Biochimie (1992) 74, 353-356

© Soci6t6 franqaise de biochimie et biologie molt~culaire / Elsevier, Paris

353

Determination of tRNA Phe recognition nucleotides for phenylalanyl-tRNA synthetase from Thermus thermophilus N Moor t, I Nazarenko 2, V Ankilova 1, S Khodyreva t, O Lavrik I Ilnstitute of Bioorganic Chemistry, 630090 Novosibirsk; 2Novosibirsk State University, 630090 Novosibirsk, h'ussia

(Received 26 August 1991; accepted 12 December 1991)

Summary - - The tRNAabe recognition nueleotides for phenylalanyl-tRNA synthetase from an extreme thermophile Thermus thermo~ h ~ s ~ e " ~" ; ~ l ~".e Uc~l?nYaenac~t~RN~rP~i 17ctrptoSCrsitPe~l)Waitrehvarious point mutations it was shown that four recognition points t important for aminoacylation at 37°C. In the case of the 73rd discriminator base A->U, but not A--->C,substitution suppresses aminoacylation. Position 20, which proved to be essential for all phenylalanyl-tRNA synthetases so far studied, is not involved in the recognition of tRNAPheby the T thermophilus enzyme. phenylalanyl-tRNA synthetase / tRNA identity points Introduction Determination of recognition nucleotides in tRNA is very important for investigation of the mechanism of interaction between tRNA and aminoacyl-tRNA synthetase (EC 6.1.1.). Significant progress in these studies has been made by both in vivo suppression assays that employ modified tRNA sequences and in vitro kinetic studies of aminoacylation of tRNA transcripts made in the T7 transcription system [ 1]. Among 20 aminoacyl-tRNA synthetases, phenylalanyl-tRNA synthetase (FRS) is one for which the tRNA recognition set is known for the organisms on different evolutionary levels: E coli [2], yeast [3, 4] and man [5]. Now it is clear that FRSs from various organisms recognize either only anticodon, like methionyl-tRNA synthetase, or only acceptor end, as alanyl-tRNA synthetase, but have an extended tRNAPherecognition set. The recognition pa~ztems for yeast and human FRSs seem to be very similar and include anticodon, G20 and A73. The only difference between them is the importance of one or two additional anticodon, stem base pairs for the recognition of tRNAPhe by human enzyme [5]. According to the in vivo investigation the E coli enzyme recognizes nucleotides located at the comer of the L-shaped tRNA molecule, where the D-loop interacts with the T-loop, a part of the anticodon stem and a variable loop [2]. Nucleotides A73 and D20 are also important for identity.

Phenylalanyl-tRNA synthetase from Thermus thermophilus is an object of the present investigation. Though it is an eubacterium like E coli, it exists at extremely high temperature (80°C) [61. We suggest that protein-nucleic acid interactions at such a temperature should have some peculiar features. Besides, this incredible thermostability may help to recognize temperature-sensitive interactions between tRNA and the enzyme. Co-crystals of tRNA phe with phenylalanyl-tRNA synthetase from T thermophilus have been obtained recently (Reshetnikova et al, in preparation), so an in vitro investigation of the recognition set together with the crystallographic data may give a detailed picture of tRNAPhe-FRS interaction.

Materials and methods T7 RNA polymerase was isolated from E coil BL21 harboring the plasmid pARI219 and purified to a specific activity of 450 000 units/mg. Phenylalanyl-tRNA synthetase with specific activity of 200 units/mg was prepared from T thermophilus HB8 according to [6]. L-[3H]phenylalanine was purchased from Isotop (Russia). T thermophilus tRNAphe (phenylalanine incorporation of 1750 pmol/A:60) was isolated as described in [7]. tRNAphe from brewer's yeast (1700 pmol/A26o)was purchased from Boehringer Mannheim, Germany. Total tRNA from human placenta was purchased from Omutninsk chemical company, Russia. Plasmids containing wild-type and mutant yeast tRNAPhe genes constructed in the T7 RNA polymerase system were kindly provided by Prof O Uhlenbeck.

354 hz vizro transcription

Results and discussion

Transcription of plasmids after cleavage with BstN1 endonuclease was carried out as described elsewhere [8], The reaction mixture (0.3 ml) contained 40 mM Tris-HCl (pH 8.1), 22 mM MgCl~, 1 mM spermidine, 5 mM DTT, 4 mM of each NTP (pH 8.0), 30 lag of BstNl digested plasmid DNA and 0.1 mg/ml of T7 RNA polymerase at 42°C for 4 h. The reaction was stopped by adding EDTA to a final concentration of 50 raM. The reaction mixture was extracted with phenol/CHCl3 (!:i) and precipitated with ethanol. The transcript was purified to a single nucleotide resolution by electrophoresis on denaturating 20% polyacrylamide gel and eluted from the gel in 200 mM NaCI for 18 h at 4°C by shaking. The purified transcript was ethanol-precipitated, washed with 70% ethanol, dried and dissolved in sterile water. The concentration of each tRNA sample was calculated from the value of absorbance at 260 nm assuming the extinction coefficient to be 5.3 x 10s M-l cm-I.

