Proc. NatL Aoad. Sci. USA
`Vol. 80, pp. 368-372, January 1983
`Biochemistry
`
`Attenuation control of pyrBI operon expression.in
`Escherichia .coli K-12
`(UTP-regulated gene expression/coupled transcription-translation/DNA sequence/in vitro transcription)
`CHARLES L. TURNBOUGH, JR., KAROL L. HICKS, AND JOHN P. DONAHUE
`Department of Microbiology, University-of Alabama in Birmingham, Birmingham, Alabama 35294
`Communicated by Bruce N. Ames, October 12, 1982
`
`The pyrBI operon of Escherichia coli K-12 en-
`ABSTRACT
`codes the subunits of the pyrimidine biosynthetic enzyme aspar-
`tate transcarbamylase (carbamoylphosphate:L-aspartate carba-
`moyltransferase,-EC 2.1.3.2). Expression of this operon apparently
`is negatively regulated by the intracellular levels of UTP. To elu-
`cidate the regulatory mechanism in which UTP functions, the nu-
`cleotide sequence of the promoter-regulatory region of the pyrBI
`operon was determined and DNA fragments containing this region
`were transcribed in vitroQ These experiments revealed a p-inde-
`pendent transcriptional terminator (attenuator) located only 23
`base pairs before the promoter-proximal end of the structural
`genes. Transcription, initiated upstream at either of two potential
`pyrBI promoters was efficiently ("-98%) terminated at this site,
`indicating that the regulation of pyrBI expression involves atten-.
`uation control. Additional features identified suggest a model for
`regulation in which the relative- rates of UTP-dependent tran-
`scription within the pyrBI leader region and coupled translation
`of the leader transcript control transcriptional termination at the
`attenuator.
`
`In Escherichia coli K-12 and closely related bacteria, de novo
`synthesis of UMP is catalyzed by six enzymes encoded by six
`unlinked pyrimidine genes and operons (1-3). The expression
`of these genes and operons appears to be noncoordinately reg-
`ulated by pyrimidine nucleotides (4), but little is known about
`the regulatory mechanisms involved. The pyrBI operon en-
`codes the catalytic (pyrB) and regulatory (pyrl) subunits of the
`pyrimidine biosynthetic enzyme aspartate transcarbamylase
`(ATCase; carbamoylphosphate:L-aspartate carbamoyltransfer-
`ase, EC 2.1.3.2) (3). Previous in vivo studies have indicated that
`pyrBI expression is negatively regulated over a several hun-
`dredfold range by the levels of a uridine nucleotide (4). Recent
`experiments using an in vitro coupled transcription-transla-
`tion system have identified UTP as the principal pyrimidine
`regulatory effector of this operon (5). This result and the ob-
`servation that ATCase synthesis is preferentially stimulated by
`sublethal concentrations of inhibitors of transcriptional elon-
`gation in Salmonella typhimurium (unpublished data) suggest
`that the rate of UTP-dependent transcription is involved in the
`regulation of pyr'BI expression. In addition, ATCase synthesis
`was shown to be selectively inhibited in. a hisT strain of S. ty-
`phimunium (6), in which the rate of translational elongation is'
`slowed (ref. 7; D. Palmer and S. Artz, personal communication).
`This result suggests that the rates of both transcription and
`translation, perhaps of a leader sequence preceding the pyrBI
`structural genes, are involved in regulation. The regulatory
`mechanism could be similar to the attenuation control mecha-
`nisms of amino acid biosynthetic operons (8).
`In this study we determined the nucleotide sequence of the
`
`The publication costs ofthis article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertise-
`ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
`
`promoter-regulatory region of the pyrBI operon ofE. coli K-12
`and characterized in vitro transcription of DNA fragments con-
`taining this region. The results indicate that pyrBI expression
`is regulated by an attenuation control mechanism (8) that could
`be sensitive to the relative rates of UTP-dependent transcrip-
`tion within the pyrBI leader region and translation ofthe leader
`transcript.
`
`MATERIALS AND METHODS
`DNA Preparations. The A specialized transducing phage Ad
`valS argI pyrB (ykl4m5) was prepared from the E. coli K-12
`lysogenic strain AD11m5 (9) provided by Akihika Kikuchi, and
`phage DNA was extracted (10). Plasmid DNA was isolated by
`the procedure of Birnboim and Doly (11) and purified further
`by CsCl/ethidium, bromide density gradient centrifugation.
`(12). DNA restriction fragments were separated by agarose gel
`electrophoresis (13) and extracted from the gel by electroelution
`(14). Fragments used as templates in in vitro transcription re-
`actions were phenol extracted. The sizes of plasmids and re-
`striction fragments were measured by agarose gel electropho-
`resis with appropriate standards.
