Autogenous regulation of RNA polymerase beta subunit synthesis in vitro

Constitutive Expression of a Cytotoxic Anticancer Protein in Tumor-Colonizing Bacteria

Bacterial cancer therapy is a promising next-generation modality to treat cancer that often uses tumor-colonizing bacteria to deliver cytotoxic anticancer proteins. However, the expression of cytotoxic anticancer proteins in bacteria that accumulate in the nontumoral reticuloendothelial system (RES), mainly the liver and spleen, is considered detrimental. This study examined the fate of the Escherichia coli strain MG1655 and an attenuated strain of Salmonella enterica serovar Gallinarum (S. Gallinarum) with defective ppGpp synthesis after intravenous injection into tumor-bearing mice (~10⁸ colony forming units/animal). Approximately 10% of the injected bacteria were detected initially in the RES, whereas approximately 0.01% were in tumor tissues. The bacteria in the tumor tissue proliferated vigorously to up to 10⁹ colony forming units/g tissue, whereas those in the RES died off. RNA analysis revealed that tumor-associated E. coli activated rrnB operon genes encoding the rRNA building block of ribosome needed most during the exponential stage of growth, whereas those in the RES expressed substantially decreased levels of this gene and were cleared soon presumably by innate immune systems. Based on this finding, we engineered ΔppGpp S. Gallinarum to express constitutively a recombinant immunotoxin comprising TGFα and the Pseudomonas exotoxin A (PE38) using a constitutive exponential phase promoter, the ribosomal RNA promoter rrnB P1. The construct exerted anticancer effects on mice grafted with mouse colon (CT26) or breast (4T1) tumor cells without any notable adverse effects, suggesting that constitutive expression of cytotoxic anticancer protein from rrnB P1 occurred only in tumor tissue.

Rifampicin can induce antibiotic tolerance in mycobacteria via paradoxical changes in rpoB transcription

Metrics commonly used to describe antibiotic efficacy rely on measurements performed on bacterial populations. However, certain cells in a bacterial population can continue to grow and divide, even at antibiotic concentrations that kill the majority of cells, in a phenomenon known as antibiotic tolerance. Here, we describe a form of semi-heritable tolerance to the key anti-mycobacterial agent rifampicin, which is known to inhibit transcription by targeting the β subunit of the RNA polymerase (RpoB). We show that rifampicin exposure results in rpoB upregulation in a sub-population of cells, followed by growth. More specifically, rifampicin preferentially inhibits one of the two rpoB promoters (promoter I), allowing increased rpoB expression from a second promoter (promoter II), and thus triggering growth. Disruption of promoter architecture leads to differences in rifampicin susceptibility of the population, confirming the contribution of rifampicin-induced rpoB expression to tolerance.

The antibiotic rifampicin inhibits transcription by targeting RpoB, a bacterial RNA polymerase subunit. Here, Zhu et al. show that certain cells in mycobacterial populations can continue to grow and divide in the presence of rifampicin due, paradoxically, to rifampicin-induced upregulation of the rpoB gene.

Regulation of Global Transcription in Escherichia coli by Rsd and 6S RNA

In Escherichia coli, the sigma factor σ⁷⁰ directs RNA polymerase to transcribe growth-related genes, while σ³⁸ directs transcription of stress response genes during stationary phase. Two molecules hypothesized to regulate RNA polymerase are the protein Rsd, which binds to σ⁷⁰, and the non-coding 6S RNA which binds to the RNA polymerase-σ⁷⁰ holoenzyme. Despite multiple studies, the functions of Rsd and 6S RNA remain controversial. Here we use RNA-Seq in five phases of growth to elucidate their function on a genome-wide scale. We show that Rsd and 6S RNA facilitate σ³⁸ activity throughout bacterial growth, while 6S RNA also regulates widely different genes depending upon growth phase. We discover novel interactions between 6S RNA and Rsd and show widespread expression changes in a strain lacking both regulators. Finally, we present a mathematical model of transcription which highlights the crosstalk between Rsd and 6S RNA as a crucial factor in controlling sigma factor competition and global gene expression.

