ABSTRACT A DNA oligomer 15 nucleotides long was used to probe the involvement of RNA secondary structure in the control of transcription termination at the attenuator of the tryptophan ( trp ) operon of Escherichia coli This 15-mer is perfectly comple - mentary to a segment of trp RNA that is thought to play a role in regulation of attenuation .
When added to an in-vitro-transcription reaction mixture containing wild-type E. coli or Salmonella typhi - murium trp operon templates , the complementary 15-mer caused a 4-fold increase in read-through transcription .
By contrast , the 15-mer did not affect attenuation when a mutant E. coli template was used that does not allow formation of a crucial RNA secondary structure .
Control experiments established that oligomers that were not complementary to E. coli trp leader RNA did not affect attenuation and that the 15-mer did not reduce termination when the transcript lacked a complementary region .
Other experiments established that the 15-mer did not increase read-through tran - scription by allowing RNA polymerase molecules that might have already stopped at the attenuator to resume transcription .
These findings provide direct support for the view that alternate base - paired structures control transcription termination at the trp attenuator .
It has been firmly established that expression of the tryptophan ( trp ) operon of enterobacterial species is controlled by a form of regulated transcription termination called attenuation ( reviewed in refs .
One model proposes that termination at the trp attenuator is modulated by mutually exclusive , alternate secondary structures that form in nascent RNA as RNA polymerase transcribes the leader region of the operon ( Fig. 1 ; refs .
According to this model , transcription termination is caused by formation of the `` terminator '' RNA secondary structure , designated 3-4 in Fig. 1 ( 2 , 4-6 ) .
Mutations and conditions that inhibit formation ofthe terminator ( Fig. 1 A ) or that increase formation of an alternate stem and loop structure , 2-3 , the `` antiterminator '' ( Fig. 1B ) , result in increased read-through transcription beyond the trp attenuator .
For example , mutations trpL1l6 and trpLl32 ( Fig. 1A ) introduce base-pair mismatches , which should destabilize the terminator ( 5 ) .
This destabilization could account for the high levels of read-through transcription observed in-vivo and in-vitro when the trp operon has either of these mutations ( 2 , 5 ) .
In contrast , mutation trpL75 ( Fig. 1A ) , which should destabilize the antiterminator , increases the frequency of transcription termination at the attenuator ( ref .
7 ; unpublished data ) ; this presumably occurs because the terminator forms more readily .
When restriction fragments containing the wild-type Esch-erichia coli trp promoter-operator-leader region and part of The publication costs ofthis article were defrayed in part by page charge payment .
This article must therefore be hereby marked `` advertisement '' in accordance with 18 U. S. C. § 1734 solely to indicate this fact .
the first trp structural gene are used as templates in a purified in-vitro-transcription system , = 95 % of the transcribing RNA polymerase molecules terminate transcription at the trp attenuator ( 4 ) .
To account for this high level of termination , it has been suggested that in the absence of translation , a third alternate secondary structure can form-structure 1-2 ( Fig. 1A ; ref .
Formation of this structure should inhibit formation of the antiterminator , which in turn should enhance formation of the terminator .
To test this model , we synthesized a 15-mer that is perfectly complementary to segment 1 ofstructure 1-2 ofE .
Addition of this complementary 15-mer to an in-vitro-transcription mixture specifically decreased transcription termination at the wild-type trp attenuator .
This result provides strong support for the view that alternate RNA secondary structures mediate transcription termination and its control at the trp attenuator .
MATERIALS AND METHODS Materials .
[ a-32P ] GTP ( specific activity , = 400 Ci/mmol ; 1 Ci = 3.7 x 1010 becquerels ) was purchased as an aqueous solution from Amersham .
E. coli Hpa 11-570 , Salmonella typhi-murium Hinfl-500 , and Serratia marcescens Hpa 11-400 restriction fragments , containing the trp promoter , leader , and wild-type or mutant attenuator regions and the initial part of the trpE gene , were used as templates in in-vitro-transcription reactions .
These restriction fragments were prepared as described ( 9 ) .
DNA concentrations were measured by spotting diluted DNA samples on buffered agarose that contained ethidium bromide ( 10 ) .
RNA polymerase holoenzyme was prepared as described ( 9 ) .