`Restriction Digests, Ligations, and Transformations.. Re-
`striction endonucleases- were obtained from New England
`BioLabs and used as recommended by the supplier. Conditions
`for ligation of restriction fragments with T4 DNA ligase (New
`England BioLabs) were essentially as described (12). The pro-
`tocol used for transformations was that of Morrison (15).
`DNA Sequence Analyses. DNA sequences were determined
`by the method of Maxam and Gilbert (16).
`In Vitro Transcription. RNA polymerase was purified from
`E. coli strain MRE 600 (17, 18) and was a gift from DavidWood
`and Jack Lebowitz. In vitro transcription assay conditions were
`as follows unless indicated otherwise in the text. Reaction mix-
`tures (0.05 ml) contained: 20 mM Tris HCl (pH 7.9); 10 mM
`MgCl2; 50 mM KCl; 0.1 mM Na2EDTA; 0.1 mM dithiothreitol;
`0.2 mM GTP; 0.2 mM CTP; 0.2 mM UTP; 0.2 mM [a-32P]ATP
`(2.5 Ci/mmol; 1 Ci = 3.7 X 10'° Bq; ICN); DNA at 21 pmol/
`ml; and RNA polymerase at 256 pmol/ml. Reaction, mixtures
`lacking ribonucleoside triphosphates were preincubated for 5
`min at 37°C. Reactions were initiated by the addition of the
`triphosphates; 30 sec later, rifampicin was added to a final con-
`centration of 4 ,uM. Incubation was continued for 30 min at
`37°C, and reactions were terminated by freezing in dry ice.
`Yeast carrier RNA (final concentration, 1 mg/ml) and 0.25 ml
`ofextraction buffer (10 mMTris-HCl, pH 7.9/10 mM Na2EDTA/
`0.1% NaDodSO4) were then added to the reaction mixtures and
`each was extracted with an equal volume of phenol. The RNA
`was precipitated with ethanol, dried, and dissolved in 99%
`formamide. Samples were placed in boiling water for 3 min,
`quick chilled in ice-water, and analyzed by electrophoresis on
`
`Abbreviations: ATCase, aspartate transcarbamylase; bp, base pair(s).
`
`368
`
`Merck Ex. 1053, pg 1332
`
`

`
`Biochemistry: Turnbough et aL
`
`Proc. Natd Acad. Sci. USA 80 (1983)
`
`369
`
`IV
`Pvu fl
`I
`
`I
`100
`
`I
`200
`
`.
`
`TDlHIHUHIM
`Hinf I Hooj
`FnDIErQdCI
`-gDLogi Dde
`Pst I
`Fn>D11 Pvul
`I~~~~~~~~~~~~~~~~~~~~~~
`...I',
`I00
`.
`I
`300
`400
`500
`600
`700
`FIG. 1.
`Strategy and restriction sites used for sequence analysis of the 758-bp Pvu II fragment containing the promoter-regulatory region of
`the pyrBI operon. All restriction fragments were 5'-end labeled (16). Arrows indicate direction and extent of each sequence determination. The
`heavy bar represents DNA encoding the NH2 terminus of the ATCase catalytic subunit.
`7% polyacrylamide/Tris/borate/EDTA, pH 8.3, gels contain-
`(22), resulting in the genetic fusion ofthe inserted promoter and
`ing 7 M urea (19). Autoradiograms of gels were scanned with
`the galK structural gene. In strains containing this recombinant
`a densitometer to quantitate transcripts.
`plasmid, the expression of the fused galK gene was regulated
`by the availability of pyrimidines in the same manner as that
`of the wild-type pyrBI operon (unpublished data).
`DNA Sequence of the pyrBI Promoter-Regulatory Region.
`The nucleotide sequence of the entire 758-bp Pvu II fragment
`was determined as summarized in Fig. 1 and is presented in Fig.
`2. The sequence encoding the NH2 terminus of the ATCase
`catalytic subunit is included in nucleotides 619-758 and is pre-
`ceded by a typical ribosome binding site (23). Computer-as-
`sisted inspection (24) ofthe entire nucleotide sequence revealed
`a number offeatures relevant to regulation ofpyrBl expression.
`Perhaps the most striking was a putative p-independent tran-
`scription termination sequence (569-595) located only 23 bp
`from the start of the ATCase catalytic subunit sequence. This
`termination sequence was identified by the G+C-rich region
`of dyad symmetry followed by eight thymidine residues in the
`antisense strand (25). Previous studies indicated that this se-
`quence is sufficient to cause transcriptional termination (26).
