Potential binding sites of the trans-activator FIS are present upstream of all rRNA operons and of many but not all tRNA operons

302

Biochimica et Biophysica Acta, 1050 (1990) 302-306 Elseviel

BBAEXP 92140

Potential binding sites of the trans-activator FIS are present upstream of all rRNA operons and of many but not all tRNA operons H a n s V e r b e e k 1, L a r s N i l s s o n

2, G a b r i e l l a

Baliko 3 and Leendert Bosch

Department of Biochemistry, Leiden University, Gorleaus Laboratories, Leiden (The Netherlands) 2 Department of Cell Biology, The University of Stockholm, Arrhenius Laboratories, Stockholm (Sweden) and 3 Institute of Biochemistry, Biological Research Center, Szeged (Hungary) (Received 16 May 1990)

Key words: Upstream activating sequence; FIS; Stable RNA synthesis

FIS, the Escherichia coil protein that stimulates the inversion of various DNA segments by binding to a recombinational enhancer, tram-activates a number of stable RNA operons and binds to the upstream activator sequence (UAS) of these operons (Nilsson et al. (1990) EMBO J. 9, 727). In a search for potential FIS-binding sites we have compared UASs of other stable RNA operons with a consensus FIS-binding sequence, compiled by comparing recombinational enhancers. Such sites can thus be recognized upstream of all rRNA and 13 tRNA operons. Matching with the consensus sequence varied, suggesting that the affinity of FIS for the sites differed. Accordingly, FIS binding to an upstream sequence of the metY(nusA) operon was found to be weaker than that to the UAS of the thrU(tu1B) oporon. No FIS binding sites were found upstream three tRNA operons.

Introduction Stable RNA synthesis has been shown to be regulated by two mechanisms: growth rate-dependent control and stringent response. Both mechanisms operate through repression of transcription initiation of stable RNA operons [1-4]. Recently we have demonstrated that a number of stable RNA operons are also regulated by trans-activation. A heat stable E. coli protein, that stimulates the inversion of various viral DNA segments by binding to a recombinational enhancer and therefore called FIS (factor for inversion stimulation), was identified as the trans-activator [5]. FIS binds specifically to a cis-acting region, upstream of the rrnB, the thrU(tufB) and the tyrT operon. The interaction with this so-called upstream activator sequence (UAS) leads to increased transcription, as was demonstrated in vivo and in vitro. FIS-dependent trans-activation is required for high growth rates and for accelerated growth upon a nutritional shift up [5]. In the present investigation we have searched for potential FIS-binding sites by comparing sequences up-

stream more than 20 stable RNA operons with the consensus sequence compiled by comparing recombinational enhancers [6]. In doing so we have asked whether FIS binding to UASs varying in affinity for the transactivator permits differential expression of tRNA genes as part of a codon-bias strategy. In the course of this study indications were obtained that FIS may also be involved in trans-activation of operons non-related to translation. Materials and Methods Plasmid YN 96, containing the metY-nusA genes, was a generous gift of Dr. Y. Nakamura. A HindlII-AluI fragment from this plasmid was isolated and labeled with a-[3zp]dCTP by filling in the HindlII site according to Maniatis [7]. The thrU(tufB) UAS fragment we used was a BamHI-BstEII fragment from - 1 7 6 to + 50 [8]. FIS was purified essentially as described by Johnson et al. [9]. The gel mobility shift experiment was performed as previously described [8]. Results

Correspondence: L. Bosch, Department of Biochemistry, Leiden University, Gorleaus Laboratories, P.O. Box 9502, 2300 RA Leiden, The Netherlands.

FIS binding to recombinational enhancers causes a conformational change resulting in bending or kinking of the DNA. This change is an essential step for recom-

0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

303 1

b i n a t i o n a l e n h a n c e r activity [10]. A c t i v a t o r sequences o f the t y r T [11], the r r n B [12], the i e u V [13] a n d the thrU(tufB) [14] o p e r o n s all show a s e q u e n c e - i n d u c e d b e n d i n g o f their D N A . F I S b i n d i n g to the U A S o f thrU(tufB) e n h a n c e s the b e n d i n g [18], suggesting t h a t this altered D N A c o n f o r m a t i o n p l a y s a role in the a c t i v a t i o n o f the o p e r o n . These d a t a p r o m p t e d a syst e m a t i c c o m p a r i s o n o f all k n o w n sequences u p s t r e a m o f stable R N A o p e r o n s with the following consensus seq u e n c e c o m p i l e d b y H i i b n e r a n d A r b e r [6] for various r e c o m b i n a t i o n a l enhancers:

