & Gilbert (1980) method. Amounts (pmol) of the 39,000 (39) Ada or 20,000 _~I~(20) protein, with (+) or without (--) 6 pmol of RNA polymerase, are shown above each column. Panel a, Non-methylated 39,000 M r Ada protein; panel b, non-methylated 20.000 M~ protein; panel c, methylated (Me) 39,000 M~ Ada protein; panel d, methylated (Me) 20,000 11/1,protein.
(Fig. 6(a)) and a region corresponding to positions - 4 9 to --31 of the lower strand (Fig. 6(b)) from DNase I digestion. On interaction with the methylated Ada or 20,000M r protein, DNase I-hypersensitive sites appeared at positions - 2 2 to - 2 0 of the upper strand, but not of the lower strand. The non-methylated Ada protein did not protect the alkA promoter region, though it promoted in vitro transcription of the alkA gene (Fig. 5). We examined the effects of the 39,000 and the 20,000 M r proteins on binding of P~NA polymerase to the alkA promoter. While no protection of the alkA promoter was observed with either RNA polymerase or the non-methylated Ada protein alone, when added together, t h e y resulted in a high degree of protection of both strands (Fig. 6). In the presence of the methylated 39,000 or 20,000 M, protein, which by itself possessed the ability to bind to the alkA promoter, RNA polymerase further extended
the region of protection. A region of the lower strand from - 5 4 to +14, including the sequence already covered by the methylated protein, was protected by RNA polymerase and a DNase I-hypersensitive site appeared at position - 5 5 of the lower strand. When the methylated 39,000 or 20,000 M, protein bound to the promoter, DNase I-hypersensitive sites were produced at - 2 2 to --20 of the upper (non-transcribed) strand (Fig. 6(a)), panels a, c and d). These sites remained sensitive even in the presence of RNA polymerase. The region 5 bp upstream from the DNase I-hypersensitive site of the upper strand could not be covered by RNA polymerase, but the corresponding region of the lower (transcribed) strand was protected. These results suggest t h a t both intact Ada and 20,000 M r proteins m a y cause a slight conformational change in this particular region of DNA; for instance, unwinding of the DNA helix or relaxation of hydrogen bond-linked base-pairs, so as to p r o m o t e inter-
H. A kimaru et al.
268
~ Alo 6 1o olo " 3
zo
R N A pol.
~ lo o OlO 3 61o o 61 Me 2o
+i--
-21 --.~
-31
!
--
41 !:
--51~
--61
Ca)
(b)
Figure 7. Interactions of the Ada protein, 20,000 Mr protein and RNA polymerase with the ada promoter. The DNase I footprinting experiment was performed as described in the legend to Fig. 6. Filled boxes indicate the regions protected by the methylated 20,000 Mr or Ada protein (left) and by both methylated Ada protein and RNA polymerase (right}. DNase I-sensitive sites are indicated by arrows. The - 3 5 and - l 0 boxes are shown by hatched boxes, and the ada regulatory sequence by an open box. Numbers are positioned from the transcription initiation site (Nakabeppu & Sekiguehi, 1986). (a) Non-methylated 39,000 Ada or 20,000 Mr protein. (b) Methylated (Me) 39,000 (39) Ada or 20,000 Mr (20) protein. action of the RNA polymerase with the promoter. We observed only a weak protection by RNA polymerase when the non-methylated 20,000 Mr protein was present in the reaction mixture. (c) Repression of ada transcription by the methylated
20,000 M, protein Footprinting analyses were made to examine interactions of the 20,000 M r protein with the ada
promoter. A HindIII-EcoRI DNA fragment containing the ada promoter sequence, radioactively labeled at the 5' end of the non-transcribed strand, was incubated with various amounts of the Ada or 20,000 Mr protein, and after a brief exposure to DNase I, the products were analyzed by denaturing gel electrophoresis (Fig. 7). The methylated but not the non-methylated form of intact Ada protein bound to the ada promoter, in agreement with reported d a t a (Sakumi & Sekiguchi, 1989). A similar
269
Transcriptional Regulation by A d a Cleavage Product o ro ¢u
Me 5 9
I,,,,,6 1 6 1 0 o.io.31
31o
Me20 I 3 6 1
I ,3 91Me20 1 1 1 Me59
~[ ada
"9[ Io c
(a)
(b)
Figure 8. Effect of the 20,000 Mr protein on ada transcription. (a) Inability of the 20,000 M~ protein to induce ada transcription. 0"3 pmo] each of the DNA fragments containing either the ada or the lacU V5 promoter, 3 pmol of RNA polymerase and various amounts of the methylated or non-methylated form of Ada (39) or 20,000 Mr (20) protein were incubated at 37°C for 60 rain, followed by run-off transcription for 6 rain. The transcripts from the ada and lacUV5 promoter are indicated by arrows. Amounts (pmol) of Ada and 20,000 M r protein added are indicated on each lane. (b) Effect of methylated (Me) 20,000 Mr protein on ada transcription. Transcription was performed in the presence of 3 pmol of RNA polymerase and l pmol of methylated Ada protein. To each tube the indicated amounts (pmol) of methylated 20,000 Mr protein were added.
but more striking effect of methylation was observed with the 20,000 Mr protein. In the presence of the methylated 20,000 M r protein, DNA regions corresponding to residues - 6 3 to - 4 5 and - 4 4 to --38 of the strand including the ada regulatory sequence, AAAGCGCA (Nakamura et al., 1988), were protected from DNase I digestion (Fig. 7(b)). On the other hand, no evident protection was observed with the non-methylated 20,000 Mr protein (Fig. 7(a)). Thus, the methylated form of the 20,000 Mr protein apparently possesses the potential to bind to the ada promoter. Because the regions protected by the methylated 20,000 M r protein were identical to those protected by the methylated Ada protein, the essential region required for the specific binding should be present in the 20,000 Mr protein. There is another indication of interaction of the methylated 20,000 Mr protein with the promoter. An intense signal for the DNase I-hypersensitive site appeared at position - 3 0 of the strand when the DNA was incubated with the methylated form of 20,000 Mr protein. Only a weak signal was detected in the same region with the non-methylated 20,000Mr protein. Although an additional
DNase I-sensitive site appeared at - 2 1 when the DNA was incubated with the methylated Ada protein, this site was not formed with the methylated 20,000 Mr protein. The disappearance of the distant DNase I-sensitive site may be caused by a decrease in volume of the protein. As reported (Sakumi & Sekiguchi, 1989), the methylated Ada protein promoted the binding of RNA polymerase to the ada promoter (Fig. 7(b)). Although the methylated 20,000 M r protein bound to the ada promoter as did the methylated Ada protein, the former apparently lacked the potential to promote the binding of RNA polymerase to the nearby region of the promoter. I t is likely that the C-terminal domain of Ada protein may make physical contact with RNA polymerase to form a stable transcription-initiation complex on the weak promoter. The observation described above raises the question of whether the 20,000 M r protein competes with the Ada protein in binding to the ada regulatory sequence, thus causing inhibition of expression of the ada gene. To examine this possibility, an in vitro transcription experiment was performed (Fig. 8).
270
H. A k i m a r u et al.
'2° I .o I00~
80
e~
"c 60 o c
~- 4 0
2O
0
4
6
of 20,O00M,
protein
2 Amounts
8 (pmol)
Figure 9. Inhibition of Ada protein-pronmted ada transcription by the methylated form of the 20,000 M, prorein. In vitro transcription was performed as described in Materials and Methods: 0"3 pmol of the template DNA, 1 pmol of methylated Ada protein and 2 pmol of RNA polymerase were incubated with 0 to 7 pmol of methylated or non-methylated forms of the protein, and then run-off transcription was performed. The autoradiogram was scanned for quantification of the transcripts. (O) Non-methylated protein; ( l ) methylated protein.
The methylated form of Ada protein efficiently promoted ada transcription while the non-methylated Ada protein exerted little effect, in agreement with our previous observations (Nakabeppu & Sekiguchi, 1986). Both non-methylated and methylated forms of the 20,000/t4, protein showed no potential to promote transcription of the ada gene (Fig. 8(a)). I t was further demonstrated that ada
transcription, promoted by the methylated Ada protein, can be repressed when the methylated form of the 20,000 Mr protein is present in the reaction mixture (Fig. 8(b)). To correlate the binding capacity of the 20,000 Mr protein to the potential to inhibit ada transcription, we compared the effects of the methylated and nonmethylated forms of the 20,000 M, protein, the latter not binding to the ada promoter. A mixture containing 0"3 pmol of the template DNA with the ada promoter region, 1 pmol of methylated Ada protein, 2 pmol of RNA polymerase and various amounts of methylated or non-methylated forms of 20,000 M, protein, was incubated and run-off transcription was performed. As shown in Figure 9, the methylated form of the 20,000 Mr protein effectively inhibited ada transcription, whereas the nonmethylated form exerted no such effect. This clearly indicates that the potential of the 20,000 Mr protein to inhibit ada transcription depends on the state of methylation, and that this inhibition is caused by competitive binding of two types of proteins, the functional intact Ada protein and the non-functional 20,000 Mr protein, to the same region of the promoter. (d) Effects of the 20.000 M, protein on expression of
the alkA and ada gene in vivo
To determine whether or not the 20,000 Mr protein regulates transcription of tile alkA and ada gene in vivo, we used systems in which the 20,000 Mr protein is supplied artificially under the control of the lac promoter. 50
o --
~
I 0
-~ 20 c
Z
P Io s m i d
~ ~
P:Ioc 20,000 ~////////////////////A
._
IO
_
.oodl..-o.-..-~ ........ di~~o..**~............ ...°=:."i ..~:.*" "..&........ &" %%%i......-+"
o
•
~ A 3-Melhylodenine-ONA ? glycosyloseD.
"~ A odo
Chromosome
)(
"~ ~E,
J 0
P: olk.4 olkA ....................