`The RNA hairpin encoded by the region ofdyad symmetry, and
`which apparently is essential for transcriptional termination
`(26), has a calculated free energy of formation of -19.1 kcal/
`mol (1 cal = 4.184 J) (27). The close proximity of the transcrip-
`tion termination sequence to the start of the pyrB structural
`gene indicates that it functions as an attenuator (8). This se-
`100
`20
`80
`40
`60
`CTGCGGGATG CACTTTTGAC GATATCATTG ATGTTACGAG CTTCCATACC GATCCAGAAA ACCAATTTGA AGACATCATG ACGGTGAAAA ATGAAATATT
`
`RESULTS
`Subcloning ofthe pyrB Gene and Identification ofthe pyrBI
`Promoter-Regulatory Region. The pyrB gene of E. coli K-12,
`which is the first structural gene in the pyrBI operon (3), was
`subdloned from the A specialized transducing phage Ad valS
`argI pyrB (ykl4m5) by ligating Pvu II or HincII restriction frag-
`ments of transducing phage and plasmid pBR322 (20) DNA.
`Recombinant plasmids were isolated from the ligation mixtures
`by transforming an E. coli K-12 pyrB auxotroph to prototrophy.
`Restriction maps of these plasmids (unpublished data) were
`used to localize the pyrB-complementing sequence within a
`758-base pair (bp) Pvu II fragment and an adjacent 1,(080-bp Pvu
`II/HincII fragment.
`The 758-bp Pvu II fragment was prepared from one of the
`recombinant plasmids, and the sequence of the DNA adjacent
`to the 1,080-bp Pvu II/HincII fragment was determined. The
`first 140 bp of this end of the fragment (Fig. 1) were found to
`encode the NH2-terminal sequence ofthe ATCase catalytic sub-
`unit (21), indicating that the promoter-regulatory region is in-
`cluded in the remaining 618 bp ofthis fragment. This conclusion
`was supported by experiments in which the 758-bp Pvu II frag-
`ment was inserted into the promoter-cloning plasmid pKO-1
`
`120
`140
`160
`180
`200
`TAGCGCCCCA CCTTATCCAA ACTGGACGGC GGTGGGTGTT ACATGGCTGG CAGGCTTTGA TTTTGAAATT AAAGTG-ATAG CGCGCATCCC TGACCAGTAA
`, pi
`220
`240
`280
`300
`GCAATAGTGT TAGCCGTTCG CTTTCACACT CCGCCCTATA AGTCGGATGA ATGGAATAAA ATGCATATCT GATTGCGTGA AAGTGAAAAA GGAAAAACA
`
`32__
`340
`360
`380
`400
`GGGAATGTCT GCAATTATTG ATACCGAAGG ACAGTTCCCC TGCAGAATCA CATCAAATAA AAATGCATAT ACCTTGACTT TTAATTCAAA TAAACCGTTT
`
`420
`440
`P2
`460
`480
`GCGCTGACAA AATATTGCAT CAAATGCFTT CGCCGCTTCT GACGATGAGT ATAATGCCGG ACAATTTGCC GGGAGGATGT ATG GTT CAG TGT GTT CGA
`MET Val Gln Cys Val Arg
`
`520
`500
`540
`560
`580
`CAT TTT GTC TTA CCG CGT CTG AAA AAA GAC GCT GGC CTG CCG TTT TTC TTC CCG TTG ATC ACC CAT TCC CAC CCC CTC AAT CGA GGG
`His Phe Val Leu Pro Arg Leu Lys Lys Asp Ala Gly Leu Pro Phe Phe Phe Pro Leu Ile Thr His Ser Gln Pro Leu Asn Arg Gly
`
`____
`
`600
`620
`640
`660
`GCT TTT TTT TGC CCA GGC GTC AGG AGA TAA MAG ATG GCT AAT CCG CTA TAT CAG AAA CAT ATC ATT TCC ATA MAC GAC CTT'AGT CGC
`Ala Phe Phe Cys Pro Gly Val Arg Arg
`MET Ala Asn Pro Leu Tyr Gln Lys His Ile Ile Ser Ile Asn Asp Leu Ser Arg
`
`680
`700
`720
`740
`758
`GAT GAC CTT AAT CTG GTG CTG GCG ACA GCG GCG AAA CTG AAA GCA AAC CCG CAA CCA GAG CTG TTG MAG CAC AAA GTC ATT GCC AG
`Asp Asp Leu Asn Leu Val Leu Ala Thr Ala Ala Lys Leu Lys Ala Asn Pro Gln Pro Glu Leu Leu Lys His Lys Val Ile Ala Ser
`
`FIG. 2.