Mapping of sequences required for the translation of the beta subunit of Escherichia coli RNA polymerase

Previous experiments using expression plasmids which overproduce the beta and beta' subunits of Escherichia coli RNA polymerase suggested that regions considerably upstream of the start of the rpoB gene, which encodes the beta subunit, are required for its efficient synthesis. To further delineate the required regions, a collection of genetic constructs that contained varying amounts of the region either upstream or downstream of the translational start of rpoB was assembled. Measurements of beta and beta' synthesis and rpoB mRNA production from a series of rpoBC expression plasmids indicated that sequences extending more than 43 bp but less than 79 bp upstream of rpoB are required for the efficient translation of rpoB mRNA. This result was confirmed by beta-galactosidase measurements from a series of rpoB-lacZ fusions that have the same set of end points upstream of rpoB as the expression plasmids. A second set of gene fusions containing differing amounts of the sequence distal to the start of rpoB fused in frame to lacZ revealed that more than 29 bp but less than 70 bp of rpoB was required for efficient translation.

An internal region of rpoB is required for autogenous translational regulation of the beta subunit of Escherichia coli RNA polymerase

In order to delineate the region involved in feedback regulation of the RNA polymerase beta subunit (encoded by rpoB), a collection of rpoB-lacZ translational fusions with different endpoints both upstream and downstream of the rpoB start site was assembled on lambda phage vectors. The extent of translational repression of beta was monitored by measuring beta-galactosidase levels in monolysogens of the fusions under conditions of increased intracellular concentrations of beta and beta' achieved via the induction of rpoBC expression from a multicopy plasmid. A construct containing as little as 29 bp upstream of the start of rpoB exhibited repression of beta-galactosidase activity to the same extent as a construct encoding the full upstream region. A construct which carried only 70 bp of the rpoB structural gene exhibited very little repression, while constructs which carried 126 or 221 bp of rpoB exhibited approximately the same degree of repression as a construct which carried 403 bp. These data suggest that the sequences important for feedback regulation of beta translation extend more than 70 bp into rpoB but are completely contained within a region which spans the sequences from 29 bp upstream to 126 bp downstream of the translational start site.

Functional specialization within the alpha-subunit of Escherichia coli RNA polymerase

The RNA polymerase core enzyme of Escherichia coli has the subunit composition alpha 2 beta beta', and when combined with one of several alternative sigma-subunits (initiation-specificity) produces holoenzyme capable of all the steps of transcription. Dimerization of the alpha-subunit and association with the beta-subunit trigger assembly of the core enzyme. Analyses of a set of deletion derivatives of rpoA (which encodes alpha) have indicated that as many as 94 carboxy-terminal amino acids (but not 153) can be removed without preventing assembly of core-like complexes in vitro. Detailed analyses of these deletion mutants have now been performed in vivo. alpha-Polypeptides truncated from the carboxy terminus to amino acid residues 235, 256 or 296 are assembled not merely into core, but also into holoenzyme-like complexes in vivo, and at least in the first two cases both of the two alpha-subunits can be replaced by the truncated versions. Nevertheless, none can complement rpoAts alleles for growth at 42 degrees C. We conclude that the domain(s) of alpha essential for the assembly of RNA polymerase (at least the major holoenzyme species) are confined to the amino-terminal 235 amino acids, while some other essential function(s) require residues close to the carboxy terminus.

In vivo analysis of overlapping transcription units in the rplKAJLrpoBC ribosomal protein-RNA polymerase gene cluster of Escherichia coli

Transcription of the rplKAJLrpoBC ribosomal protein (rpl) RNA polymerase (rpo) gene cluster is governed by a complex set of signals. To dissect the transcription units active in vivo and to quantify the relative contribution of each, an extensive array of rplKAJLrpoB/lacZ gene fusions were constructed on lambda phage derivatives and introduced in single copy into the chromosomes of lac- cells. Measurements of beta-galactosidase production from fusions containing wild-type and/or mutagenized rplrpo DNA fragments permitted the establishment of high-resolution transcription profiles of the gene cluster. The results show that transcription initiated at the upstream rplKp promoter (located just before rplK) does not terminate before the rplJp promoter (located upstream from rplJ), but instead reads through into the distal genes. In addition, rplJp continues to function efficiently in the presence of readthrough transcription from rplKp. As a result the rplJL genes are transcribed at almost twice the frequency of the upstream rplKA genes. However, the transcription of rpoB, which is situated downstream from the previously identified attenuator (rpoBa), is only marginally increased (20%) when both promoters are present. This suggests that although both transcription units overlap, transcriptional termination at rpoBa is modulated in response to the frequency of initiation from both promoters.