Aminoacylation of tRNA Kinetic reactions were carried out in 30 lal of 50 mM Tris-HCI (pH 8.5), containing either 2 mM ATP and 9 mM MgCI2 (buffer A) or 10 mM ATP and 50 mM MgCI, (buffer B). The reaction mixture contained 5 laM L-[3H]phenylalanine and 3 lag/ml of FRS in both assays. The tRNA samples were heated to 90°C for 2-3 min and cooled to 25°C prior to addition to the aminoaeylation reaction mixture. Six tRNA concentrations were used for determination of Kr~ and V,, values. Reaction mixtures were incubated at 37°C. At 30-s intervals 5-lal aliquots were spotted on FN-I 1 paper filters impregnated with 5% trichloroacetic acid. KM and Vmvalues were calculated using an Enz Fitter program. Standard errors were within 10% of the indicated values. The final aminoacylation level for each tRNA was determined in the same reaction mixture with 0.1-0.2 laM tRNA and 3-30 lag/ml of synthetase.

The ability o f T thermophilus FRS to aminoacylate tRNAPh~ from E coil, yeast and human placenta was investigated. VJKM ratio was used to measure the efficiency o f aminoacylation. As shown in table I, E coli tRNA Phe is an efficient substrate for this e n z y m e under the aminoacylation conditions optimal for the therrnophilic system, ie 5 m M ATP and 9 m M MgCI2 (buffer A). Yeast tRNA vhe has a VJKM ratio 30 times lower than the cognate tRNAphe under the same conditions. T h e increase in ATP and Mg2+ concentrations up to 10 m M and 50 raM, respectively (buffer B), yields a four-fold higher Vm for the yeast tRNAphe and a 4.5-fold lower VJI(,M for the T thermophilus tRNAPhe. So, in buffer B, V J K u ratios for these two tRNA Phe are quite comparable. The aminoacylation o f the yeast tRNAPhe transcript was also investigated at different Mg2+ concentrations. It proved to be a poor substrate for T thermophUus FRS in the buffer A but, in the buffer B, its KM and V,,, are very similar to the corresponding values for the T thermophilus tRNAphe. So, in the buffer B, the unmodified yeast tRNA w~ transcript is a better substrate than its modified counterpart. The analogous results were obtained for the human tRNA~e and the corresponding transcript: the latter has a KM 4.5 times lower than the native tRNA (table I). The only explanation o f these facts m a y be a negative role o f modifications in the structure o f yeast and

Table 1. Aminoacylation of tRNAs with T thermophilus FRS.

tRNA PI'"

KMf p2ff)

Vm (nmol rain-I mg-O

Normalized VJKM

Aminoacylation assay~

T thermophilus

0.16

40

(1.0)

A

E coli

0.28

36

0.5 !

A

Yeast

2. i

14

0.03

A

-

2

0.04

A

Yeast tRNAE'hetranscript wdd-type

T thermoptzilus

0.22

12

0.22

B

Yeast

2.1

60

0.11

B

Yeast tRNAPhe transcript

0.20

25

0.50

B

Human

0.82

25

0.12

B

Human t R N A ~ transcript

0.18

16

0.36

B

~See Materials and methods.

355 human tRNA phe which prevent them from efficient recognition by the thermophilic FRS. A collection of yeast tRNAPhe mutant transcripts from prof Uhlenbeck's lab was used for investigation on a T thermophilus FRS recognition set. These tRNA phe transcripts with different point mutations were studied in aminoacylation reaction with T thermophilus FRS. Though the optimal temperature for tRNA aminoacylation catalyzed by the thermophilic FRS is 80°C [6], at the first steps of our investigation we used a lower temperature (37°C) in the kinetic experiments, to avoid possible melting of the transcripts. The data in table II show that all the three bases of the anticodon are very important for tRNAPh~ aminoaeylation. The substitution of any of the anticodon nucleotides G34, A35 or A36 resulted in a 40- to 250fold drop in VJKM. Another significant point is A73, the so-called discriminator base for a great number of tRNAs [1]. However, in our case the substitution of A73 to U73 was the only one to inhibit aminoacylation, A73-C73 replacement having a much smaller effect. The 'specificity' of this point mutation is in agreement with the data obtained for a tyrosyl-tRNA synthetasetRNAryr system from E coli [9]. The 20th position, an important recognition point for all phenylalanyl-tRNA synthetases so far studied, does not participate in the recognition of tRNAPhe by the thermophilic enzyme at all. The two tRNAphe mutants G20U (G replaced for U at the 20th position) and G20A show the same Vm/KM values when compared with the wild-type transcript. A high substrate efficiency of the natural E coli tRNAPh~, which has D in the 20th position, with the thermophilic enzyme supports this observation. The substitutions in the 59th, 60th positions and in the 26--44 base pair involved in tertiary interactions, which are important for the E coli FRS-tRNAP~e recognition 12], appeared to have a very modest, if any, effect in the ease of the T thermophilus enzyme. The same result was obtained for the anticodon base pairs 30--40 and 31-39, substantial recognition elements for the human FRS [51. The transcript with G30, C40, A31 and U39 substituted to C30, G40, C31 and G39 was aminoacylated with the efficiency comparable with that for the wild-type transcript. So, it may be concluded that at 37°C four out of five recognition points common to E coli, yeast and human FRSs, G34, A35, A36 and A73, are important for the recognition of tRNA phe b y the T thermophilus enzyme. This enzyme gives us the first example of a phenylalanyl-tRNA synthetase, for which a nucleotide in the 20th position does not participate in the recognition process. To identify other possible recognition points, different yeast tRNA transcripts (tRNATy r~phe, t R N A Met~phe, tRNAArg-~phe) were examined in aminoacylation reac-