`Nucleotide sequence and encoded polypeptides of the 758-bp Pvu II fragment containing the pyrBI promoter-regulatory region. Only
`the sequence of the antisense strand is shown; numbering is from the 5' end. Dyad symmetries are indicated by the arrows with the center of sym-
`metry shown by the dots. Pribnow box and -35 region sequences of promoters P1 and P2 are indicated by brackets. Further details are in the text.
`
`Merck Ex. 1053, pg 1333
`
`

`
`370
`
`Biochemistry: Turnbough et aL
`
`quence is nearly identical to that of the tryptophan attenuator
`of E. coli K-12 (8).
`Two potential pyrBI promoters showing homology with the
`consensus sequence for prokaryotic promoters (25) were located
`(designated P1 and P2 in Fig. 2). Transcription appears to be
`initiated at both of these promoters in vitro. In promoter P1,
`the Pribnow box (257-263) is similar to the consensus sequence,
`but the -35 region is not. In promoter P2, the Pribnow box
`(450-456) is identical to the consensus sequence and the highly
`conserved TTG sequence is present in the -35 region (428-
`430). The transcripts initiated at either promoter apparently
`could encode small polypeptides in addition to ATCase. Within
`the leader region of the shorter P2 transcript is an open reading
`frame (481-612), preceded by an apparent ribosome binding
`site, that could encode a 44-amino acid polypeptide (Fig. 2). The
`termination codon for this polypeptide is only three nucleotides
`before the start of the ATCase sequence. The leader region of
`the P1 transcript could encode two polypeptides-a 45-amino
`acid polypeptide and the 44-amino acid polypeptide included
`in the P2 transcript. The 45-amino acid polypeptide (not shown
`in Fig. 2) would be encoded by nucleotides 305-439. The ri-
`bosome binding site for the synthesis of this polypeptide ap-
`parently would be included in a stable RNA hairpin (calculated
`free energy, -18.2 kcal/mol), which may inhibit translation
`(23).
`One other region of dyad symmetry (511-533) capable of
`encoding a stable RNA hairpin (calculated free energy, -12.2
`kcal/mol) was identified. This region and the adjacent thymi-
`dine-rich sequence included in nucleotides 541-548 appear to
`comprise a strong transcription pause site when the rate oftran-
`scription is limited by low levels of UTP (see below).
`In Vitro Transcription of DNA Fragments from the pyrBI
`Promoter-Regulatory Region. To localize sites of transcrip-
`tional initiation and termination, restriction fragments from the
`pyrBI promoter-regulatory region were transcribed in vitro and
`the resulting transcripts were analyzed by polyacrylamide gel
`electrophoresis. Transcription of the 758-bp Pvu II fragment
`produced two major transcripts approximately 328 and 135 nu-
`cleotides long (Fig. 3, lane 1). Transcripts of the same size also
`were synthesized from the 474-bp Dde I fragment (lane 3),
`which is an internal segment of the Pvu II fragment (Fig. 3
`Lower). These results suggested that the two transcripts were
`initiated at the previously identified promoters P1 and P2 with
`termination of both transcripts occurring at the putative atten-
`uator preceding the pyrB structural gene. To examine this pos-
`sibility, the 758-bp Pvu II fragment was transcribed in a reaction
`in which GTP was replaced by the analog ITP. Substituting ITP
`for GTP has been shown to substantially decrease transcription
`termination at attenuators (28). If the above assignments are
`correct, this substitution should result in the synthesis of run-
`off transcripts approximately 491 and 298 nucleotides long.
`Transcripts of the predicted size were detected, although the
`amount of the 298-nucleotide transcript was less than expected
`(Fig. 3, lane 2). No 135-nucleotide transcript was detected, and
`only a small amount of a 328-nucleotide transcript was synthe-
`sized. There appeared to be only slight differences in the elec-
`trophoretic mobilities of transcripts containing either ITP or
`GTP. To confirm the location ofthe apparent pyrBI promoters,
`the 304-bp Dde I/Taq I fragment (Fig. 3 Lower) was tran-
`scribed. Transcription initiated at promoters P1 and P2 should
`produce run-off transcripts approximately 230 and 37 nucleo-
`tides in length, respectively, and transcripts of this size were
`detected (Fig. 3, lane 4). From the data presented, it is pre-
`sumed that transcription is initiated at promoters P1 and P2.
