Suppressor and novel mutants of bacteriophage T4 tRNAGly

./. .l~ol. Hiol. (1987) 193. 223-326

Suppressor and Novel Mutants of Bacteriophage T4 tRNAGly \Ve have isolated a weak UGA suppressor of phage T4 tRNAGly in which the ant)icodon is changed from ITCC to LTCA. Two secondary mutants lacking suppressor activity are atypical change the T stem of the cloverleaf model. One in accumulating tRNA G1y. Both mutations involves a G to A change at the 5’ base position of the middle base-pair: the second involves a (’ to I’ change at a constant, base position next to the T loop. The precursor RKAs of the mutants were cleaved in vitro with the catalytic RNA subunit of RNase P. Relative to normal precursor RNA, t,he precursor mutat,ed at the middle base-pair position of the T stem was cleaved more rapidly, whereas the precursor mutated at the base-pair position next to t)he T loop was cleaved more slowly.

The t RNA system of bacteriophage T4 has provided information on tRNA transcription and and on the relationship of tRNA processing. structure to function (Broida CYEAbelson, 1985). Suppressor derivatives of the tRNAs have played a key role in t,hesr studies. Thus far, four of eight tRN.4s have been obtained as suppressors. These include tRNASe’. tRNAG’“, tRNAL’” and tRNAArg (e.g. see Foss rf nl.. 1979). We now report’ on the isolation of a suppressor of tRNAGly. A preliminary analysis of mutants derived from this suppressor allows us t,o speculate on the importance of a region on the 1’ strm of t#he cloverleaf model for tRNAGly synthesis and stabilit’y. Previous sequencing studies established that the anticodon of T.4 tRNAGty is UCC (Barrel1 et al., 1973: Stahl lc(“lain. 1!177). This revertant retained eL1G and A53 markers of the st,arting strain. but lost the Op(“23 allele. Qualitative spot tests showed that at the rL1 coclon. IJGA was suppressed, but LJAA and I’;\(: were not. The suppressor proved weak, with an rffic+nc~y of about lo,, (for met)hods, see Foss et

al.. 1979; M&lain & Foss, 1984). Weak suppression is consistent with previous results with cell-coded performed suppressors of tRNAGly, which generally at efficiencies of 0.5 to S.Oo/o, though values of 2700 have been obtained (e.g. see Hill, 1975). After further studies described below. the new suppressor was designated psu&,: it is the fifth 7’4 suppressor, and it is specific for opal nonsense mutations. The fourth suppressor was described by Kao & M&lain (1977). To confirm that the suppressor was tR’NAGly. we obtained suppressor-negative derivatives of psu&, and examined these for reduced tRNAGly synthesis. It has been demonstrated that such mutants accumulate little mature tRNA as the result of defects in precursor RNA processing and/or stability (Abelson et al., 1970). Mut,ants were produced by hydroxylamine mutagenesis of psu&,eLlG-A53. (‘ells were infected with no more t!han one mutagenized phage each, and incubated to lysis. Non-lysed cells, arising from failure to suppress the eLlG UGA mutation, were collected and the phage purified. The resulting phage were then screened genetically to ident,ify the desired types (Comer et aE., 1974). Ten suppressor-negative tnutants were isolated; an additional two of spontaneous origin were also obtained. ,411 mutants were absolute, reducing the plating efficiency on li:. coli strain B/5 from approximately 1 to less than 1w4. Seven hydroxylamine mutants and both sponta,neous mutants synt.hesized no t,RNAGly (Fig. 1). confirming previous result’s These mut’ants were not studied further. The three remaining hydroxplamine mutants unexpectedly produced substantial t.RNiiZ ci’y (Fig. 1). The levels of tRNAGtY in these three mutants relative to the parent phage were 82:,, to 9500. The three mutants and the parent psu,+,, were then sequenced. using as a guide t,he previously determined sequence of wild-type tRNAGty (Barrel1 et nl., 1973; Stahl of /II.. 1973). The psu&, strain was analyzed first. IJniformly

224

W. H. McClain

tRN

Figure 1. Gel profiles of strain p&,-eLlG-A53 (lane 3), and 2 (lane l), strain psu&,- A49-eLlG-A53 typical suppressor-negative mutant strains (lanes 2 and 4). The position of tRNAG’y is indicated. E. coli strain B/5 was labeled with 32P 4 to 15 min after infection. RNAs were fractionated on a 10% polyacrylamide gel.

et’ al.