(o)
I
20
,
u
I
,
40
u
I
60
i
I
80
Time (rain) (b)
Figure 10. Promotion of expression of the alkA gene by the 20,000 M, protein in vivo. (a) The system for measuring transcription-activator activity of the 20,000 M~ protein. P:alkA and P:lac indicate the alkA and lac promoter, respectively. (b) Promotion of the alkA expression by the 20,000 M, protein. E. coli YN180 (A(ada-alkB) : : kan) cells harboring either pMF511 or pMF3 plasmid were grown at 37°C to A6~o = 0"3 in M9 medium (pH 7"4), and treated with or without 5/~g MNNG/ml at 37°C for 10 rain in M9 (pH 6-0). Plasmid pMF511 carries the truncated ada gene, coding for" the 20,000 Mr protein, attached to the lac promoter, while the control plasmid pMF3 does not. The unadapted or adapted cells were resuspended in M9 medium (pH 7"4), removed at the time indicated and crude extracts prepared with buffer A: 400 pg of protein of the crude extracts were used for assay of 3-methyladenine-DNA glycosylase II activity. (O) Unadapted cells with pMF511 (20,000 M,); (O) adapted cells with pMF511 {20,000 Mr); (/k) unadapted ceils with pMF3; (A) adapted cells with pMF3.
271
Transcriptional Regulation by A d a Cleavage Product
We constructed a low copy plasmid carrying the region coding for the 20,000 Mr protein. A 0'8 kb P v u I I - P v u I I fragment containing the lac promoter region and the coding region for the 20,000 Mr protein was inserted into the EcoRI site of pMF3, and the resulting plasmid, pMF511, was used to transform E. coti strain YN180 (Aada). Cells harboring pMF511 were grown in M9 medium containing 0"5 mM-IPTG at 37°C for two hours to produce the 20,000 M, protein. The cells were then treated with 5 #g of MNNG/ml for ten minutes and incubated in M9 medium. As illustrated in Figure 10(a), the effect of the 20,000 11/, protein on the alkA promoter is monitored by an activity of 3-methyladenineDNA glycosylase II, a product of the alkA gene. At various times, the cells were collected for assay of 3-methyladenine-DNA glycosylase II activity (Fig. 10(b)). We found that glycosylase II activity increased when cells carrying the 20,000 Mr protein were treated with MNNG, but without this treatment there was no significant increase in enzyme activity. Control cells carrying the vector alone exhibited no increase in the activity, even after MNNG treatment. Thus, the result obtained in vitro is in good accord with the finding in vivo, To investigate effects of the 20,000 Mr protein on ada transcription in vivo, plasmid pCM51l was constructed. It carries the chloramphenicoI acetyltransferase gene attached to the ada promoter and the gene coding for the 20,000 M, protein, under the control of the lac promoter (Fig. l l(a)). E. coli ABll57 cells harboring pCM51 or vector plasmid pKT410 were treated with 50 #g of MMS/ml (adaptive treatment). The cells were grown in the presence of 1 mM-IPTG and acetyltransferase activity was assayed at various times. As shown in Figure I 1(b), expression of the ada promoter, as measured by the acetyltransferase activity, was considerably lower in cells harboring pCM51 l, in which formation of the 20,000 M~ protein is induced, as compared with the level in cells harboring pKT410 that does not contain the 20,000 Mr protein-coding gene. Without adaptive treatment, the enzyme activity in both types of cells was very low and levels were not affected by induction with IPTG (data not shown), It would thus appear that the methylated form of 20,000 Mr protein tunctions in vivo as a repressor for ada gene expression.
A A
A A
t
P:c/do
P:/oc
co/
20,000
Plosmid
•
P:oda
Chromosome
•
odo
¢ ! (a)
80 70
o C"
'- ~. 6 0
50 "E E >"
40
¢J
2C
~
0 "~ IPTG
I
20
I
I
40
!
I
60
!
80
Time (mini ( b)
4. Discussion
Figure 11. Repression of ada expression by the 20,000 Mr protein in vivo. (a)The system for measuring repressor activity of the 20,000 M, protein in vivo. P : ada and P:/ac indicate the ada and lac promoter, respectively. The reporter gene, cat, and the gene coding for the 20,000 M, protein are placed on plasmid pCM511 while the intact Ada protein is produced from the ada gene on the chromosome. -% Positive effects of the intact Ada protein; - ~ , negative action of 20,000 Mr protein. (b) Repression of the ada expression by the 20,000 M~ protein. E. coil ABI157 cells harboring vector piasmid pKT410 (0) or plasmid pCM511 coding for the 20,000 Mr protein (O) were grown in M9 medium to A66o -- 0"3. The cells were incubated in the presence of 50 ~tg MMS/ml for 60min and then transferred to medium containing l mM-IPTG and 0-005~'o MMS. The cells were collected at various times after addition of IPTG, and chloramphenicol acetyltransferase activity was determined.
The 39,000 Mr Ada protein is cleaved to smaller polypeptides in in vitro proteolytic processes (Teo et al., 1984; Nakabeppu et al., 1985; Teo, 1987). The cleavage site is between Lys178 and Gtn179 of Ada protein, as deduced from the N-terminal amino acid sequence of the cleavage products and the nucleotide sequence of the ada gene (Teo et al., 1984; Demple et al., 1985). On incubation of Ada protein with a proteinase in vitro, the activity that promoted transcription of the ada gene decreased while the transcription-promoting activity on the alkA gene remained (Yoshikai et al., 1988). This
suggests that the cleavage prod.uct of Ada protein might act as a transcriptional activator for expression of alkA, but not of the ada gene. In the present work, we investigated the activity of the N-terminal half of Ada protein as the transcriptional regulator both in in vitro and in vivo systems. Using site-directed mutagenesis of the cloned ada gene, the 20,000 Mr protein was overproduced and purified. The protein carried a methyl-
H. Akimaru et al.
272
-35 Non-transcribed --..,, strand ACAAC C G G ~ G ~ , . ~ A ~ Q ~ t ~ G I C G
"I~TrG G C C T I I ~ " ~ ( ~ / C ~ Transcribed -50 -40 strand
-10 .° * * ;GC;GAAAGC CA~CGTCTGA ATAAC~TTTA GGATGAA~AAJ GGAGA
~J ~CGC GTCGCAGACT TATT GCAAAT ACGACTI-I'CG CCTAC'I-I'A'I-rl CCTCT -30 -20 -10 +I +10
Figure 12. Regions of the alkA promoter protected by the methylated form of the 39,000 Ada or 20,000 Mr protein and RNA polymerase. The region protected by 39,000 Ada or 20,000 Mr protein is indicated by shaded boxes, while the region protected by RNA polymerase by open boxes. Bold letters indicate DNase I-hypersensitive sites. Filled circles correspond to the - 3 5 and - l 0 boxes, and + 1 shows the transcription initiation site.
transferase activity that catalyzes transfer of a methyl group from methylphosphotriester of alkylated DNA to its own molecule. (a) Transcriptional regulator activity of the 20,000 M, protein In in vitro transcription experiments, only a methylated form of the 20,000 M r protein exhibited activity to promote transcription of the alkA gene, whereas the intact Ada protein, irrespective of methylation, promoted transcription, in agreement with our findings with the in vitro proteolytic product (Yoshikai et al., 1988). We also found that • the 20,000M r protein, once methylated, can promote alkA expression in vivo. Binding of the intact Ada and 20,000 M r proteins to the alkA promoter has been studied using DNase I footprinting techniques. Since at least 4 bp to the 5' side and 6 bp to the 3' side of the cutting site are covered by the DNase I molecule on one side of DNA double helix (Suck & Oefner, 1986), the protected regions seen with footprinting are larger than that actually covered by the protein. Taking this into account, a region covered by the 20,000 M r domain of Ada protein was estimated, as shown in Figure 12. Site-directed mutagenesis studies revealed that a sequence AAAGCGCA is required for the regulated expression of the ada gene (Nakamura et al., 1988). It is revealed here that this sequence is indeed the binding site for the 20,000 M r protein as well as for the intact Ada protein. An identical sequence exists in the alkA promoter but it overlaps with the predicted - 3 5 box, the putative binding site of RNA polymerase. There is a closely related sequence, AAAGCAAA, in the upstream region in the alkA promoter, and this sequence was indeed protected by binding of either the intact Ada or 20,000 M r protein. Thus, the AAAGCAAA sequence may be the preferred binding site for Ada protein. To gain support for this notion, studies using mutant DNA sequences are in progress in our laboratory. The region protected by the methylated form of 20,000 M r protein was the same as that covered by the methylated intact Ada protein, thereby indicating that the 19,000 Mr domain of Ada protein does not attach to the DNA, and only the 20,000 M r domain is directly involved in binding to the alkA promoter. Essential regions of Ada protein for specific DNA binding must reside on the N-terminal
half of the protein. Since neither the non-methylated form of intact Ada nor the 20,000 Mr protein was capable of binding to the alkA and ada promoter region, methylation of the cysteine residue at position 69 of the protein would cause a subtle and sufficient conformational change of the N-terminal half of Ada protein so as to gain access to the proper site of the promoter. Information of the three-dimensional structures of the methylated and non-methylated forms of Ada protein, particularly those of the N-terminal domains, would help us to understand the molecular mechanisms for the specific binding and subsequent promoter activation. Although the non-methylated form of intact Ada protein could not in itself bind tightly to the promoter, it could promote binding of RNA polymerase to the promoter and as a result the proper ternary complex formed. This is in accord with our finding that the non-methylated form of Ada protein is as active as the methylated form in promoting the alkA transcription in vitro (Nakabeppu & Sekiguchi, 1986). Though the binding potential of the non-methylated form of Ada protein is weaker than that of the methylated form, it appears to be high enough to form a stable initiation complex when RNA polymerase is present. (b) Turn off of the adaptive response by cleavage of Ada protein Ada protein is a positive regulator for the adaptive response to alkylating agents in E. coli. When the protein is methylated, it acquires the potential to bind tightly to the ada regulatory sequence, thereby facilitating further binding of RNA polymerase to the ada promoter (Sakumi & Sekiguchi, 1989). This would turn on the adaptive response to ensure that the rates of synthesis of DNA repair enzymes were accelerated. With regard to turn-off of the adaptive response, the methylated Ada protein is not actively demethylated, thus there has to be a mechanism that will decrease the amount of activated Ada protein. The cleavage of Ada protein might be related to down-regulation of the adaptive response. Neither of a 20,000 and a 19,000 M r protein, derived from the N and C-terminal halves of the Ada protein, respectively, had the potential to promote ada transcription, yet they still carried
Transcriptional Regulation by Ada Cleavage Product Intoct AdO protein
273
formed more abundantly than others repaired by other classes of enzymes (Lindahl et al., 1988). odo
-35
-10
t'IIIA
~//IA
.35
-10
,,,
We thank M. Ohara for comments on the manuscript. This work was supported by grants from the Ministry of Education, Science and Culture of Japan (61065007) and from the Institute of Protein Engineering, Osaka, Japan.