`The relative frequency of transcription from promoters P1
`and P2 was estimated by measuring the radioactivity in the 328-
`
`Proc. Natl. Acad. Sci. USA 80 (1983)
`
`1
`
`2
`
`3
`
`4
`
`_
`
`_ 230
`
`491+
`328
`298
`
`135 ,_
`
`V
`
`37
`
`Pvu I
`
`Dde I
`
`go ili
`
`I
`
`P,
`
`268
`
`P2
`
`anNH2-terminus
`
`461
`
`595 619
`
`758
`
`194
`
`194
`
`I
`
`667
`
`I
`
`I
`497
`
`FIG. 3. In vitro transcription of the pyrBIpromoter-regulatory re-
`gion. (Upper) Autoradiogram of a polyacrylamide gel used to separate
`transcripts from: lane 1, 758-bp Pvu II fragment; lane 2, 758-bp Pvu
`II fragment with ITP substituted for GTP; lane 3, 474-bp Dde I frag-
`ment; and lane 4, 304-bp Dde I/Taq I fragment. Restriction fragment
`sizes represent the lengths of the sense strand. The lower part of the
`autoradiogram was enhanced using an intensifying screen to show the
`37-nucleotide transcript, which has a relatively low specific activity.
`The major transcript above the 328-nucleotide transcript in lanes 1 and
`3 and those above the 230-nucleotide transcript in lane 4 are approx-
`imately the same length as or larger than the respective templates and
`are presumed to result from nonspecific transcription. (Lower) DNA
`templates with probable transcriptional initiation sites for promoters
`P1 and P2, probable termination site at the attenuator, and NH2 ter-
`minus of the ATCase catalytic subunit indicated. Transcript lengths
`were estimated from the gel by using single-stranded DNA standards.
`
`and 135-nucleotide transcripts synthesized from the Pvu II frag-
`ment (Fig. 3, lane 1). After correcting for the differences in the
`specific activities of the transcripts, it was determined that ap-
`proximately the same number of transcripts were initiated at
`each promoter.
`The efficiency of transcriptional termination at the putative
`pyrBI attenuator was determined by comparing the levels of
`attenuated and run-off transcripts synthesized from the Pvu II
`fragment (Fig. 3, lane 1). Approximately 98% of the transcripts
`initiated at promoters PI and P2 apparently were terminated at
`the attenuator.
`To examine UTP-specific effects on transcription ofthe pyrBI
`promoter-regulatory region, the 758-bp Pvu II fragment was
`transcribed in a series of reactions in which one ribonucleoside
`triphosphate at a time was varied from the near optimal con-
`centration of 200 tLM (29) to 5 1LM. The three other ribonu-
`
`Merck Ex. 1053, pg 1334
`
`

`
`Biochemistry: Turnbough et aL
`cleoside triphosphates were maintained at 200 AuM. Reaction
`mixtures were incubated for 5 min after the addition ofrifampi-
`cin and the transcripts were analyzed as above (data not shown).
`Reducing the concentration ofUTP or any other ribonucleoside
`triphosphate did not increase the levels of the run-off tran-
`scripts. The most striking effect observed was that the reduction
`in the level of the 328-nucleotide transcript was much greater
`with a decrease in UTP level than with a decrease in the con-
`centration ofany other ribonucleoside triphosphate. Decreases
`in the level of this transcript caused by lowering the UTP con-
`centration were accompanied by increases in the level of an
`approximately 280-nucleotide transcript, suggesting transcrip-
`tional pausing or termination before the putative attenuator.
`UTP-sensitive transcriptional pausing was examined by tran-
`scribing the Pvu II fragment in four reactions in which the con-
`centration of one ribonucleoside triphosphate was reduced to
`20 jxM, a suboptimal concentration (30), while the others were
`maintained at 200 ,uM. Reactions were initiated by the simul-
`taneous addition ofthe ribonucleoside triphosphates and rifam-
`picin (10 ,uM), which permits only a single round of transcrip-
`tion. Aliquots were removed at different times and transcripts
`were analyzed (Fig. 4). Only at suboptimal UTP, substantial
`levels of transcripts approximately 81 and 280 nucleotides long
`were detected at early time points. The pattern of disappear-
`ance of these two transcripts with the accumulation of the 135-
`
`UTP
`
`-
`
`I
`
`CTP
`
`-
`
`328
`280
`
`P
`P-
`
`do
`
`.*I*.
`V&
`
`0
`
`_.4 aj
`
`169
`160 *'_
`135 -4._
`
`81
`
`0.5
`
`1
`
`1.5 2 5
`0.5 1
`MINUTES
`
`1.5 2
`
`5
`
`In vitro transcription of the 758-bpPvu II fragment at sub-
`FIG. 4.