Oligomer concentrations were determined by measuring A2w and assuming molar extinction coefficients of 1.48 X 105 , 1.59 x 105 , and 1.60 X 105 M - ` cm ' - forthe 14-mer , 15-mer , and 16-mer , respectively .
The 14-mer , d ( G-C-G-C-A-G-T-A-C-A-T-C-T-C ) , the 15-mer , d ( A-G-T-G-C-G-C-C-A-C-C-A-A-C-C ) , and the 16-mer , d ( C-G-C-C-G-C-T-T-T-G-T-G-A-A-C-C ) , were synthesized using standard triester methods ( 11 ) .
Intermediates were purified by using normal phase silica gel ( Sil G-25 UV254 , Brinkmann ) with MeOH/CHCl3 , 1:9 ( voVvol ) , and then re-verse-phase silica gel ( silica gel 60 F-254 silanized ; EM Re-agents ) with acetone/water , 65:35 ( vol/vol ) .
Purity of fully protected intermediates was monitored by comparing the absorbance in HC1OJEtOH , 1:1 ( vol/vol ) , at 500 , 330 , 300 , 280 , 260 , and Abbreviation : Hpa 11-570 , Hinfl-500 , and Hpa 11-400 , designations for specific restriction fragments in which the numbers specify length in base pairs .
t To whom reprint requests should be addressed .
250 nm , with absorbances predicted from the base composition by a computer program that treats the composite absorbance of a given sequence as a linear combination of the absorbances of the appropriate monomers .
§ Deprotected oligonucleotides were initially isolated by high performance liquid chromatography on a Spherogel TSK2000SW ( 7.5 X 600 mm ) exclusion column with 20 % ( vol/vol ) acetonitrile in 1 % ammonium ace-tate ( pH 6.0 ) and further purified by using a 7-13 % acetonitrile gradient in 1 % ammonium acetate ( pH 6.0 ) and an Alltech C8 ( 4.6 X 250 mm ) reverse-phase column .
The sequence of purified oligomers was confirmed by 32P labeling with T4 polynucleotide kinase ( Bethesda Research Laboratories ) , excising the appropriate band from a 25 % polyacrylamide gel ( no urea ) , and subjecting the eluted material to a modified Maxam-Gilbert procedure ( unpublished data ) .
Transcription Experiments and Gel Electrophoresis .
Reaction mixtures ( 21.5 , l ) containing 36 mM Tris acetate , pH 7.8 ( at 22 °C ) , Na2EDTA , 0.1 mM dithiothreitol , 4 mM Mg 0.1 mM acetate , 50 mM KC1 , 5 % ( vol/vol ) glycerol , 20 AM [ a-32P ] GTP ( 15 , uCi ) , 150 AM ATP , 0.025 , g ofrestriction fragment ( 0 .
075 pmol ) , and 0.072 , g of RNA polymerase holoenzyme ( -0 .
15 pmol ) preincubated for 37 °C .
Either 1 , ul of 10 were 10 min at Tris-HCl/0 .1 mM EDTA , pH 7.9 , or 1 , ul of that buffer mM containing oligomer ( generally 37.5 pmol ) was added to the reaction mixtures .
Preincubation was continued for another 10 min at 37 °C , and transcription was started by adding 2.5 , ul of a prewarmed solution of 1.5 mM CTP and 1.5 mM UTP to each reaction mixture .
After 20 min at 37 °C , transcription reactions were stopped ( 9 ) , and the reaction mixtures were loaded directly ( without heating ) onto 6 % polyacrylamide/7 M urea denaturing slab gels containing 0.09 M Tris borate/2 .5 mM EDTA , pH 8.3 ( 4 , 8 ) and were electrophoresed .
Transcription products were detected by autoradiography , cut from the gels , and assayed by Cerenkov radiation .
For each transcription reaction , the amounts ofterminated leader RNA and read-through RNA were used to calculate the molar percentage of read-through transcription .
Transcription in the Presence of Oligomers .
E. of a typical transcription experiment in which the wild-type coli Hpa II-570 restriction fragment was used as template are 1 .
shown in Fig. 2 ( lane 1 ) and quantitated in Table RNA polymerase molecules that initiate transcription at the trp promoter either terminate transcription at the trp attenuator to produce continue a 140-nucleotide terminated-leader transcript ( 4 ) or transcription to the end of the restriction fragment and synthesize a 260-nucleotide read-through transcript ( 4 ) .