,
20,000 Mr protein
References aria
,, ,
I
Figure 13. Model for promotion of transcription by the Ada protein (above) and repression of transcription by the 20,000 Mr protein (below). ada indicates the ada regulatory sequence of the ada gene. k = l0 a Mr.
distinct methyltransferase activities. Thus, cleavage of the Ada protein would cause a decrease in the level of a positive gene regulator. If the cleavage product binds to the ada promoter without facilitating further binding of RNA polymerase to the nearby region, it would provide a good model for negative modulation of the adaptive response (Karran & Hall, 1988). We have overproduced the N-terminal 20,000 Mr protein and examined its ability to interact with the ada and alkA promoter. The methylated 20,000 Mr protein indeed bound to the ada regulator), sequence, but this binding did not facilitate binding of RNA polymerase nor did it promote ada transcription (see Fig. 13). I n vitro studies revealed that the methylated but not the non-methylated form of the 20,000 Mr protein competes with the methylated Ada protein for the particular binding site within the ada promoter sequence, thus repressing ada transcription. By using an artificially constructed gene expression system, we have shown that the methylated 20,000 M r protein prevents ada transcription also in vivo. However, more rigorous tests may be necessary to gain support for the notion that proteolysis is indeed involved in cessation of the adaptive response. Appropriate mutants either defective in protease activity or with altered amino acid sequences near or at the cleavage sites of Ada protein will be useful. It is of interest to note that the cleavage of Ada protein causes differential effects on expression of the two genes, ada and alkA, belonging to the same regulon, a finding which may relate to the fact that certain types of DNA lesions, which can be repaired by one class of enzymes, are
Barnes, W. M., Bevan, M. & Son, P. H. (1983). Methods Enzymol. 101, 98-122. Clarke, N. D., Kraal, M. & Seeberg, E. (1984). Mol. Gen. Genet. 197, 368-372. Demple, B., Jacobsson, A., OIsson, M., Robins, P. & Lindahl, T. (1982). J. Biol. Chem. 257, 13776-13780. Demple, B., Sedgwick, B., Robins, P., Totty, N., Waterfield, M.D. & Lindahl, T. (1985). Proc. Nat. Acad. Sci., U.S.A. 82, 2688-2692. Evensen, G. & Seeberg, E. (1982). Nature (London), 296, 773-775. Karran, P. & Hall, J. (1988). In Nucleic Acids and Molecular Biology (Eekstein, F. & Lilley, D. M. J., eds), vol. 2, pp. 188-t97, Springer-Verlag, Berlin. Karran, P., Hjelmgren, T. & Lindahl, T. (1982). Nature (London), 296, 770-773. Lindahl, T., Sedgwick, B., Sekiguchi, M. & Nakabeppu, Y. (1988). Annu. Rev. Biochem. 57, 133-157. Manis, J. J. & Kline, B. C. (1977). Mol. Gen. Genet. 152, 175-182. Maxam, A. M. & Gilbert, W. (1980). Methods Enzymol. 65, 499-560. Nakabeppu, Y. & Sekiguchi, M. (1986). Proc. Nat. Acad. Sci., U.S.A. 83, 6297-6301. Nakabeppu, Y., Kondo, H. & Sekiguchi, M. (1984a). J. Biol. Chem. 259, 13723-13729. Nakabeppu, Y., Miyata, T., Kondo, H., Iwanaga, S. & Sekiguchi, M. (1984b). J. Biol. Chem. 259, 13730-13736. Nakabeppu, Y., Kondo, H., Kawabata, S., Iwanaga, S. & Sekiguchi, M. (1985). J. Biol. Chem. 260, 7281-7288. Nakamura, T., Tokumoto, Y., Sakumi, K., Koike, G., Nakabeppu, Y. & Sekiguchi, M. (1988). J. Mol. Biol. 202, 483-494. Sakumi, K. & Sekiguchi, M. (1989). J. Mol. Biol. 205, 373-385. Sanger, F., Coulson, A. R., Barre]l, B. G., Smith, A. J. H. & Roe, B. A. (1980). J. Mol. Biol. 143, 161-178. Sekiguchi, M. & Nakabeppu, Y. (1987). Trends Genet. 3, 51-54. Shaw, W. V. (1975). Methods Enzymol. 43,737-755. Suck, D. & Oefner, C. (1986). Nature (London), 321, 620-625. Takano, K., Nakabeppu, Y. & Sekiguchi, M. (1988). J. Mol. Biol. 201,261-271. Teo, I. (1987). Murat. Res. 183, 123-127. Teo, I., Sedgwick, B., Demple, B., Li, B. & Lindahl, T. (1984). EMBO J. 3, 2151-2157. Yoshikai, T., Nakabeppu, Y. & Sekiguehi, M. (1988). J. Biol. Chem. 263, 19174-19180.
Edited by R. Schleif
Positive and Negative Regulation of Transcription by a Cleavage Product of Ada Protein Hiroshi Akimaru, Kunihiko Sakumi, Tomonori Yoshikai, Motoaki Anai~ and Mutsuo Sekiguchi:~ Department of Biochemistry, Faculty of Medicine K y u s h u University 60, F u k u o k a 812, J a p a n (Received 2 March 1990; accepted 16 J u l y 1990) The 39,000 M r Ada protein of Escherichia coli that carries two distinct methyltransferase activities and activity to promote transcription of the ada and the alkA genes is cleaved by a cellular proteinase. As a result, the 20,000 and the 19,000 M r proteins are formed, which are derived from the N-terminal and the C-terminal halves of the protein, respectively. To elucidate the molecular mechanism of transcriptional control by Ads protein, the N-terminal 20,000 Mr protein was overproduced by manipulating the cloned ada gene. The protein possessed an activity to transfer a methyl group from the methylphosphotriester of the alkylated DNA to its own molecule and retained the potential to promote transcription of the alkA gene. The methylated form of the 20,000 Mr proteins binds to the proper allcA regulatory sequence, as does the intact Ada protein, and facilitates further binding of RNA polymerase to the promoter, thus forming an active transcription initiation complex. The non-methylated 20,000 M r protein was incapable of binding itself or supporting RNA polymerase binding to the alkA promoter. When the 20,000 M r protein was produced under the control of the lac promoter in E. coli and then exposed to a methylating agent, a considerable amount of 3-methyladenine-DNA gtycosylase II, the product of the alkA gene, was formed. Thus, the results obtained in in vitro experiments were confirmed by the events observed in vivo. The methylated 20,000 M r protein also binds to the ada promoter; however, it does not facilitate further binding of RNA polymerase to the promoter nor does it promote ada transcription in vitro. These findings indicate that the N-terminal half of Ada protein is mainly responsible for recognition of and binding to alkA and the ada regulatory sequence. The methylated 20,000 M r protein occupies the same region of the ada promoter to which the intact Ada protein would bind, thereby suggesting that it acts as a repressor for expression of the ada gene. The ada transcription promoted by the Ada protein was greatly inhibited by the methylated, but not the non-methylated, form of the 20,000 M r protein. In an in vivo system, formation of the 20,000 M r protein leads to inhibition of transcription from the ada promoter. We suggest that termination of the adaptive response may come about by proteolytic cleavage of the Ada protein, the result being a loss of the activator as well as formation of the repressor for ada transcription.