`optimal ribonucleoside triphosphate concentrations. The figure shows
`an autoradiogram of a polyacrylamide gel used to separate transcripts
`synthesized during a single round of transcription at suboptimal con-
`centrations (20 pM) of UTP or CTP. Transcription at suboptimal ATP
`or GTP was similar to that at suboptimal CTP. Aliquots for analysis
`were removed from the reaction mixtures at the indicated times. The
`RNA polymerase concentration was 84 pmol/ml in this experiment
`and other changes in the transcription protocol described in Materials
`and Methods are indicated in the text. Transcript lengths were esti-
`mated as in Fig. 3.
`
`Proc. Natd Acad. Sci. USA 80 (1983)
`
`371
`
`and 328-nucleotide transcripts indicated that transcription in-
`itiated at promoters P1 and P2 paused before the putative at-
`tenuator at sites within the thymidine-rich sequence included
`in nucleotides 541-548 (Fig. 2). Additional transcripts approx-
`imately 160 and 169 nucleotides long also were detected at early
`time points and indicated pausing near promoter P2 of tran-
`scription initiated at promoter P1. This pausing was not ribonu-
`cleoside triphosphate-specific. Two other effects on transcrip-
`tion observed only at suboptimal UTP were a severalfold in-
`crease in the level of the 135-nucleotide transcript and a lag in
`the appearance oftranscripts initiated at P1. The cause and reg-
`ulatory significance of these effects are not known.
`DISCUSSION
`The presence of a strong p-independent transcriptional termi-
`nator immediately preceding the pyrBl structural genes indi-
`cates that the expression of this operon is regulated by an at-
`tenuation control mechanism (8). The apparent involvement of
`transcription and translation in the regulation of pyrBI expres-
`sion suggests that this mechanism may be similar to the atten-
`uwtion control mechanisms described for many amino acid bio-
`synthetic operons in enteric bacteria (8). Unlike these operons,
`however, the pyrBI operon does not contain sequences pre-
`ceding the G+C-rich region ofdyad symmetry in the attenuator
`that permit the formation ofalternative stem-loop structures in
`the leader transcript as a means of regulating transcriptional
`termination. A mechanism that could function in regulating
`transcriptional termination at the pyrBI attenuator is suggested
`from studies on attenuation control of histidine operon expres-
`sion in S. typhimurium (7). These studies indicate that ifa trans-
`lating ribosome is allowed to follow closely behind RNA poly-
`merase during the transcription of the attenuator region, then
`it can disrupt or prevent the formation of the attenuator-en-
`coded RNA hairpin necessary for termination and thereby per-
`mit transcription to continue into the structural genes. Because
`the pyrBI attenuator is included in a leader polypeptide-en-
`coding sequence (Fig. 2), regulating the relative rates of tran-
`scription and translation of this sequence could control tran-
`scriptional termination and the expression of the structural
`genes.
`Based on the results presented in this study and other data,
`we propose the following model for regulation of the pyrBI op-
`eron. Transcription is initiated at promoter P1 or P2. When the
`Jatracellular level ofUTP is low, RNA polymerase slows or stops
`temporarily within the run of thymidine residues in the UTP-
`sensitive pause site located approximately 20 bp before the at-
`tenuator. This pause provides enough time for a ribosome to
`initiate translation ofthe 44-amino acid leader polypeptide and
`translate up to the stalled RNA polymerase. When the poly-
`merase eventually passes this pause site and transcribes the
`attenuator, the formation of the attenuator-encoded RNA hair-
`pin is disrupted or precluded by the adjacent translating ribo-
`some. In the absence of the termination hairpin, RNA poly-
`merase continues transcribing into the pyrBI structural genes.
`Translation of the leader polypeptide is terminated within the
`ribosome binding site preceding the ATCase catalytic subunit
`cistron, which should ensure the immediate initiation of trans-
`lation of this protein. When the intracellular level of UTP is
`high, RNA polymerase does not pause during the transcription
`of the leader region. This does not provide enough time for a
`ribosome to bind to the transcript and catch up to RNA poly-
`merase before the formation of the attenuator-encoded hairpin
`and subsequent termination of transcription before the struc-
`tural genes.