Under the conditions used , about 5 % of the transcribing RNA polymerase molecules continued transcription beyond the trp attenuator ( Table 1 ) .
On a molar basis , the sum of the terminated-leader and read-through RNA species obtained represents 6 % of the amount of restriction-fragment template added to the reaction mixture .
When a 15-mer complementary to segment 1 ( Fig. 1C ) was added to the transcription reaction mixture , significantly more read-through transcript and significantly less terminated-leader transcript were observed ( Fig. 2 , lane 2 ; Table 1 ) .
At a 500-fold molar excess of 15-mer to restriction fragment at 37 °C , = 21 % of the transcribing RNA polymerase molecules continued tran scription beyond the attenuator .
At this concentration , the 15-mer caused only a modest reduction of about 14 % in total trp RNA synthesis .
Several control experiments established that the reduction in transcription termination observed in the presence of the 15-mer depended on complementarity between the 15-mer and segment 1 of trp RNA .
A 14-mer and a 16-mer with approximately the same G + C content as the 15-mer , but which are not complementary to trp RNA , were added in transcription experiments ( Fig. 2 , lanes 5-8 ; Table 1 ) ; neither ofthese oligomers affected transcription termination at the trp attenuator .
The effect of the complementary 15-mer on transcription with mutant E. coli templates was also examined ( Table 1 ) .
The 15-mer did not reduce termination with the trpL75 mutant template ( Fig. 2 , lanes 3 and 4 ; Table 1 ) .
This result strongly suggests that the decrease in transcription termination observed with the wild-type template is dependent upon formation of alternate RNA secondary structure 2-3 ( see Fig. 1 and Discussion ) .
The 15-mer did not detectably affect termination with either the trpL116 or trpL132 mutant template ( Table 1 ) .
These templates normally allow appreciable read-through transcription because the trpL116 and trpLl32 mutations destabilize the terminator ( see Discussion ) .
Attenuation also was examined by using trp operon templates isolated from S. typhimurium and Ser .
The 15-mer complementary to segment 1 in E. coli trp RNA ( Fig. 1C ) is also complementary to an analogous segment 1 in S. typhimurium trp RNA ( 12 ) .
The 15-mer reduced transcription termination at the trp attenuators of E. coli and S. typhi-murium almost equally ( Table 1 ) .
In contrast , wild-type trp RNA of Ser .
marcescens is not perfectly complementary to the 15-mer ( two mismatches ; ref .
13 ) , and the base pairing that is possible is not in a region of Ser .
marcescens trp RNA analogous to segment 1 in E. coli trp RNA ( unpublished data ) .
The 15-mer did not affect termination with the wild-type Ser .
Consistent with this result , the 15-mer did not relieve termination when we employed a Ser .
marcescens mutant template containing a deletion that reduces transcription read-through and removes all regions complementary to the 15-mer ( trp4425 , Table 1 ) .
Because some preliminary evidence suggests that RNA polymerase molecules may remain associated with terminated-leader RNA and the DNA template after attenuation in vitr ( refs .
14 and 15 ; unpublished data ) , we determined whether the 15-mer exerts its effect as the polymerase approaches the attenuator or whether it acts by allowing polymerase molecules that have stopped at the attenuator to resume transcription .
In this experiment , a transcription reaction mixture lacking the 15-mer was incubated for 10 min at 370C .
Transcription initiation was then blocked by addition of rifampicin ( to 10 Ag/ml ) ; after 3 min , the 15-mer was added , and incubation was continued for 15 min .
The 15-mer did not increase read-through transcription .
Analysis of a parallel mixture lacking rifampicin proved that a substantial fraction of the RNA polymerase molecules added remained active throughout the incubation period .
Finally , it has been observed that RNA polymerase molecules pause in the vicinity of base pair 90 as they transcribe the trp leader region ( 9 , 16 ) .
Transcription experiments performed with nucleotide analogues suggest that the pause near base pair 90 in the trp leader region may be caused by formation of structure 1P2 ( 16 ) .
Consequently , we examined pausing on the wild-type E. coli template in the presence of the complementary 15-mer .
The 15-mer did not detectably alter the formation or half-life of the paused-leader transcript at 37TC .
However , if the 15-mer reduced the formation of the paused species by only 20 % , our measurements would not have been sensitive enough to detect it .
Conditions Affecting Transcription Termination in the Presence of the 15-mer .