Autogenous regulation of the RNA polymerase beta subunit of Escherichia coli occurs at the translational level in vivo

A series of transcriptional and translational fusions of the gene for the beta subunit of RNA polymerase (rpoB) to the lacZ reporter gene have been constructed on lambda vectors. Both transcriptional and translational fusions carry the upstream rplKAJL ribosomal protein gene region, which contains the two strong promoters rplKp and rplJp responsible for the transcription of rpoBC. Monolysogens carrying either the transcriptional translational fusion were assayed for beta-galactosidase, providing a measure of the transcription or of both transcription and translation of rpoB, respectively. Translational fusion monolysogens which also carried a multicopy plasmid containing the beta and beta' genes (rpoBC) under the control of a regulatable promoter, exhibited a substantial decrease in the beta-galactosidase levels upon overproduction of beta and beta'. No significant effect was seen in comparable experiments with the transcriptional fusions. These results argue that in vivo, the synthesis of the RNA polymerase beta subunit is autogenously regulated by a translational mechanism. Furthermore, experiments with the overexpressing plasmids confirm the requirement for a portion of the rplL-rpoB intercistronic region in the vicinity of the RNaseIII processing site for the efficient translation of the beta subunit mRNA.

Overexpression and purification of a biologically active rifampicin-resistant beta subunit of Escherichia coli RNA polymerase

The gene rpoB (rifD 18), which encodes rifampicin-resistant beta subunit of Escherichia coli RNA polymerase, has been placed on an overexpression plasmid under the control of bacteriophage T7 promoter. Induction of the T7 RNA polymerase gene in the host cells resulted in extensive overproduction of the beta polypeptide. Most of the overproduced material was recovered from cell lysates in insoluble form and was solubilized by extraction with 6 M urea. Purified overproduced beta subunit was added, in molar excess, to urea-denatured rifampicin-sensitive RNA polymerase. Upon removal of urea by dialysis, the reconstituted enzyme became rifampicin-resistant, indicating that overproduced beta subunit can be efficiently assembled into functional holoenzyme.

Feedback regulation of RNA polymerase subunit synthesis after the conditional overproduction of RNA polymerase in Escherichia coli

The beta and beta' subunits of RNA polymerase are thought to be controlled by a translational feedback mechanism regulated by the concentration of RNA polymerase holoenzyme. To study this regulation in vivo, an inducible RNA polymerase overproduction system was developed. This system utilizes plasmids from two incompatibility groups that carry RNA polymerase subunit genes under lac promoter/operator control. When the structural genes encoding the components of core RNA polymerase (alpha, beta and beta') or holoenzyme (alpha, beta, beta' and sigma 70) are present on the plasmids, induction of the lac promoter results in a two fold increase in the concentration of functional RNA polymerase. The induction of RNA polymerase overproduction is characterized by an initial large burst of beta beta' synthesis followed by a gradual decrease as the concentration of RNA polymerase increases. Overproduction of RNA polymerase in a strain carrying an electrophoretic mobility mutation in the rpoB gene results in the specific repression of beta beta' synthesis off the chromosome. These results indicate that RNA polymerase feedback regulation controls beta beta' synthesis in vivo.

Prokaryotic gene expression in vitro: transcription-translation coupled systems

Transcription-translation coupled systems have been developed to study prokaryotic gene expression. Several types of expression system have been described. The original system consists of a crude unfractionated Escherichia coli extract, which supports protein synthesis directed by a template DNA. Control of gene expression at the transcriptional stage has been studied using this unfractionated system. In this respect, two examples of particular interest, lactose and tryptophan operons, are described. Other systems are either partially reconstituted or highly defined, containing up to 30 purified factors necessary for transcription (RNA polymerase) and translation (aminoacyl-tRNA synthetases, initiation, elongation and release factors). Additional differences between the various systems relate to the analysis of the gene products. Whereas most methods involve analysis of the totally synthesized protein, a particular system implies the formation of only the N-terminal di- or tripeptide of the gene product. Reconstituted systems have proved useful in studies on transcriptional, e.g., discovery and role of L factor, as well as translational regulation of gene expression, e.g., autogenous control of ribosomal protein synthesis.