1. I n t r o d u c t i o n
Treatment of Escherichia coli cells with low doses of simple aikylating agents, such as MNU§, results Visiting scientist from the Department of Medical Technology, School of Health Sciences, Kyushu University, Japan. :~Author to whom all correspondence should be addressed. § Abbreviations used: MNU, N-methyl-N-nitrosourea; MNNG, N-methyl-N'-nitro-N-nitrosoguanidine; MMS, methyl methanesulibnate; IPTG, isopropylfl-D-thiogalactopyranoside; X-gal, 5-bromo-4-chloro-3indolyl-fl-D-galactopyranoside; kb, 103 base-pairs; bp, base-pair(s). 261 0022-2836/90/220261-13 $03.00/0
in induction of formation of a series of enzymes that repair alkylation lesions in DNA (Karran et al., 1982; Evensen & Seeberg, 1982; Demple et al., are 0 -methylguanine-DNA 1982). Included methyltransferase and 3-methyladenine-DNA glycosylase II, coded by the ada (Teo et al., 1984; Nakabeppu et al., 1985) and alkA genes (Nakabeppu et al., 1984a,b; Clarke et al., 1984), respectively. In this process Ada protein, the product of the ada gene, plays a central role; it acts as a transcriptional regulator for its own gene, ada, and other g e n e s belonging to the ada regulon (for a review, see Sekiguchi & Nakabeppu, 1987; Lindahl et al., 1988). © 1990 Academic Press Limited
262
H. A k i m a r u eta;l:
When Ada protein accepts a methyl group- from m e t h y l p h o s p h o t r i e s t e r of alkylated D N A to the residue Cys69 of its own molecule, the protein acquires the potential to bind to the ada p r o m o t e r to enhance transcription of the gene (Sakumi & Sekiguchi, 1989). Once Ada protein is overproduced in this manner, it further p r o m o t e s transcription of the alkA and other genes. Ada protein is cleaved to smaller polypeptides b y a cellular proteinase(s) in vitro (Teo et al., 1984; N a k a b e p p u et al., 1985). On the basis of the N-terminal amino acid sequence of the products and the nucleotide sequence of the ada gene, the m a j o r cleavage site was found to be between L y s l 7 8 and Glnl79 of Ada protein (Demple et al., 1985). The two cleavage products possessed distinct methyltransferase activities, b u t neither p r o m o t e d transcription of the ada gene. Thus, proteolysis of the Ada protein m a y be related to cessation of the a d a p t i v e response. I t was found, however, t h a t the m e t h y l a t e d N-terminal half of protein, b u t not the m e t h y l a t e d C-terminal half, is capable of p r o m o t i n g transcription of the a l k A gene (Yoshikai et al., 1988), t h e r e b y implying t h a t the potential transcription-promoting a c t i v i t y resides on the N-terminal half of the protein. To elucidate the molecular mechanism involved in transcriptional regulation b y Ada protein, we investigated potentials of the N-terminal half of the Ada protein, as a transcriptional regulator. We used footprinting techniques to examine the potential of the intact Ada and 20,000 M r proteins, with or without methylation, to form a transcription initiation complex. Effects of the 20,000 M r protein on expression of the a l k A and ada gene both in vitro and in vivo were also investigated. D a t a obtained support the view t h a t the 20,000 M r protein acts as a repressor for ada transcription while acting as an a c t i v a t o r for alkA transcription.
2. Materials and Methods (a) Enzymes and chemicals Restriction enzymes, bacteriophage T4 polynucleotide kinase, calf intestine alkaline phosphatase, T4 DNA ligase, and the Klenow fragment of DNA polymerase I were purchased from Takara Shuzo Co., Ltd (Kyoto). E. coli RNA polymerase holoenzyme, bovine serum albumin, poly(dA), poly(dT) and calf thymus DNA were obtained from Pharmacia. Heparin, rifampicin, IPTG and X-gal were products from Wako Pure Chemical Industries, L t d (Osaka). Ribosomal RNA was from Boehringer-Mannheim Bioehemieals. [a-a2p]UTP 3 [a- 3 2 p ]d CT P , [~-3 2 P]ATP and [methyl-H]MNU were' purchased from New England Nuclear and Amersham. MNNG and MMS were obtained from Nakarai Chemicals Co., Ltd (Tokyo). (b) Bacteria, and plasmids E. coli strain YN180 (A(ada-alkB) : : kan), constructed by Takano et al. (1988), was used for overproduction of the 20,000 M r protein. Strain WB373 (Barnes et al., 1983) was used for site-directed mutagenesis. Plasmid pYN3089
(Takano et al., 1988) was introduced into appropriate bacterial strains. To produce a truncated ads gene for the N-terminal half of the Ada protein, Ml3 single-stranded DNA carrying the entire ada gene was annealed with a synthetic primer containing the termination codon (TGA) and the EcoRV site, and the complementary strand was synthesized by the DNA polymerase I reaction. After transfection of E. coil WB373, mutant phage were selected and sequenced using the dideoxy method (Sanger et al., 1980). The cloned DNA carrying the termination codon within the ada gene was digested with SalI and EcoRV, and the 263 bp fragment containing the termination codon was isolated and joined to a SalI-SmaI fragment derived from pYN3089. This plasmid, in which the truncated gene can be transcribed under the control of the lac promoter, was termed pHA511. To investigate effects of the 20,000 M r protein on alkA and ada transcription in vivo, plasmids pMF511 and pCM511 were constructed. Plasmid pMF511 was constructed by insertion of a 0"8 kb P v u I I - P v u I I fragment, derived from pHA511 and containing the lac promoter and the truncated gene coding for the 20,000 M r protein, into the EcoRI site of plasmid pMF3 (Manis & Kline, 1977). The same fragment was inserted into the EcoRV site of plasmid pKT410, which contains the chloramphenicol acetyltransferase gene with the ada promoter (Takano et al., 1988). The resulting plasmid was termed pCM511. (c) Site-directed mutagenesis Oligonucleotide-directed in vitro mutagenesis was performed using Amersham's system. Oligonucleotides were chemically synthesized by an Applied Biosystem DNA synthesizer 381A. (d) Purification of 20,000 M, protein The 20,000 M r protein was purified from YNI80 cells harboring plasmid pHA511, by a modification of the procedure for purification of Ada protein (Nakabeppu el at., 1985). Transformed cells were cultured in 1 l of PYG broth (1% (w/v) polypeptone, 0"5% (w/v) yeast extract, 0"01% (w/v) MgSOa. TH20, 0"25% (v/v) glycerol, 0"l M-potassium phosphate (pH 7'0)) containing 50 pg/ml each of ampieillin and kanamycin, at 300C to A66o = 0"3. After addition of IPTG (a final concentration, 5 x l0 -4 M) the culture was incubated at 37 °C for 2 h. The cells were harvested, washed and then suspended in 25 ml of buffer A (20mM-Tris'HCl (pHS'5), 1 mM-EDTA, l mM-2mercaptoethanol, 5% glycerol) and disrupted in an ultrasonic disintegrator. The lysate was centrifuged at 12,000 g for 30 min, and the supernatant was taken as the crude extract (fraction I). After treatment with 0"5 % Polymin-P to remove nucleic acids, the sample was precipitated by addition of (NH4)2SO 4 to 52 % saturation. The precipitate was dissolved in 15 ml of buffer A (fraction II). Fraction I I was applied to a phosphoeellulose P-11 column, and eluted with a linear gradient from 0"15 M to 0"5 M-NaCI in buffer A. Fractions showing a peak of the activity were pooled and concentrated (fraction III), and applied to a Superose-12 column. The eluted fractions were collected and concentrated by Centricon 10 (fraction
IV). (e) Transcription in vitro In vitro transcription experiments were performed as described by Sakumi & Sekiguchi (1989). Purification of
Transcriptional Regulation by Ada Cleavage Product Ada protein and preparation of methylated DNA were performed as described by Takano et al. (1988). DNA templates for in vitro transcription were a 265bp MluI-AccI fragment containing the alkA promoter, a 176bp EcoRI-HindIII fragment containing the ada promoter and a 205bp EcoRI-EcoRI fragment containing the laeUV5 promoter, prepared from pYN3072, pYN3066 and pYN3077, respectively (Nakabeppu & Sekiguchi, 1986). The concentration of Ada protein was determined spectrophotometrically {Nakabeppu & Sekiguchi, 1986). To quantify the 20,000 Mr protein, dye-strained SDS/polyacrylamide gels were scanned, using a chromatoscanner (Shimadzu CS-930). The reaction mixture (35 ~l) contained 50 mMTris'HCI (pH 7"8), 3 mM-magnesium acetate, 0"1 mMEDTA, 0"l mM-dithiothrcitol, 50 mM-NaCl, 25 ~g nuclease-free bovine serum albumin/ml, 3pmol RNA polymerase, 0"3 pmol each of template DNA fragments, methylated or non-methylated 20,000 Mr or Ada protein. Amounts of the 20,000 -SIr and Ada protein are given in the Figure legends. After incubation at 37 °C for 30 rain, 15pl of a pre-warmed solution containing 50pM[~-32P]UTP (2/~Ci/reaetion), 160 pM each of ATP, CTP and GTP, and 200pg heparin]ml were added. The mixture was incubated at 37°C for l0 rain and the reaction was terminated by adding 50~1 of stop solution containing 40 mM-EDTA, 100 #g E. coli ribosomal R,NA/m[. The RNA products were collected by precipitation with ethanol and analyzed by denaturing 10% polyacrylamide gel electrophoresis. In the case of competition experiments, l pmol of the methylated Ada protein was added to a mixture (35 pl) containing 0-3pmol of DNA fragment with the ada promoter, 0 to 7 pmol methylated or non-methylated 20,000 Mr protein and 2 pmo] RNA polymerase. After incubation at 37°C for 30 min, run-off transcription was performed. The transcripts were analyzed by denaturing 10~/o polyacrylamide gel electrophoresis, followed by autoradiography, and quantified using a chromatoscanner. (f) DNase 1 protection assay (footprintinff ) Plasmid pYN3072 containing the alkA promoter region was digested with CfrlOI. For the 5' end-labeling, the DNA fragments were treated with calf intestinal alkaline phosphatase and then with phage T4 polynucleotide kinase and [T-a2P]ATP (:>3000 Ci/mmol). After digestion with AcclII, the 276 bp fragment was purified using 80/o polyacrylamide gel etectrophoresis and recovered with DE81 paper. The 3' end-labeling was performed by filling the (::frl0I site of the DNA fragment with the Klenow fragment and [a-a2P]dCTP. Preparation of the probe for the ada promoter and DNase I protection assay (footprinting) were performed as described by Sakumi & Sekiguchi (1989). (g) Adaptive treatment and in vivo ffene expression To observe in vivo expression of the alkA gene, E. coti YN180 (A(ada-alkB) : : kan) cells harboring plasmid pMF511 were incubated in 100 ml of M9 salt medium (pH 7"4) (Takano et al., 1988) containing ampicillin and kanamycin (50/~g/ml each) at 37°C to A66o = 0"3. After addition of 0"5 mM-IPTG, the cells were cultured at 37°C for 2 h, collected, washed with M9 salt (pH 6-0), and then suspended in 100 ml of M9 salt (pH 6"0) with or without MNNG. For adaptation, the cells were treated with 5 #g MNNG/ml for 10min at 37°C. The adapted and