`In this attenuation control mechanism there is no require-
`ment for a purely regulatory protein. Several attempts to isolate
`
`Merck Ex. 1053, pg 1335
`
`

`
`372
`
`Biochemistry: Turnbough et aL
`
`Proc. Natl. Acad. Sci. USA 80 (1983)
`
`mutants with altered pyrimidine regulatory protein activity
`have been unsuccessful (31-33). Recently, a mutant of S. ty-
`phimurium containing increased levels of ATCase and several
`other pyrimidine biosynthetic enzymes was isolated (34). The
`mutation responsible for these effects was mapped within the
`gene cluster encoding the ( and /3' subunits of RNA polymer-
`ase, indicating a regulatory role for this enzyme in the control
`ofpyrimidine gene expression. This observation is entirely con-
`sistent with the proposed model in which an alteration in RNA
`polymerase reducing the rate of transcriptional elongation or
`the efficiency of attenuation should result in increased expres-
`sion of the pyrBI operon. The observation that other pyrimidine
`genes are affected by this mutation suggests that they may be
`regulated by similar attenuation mechanisms.
`An important feature of the proposed model is the UTP-sen-
`sitive transcription pause site preceding the attenuator. Be-
`cause regions of dyad symmetry are frequently associated with
`pausing (35), it is likely that this site includes the sequence of
`dyad symmetry preceding the thymidine-rich region within
`which pausing occurs (Fig. 2). We presume that these se-
`quences constitute a uniquely strong pause site when the rate
`oftranscription through the thymidine-rich region is limited by
`the availability of UTP. It has been shown that substantial de-
`repression of pyrBI expression occurs only when the concen-
`tration of UTP is below an apparent threshold of approximately
`0.2 mM (5, 36), which is the UTP concentration below which
`the rate of transcriptional elongation is reduced in vitro (29, 30).
`Although UTP appears to be the principal nucleoside triphos-
`phate effector of pyrBi expression, it has been suggested that
`GTP also is involved in regulation, based on limited stimulation
`of expression that occurs upon guanine starvation (33). It is pos-
`sible that low levels of GTP cause limited transcriptional paus-
`ing within the leader region. The physiological significance of
`this regulation is unclear.
`The proposed model can accommodate transcriptional ini-
`tiation at either promoter P1 or P2, but there is no requirement
`that both function. The only significant difference between the
`leader transcripts initiated at the two promoters appears to be
`that the P1 transcript encodes a second leader polypeptide as
`described in Results. We have not assigned a regulatory function
`for this polypeptide, and it is possible that its translation would
`be precluded by secondary structure in the transcript. If, in fact,
`there is only one pyrBI promoter in vivo, the better candidate
`is promoter P2. It is closer to the structural genes and shows
`greater homology with the consensus sequence for prokaryotic
`promoters. In addition, in vitro transcription initiated at this
`promoter is selectively inhibited by guanosine tetraphosphate
`(unpublished data), which has been shown to be a negative ef-
`fector ofATCase synthesis in vivo and in vitro in a coupled tran-
`scription-translation system (5). Promoters negatively affected
`by guanosine tetraphosphate typically contain a G+C-rich se-
`quence between the Pribnow box and the site oftranscriptional
`initiation (37), and such a sequence is present only in promoter
`P2. In formulating the model, we have assumed that regulation
`involving guanosine tetraphosphate occurs at the level of tran-
`scriptional initiation and is independent of the attenuation con-
`trol mechanism.
`Although the proposed attenuation control mechanism is
`consistent with the known features of pyrBI expression, addi-
`tional experiments are clearly required to confirm the model.
`We also think that similar mechanisms of regulation may func-
`tion in the expression of other pyrimidine genes and similar
`studies should provide useful information.
`
`2.
`3.
`
`4.
`
`5.
`6.
`
`7.
`
`8.
`9.
`10.
`
`11.
`
`12.
`
`We thank John Roth for running the computer program used to ana-
`lyze our DNA sequence data and for valuable discussions. We acknowl-
`edge the efforts of David Bole, who contributed to early DNA sequence
`experiments. This work was supported by National Institutes of Health
`Grant GM29466.
`O'Donovan, G. A. & Neuhard, J. (1970) Bacteriol. Rev. 34, 278-
`1.
`343.
`Bachmann, B. J. & Low, K. B. (1980) Microbiol Rev. 44, 1-56.
`Pauza, C. D., Karels, M. J., Navre, M. & Schachman, H. K.
`(1982) Proc. Natl Acad. Sci. USA 79, 4020-4024.
`Pierard, A., Glansdorff, N., Gigot, D., Crabeel, M., Halleux, P.
`& Thiry, L. (1976)J. Bacteriol 127, 291-301.
`Turnbough, C. L., Jr. (1983) J. BacterioL, in press.
`Neill, R. J. (1979) Dissertation (Univ. of California, Berkeley,
`CA).