Table IL Aminoacylation of yeast tRNAPhe transcripts with T thermophilus FRS. tRNA

KM (laM)

Normalized a

v~

V~'KM

Yeast tRNA phe transcripts: Wild-type

0.20

(100)

( 1.0)

G20U

0.17

80

0.94

G20A

0.21

150

1.4

A73U

13

30

0.005

A73C

0.57

120

0.40

G34A

14

30

0.004

A35U

5.2

40

0.015

A36C

4.8

60

0.025

U59C

0.59

140

0.47

C60U

0.27

160

!.2

G26A, A44G

0.20

100

1.0

G30C, C40G, A31C, U39G

0.57

130

0.46

tRNA Met-*l'he

0.12

83

1.4

tRNATyr-*Phe

0.36

200

I. 1

tRNAArg--*phe

0.36

53

0.30

aVmand VJKM are normalized to the wild-type transcript. tion with thermophilic enzymes. Due to multiple nucleotide changes these tRNA sequences have all recognition elements needed for the yeast FRS [3]. The tRNATyr~phc and tRNA Mct--'Phe exhibit normal aminoacylation kinetics with T thermophilus FRS (table II). The tRNAarg-*Phe has a three-fold lower V,./KM than the yeast tRNAPh~ transcript. Kinetic experiments with the same three tRNAs at 50°C revealed a very small effect of temperature on the recognition process. The data in table III show a two-fold increase in Km only in the case of tRNAArg-*ph¢. Therefore, we assume that the four single-stranded nucleotides in the 34-, 35-, 36th and 73rd positions of tRNA phe constitute the main part of the recognition set for the T thermophilus FRS. However, we do not exclude the existence of other recognition elements in tRNA phe which may be discovered under conditions optimal for this highly thermophilic enzyme.

356 Table IH. Temperature influence on the aminoacylation of tRNA phe transcripts.

2

Yeast tRNA transcripts

3

37°C

SO°C

K,, ( ~ )

VJK,,

K,, (p.M) Vm/K,,

tRNAVh~

0.20

(1.0)

0.22

(1.0)

tRNATyr~phe

0.36

1.1

0.37

1.4

tRNA Met~phe

0. ! 2

1.4

0.18

1.1

tRNAArg~phe

0.36

0.30

0.82

0.20

4

5 6

Acknowledgment

7

We would like to thank Prof 0 Uhlenbeck for a collection of plasmids containing wild-type and mutant yeast tRNA phe genes.

8

References

9

1

Normanly J, Abelson J (1989) tRNA identity. Annu Rev Biochem 58, 1029-1049

Mc Clain W, Foss K (1988) Nucleotides that contribute to the identity of Escherichia coli tRNA phe. J Mol Biol 202, 697-709 Sampson J, Direnzo A, Behlen L, Uhlenbeck O (1989) Nucleotides in yeast tRNAphe required for the specific recognition by its cognate synthetase. Science 243, 1363-1366 Sampson J, DiRenzo A, Behlen L, Uhlenbeck O (1990) Role of the tertiary nucleotides in the interaction of yeast phenylalanine tRNA with its cognate synthetase. Biochemistry 29, 2523-2532 Nazarenko I, Tinkle Peterson E, Zakharova O, Lavrik O, Uhlenbeck O (1992) Recognition nucleotides for human phenylalanyl-tRNA synthetase, in press Ankilova V, Reshetnikova L, Chemaya M, Lavrik O (1988) Phenylalanyl-tRNA synthetase from Thermus thermophilus HB8. Purification and properties of the crystallizing enzyme. FEBS Lett 227, 9-13 WatanabeK, Oshima T, lijima K, Yamaizumi Z, Nishimura S (1980) Purification and thermal stability of several amino acid-specific tRNAs from an extreme thermophile Thermus thermophilus HB 8. J Biochem 87, 1-13 Sampson J, Uhlenbeck O (1988) Biochemical and physical characterization of an unmodified yeast phenylalanine transfer RNA transcribed in vitro. Proc Natl Acad Sci USA 85, 1033-1037 Himeno H, Hasegawa T, Veda T, Watanabe K, Shimizu M (1990) Conversion of aminoacylation specificity from tRNA/~ r to tRNA set in vitro. Nucleic Acids Res 18, 6815-6819