263
unadapted cells were collected by centrifugation, washed with M9 salt (pH 7-4), suspended in 100 m l o f M9 medium (pH 7"4) and incubated at 37°C. Portions of the cultures were taken at various times and the cells collected and sonicated in 1 ml of buffer A. The iysate was centrifuged at 12,000 g for 30 min and the supernatant was taken as the crude extract. Using 400 pg protein of the crude extract, methyltransferase and 3-methyladenine-DNA glycosylase II activities were determined as described (Nakabeppu et al., 1984a). In vivo expression of the ada gene was investigated as follows. E. coli ABlI57 (ada +) dells harboring plasmid pCM511 or pKT410 were incubated at 37°C to A660 -~ 0"3 in 100 ml of M9 medium (pH 7"4) containing 50 ]~g ampicillin/ml. After addition of 0"005% MMS, the cultures were incubated at 37 °C for 1 h (adaptive treatment). The cells were collected by centrifugation, washed with M9 salt (pH 7-4), and suspended in 100 ml of M9 medium (pH 7"4) containing 1 mM-IPTG, 0-005% MMS. The cells were taken at various times, collected and sonicated in l ml of buffer A. After centrifugation at 12,000g for 30 min, the supernatant was taken as a crude extract. Using 50/~g protein of the crude extract, chloramphenicol acetyltransferase activity was determined as described by Shaw (1975). 3. Results (a) Preparation and characterization of the 20,000 M, protein The N-terminal half of Ada protein (a 20,000 M, protein) has been generated b y cleavage of Ada protein with a cellular proteinase, as used in other studies (Yoshikai et al., 1988). Since the proteolytic cleavage site is between Lys178 and Gln179 of the protein (Demple et al., 1985), we introduced a termination codon (TGA) into the site corresponding to the 179th residue of Ada protein (Fig. 1). To confirm t h a t the plasmid constructed, pHA511, carries the desired sequence, a b o u t a 500 bp region coding for the 20,000 M, protein was sequenced. Cells harboring pHA511 were incubated a t 37°C in medium containing 0"5 mM-IPTG and a large a m o u n t of the 20,000 M r protein was produced. The crude e x t r a c t was treated with P o l y m i n - P to remove nucleic acids, after which column chromatog r a p h y on phosphocellulose was performed. The protein was eluted from the column with a gradient of 0"15M to 0-5M-NaCI and t h e n subjected to Superose-12 gel filtration. Purification was followed by SDS/polyacrylamide gel electrophoresis (Fig. 2) as well as assaying for methyltransferase activity. Table 1 summarizes the result of a typical purification. The enzyme was purified a b o u t 150-fold a t the final step (fraction IV). SDS/polyacrylamide gel electrophoresis revealed t h a t this fraction contained a polypeptide with a molecular mass of 20,000, as the major component. About 0.17 mg of the homogeneous protein was obtained from 3"4 g of cells, wet weight. Among the two methyltransferase activities associated with the intact 39,000 M r Ada protein, the 20,000 M r protein would carry o n l y one of the activities, one to transfer a m e t h y l group from methylphosphotriester of m e t h y l a t e d DNA. To
H. A kimaru et al.
264
M~orproteasedeavage
(a)
1 NH2-Met
--
-- --
Met -ThrATG
(b)
1 N H 2 - Met - - - - - -
178~17g
Ala- Lys-GIn-Phe-Arg
ACG GCT AAA
CAA
TIC
s~e
-His-Gly-Gly
--
--
--
354
Arg - C O O H
CGT CAC GGT GGC
178 Met - T h r - Ala - Lys - COOH ATG
ACG
GCT AAA ~
TAT CGA TAC GGT GGC
t
EcoRV
Figure 1. Construction of a truncated gene coding for the N-terminal half of the Ada protein. To introduce a termination eodon (TGA) next to the codon for the 178th lysine residue, which is located at the major proteolytic cleavage site of Ada protein, a 30-mer oligonucleotide was synthesized, annealed with M13 single-strand DNA carrying the o~a ceding sequence, and the complementary strand synthesized. As a result, the modified gene carried the EeoRV site near the termination codon. (a) The 39,000 _Mr Ada protein and a part of its coding sequence. (b) The oligonucleotide primer used for production of a truncated ada gene, coding for the N-terminal half of the Ada protein. confirm this, we incubated the two types of proteins with [3H]MNU-treated DNA or [aH]MNU-treated poly(dA) annealed with poly(dT), the former carrying 06-methylguanine and methylphosphotriester while the latter only methyiphosphotriester as the major methyl donors. When the reaction products
M
!
D
Trr
Nr (x m"s) 94 ~
67l=
43
30
20
were analyzed by SDS/polyacrylamide gel electrophoresis followed by fluorography, the result shown in Figure 3 was obtained. In all cases, a band corresponding to each protein was detected, but the relative ratios of transfer of methyl groups to the 20,000 Mr and to the intact Ada protein differed. Densitometric scanning revealed that the ratio with methylated poly(dA)-poly(dT) was approximately 1 while the value with methylated DNA was 0-5, as deduced from the supposition that the 20,000 M r protein carries only methylphosphotriester-DNA methyltransferase activity. Further quantitative determinations of methyl transfer to the two types of proteins were made, using methylated DNA as the substrate. From the result shown in Figure 4, it is evident that the 20,000 M r protein accepts the methyl group to the extent of about half of that accepted by the intact Ada protein. (b) Promotion of alkA transcription by the 20,000 M, protein in vitro
"q
1 4 . 4 "~
Figure 2. SDS/polyacrylamide gel electrophoresis of the 20,000 Mr protein. Samples were run in 12-5% polyacrylamide gels containing SDS at 25 mA for 4 h and stained with Coomassie brilliant blue. Each lane contained the following: lanes I and II, 40/~g protein of fraction I and II; lanes III and IV, l0 pg protein of fraction III and IV; lane M, molecular mass markers: phosphorylase b (94,000Mr), bovine serum albumin (67,000Mr), ovalbumin (43,000Mr), carbonic anhydrase (30,000Mr), soybean trypsin inhibitor (20,000 Mr), and ce-lactalbumin (14,400 Mr). An arrow indicates the 20,000 M, protein.
We performed an in vitro transcription experiment to measure the transcription-promoting activity of the 20,000 M r protein. A DNA fragment containing the alkA promoter region, together with a fragment containing the lacUV5 promoter as a control, were incubated with E. coli RNA polymerase holoenzyme and the purified preparations of Ada and 20,000 M r proteins. To methylate the proteins, MNU-treated DNA was included in the reaction mixture. Transcription was initiated by adding a mixture of ATP, GTP, CTP and [a-32P]UTP, and the transcription products were analyzed by polyacrylamide gel electrophoresis followed by autoradiography. The result of the in vitro transcription experiment is shown in Figure 5. In agreement with previous data (Nakabeppu & Sekiguehi, 1986), both nonmethylated and methylated forms of Ada protein had an almost equal potential to promote transcription of the alkA gene, yet with the 20,000 Mr protein
Transcriptional Regulation by Ada Cleavage Product
265
Table 1
Purification of the 20,000 M, protein from an extract of E. coli YN180 harboring pHA511 Fraction I II Ill IV
Crude extract Polymin-P Phosphocellulose Superose-I2
Total activity (units)
Total protein (mg)
Specific activity (units/rag)
Purification (fold)
Recovery (%)
14,000 13,900 7670 2380
426 305 2-12 0"48
32"8 45-4 3620 4950
l'O0 1-38 110 151
100 98"6 54.8 17"0
One unit is defined as the activity required to transfer 1 pmol methyl group//Jg protein.
different results were obtained. The m e t h y l a t e d form of the 20,000 Mr protein was as active as the intact Ada protein in promoting transcription of the alkA gene, but the transcription-promoting activity of the non-methylated form of 20,000 M r protein remained at a low level. To investigate the interactions of the 20,000 M r protein with the alkA promoter, we used the footprinting technique. A CfrlOI-AccIII fragment M
I
2
3
4
M r ( x l O "s) 94 67
containing the alkA p r o m o t e r sequence, radioactively labeled in either the lower (transcribed) or the upper (non-transcribed) strand, was incubated with various amounts of the non-methylated or methylated form of the 20,000 M r protein. After a brief exposure to DNase I, the products were analyzed by polyacrylamide gel electrophoresis followed b y autoradiography (Fig. 6). No detectable protection of both upper and lower strands of the alkA promoter region was observed with the non-methylated form of the 20,000 M r protein. On the other hand, the m e t h y l a t e d forms of both intact 39,000 and 20,000 M r proteins protected DNA regions corresponding to positions - 5 2 to - 4 9 and - 4 5 to - 2 9 of the upper strand
4;5
39,000
9
b8 x C
~ 9 1 ' - - 2 0 t 000
m
14"4
i
Figure 3. Methyltransferase activities of the 39,000 Ada and 20,000 Mr proteins. Purified preparations of the 2 types of proteins were incubated with [3H]MNU-treated calf thymus DNA (5"4 × 10a disints/min) or [3H]MNU-treated poly(dA) (7 x 104 disints/min) annealed with poly(dT), at 37°C for 15 rain in 50~1 of MT buffer (70 mM-Hepes- KOH (pH 7"8), 1 mM-dithiothreitol, 5 mM-EDTA). The reaction was terminated by heating at 90°C for 3 min after addition of 20/~l of 250 mM-Tris" HCI (pH 6"8), 20% (v/v) 2-mercaptoethanol, 9"2% (w/v) SDS, 40% (v/v) glycerol. The samples were applied to 15% polyacrylamide/SDS gels. After electrophoresis, the gels were fixed in 30% methanol, 10% acetic acid for 15 min, and [3H]methyl-accepted proteins were detected by fluorography. Lane 1, 20,000M r protein incubated with methylated DNA; lane 2, 39,000 Mr Ada protein with methylated DNA; lane 3, 20,000 M~ protein with methylated poly(dA)'poly(dT); lane 4, 39,000 Mr Ada protein with methylated poly(dA)'poly(dT); M, molecular mass markers as described in the legend to Fig. 2.
o
O
0
8
I 16
I
I
I
24 32 40 Protein (pmol)
I
!