32P-labeled tRNAGiy, obtained from infected cells and purified by gel electrophoresis, was fingerprinted and the products analyzed. All RNase A and T, oligonucleotides were then digest,ed with complementary RNases (see McClain & Foss, 1984). In t,he RNase A fingerprint, one mole of AAU was replaced by AAAU (not shown; see below). One of the A residues in AAAU was modified to ms2i6A (and i6A); presumably. it was the middle A residue (see Fig. 2). This base modification is not present in wild-type T4 t’RNAGly. Similar msZi6A modification accompanying formation of a cell-coded tRNAGiy suppressor has been reported (Prather et al., 1983). In t’he RNase T, fingerprint,, wild-type ACUUCCAAUCUG was replaced by a product’ yielding AC, C, U, C: AAAU. C, U, G upon RNase A digestion. The indicated A modification was again noted. All other products were consistent with the wild-type sequence of t,RNAGly, including modified bases fT and Y’. Thus, the psu&, mutation causes a C to A change in the ant,icodon of tRNAGly (Fig. 3). Similar sequence analysis of the three mutants revealed the same C to A anticodon change and an additional change. Two had a G to A change at position 49 from the 5’ t)erminus; one of t,hese was saved as a representative, and named psu&A49 (Fig. 3). The RNase A fingerprint of this mutant showed that GAGT was replaced by AAGT. The T, fingerprint showed that UG and AG were replaced by UAAG (Fig. 2(a) and (b)). The change in the third mutant (psu&-U59) was established from RNA and DNA sequence determinat’ions. The RNase A pattern was similar to the ~su&~ parent. However, in the RNase T, pattern, fragment AUUCUCAUUAUCCG migrated more rapidly in the first dimension (see Fig. 2(b)), indicating a C to U change in it’. RNase V-Z digestion of this T, fragment, indicat,ed that UUCUCA was replaced by CU(l-1. F, C)A: the struct’ure indicated for the latter U, product, was det,ermined by partial digestion with spleen phosphodiesterase. DNA sequencing (Mazzara et al.. 1981; Sanger et al., 1977) was used t,o determine the exact location of the change. DNA prepared from virions of psu&- U59-eLlG-A53 was digested with TaqI and cloned into the Clal sit’e of plasmid pBR322. The DNA was then recloned into the replicative forms of vectors mp8 and mp9 following double digestions of each with EcoRI and HindIII. The sequence in mp8 was read from t,RNAGly position 1 to 56, and in mp9 from t,RNAGly position 53 to 74. Figure 3 shows the location of this mutation. We then examined the mutants for secondary alterations associated with the base changes not,ed above. The levels of certain modified nucleotides in tRNAGly were determined for mutant A49. RNase A or T, fragments containing residues fi33, ms2i6A36, Y38 and T52 were isolated, digested to mononucleotides with RNase T,, then fractionated on thin-layer plates (Barrel1 et al., 1973). The modification levels were reduced t)wo- to fourfold.

Letters to the Editor

(a)

(b)

Figure 2. (a) RKase A fingerprint and (b) RNase T, fingerprint of t’RNAG’y produced by strain psn:,,-A49-rLlG-A53. Fingrrprint origins are at the upper right (for methods, see Barrel1 et al.. 1973). The arrow in (b) tnarks the oligonuclrotide altered in strain psu&- U59-eLlG-A53; the mutant form migrated nearer the t,ail of the arrow.