Direct evidence for autogenous regulation of the Escherichia coli genes rpoBC in vivo

We have fused the rpoBC genes to the strong controllable promoter PL in phage lambda while deleting most of the intercistronic regulatory DNA and ribosomal protein genes upstream of rpoB. Induction of a lysogen carrying the recombinant prophage gave rise to a 2-3-fold oversynthesis of beta beta' in the cell whereas rpoBC-mRNA levels rose by at least 10-fold. Similar observations were made when these sequences were present in the prophage, indicating that the removal of DNA sequences up to 26 base pairs before rpoB does not affect post-transcriptional autogenous regulation of beta beta' synthesis. Overexpression of beta beta' also autogenously regulated the synthesis of the beta polypeptide from the chromosome in two strains carrying electrophoretic mobility mutations in rpoB. S1 nuclease mapping experiments indicated that this regulation was also post-transcriptional, and confirmed that phage beta-mRNA synthesis exceeded chromosomal beta-mRNA synthesis by 20-fold. The provision of excess beta alone in the cell caused autoregulation of chromosomal beta, but not beta' synthesis, indicating that beta and beta' are regulated independently.

Mutant of Escherichia coli with unusual patterns of rpoB,C expression in response to rifampicin and acridine orange

A mutation located near rpoB (89') in E. coli is responsible for unusual patterns of beta and beta ' (but not L7/L12) synthesis in response to the drugs rifampicin and acridine orange.

An in vitro system to measure gene expression based on dipeptide synthesis

A simplified E. coli in vitro system has been developed to study gene expression based on the synthesis of the first di- or tripeptide of the gene product. Plasmids containing bacterial and chloroplast genes have been used as templates in this system. The expression of the E. coli L10 operon, which is under both transcriptional and translational control, has been investigated in some detail using the dipeptide system. A similar system has been developed, using eukaryotic translation components, to measure the expression of eukaryotic mRNA based on dipeptide formation.

Assembly-controlled autogenous modulation of bacteriophage P22 scaffolding protein gene expression

In the assembly of bacteriophage P22, precursor particles containing two major proteins, the gene 5 coat protein and the gene 8 scaffolding protein, package the DNA molecule. During the encapsidation reaction all of the scaffolding protein molecules are released intact and subsequently participate in further rounds of DNA encapsidation. We have previously shown that even though it lies in the center of the late region of the genetic map, the scaffolding protein gene is not always expressed coordinately with the remainder of the late proteins and that some feature of the phage assembly process affects its expression. We present here in vivo experiments which show that there is an inverse correlation between the amount of unassembled scaffolding protein and the rate of scaffolding protein synthesis and that long amber fragments of the scaffolding protein can turn down the synthesis of intact scaffolding protein in trans. These results support a model for scaffolding protein regulation in which the feature of the assembly process which modulates the rate of scaffolding protein synthesis is the amount of unassembled scaffolding protein itself.

Translational control of the expression of the beta subunit gene of E. coli RNA polymerase

Using the plasmid pNF1337 as template, a mRNA preparation has been obtained that directs the in vitro synthesis of fMet-Val, the N-terminal dipeptide of the beta subunit of RNA polymerase. RNA polymerase holoenzyme specifically inhibits the mRNA-directed synthesis of fMet-Val showing that the autoregulation by RNA polymerase of beta, beta 1 synthesis is at the level of translation. L factor (nusA gene product) stimulates the synthesis of fMet-Val from a DNA template but not from mRNA. Rifampicin has no effect on the mRNA-directed synthesis of fMet-Val or the ability of RNA polymerase to inhibit fMet-Val synthesis.

Genetic analysis of two bacterial RNA polymerase mutants that inhibit the growth of bacteriophage T7

The Escherichia coli mutants 7009 and BR3 are defective in the growth of bacteriophage T7. We have previously shown that both of these mutant hosts produce an altered RNA polymerase which is resistant to inhibition by the T7 gene 2 protein (De Wyngaert and Hinkle 1979). In both strains, the mutation which prevents T7 growth is closely linked to rifA (rpoB). Both mutants are complemented by transformation with a multicopy plasmid carrying rpoB and rpoC but not by a plasmid carrying only rpoB. This indicates that the mutations reside in rpoC, the structural gene for the beta' subunit of RNA polymerase. When a single copy of the wildtype rpoC allele is introduced into the mutant using the transducing phage lambda drifd18, the mutant allele is dominant over wildtype. The lambda drifd18 transductant also remains unable to support the growth of T7 in the presence of rifampin. This supports our conclusion that the mutation is in rpoC. We have measured the growth of T7 phage, the kinetics of phage DNA synthesis, and the structure of replicative DNA intermediates in several transductants, and compared these results with those obtained in the original mutant strains.