48
56
64
F i g u r e 4. Transfer of m e t h y l groups to the 20,000 and
39,000 M, Ada proteins. Various amounts of purified preparations of the 20,000 and 39,000 Mr Ada proteins were incubated with [3H]MNU-treated calf thymus DNA (5"4x 104 disints/min) at 37°C for 15 rain in 50gl of MT buffer. The reaction mixture was heated at 90°C for 15 min after addition of 100/~l of 1 mg bovine serum albumin/ml and 400 ~1 of 0"8 M-trichloroacetic acid, and the proteins were precipitated. The pellet was rinsed with 500 gl of 5O/o (w/v) trichloroacetic acid and dissolved in 200 gl of 0"1 M-NaOH. The radioactivity was determined in a liquid scintillation counter. ( 0 ) 20,000 M, protein; (O) 39,000 Mr intact Ada protein.
266
H. A kimaru et ~1.
Non-methylated 39 20 IO'~--~l~(pmol)
39 '0
Methytoted 20 I 3 6' '0 I 3
6' (pmol)
<] -{r i ],'
<~] /oc
:
~] ..
:
.
:.::
4 (o)
.i-! 14
. • ,j
o/kA
(b)
Figure 5. Promotion of transcription of the alkA gene in "vitro by the 20,000 M~ protein. For the reaction, 3 pmol of RNA polymerase holoenzyme and various amounts of non-methylated or methylated forms of the 39,000 Ada and 20,000 M~ proteins were incubated at 37°C for 15 min in the mixture described by Nakabeppu & Sekiguchi (1986). The labeled products were analyzed by denaturing 10°/o polyacrylamide gel electrophoresis, followed by autoradiography. The filled and open triangles indicate the 23-nucleotide alkA and the 63-nucleotide lac transcripts, respectively. Amounts (pmol) of the non-methylated or methylated 39,000 (39) Ada and 20,000 (20) proteins added are shown above each column. (a) Non-methylated 39,000 Ada and 20,000 M~ proteins. (b) Methylated 39,000 Ada and 20,000 M r proteins. The number above each lane indicates the amount of protein used (pmol).
a
b
39
20
c
d
Me .'59
Me 20
A•
-I-I- I+l+J
+1,
toe
35 [
1
+11
+I.
lop
-10l =
-21"
-21"
t5 r
-41 '
-41"
-51'
-51"
-61"
-61"
-35 i ~
(e) Fig. 6.
Transcriptional Regulation by Ada Cleavage Product
267
o
b
c
d
59
20
Me 59
Me 2 0
c°l l
['l-I-I+l+l
TA""
iI
+1
+1
-11
-11
-21
-21 -31
(b) Figure 6. Binding of the 20,000 M, protein and RNA polymerase to the alkA promoter. (a) Upper strand. (b) Lower strand. 0"3 pmol of a 276 bp alkA promoter DNA fragment, labeled at (a) the 5' end or (b) the 3' end, was incubated with various amounts of the non-methylated or methylated forms of 39,000 Ada or 20,000 Mr proteins at 22°C for 60 rain, and treated with 20 ng of DNase I for 30 s. The samples were analyzed by denaturing 8% polyacrylamide gel electrophoresis. Positioned numbers at the right sides of the gel patterns indicate base numbers from the transcription initiation site (Nakabeppu & Sekiguchi, 1986). The - l0 and - 3 5 boxes are indicated by open boxes, DNA sequences protected by the proteins are shown by filled boxes, and RNA polymerase-binding sequences by hatched boxes. DNase I-hypersensitive sites are indicated by arrows. G'A and T" C show the sequences determined by the Maxam & Gilbert (1980) method. Amounts (pmol) of the 39,000 (39) Ada or 20,000 _~I~(20) protein, with (+) or without (--) 6 pmol of RNA polymerase, are shown above each column. Panel a, Non-methylated 39,000 M r Ada protein; panel b, non-methylated 20.000 M~ protein; panel c, methylated (Me) 39,000 M~ Ada protein; panel d, methylated (Me) 20,000 11/1,protein.
(Fig. 6(a)) and a region corresponding to positions - 4 9 to --31 of the lower strand (Fig. 6(b)) from DNase I digestion. On interaction with the methylated Ada or 20,000M r protein, DNase I-hypersensitive sites appeared at positions - 2 2 to - 2 0 of the upper strand, but not of the lower strand. The non-methylated Ada protein did not protect the alkA promoter region, though it promoted in vitro transcription of the alkA gene (Fig. 5). We examined the effects of the 39,000 and the 20,000 M r proteins on binding of P~NA polymerase to the alkA promoter. While no protection of the alkA promoter was observed with either RNA polymerase or the non-methylated Ada protein alone, when added together, t h e y resulted in a high degree of protection of both strands (Fig. 6). In the presence of the methylated 39,000 or 20,000 M, protein, which by itself possessed the ability to bind to the alkA promoter, RNA polymerase further extended
the region of protection. A region of the lower strand from - 5 4 to +14, including the sequence already covered by the methylated protein, was protected by RNA polymerase and a DNase I-hypersensitive site appeared at position - 5 5 of the lower strand. When the methylated 39,000 or 20,000 M, protein bound to the promoter, DNase I-hypersensitive sites were produced at - 2 2 to --20 of the upper (non-transcribed) strand (Fig. 6(a)), panels a, c and d). These sites remained sensitive even in the presence of RNA polymerase. The region 5 bp upstream from the DNase I-hypersensitive site of the upper strand could not be covered by RNA polymerase, but the corresponding region of the lower (transcribed) strand was protected. These results suggest t h a t both intact Ada and 20,000 M r proteins m a y cause a slight conformational change in this particular region of DNA; for instance, unwinding of the DNA helix or relaxation of hydrogen bond-linked base-pairs, so as to p r o m o t e inter-
H. A kimaru et al.
268
~ Alo 6 1o olo " 3
zo
R N A pol.
~ lo o OlO 3 61o o 61 Me 2o
+i--
-21 --.~
-31
!
--
41 !:
--51~
--61
Ca)
(b)
Figure 7. Interactions of the Ada protein, 20,000 Mr protein and RNA polymerase with the ada promoter. The DNase I footprinting experiment was performed as described in the legend to Fig. 6. Filled boxes indicate the regions protected by the methylated 20,000 Mr or Ada protein (left) and by both methylated Ada protein and RNA polymerase (right}. DNase I-sensitive sites are indicated by arrows. The - 3 5 and - l 0 boxes are shown by hatched boxes, and the ada regulatory sequence by an open box. Numbers are positioned from the transcription initiation site (Nakabeppu & Sekiguehi, 1986). (a) Non-methylated 39,000 Ada or 20,000 Mr protein. (b) Methylated (Me) 39,000 (39) Ada or 20,000 Mr (20) protein. action of the RNA polymerase with the promoter. We observed only a weak protection by RNA polymerase when the non-methylated 20,000 Mr protein was present in the reaction mixture. (c) Repression of ada transcription by the methylated
20,000 M, protein Footprinting analyses were made to examine interactions of the 20,000 M r protein with the ada
promoter. A HindIII-EcoRI DNA fragment containing the ada promoter sequence, radioactively labeled at the 5' end of the non-transcribed strand, was incubated with various amounts of the Ada or 20,000 Mr protein, and after a brief exposure to DNase I, the products were analyzed by denaturing gel electrophoresis (Fig. 7). The methylated but not the non-methylated form of intact Ada protein bound to the ada promoter, in agreement with reported d a t a (Sakumi & Sekiguchi, 1989). A similar
269
Transcriptional Regulation by A d a Cleavage Product o ro ¢u
Me 5 9
I,,,,,6 1 6 1 0 o.io.31
31o
Me20 I 3 6 1
I ,3 91Me20 1 1 1 Me59
~[ ada
"9[ Io c
(a)
(b)
Figure 8. Effect of the 20,000 Mr protein on ada transcription. (a) Inability of the 20,000 M~ protein to induce ada transcription. 0"3 pmo] each of the DNA fragments containing either the ada or the lacU V5 promoter, 3 pmol of RNA polymerase and various amounts of the methylated or non-methylated form of Ada (39) or 20,000 Mr (20) protein were incubated at 37°C for 60 rain, followed by run-off transcription for 6 rain. The transcripts from the ada and lacUV5 promoter are indicated by arrows. Amounts (pmol) of Ada and 20,000 M r protein added are indicated on each lane. (b) Effect of methylated (Me) 20,000 Mr protein on ada transcription. Transcription was performed in the presence of 3 pmol of RNA polymerase and l pmol of methylated Ada protein. To each tube the indicated amounts (pmol) of methylated 20,000 Mr protein were added.
but more striking effect of methylation was observed with the 20,000 Mr protein. In the presence of the methylated 20,000 M r protein, DNA regions corresponding to residues - 6 3 to - 4 5 and - 4 4 to --38 of the strand including the ada regulatory sequence, AAAGCGCA (Nakamura et al., 1988), were protected from DNase I digestion (Fig. 7(b)). On the other hand, no evident protection was observed with the non-methylated 20,000 Mr protein (Fig. 7(a)). Thus, the methylated form of the 20,000 Mr protein apparently possesses the potential to bind to the ada promoter. Because the regions protected by the methylated 20,000 M r protein were identical to those protected by the methylated Ada protein, the essential region required for the specific binding should be present in the 20,000 Mr protein. There is another indication of interaction of the methylated 20,000 Mr protein with the promoter. An intense signal for the DNase I-hypersensitive site appeared at position - 3 0 of the strand when the DNA was incubated with the methylated form of 20,000 Mr protein. Only a weak signal was detected in the same region with the non-methylated 20,000Mr protein. Although an additional
DNase I-sensitive site appeared at - 2 1 when the DNA was incubated with the methylated Ada protein, this site was not formed with the methylated 20,000 Mr protein. The disappearance of the distant DNase I-sensitive site may be caused by a decrease in volume of the protein. As reported (Sakumi & Sekiguchi, 1989), the methylated Ada protein promoted the binding of RNA polymerase to the ada promoter (Fig. 7(b)). Although the methylated 20,000 M r protein bound to the ada promoter as did the methylated Ada protein, the former apparently lacked the potential to promote the binding of RNA polymerase to the nearby region of the promoter. I t is likely that the C-terminal domain of Ada protein may make physical contact with RNA polymerase to form a stable transcription-initiation complex on the weak promoter. The observation described above raises the question of whether the 20,000 M r protein competes with the Ada protein in binding to the ada regulatory sequence, thus causing inhibition of expression of the ada gene. To examine this possibility, an in vitro transcription experiment was performed (Fig. 8).