2

3

4

5

6

7

8

9

10

GnnYRnnAnnYRnnC T T A A s m a y be c o n c l u d e d f r o m T a b l e I, F I S b i n d i n g ups t r e a m stable R N A promoters m a y b e a general p h e n o m e n o n . F i v e o f the seven r r n o p e r o n s have a seq u e n c e of fifteen b a s e p a i r s centering at p o s i t i o n - 7 1 that fully m a t c h e s the consensus. O n e m i s m a t c h is f o u n d in the c o r r e s p o n d i n g sequence of r r n D a n d two in that of r r n E . Seven o u t o f sixteen t R N A sequences c o m p l e t e l y m a t c h the consensus, eight have o n e a n d the metY(nusA) and pheU have two mismatches. These sequences d o n o t all c e n t e r at - 7 1 . C o n c e i v a b l y , deviations f r o m the c o n s e n s u s sequence m a y b e r e l a t e d to lowering of the affinity for F I S . I n a c c o r d a n c e with this a s s u m p t i o n are the e x p e r i m e n t s o f Figs. 1 a n d 2, showing F I S b i n d i n g in vitro to a D N A f r a g m e n t d e r i v e d

Fig 1. In vitro binding of FIS to UAS-containing fragments of the and the thrU(tufB) operons studied with electrophoretic retardation. Lanes 1-5 contain 50 ng of the H i n d l I I - A l u l fragment of the m e t Y ( n u s A ) U A S . ( T h e A l u I cleavage site is position +30. The H i n d l I I cleavage position is not known. The fragment comprises approximately 250 bp.) Lanes 6-10 contain 50 ng of an end-labeled UAS fragment of the thrU(tufB) operon comprising positions - 176 to + 50. Amounts of FIS added: Lanes 1 and 6, none; lanes 2 and 7, 0.02 ng; lanes 3 and 8, 0.2 ng; lanes 4 and 9, 2 ng; lanes 5 and 10, 20 ng. metY(nusA)

f r o m the sequence u p s t r e a m the m e t Y ( n u s A ) A l t h o u g h this f r a g m e n t is a b l e to f o r m three with F I S (of. Fig. 1) like c o r r e s p o n d i n g U A S o f o t h e r stable R N A o p e r o n s [5], c o m p e t i t i o n

promoter. complexes fragments between a

TABLE I Compilation o f potential and established F1S-binding sites upstream stable R N A operons

Potential FIS-binding sites have been searched upstream tRNA and rRNA operons, the sequences of which are known. The location of the site is indicated by the position of its center. Sequences of 15 bp best matching the consensus are listed. Mismatches are indicated in bold. Sequences are taken from Refs. 1, 20, 24, 25 and 26.

operon

position

FIS binding site

tRNAs

rrn A rrn B

71 71 - 72 - 71 - 71 - 71 - 71 - 71 - 72 - 62 - 71 - 71 - 107 - 82 - 71 - 71 - 71 - 71 - 71 - 73 - 73 - 74 - 72 - 71

GaaTAaaAaaTGcgC GgtTGaaTgtTGcgC GgcTGaaAaaTGcgC GggTGaaAaaaCAaC GccTGaaAagTGagC GatTAaaAgaTGagC CtgGGgtAtgCAgcA GctTAaaAttTAcgA GtgCGaaTcaTAagC GgcTGttTttCAggC GcaTAaaAtgTGacC GgaTGaaAatTAcgC GttTGttTgcCGctA GttCGcaAceTAacC GccTGatTttTCagC GcaTAaaTggCGagG GaaCAttTttCCagC TtgCGgaTtaTTacC TatCGctAacTGatT TgcTGacTttCTcgC TgaTAgaTtgTGcgG CtgCAgaTttTAcgT GcaCAaaCcgTAacC TgcCGcaAtcTTaaG