Fig. 3 shows the dependence of read-through transcription on the amount of 15-mer added to reaction mixtures containing the wild-type E. coli template at 370C .
At amounts below 1 pmol ( molar ratio of 15-mer to template , = 10 ) , the 15-mer did not affect attenuation appreciably .
At amounts above 23 pmol ( molar ratio of 15-mer to template , -300 ) , the 15-mer maximally stimulated read-through transcription beyond the trp attenuator .
In view of this concentration dependence , experiments were generally performed with 37.5 pmol of 15-mer per 25 1 , u of reaction mixture ( molar ratio of 15-mer to template , -500 ) .
The ability of the complementary 15-mer to influence transcription termination on the wild-type E. coli template was strictly a function of the temperature of the transcription reaction .
At 12 °C and 220C , the complementary 15-mer did not significantly increase read-through transcription .
Increasing the KC1 concentration in the reaction mixture from 50 mM to 150 mM or decreasing the [ a-32P ] GTP concentration from 20 / .
M to 1 IuM did not alter the effect of the 15-mer on read-through transcription at 370C .
DISCUSSION The results show that a 15-mer perfectly complementary to what we designate segment 1 of E. coli trp RNA ( Fig. 1C ) significantly decreased transcription termination at the wild-type attenuators ofE .
coli and S. typhimurium ( Fig. 2 ; Table 1 ) .
This effect was sequence specific ; the 15-mer did not reduce attenuation with templates whose transcripts lack complementarity to the 15-mer .
In addition , a 14-mer and a 16-mer that are not complementary to E. coli trp RNA did not affect read-through transcription on the wild-type E. coli template ( Fig. 2 ; Table 1 ) .
Rifampicin control experiments indicated that the 15-mer did not act by allowing polymerase molecules that might have stopped at the attenuator to resume transcription .
The decrease in attenuation caused by the complementary 15-mer supports our proposed model in which alternate RNA secondary structures mediate transcription termination and its control at the trp attenuator ( see Introduction ) .
It is important to note that , although the 15-mer is perfectly complementary to RNA segment 1 ( Fig. 1C ) , it is not complementary to any other segment of trp leader RNA .
When the 15-mer is included in the transcription reaction mixture , it presumably base pairs with RNA segment 1 , possibly before RNA segment 2 is synthesized .
This pairing between the 15-mer and segment 1 could prevent formation of structure 1-2 ( Fig. 1A ) and allow structure 2-3 , the antiterminator , to form ( Fig. 1C ) .
Formation of the antiterminator presumably prevents formation of structure 3-4 , the terminator , ( Fig. 1A ) and thereby permits increased read-through transcription beyond the attenuator .
Transcription experiments performed with mutant E. coli templates further support this interpretation .
The 15-mer did not increase read-through transcription on a template containing the trpL75 mutation ( Fig. 1 ; Table 1 ) .
This result can be explained by noting that the trpL75 change destabilizes the antiterminator ( 7 ) .
Therefore , even if the 15-mer and segment 1 formed a paired structure , the antiterminator could not form and interfere with formation of the terminator .
In addition , the 15-mer did not increase read-through transcription on templates containing either the trpLl16 or trpLl32 mutation ( Table 1 ) .
Each of these mutations partially destabilizes the terminator and causes high level read-through transcription ( Table 1 ) .
With the trpL116 and trpL132 mutant templates , increased antiterminator formation , presumably caused by pairing ofthe 15-mer to RNA segment 1 , apparently is not sufficient to significantly reduce formation of the destabilized terminator .
This is not unexpected in view of the modest increase in read-through transcription obtained with the wild-type template ( Table 1 ) .
The effect of the 15-mer on attenuation depends on the reaction temperature and on the concentration of 15-mer added to the transcription reaction mixture ( Fig. 3 ) .
In other studies , it has been shown that the extent of transcription termination at the wild-type E. coli trp attenuator is markedly influenced by temperature ( ref .
At present , there are no experimental findings that explain fully either the temperature dependence or the concentration dependence of the effect of the 15-mer on attenuation .
We expect that these dependencies are related to the rate of formation of different secondary structures and the availability of RNA segment 1 and its rate of pairing with the 15-mer .
Studies on attenuation suggest that a site contributed in part by the/3 subunit of RNA polymerase recognizes the RNA terminator as the signal to terminate transcription ( 17 ) .