270
H. A k i m a r u et al.
'2° I .o I00~
80
e~
"c 60 o c
~- 4 0
2O
0
4
6
of 20,O00M,
protein
2 Amounts
8 (pmol)
Figure 9. Inhibition of Ada protein-pronmted ada transcription by the methylated form of the 20,000 M, prorein. In vitro transcription was performed as described in Materials and Methods: 0"3 pmol of the template DNA, 1 pmol of methylated Ada protein and 2 pmol of RNA polymerase were incubated with 0 to 7 pmol of methylated or non-methylated forms of the protein, and then run-off transcription was performed. The autoradiogram was scanned for quantification of the transcripts. (O) Non-methylated protein; ( l ) methylated protein.
The methylated form of Ada protein efficiently promoted ada transcription while the non-methylated Ada protein exerted little effect, in agreement with our previous observations (Nakabeppu & Sekiguchi, 1986). Both non-methylated and methylated forms of the 20,000/t4, protein showed no potential to promote transcription of the ada gene (Fig. 8(a)). I t was further demonstrated that ada
transcription, promoted by the methylated Ada protein, can be repressed when the methylated form of the 20,000 Mr protein is present in the reaction mixture (Fig. 8(b)). To correlate the binding capacity of the 20,000 Mr protein to the potential to inhibit ada transcription, we compared the effects of the methylated and nonmethylated forms of the 20,000 M, protein, the latter not binding to the ada promoter. A mixture containing 0"3 pmol of the template DNA with the ada promoter region, 1 pmol of methylated Ada protein, 2 pmol of RNA polymerase and various amounts of methylated or non-methylated forms of 20,000 M, protein, was incubated and run-off transcription was performed. As shown in Figure 9, the methylated form of the 20,000 Mr protein effectively inhibited ada transcription, whereas the nonmethylated form exerted no such effect. This clearly indicates that the potential of the 20,000 Mr protein to inhibit ada transcription depends on the state of methylation, and that this inhibition is caused by competitive binding of two types of proteins, the functional intact Ada protein and the non-functional 20,000 Mr protein, to the same region of the promoter. (d) Effects of the 20.000 M, protein on expression of
the alkA and ada gene in vivo
To determine whether or not the 20,000 Mr protein regulates transcription of tile alkA and ada gene in vivo, we used systems in which the 20,000 Mr protein is supplied artificially under the control of the lac promoter. 50
o --
~
I 0
-~ 20 c
Z
P Io s m i d
~ ~
P:Ioc 20,000 ~////////////////////A
._
IO
_
.oodl..-o.-..-~ ........ di~~o..**~............ ...°=:."i ..~:.*" "..&........ &" %%%i......-+"
o
•
~ A 3-Melhylodenine-ONA ? glycosyloseD.
"~ A odo
Chromosome
)(
"~ ~E,
J 0
P: olk.4 olkA ....................
(o)
I
20
,
u
I
,
40
u
I
60
i
I
80
Time (rain) (b)
Figure 10. Promotion of expression of the alkA gene by the 20,000 M, protein in vivo. (a) The system for measuring transcription-activator activity of the 20,000 M~ protein. P:alkA and P:lac indicate the alkA and lac promoter, respectively. (b) Promotion of the alkA expression by the 20,000 M, protein. E. coli YN180 (A(ada-alkB) : : kan) cells harboring either pMF511 or pMF3 plasmid were grown at 37°C to A6~o = 0"3 in M9 medium (pH 7"4), and treated with or without 5/~g MNNG/ml at 37°C for 10 rain in M9 (pH 6-0). Plasmid pMF511 carries the truncated ada gene, coding for" the 20,000 Mr protein, attached to the lac promoter, while the control plasmid pMF3 does not. The unadapted or adapted cells were resuspended in M9 medium (pH 7"4), removed at the time indicated and crude extracts prepared with buffer A: 400 pg of protein of the crude extracts were used for assay of 3-methyladenine-DNA glycosylase II activity. (O) Unadapted cells with pMF511 (20,000 M,); (O) adapted cells with pMF511 {20,000 Mr); (/k) unadapted ceils with pMF3; (A) adapted cells with pMF3.
271
Transcriptional Regulation by A d a Cleavage Product
We constructed a low copy plasmid carrying the region coding for the 20,000 Mr protein. A 0'8 kb P v u I I - P v u I I fragment containing the lac promoter region and the coding region for the 20,000 Mr protein was inserted into the EcoRI site of pMF3, and the resulting plasmid, pMF511, was used to transform E. coti strain YN180 (Aada). Cells harboring pMF511 were grown in M9 medium containing 0"5 mM-IPTG at 37°C for two hours to produce the 20,000 M, protein. The cells were then treated with 5 #g of MNNG/ml for ten minutes and incubated in M9 medium. As illustrated in Figure 10(a), the effect of the 20,000 11/, protein on the alkA promoter is monitored by an activity of 3-methyladenineDNA glycosylase II, a product of the alkA gene. At various times, the cells were collected for assay of 3-methyladenine-DNA glycosylase II activity (Fig. 10(b)). We found that glycosylase II activity increased when cells carrying the 20,000 Mr protein were treated with MNNG, but without this treatment there was no significant increase in enzyme activity. Control cells carrying the vector alone exhibited no increase in the activity, even after MNNG treatment. Thus, the result obtained in vitro is in good accord with the finding in vivo, To investigate effects of the 20,000 Mr protein on ada transcription in vivo, plasmid pCM51l was constructed. It carries the chloramphenicoI acetyltransferase gene attached to the ada promoter and the gene coding for the 20,000 M, protein, under the control of the lac promoter (Fig. l l(a)). E. coli ABll57 cells harboring pCM51 or vector plasmid pKT410 were treated with 50 #g of MMS/ml (adaptive treatment). The cells were grown in the presence of 1 mM-IPTG and acetyltransferase activity was assayed at various times. As shown in Figure I 1(b), expression of the ada promoter, as measured by the acetyltransferase activity, was considerably lower in cells harboring pCM51 l, in which formation of the 20,000 M~ protein is induced, as compared with the level in cells harboring pKT410 that does not contain the 20,000 Mr protein-coding gene. Without adaptive treatment, the enzyme activity in both types of cells was very low and levels were not affected by induction with IPTG (data not shown), It would thus appear that the methylated form of 20,000 Mr protein tunctions in vivo as a repressor for ada gene expression.
A A
A A
t
P:c/do
P:/oc
co/
20,000
Plosmid
•
P:oda
Chromosome
•
odo
¢ ! (a)
80 70
o C"
'- ~. 6 0
50 "E E >"
40
¢J
2C
~
0 "~ IPTG
I
20
I
I
40
!
I
60
!
80
Time (mini ( b)
4. Discussion
Figure 11. Repression of ada expression by the 20,000 Mr protein in vivo. (a)The system for measuring repressor activity of the 20,000 M, protein in vivo. P : ada and P:/ac indicate the ada and lac promoter, respectively. The reporter gene, cat, and the gene coding for the 20,000 M, protein are placed on plasmid pCM511 while the intact Ada protein is produced from the ada gene on the chromosome. -% Positive effects of the intact Ada protein; - ~ , negative action of 20,000 Mr protein. (b) Repression of the ada expression by the 20,000 M~ protein. E. coil ABI157 cells harboring vector piasmid pKT410 (0) or plasmid pCM511 coding for the 20,000 Mr protein (O) were grown in M9 medium to A66o -- 0"3. The cells were incubated in the presence of 50 ~tg MMS/ml for 60min and then transferred to medium containing l mM-IPTG and 0-005~'o MMS. The cells were collected at various times after addition of IPTG, and chloramphenicol acetyltransferase activity was determined.