Ile 1, Ala 1B Gin 2 Trp, Gin 2 Ile 1, Ala 1B, Thr 1 Gin 2 lie 1, Aia 1B, Asp 1 Gin 2 Val 1, Lys Val 1, Lys Ser 1 Thr 3, Thr4, Tyr2, Gly 2 Tyr 1 Arg 4 Met m, Leu 3, Gin 1, Gin 2 Ala 2 Arg 5 Leu 1 Asn Pro 2 Pro 1 Phe fMet 2 Arg 3, His, Leu 1, Pro 1 Phe

rrn rrn rrn rrn

C D G H

rrn E val T lys T ser T thr U tyr T dna Y met T ala arg leu V

ash T pro pro phe V met Y arg T phe U

-

304 1

2

3

4

5

6

Fig. 2. Competition for FIS binding between UAS fragments of the operons. 50 rig of the end-labeled UAS fragment of the thrU(tufB) operon and increasing amounts of the non-labeled UAS fragment of the metY(nusA) operon were incubated and protein/DNA complex formation was analysed as in Fig. 1. Amounts of FIS added: lane 1, none; lanes 2 to 6, 20 rig. Amounts of the metY(nusA) UAS fragment added: lanes 1 and 2, none; lane 3, 50 ng; lane 4, 150 ng; lane 5, 300 ng; lane 6, 500 ng.

mctY(m~l) and the thrU(tufB)

labeled UAS fragment from the thrU(tufB) operon and an unlabeled fragment from the metY(nusA) operon shows that the affinity of the latter fragment for FIS is approx. 10-times lower (Fig. 2). In Table II the tRNA genes that we assume to be under FIS control are summarized. For each tRNA, the total number of genes on the E. coil chromosome is given. We have attempted to classify these genes by their number carrying potential promoter-proximal FIS binding sites (column 3) and the matching of these sites with the consensus sequence (column 5) in an attempt to find a correlation with the extent of gene expression (according to Ikemura [15], column 6) and the codon usage (according to Grosjean and Fiers [16], column 7). Although the majority of the sites mat_ch the consensus fully or partially, Table II does not show any obvious relationship between their putative strength of FIS binding and the extent of expression or the codon usage in strongly or weakly expressed genes. Probably not all tRNA genes are regulated by FIS-

TABLE II

Expression and codon usage of tRNA's assumed to be under FIS control. For each tRNA the total number of genes on the E. coil chromosome and the number of these genes carrying potential FIS-binding sites are listed in columns 2 and 3, respectively. Matching of potential FIS-binding sites with the consensus is indicated as + + , perfect; + , single mismatch; + / - , two mismatches. The extent of gene expression is taken from Ikemura [15]. Codon usage is taken from Grosjean and Fiefs [16]. tRNA

Ala 1B Ala 2 Arg 3 Arg 4 Arg 5 Asn Asp 1 Gin I Gin 2 Glu 2 Gly 2

His Ile I Leu 1 Len 3

Lys Met m Met f2 Pbe Pro I Pro 2 Pro 3 Ser 1 Ser 5 Thr I Thr 3 Thr 4

Trp Tyr I Tyr 2 Val 1

No. of genes

FIS reg. genes

anticodon

fis bind. sites

expr./ gene

codon usage S

W

3 2 1 1 1 3 3 2 2 4 1 1 3 4 1 3 2 1 2 1 1 1 1 2 1 1 1 1 2 1 3

3 2 1 1 1 1 2 2 2 4 1 1 3 4 1 3 2 1 1 1 1 1 1 1 1 1 1 1 2 1 3

UGC GGC CCG UCU CCU GUU GUC UUG CUG UUC UCC GUG GAU CAG UAG UUU AUG AUG GAA CGG GGG UGG UGA GGA GGU GGU UGU CCA GUA GUA UAC

+ + + +/ + +/ + + + + + + + + + +/ + + + + + + + + +/ + +/+/ +/ +/ +/ + + + + + + + + + + + + + +

major major minor minor minor 0.21 0.27 0.15 0.2 0.23 0.2 0.4 0.36 0.25 ? 0.34 0.14 0.1 0.17 major minor

48 42 0.2 1 0.2 32 61 7 39 83 4 18 28 66 1 69 27 27 29 31 4

48 51 8 5 3 38 55 17 49 59 21 29 58 56 4 39 25 25 48 19 15

major

5

9

0.23 ? minor 0.2

1 35 26 26 3 5 25 25 39

7 16 23 23 6 13 30 30 33

minor 0.31 0.16 0.16 0.35

305 TABLE III

tRNA genes assumed to be FIS-independent For details see legend to Table II. tRNA