The 5’ and 3’ termini of tRNAGly were then examined. RNaxe T, digestion of tRNAGly normally produces p(: a,nd CUCCA,, (see Fig. 3). These fragments were reduced in yield about twofold in

mutant A49. and were absent in mutant U59; new termini were not identified. We next, investigated tRNAGly synthesis. RNase P is an enzyme that cleaves precursor RNAs to generate the 5’ termini of tRNAs. This enzyme is composed of two subunits, an RNA and a protein; the R.XA subunit (Ml RKA) is responsible for catalysis (Guerrier-Takada et al., 1983). Mutant precursor RSAs were isolated and cleaved in vitro by Ml RXX. Relative to the precursor of psu&,,, we found that precursor of mutant psu&,-A49 was a

U pG

0

C

COG GOC G@C A@U

psu;-u59

lJ@A AOU

?“”

A U G ImL-’ G DA

C

A

0.0..

UAUG

G U G\A Cl T yrc U “‘\ A A u GA p~u;,~-A49

0.0 “UAC ’

u 8 COG A@ U G@C Am’+’

twofold better substrate for Ml RNA cleavage (Fig. 4). The precursor of mutant psu&,,-U59 was

A A

C

Figure 3. Seyurhnce of phage T4 tRNAG”’ (Barrel1 et al., 1!173) giving locations of the anticodon change in strain psu&, and the changes in suppressor-negative mutants. Partial modification to msZi6A (and i6A) was noted for an ;I rrsidue adjacent t,o the antirodon.

’

A

UACUC

U

U

;cc \ A l P%op

G

226

W. H.

McClain

et al. RNase P. The latter observation suggests that the distortion (Hunter et al.. 1986) and destabilization (de Bruijn & Klug. 1983) of the helix containing the mutant AC opposition facilitates precursor RNA cleavage in some manner. However, additional work is needed to determine the basis of this rate enhancement, as well as the behavior of RNase P holoenzyme on the mutant substrate. This work was support’ed by grant AI10257 from t,hr National Tnst,itutes of Health.

William H. McClain K. Foss Jay Schneider Department of Bacteriology CTniverxity of i$%consin Madison WI 53706. I’.S.A.

Cecilia Guerrier-Takada Sidney Altman I 2

I 4

I 6

Department of Biology Yale University New Haven. CT 06520. 1T.S.A.

I 0

Time (mln)

Figure 4. Time-course of precursor RNA cleavage by the Ml subunit of RNase P. + , precursor RNA of parent phage, strain psu& -eLlG-653. A49, precursor RNA of mutant phage. strain psu&,-A49-eLlG-A53. RNAs from infected E. coli strain A49 growing at 42°C (McClain & Foss, 1984) were fractionated on a polyacrylamide gel; purified precursor RNAs were electroeluted from the gel: 12 ng of Ml RNA was preincubated for 5 min at 37°C in 20 ~1 of buffer containing 5OmM-Tris. HCI (pH 7.5). 100 mM-NH,Cl and 100 m&I-MgCl, (Guerrier-Takada et al., 1983). Precursor RNA was added and samples were removed at the indicated times. Products were fractionated on a 60/o polyacrylamide gel containing 7 M-urea. cleaved more slowly by Ml RNA, and we estimate that it was a two- to threefold poorer substrate (not’

shown). The gel mobility of the tRNA”lY product liberated from each precursor was normal, indicating site-specific cleavage, though specificit,y was not confirmed by product analysis. We do not know if the rate differences observed with Ml RNA would also be obtained with the RNase P holoenzyme. Several novel findings are reported above. That certain mutant changes on the T stem allow appreciable synthesis and stability of tRNAGIY was not, expected. Despite these substantial levels, the protein synthetic functions of the mutant’ molecules are affected as the suppressor activities are reduced. Also noteworthy is the accelerated cleavage of the mutant A49 precursor RNA by the Ml subunit of

Edited

Received 14 July 1986

References Abelson, J N.. Geftrr. M. L., Barnett. L., Landy. A., Russell. R. J,. & Smith, ,J. D. (1970). .J. Mol. Biol. 47. 15-28. Barrel], H. G..
by S. Brenner