The 39,000 Mr Ada protein is cleaved to smaller polypeptides in in vitro proteolytic processes (Teo et al., 1984; Nakabeppu et al., 1985; Teo, 1987). The cleavage site is between Lys178 and Gtn179 of Ada protein, as deduced from the N-terminal amino acid sequence of the cleavage products and the nucleotide sequence of the ada gene (Teo et al., 1984; Demple et al., 1985). On incubation of Ada protein with a proteinase in vitro, the activity that promoted transcription of the ada gene decreased while the transcription-promoting activity on the alkA gene remained (Yoshikai et al., 1988). This
suggests that the cleavage prod.uct of Ada protein might act as a transcriptional activator for expression of alkA, but not of the ada gene. In the present work, we investigated the activity of the N-terminal half of Ada protein as the transcriptional regulator both in in vitro and in vivo systems. Using site-directed mutagenesis of the cloned ada gene, the 20,000 Mr protein was overproduced and purified. The protein carried a methyl-
H. Akimaru et al.
272
-35 Non-transcribed --..,, strand ACAAC C G G ~ G ~ , . ~ A ~ Q ~ t ~ G I C G
"I~TrG G C C T I I ~ " ~ ( ~ / C ~ Transcribed -50 -40 strand
-10 .° * * ;GC;GAAAGC CA~CGTCTGA ATAAC~TTTA GGATGAA~AAJ GGAGA
~J ~CGC GTCGCAGACT TATT GCAAAT ACGACTI-I'CG CCTAC'I-I'A'I-rl CCTCT -30 -20 -10 +I +10
Figure 12. Regions of the alkA promoter protected by the methylated form of the 39,000 Ada or 20,000 Mr protein and RNA polymerase. The region protected by 39,000 Ada or 20,000 Mr protein is indicated by shaded boxes, while the region protected by RNA polymerase by open boxes. Bold letters indicate DNase I-hypersensitive sites. Filled circles correspond to the - 3 5 and - l 0 boxes, and + 1 shows the transcription initiation site.
transferase activity that catalyzes transfer of a methyl group from methylphosphotriester of alkylated DNA to its own molecule. (a) Transcriptional regulator activity of the 20,000 M, protein In in vitro transcription experiments, only a methylated form of the 20,000 M r protein exhibited activity to promote transcription of the alkA gene, whereas the intact Ada protein, irrespective of methylation, promoted transcription, in agreement with our findings with the in vitro proteolytic product (Yoshikai et al., 1988). We also found that • the 20,000M r protein, once methylated, can promote alkA expression in vivo. Binding of the intact Ada and 20,000 M r proteins to the alkA promoter has been studied using DNase I footprinting techniques. Since at least 4 bp to the 5' side and 6 bp to the 3' side of the cutting site are covered by the DNase I molecule on one side of DNA double helix (Suck & Oefner, 1986), the protected regions seen with footprinting are larger than that actually covered by the protein. Taking this into account, a region covered by the 20,000 M r domain of Ada protein was estimated, as shown in Figure 12. Site-directed mutagenesis studies revealed that a sequence AAAGCGCA is required for the regulated expression of the ada gene (Nakamura et al., 1988). It is revealed here that this sequence is indeed the binding site for the 20,000 M r protein as well as for the intact Ada protein. An identical sequence exists in the alkA promoter but it overlaps with the predicted - 3 5 box, the putative binding site of RNA polymerase. There is a closely related sequence, AAAGCAAA, in the upstream region in the alkA promoter, and this sequence was indeed protected by binding of either the intact Ada or 20,000 M r protein. Thus, the AAAGCAAA sequence may be the preferred binding site for Ada protein. To gain support for this notion, studies using mutant DNA sequences are in progress in our laboratory. The region protected by the methylated form of 20,000 M r protein was the same as that covered by the methylated intact Ada protein, thereby indicating that the 19,000 Mr domain of Ada protein does not attach to the DNA, and only the 20,000 M r domain is directly involved in binding to the alkA promoter. Essential regions of Ada protein for specific DNA binding must reside on the N-terminal
half of the protein. Since neither the non-methylated form of intact Ada nor the 20,000 Mr protein was capable of binding to the alkA and ada promoter region, methylation of the cysteine residue at position 69 of the protein would cause a subtle and sufficient conformational change of the N-terminal half of Ada protein so as to gain access to the proper site of the promoter. Information of the three-dimensional structures of the methylated and non-methylated forms of Ada protein, particularly those of the N-terminal domains, would help us to understand the molecular mechanisms for the specific binding and subsequent promoter activation. Although the non-methylated form of intact Ada protein could not in itself bind tightly to the promoter, it could promote binding of RNA polymerase to the promoter and as a result the proper ternary complex formed. This is in accord with our finding that the non-methylated form of Ada protein is as active as the methylated form in promoting the alkA transcription in vitro (Nakabeppu & Sekiguchi, 1986). Though the binding potential of the non-methylated form of Ada protein is weaker than that of the methylated form, it appears to be high enough to form a stable initiation complex when RNA polymerase is present. (b) Turn off of the adaptive response by cleavage of Ada protein Ada protein is a positive regulator for the adaptive response to alkylating agents in E. coli. When the protein is methylated, it acquires the potential to bind tightly to the ada regulatory sequence, thereby facilitating further binding of RNA polymerase to the ada promoter (Sakumi & Sekiguchi, 1989). This would turn on the adaptive response to ensure that the rates of synthesis of DNA repair enzymes were accelerated. With regard to turn-off of the adaptive response, the methylated Ada protein is not actively demethylated, thus there has to be a mechanism that will decrease the amount of activated Ada protein. The cleavage of Ada protein might be related to down-regulation of the adaptive response. Neither of a 20,000 and a 19,000 M r protein, derived from the N and C-terminal halves of the Ada protein, respectively, had the potential to promote ada transcription, yet they still carried
Transcriptional Regulation by Ada Cleavage Product Intoct AdO protein
273
formed more abundantly than others repaired by other classes of enzymes (Lindahl et al., 1988). odo
-35
-10
t'IIIA
~//IA
.35
-10
,,,
We thank M. Ohara for comments on the manuscript. This work was supported by grants from the Ministry of Education, Science and Culture of Japan (61065007) and from the Institute of Protein Engineering, Osaka, Japan.
,
20,000 Mr protein
References aria
,, ,
I
Figure 13. Model for promotion of transcription by the Ada protein (above) and repression of transcription by the 20,000 Mr protein (below). ada indicates the ada regulatory sequence of the ada gene. k = l0 a Mr.
distinct methyltransferase activities. Thus, cleavage of the Ada protein would cause a decrease in the level of a positive gene regulator. If the cleavage product binds to the ada promoter without facilitating further binding of RNA polymerase to the nearby region, it would provide a good model for negative modulation of the adaptive response (Karran & Hall, 1988). We have overproduced the N-terminal 20,000 Mr protein and examined its ability to interact with the ada and alkA promoter. The methylated 20,000 Mr protein indeed bound to the ada regulator), sequence, but this binding did not facilitate binding of RNA polymerase nor did it promote ada transcription (see Fig. 13). I n vitro studies revealed that the methylated but not the non-methylated form of the 20,000 Mr protein competes with the methylated Ada protein for the particular binding site within the ada promoter sequence, thus repressing ada transcription. By using an artificially constructed gene expression system, we have shown that the methylated 20,000 M r protein prevents ada transcription also in vivo. However, more rigorous tests may be necessary to gain support for the notion that proteolysis is indeed involved in cessation of the adaptive response. Appropriate mutants either defective in protease activity or with altered amino acid sequences near or at the cleavage sites of Ada protein will be useful. It is of interest to note that the cleavage of Ada protein causes differential effects on expression of the two genes, ada and alkA, belonging to the same regulon, a finding which may relate to the fact that certain types of DNA lesions, which can be repaired by one class of enzymes, are
Barnes, W. M., Bevan, M. & Son, P. H. (1983). Methods Enzymol. 101, 98-122. Clarke, N. D., Kraal, M. & Seeberg, E. (1984). Mol. Gen. Genet. 197, 368-372. Demple, B., Jacobsson, A., OIsson, M., Robins, P. & Lindahl, T. (1982). J. Biol. Chem. 257, 13776-13780. Demple, B., Sedgwick, B., Robins, P., Totty, N., Waterfield, M.D. & Lindahl, T. (1985). Proc. Nat. Acad. Sci., U.S.A. 82, 2688-2692. Evensen, G. & Seeberg, E. (1982). Nature (London), 296, 773-775. Karran, P. & Hall, J. (1988). In Nucleic Acids and Molecular Biology (Eekstein, F. & Lilley, D. M. J., eds), vol. 2, pp. 188-t97, Springer-Verlag, Berlin. Karran, P., Hjelmgren, T. & Lindahl, T. (1982). Nature (London), 296, 770-773. Lindahl, T., Sedgwick, B., Sekiguchi, M. & Nakabeppu, Y. (1988). Annu. Rev. Biochem. 57, 133-157. Manis, J. J. & Kline, B. C. (1977). Mol. Gen. Genet. 152, 175-182. Maxam, A. M. & Gilbert, W. (1980). Methods Enzymol. 65, 499-560. Nakabeppu, Y. & Sekiguchi, M. (1986). Proc. Nat. Acad. Sci., U.S.A. 83, 6297-6301. Nakabeppu, Y., Kondo, H. & Sekiguchi, M. (1984a). J. Biol. Chem. 259, 13723-13729. Nakabeppu, Y., Miyata, T., Kondo, H., Iwanaga, S. & Sekiguchi, M. (1984b). J. Biol. Chem. 259, 13730-13736. Nakabeppu, Y., Kondo, H., Kawabata, S., Iwanaga, S. & Sekiguchi, M. (1985). J. Biol. Chem. 260, 7281-7288. Nakamura, T., Tokumoto, Y., Sakumi, K., Koike, G., Nakabeppu, Y. & Sekiguchi, M. (1988). J. Mol. Biol. 202, 483-494. Sakumi, K. & Sekiguchi, M. (1989). J. Mol. Biol. 205, 373-385. Sanger, F., Coulson, A. R., Barre]l, B. G., Smith, A. J. H. & Roe, B. A. (1980). J. Mol. Biol. 143, 161-178. Sekiguchi, M. & Nakabeppu, Y. (1987). Trends Genet. 3, 51-54. Shaw, W. V. (1975). Methods Enzymol. 43,737-755. Suck, D. & Oefner, C. (1986). Nature (London), 321, 620-625. Takano, K., Nakabeppu, Y. & Sekiguchi, M. (1988). J. Mol. Biol. 201,261-271. Teo, I. (1987). Murat. Res. 183, 123-127. Teo, I., Sedgwick, B., Demple, B., Li, B. & Lindahl, T. (1984). EMBO J. 3, 2151-2157. Yoshikai, T., Nakabeppu, Y. & Sekiguehi, M. (1988). J. Biol. Chem. 263, 19174-19180.
Edited by R. Schleif