No. of genes

anticodon

expr./ gene

codon usage S W

Gly 1 Leu 2 Leu 6

1 1 1

CCC GAG CAA

0.1 0.3 minor

3 11 3

13 27 12

dependent trans-activation. In Table III tRNA genes are listed lacking a promoter upstream sequence that matches the consensus. FIS-dependent regulation may not be restricted to stable RNA operons. Table IV lists a number of operons with promoter upstream sequences of fifteen base pairs, matching the consensus to some extent. Discussion Recent screening of a large number of UAS mutations of the rrnB operon by Gaal and co-workers [17] shows pronounced effects on gene expression by deviations from the consensus as compiled by Hiibner and Arber [6]. Also Bauer et al. [13] demonstrated that a double T to G mutation at positions - 7 1 and - 7 2 of the leuV UAS strongly reduces the expression of the operon. Since these mutations affect base pairs in established (R. Gourse, personal communication) or putative FIS binding sites, respectively, they strengthen our assumptions concerning the optimal FIS binding sequence. Comparison of all proposed FIS-binding sites, both of recombinational enhancers and of stable RNA operon UASs, may thus lead to an improved consensus sequence: GnnYRaaAaaYRnnC t

ttTtt

a

with preferred A / T base pair stretches in the center and with G and C being preferably the first and the last nucleotide, respectively. We assume the ideal activator sequences to be rich in A / T base pairs causing bending of the DNA helix and to have multiple optimal FIS binding sites. Furthermore position - 7 1 seems to be the optimal center of the promoter proximal FIS binding site. This site seems to play a dominant role in activation of the operon (Verbeek et al., unpublished results). The strong UASs of rrnA [18], rrnB [12], thrU(tufB) [14], tyrT [11] and leuV [13] all meet these criteria. The three tRNA species Metm, His and Trp, decoding codons that are unique for the amino acids concerned, seem to be under FIS control. The data of Figs. 1 and 2 show that the UAS of the metY(nusA) operon,

specifying the initiator tRNA Metf2, binds FIS albeit weakly. No sequences upstream of .the tRNA operon metZ, specifying the more abundant initiator tRNA Metfl, have been published. An interesting family is that of the tRNA species accepting leucine. The operons specifying the tRNA's Leu 1 and 3 seem to be regulated by FIS (Table I), the operons specifying Leu 2 (sometimes called Leu 3 [19]) and Leu 6 (also .called Leu X) are not (Table III). The latter two operons also seem to lack the G / C rich discriminator region just upstream of the transcription start, reported [20] to be required for stringent control. The operon specifying Leu 2 furthermore has a very unusual spacing between the - 3 5 and the - 1 0 sequences [19]. Together these features make the relatively high expression level of this operon very surprising. One may expect F1S-activated operons to be stronger than FIS-independent operons. However, our data do not show a clear relationship between the consensus matching of the putative promoter proximal FIS-binding sites and the expression level of the tRNA genes. A few examples only can be found that do show such a relationship such as the highly expressed operon phel/ which has a FIS-binding site at position - 7 3 and the poorly expressed operon pheU that has a very poor FIS-binding site at -71. Also the operon valT specifying the tRNAs Vail and Lys and the operon lysT specifying the tRNAs Lys and Vall are highly expressed and are endowed with perfect FIS-binding sites. The high expression of tRNA genes cotranscribed with rRNA is due to FIS-dependent trans-activation of the rrn operons. Genes of several major tRNAs, however, seem to have weak FIS-binding sites or no binding sites at all. Furthermore, several minor and rare tRNAs are specified by operons that presumably are FIS-controlled such as the tRNAs Arg4 and Arg5. Clearly, additional factors are involved in regulating the tRNA levels of the cell. This also becomes apparent in the significant differences in expression level and codon usage of different tRNA genes of one and the same operon such as thrU and thrT. A dominant role of FIS-dependent trans-activation becomes manifest at

TABLE IV

Potential FIS-binding sites upstream of operons non-related to translation For details see legend to Table I Operon