`Johnston, H. M., Barnes, W. M., Chumley, F. G., Bossi, L. &
`Roth, J. R. (1980) Proc. Natl. Acad. Sci. USA 77, 508-512.
`Yanofsky, C. (1981) Nature (London) 289, 751-758.
`Kikuchi, A. & Gorini, L. (1975) Nature (London) 256, 621-624.
`Zubay, G., Chambers, D. A. & Cheong, L. C. (1970) in The Lac-
`tose Operon, eds. Zipser, D. & Beckwith, J. R. (Cold Spring Har-
`bor Laboratory, Cold Spring Harbor, NY), pp. 375-391.
`Birnboim, H. C. & Doly, J. (1979) Nucleic Acids Res. 7, 1513-
`1523.
`Davis, R. W., Botstein, D. & Roth, J. R. (1980) Advanced Bac-
`terial Genetics (Cold Spring Harbor Laboratory, Cold Spring
`Harbor, NY).
`Jagusztyn-Krynicka, E. K., Smorawinska, M. & Curtiss, R., III
`(1982)J. Gen. Microbiol 128, 1135-1145.
`McDonell, M. W., Simon, M. N. & Studier, F. W. (1977)J. Mol
`Biol 110, 119-146.
`Morrison, D. A. (1979) Methods Enzymol 68, 326-331.
`Maxam, A. M. & Gilbert, W. (1980) Methods Enzymol 65, 499-
`560.
`Burgess, R. R. & Jendrisak, J. J. (1975) Biochemistry 14, 4634-
`4638.
`Gonzalez, N., Wiggs, J. & Chamberlin, M. J. (1977) Arch.
`Biochem. Biophys. 182, 404-408.
`Maniatis, T., Jeffrey, A. & van deSande, H. (1975) Biochemistry
`14, 3787-3794.
`Bolivar, F., Rodriguez, R. L., Greene, P. J., Betlach, M. C.,
`Heyneker, H. L. & Boyer, H. W. (1977) Gene 2, 95-113.
`Gigot, D., Glansdorff, N., Legrain, C., Pierard, A., Stalon, V.,
`Konigsberg, W., Caplier, I., Strosberg, A. D. & Herve, G.
`(1977) FEBS Lett. 81, 28-32.
`McKenney, K., Shimatake, H., Court, D., Schmeissner, U.,
`Brady, C. & Rosenberg, M. (1981) in Gene Amplification and
`Analysis, eds. Chirikjian, J. G. & Papas, T. S. (Elsevier/North-
`Holland, New York), Vol. 2, pp. 383-415.
`Gold, L., Pribnow, D., Schneider, T., Shinedling, S., Singer, B.
`S. & Stormo, G. (1981) Annu. Rev. Microbiol 35, 365-403.
`Brutlag, D. L., Clayton, J., Friedland, P. & Kedes, L. H. (1982)
`Nucleic Acids Res. 10, 279-294.
`Rosenberg, M. & Court, D. (1979) Annu. Rev. Genet. 13, 319-
`353.
`Christie, G. E., Farnham, P. J. & Platt, T. (1981) Proc. Nati
`Acad. Sci. USA 78, 4180-4184.
`Salser, W. (1977) Cold Spring Harbor Symp. Quant. Biol. 42,
`985-1002.
`Lee, F. & Yanofsky, C. (1977) Proc. Natl Acad. Sci. USA 74,
`4365-4369.
`Chamberlin, M. J., Nierman, W. C., Wiggs, J. & Neff, N. (1979)
`J. Biol Chem. 254, 10061-10069.
`Kassavetis, G. A. & Chamberlin, M. J. (1981)J. Biol Chem. 256,
`2777-2786.
`O'Donovan, G. A. & Gerhart, J. C. (1972) J. Bacteriol 109,
`1085-1096.
`Justesen, J. & Neuhard, J. (1975) J. Bacteriol 123, 851-854.
`Jensen, K. F. (1979)J. Bacteriol 138, 731-738.
`Jensen, K. F., Neuhard, J. & Schack, L. (1982) EMBO J. 1, 69-
`74.
`Farnham, P. J. & Platt, T. (1981) Nucleic Acids Res. 9, 563-577.
`Schwartz, M. & Neuhard, J. (1975)J. BacterioL 121, 814-822.
`Travers, A. A. (1980)J. BacterioL 141, 973-976.
`
`13.
`
`14.
`
`15.
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`23.
`
`24.
`
`25.
`
`26.
`
`27.
`
`28.
`
`29.
`
`30.
`
`31.
`
`32.
`33.
`34.
`
`35.
`36.
`37.
`
`Merck Ex. 1053, pg 1336