Potential FIS-bind/ng site

Center position

Reference

spf

AtgCAacAacTGaaA TatTAgaAatTCctA GgaCGaaAgcTAaaT AtaCGctTgaCGgaA

-

21

bol A dna A

AgcCAatTttGtcT

62 76 54 77 76

22 23

306 very high growth rates and during growth acceleration upon a nutritional shift [5]. Finally, an involvement of FIS in the regulation of operons non-related to translation may be considered. As indicated in Table IV, sequences of fifteen base pairs with characteristics of FIS binding are present upstream of the operons spf, dnaA and bolA, although not centering at position -71. The gene spf, specifying spot 42 RNA [21], possibly is a stable RNA gene but this remains to be established. The operons dnaA and bolA are regulated in a growth phase-dependent manner [22,23] which may also point to regulation by FIS. Experimental evidence supporting FIS control is lacking for all three operons.

Acknowledgements Plasmid YN 96, containing the metY(nusA) genes, was a generous gift of Dr. Y. Nakamura. The investigation was supported in part by the commission of the European Communities, Biotechnology Action Programme (BAP), Directorate-General 'Science, Research and Development', Brussels. L.N. was the recipient of a long-term EMBO fellowship and was supported by the Swedish Science Research Council. G.B. was the recipient of a FEBS grant.

References 1 2 3 4

Lindahl, L. and Zengel, J.M. (1986) Ann. Rev. Genet. 20, 297-326. Gausing, K. (1974) Mol. Gen. Genet. 129, 61-75. Gausing, K. (1977) J. Mol. Biol. 115, 335-354. Dennis, P.P. and Bremer, H. (1974) J. Mol. Biol. 84, 407-422.

5 Nilsson, L., Vanet, A., Vijgenboom, E. and Bosch, L. (1990) EMBO J. 9, 727-734. 6 Hubner, H.E. and Arber, W. (1988) EMBO J. 8, 577-585. 7 Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. 8 Vijgenboom, E., Nilsson, L. and Bosch, L. (1988) Nucl. Acid Res. 16, 10183-10197. 9 Johnson, R.C., Bruist, M.F. and Simon, M.I. (1986) Cell 46, 531-539. 10 Hubner, P., Haffter, P., Iida, S. and Arber, W. (1989) J. Mol. Biol. 205, 493-500. 11 Lamond, A.I. and Travers, A.A. (1983) Nature 305, 248-250. 12 Gourse, R.L., de Boer, H.A. and Nomura, M. (1986) Cell 44, 197-205. 13 Bauer, B.F., Kar, E.G., Elford, R.M. and Holmes, W.M. (1988) Gene, 63, 123-134. 14 Van Delft, J.H.M., Marinon, B., Schmidt, D.S. and Bosch, L. (1987) Nucl. Acid Res. 15, 9515-9530. 15 Ikemura, T. (1981) J. Mol. Biol. 146, 1-21. 16 Grosjean, H. and Fiers, W. (1982) Gene 18, 199-209. 17 Gaal, T., Barkel, J., Dickson, R.R., De Boer, H.A., De Haseth, P.L., Aiavi, H. and Gourse, R.L. (1989) J. Bact. 171, 4852-4861. 18 Nachaliel, N., Melnick, J., Gafny, R and Glaser, G. (1989) Nu¢l. Acid Res. 17, 9811-9822. 19 Wahab, S., Elford, R. and Holmes, M. (1989) Gene 81, 193-194. 20 Travers, A.A. (1980) J. Bact. 141,973-976. 21 Polayes. D.A., Rice, P.W., Garner, M.M. and Dahlberg, J.E. (1988) J. Bact. 170, 3110-3114. 22 Aide,a, M., Garrido, T., I-Iernandez-Chieo, C., Vincente M. and Kushner, S.R. (1989) EMBO J. 8, 3923-3931. 23 Atlung, T., Clausen, E.S. and Hansen, F.G. (1985) Mol. Gen. Genet. 200, 442-450. 24 Komine, Y., Adaehi, T., Inokuchi, H. and Ozeki, H. (1990) J. Mol. Biol. 212, 579-598. 25 Muramatsu, M. and Mizuno, T. (1990) Mol. Gen. Genet. 220, 325-328. 26 Ishii, S., Ihara, M., Maekawa, T., Nakamura, Y., Uchida, H. and F. Imamoto (1984) NAR 12, 3333-3342.