An amber mutation in the gene encoding the beta' subunit of Escherichia coli RNA polymerase

An Escherichia coli strain carrying an amber mutation (UAG) in rpoC, the gene encoding the beta prime subunit of RNA polymerase, was isolated after mutagenesis with nitrosoguanidine. The mutation was moved into an unmutagenized strain carrying the supD43,74 allele, which encodes a temperature-sensitive su1 amber suppressor, and sue alleles, which enhance the efficiency of the suppressor. In this background, beta prime is not synthesized at high temperature. Suppression of the mutation by the non-temperature-sensitive amber suppressor su1+ yields a protein which is functional at all temperatures examined (30, 37, and 42 degrees C).

Escherichia coli merodiploids with altered ratios of episomal to chromosomal RNA polymerase activity

Starting with an E. coli merodiploid strain (rpoB+ (rifS)/rpoB- (rifR)) containing equal amounts of rifampicin-sensitive and resistant RNA polymerase activities, mutants were isolated that had a disproportionately high ratio of the rifampicin-resistant enzyme activity. With one strain it is shown that the mutation responsible for this phenotype is closely linked to the rifR (rpoB-) allele.

Negative regulation of beta and beta' synthesis by RNA polymerase

The genes for the beta and beta' subunits of RNA polymerase, rpoB and rpoC, and the genes for the two ribosomal proteins, rplL and rplJ, are part of the beta operon. Although this operon and contains a single strong promoter, the genes of the operon are not always coordinately expressed in vivo. This has now been confirmed in vitro where the lack of coordinate expression has been shown to be correlated with the selective inhibition of rpoB and rpoC gene expression by RNa polymerase. Rifampicin, which stops the initiation of transcription, also relieves this autogenous inhibition of beta and beta' (beta beta') synthesis. The inhibitory action of RNA polymerase and its reversal by rifampicin most likely occurs at a posttranscriptional or translation level.

Transcriptional and post-transcriptional control of ribosomal protein and ribonucleic acid polymerase genes

A partial restriction of ribonucleic acid (RNA) polymerase activity has been used to dissociate the coordinate synthesis of ribosomal proteins and subunits of RNA polymerase and to identify transcriptional and post-transcriptional control signals which regulate the expression of these component genes. Within the beta operon [which has the genetic organization: promoter (p beta), rplJ (L10), r;lL (L7/L12), attenuator, rpoB (beta), rpoC (beta'), terminator], the restriction caused a disproportionate increase between proximal and distal gene transcriptions; the transcriptional intensities of the proximal ribosomal protein genes and the distal RNA polymerase genes were elevated about two- and fourfold, respectively. Transcription within the operon containing four ribosomal protein genes and the RNA polymerase alpha gene was also enhanced, whereas transcription within operons containing only ribosomal protein genes was virtually unaffected by the restriction. It was thus concluded that the mechanisms controlling transcription initiation or attenuation or both in operons containing RNA polymerase subunit genes are coupled to the global rate of RNA synthesis. By introducing the composite ColE1 plasmid pJC701 carrying the proximal portion of the L10 operon, including the beta subunit gene, it was possible to achieve a 10- and a 30-fold range in the transcriptional intensities of the genes specifying L10 and L7/L12 and beta, respectively. Under these conditions, the relative synthesis rates of L7/L12 and beta protein varied by less than 2-fold and by about 15-fold, respectively. These observations corroborate the existence of a post-transcriptional mechanism which severely restricts translation of excess L7/L12 and L10 ribosomal protein messenger RNA; this mechanism is probably important in maintaining the balanced synthesis of ribosome components under conditions in which their messenger RNA levels are dissociated. Furthermore, the observed reduction in the translation efficiency of beta subunit messenger RNA may be related to an inhibitory effect caused by accumulation of RNA polymerase assembly intermediates.

Chemistry and biology of E. coli ribosomal protein L12

E. coli ribosomal protein L12, because of its unique features, has been studied in more detail than perhaps any of the other ribosomal proteins. Unlike the other ribosomal proteins that are generally present in stoichiometric amounts, there are four copies of L12 per ribosome, some of which are acetylated on the N-terminal serine. The acetylated species, referred to as L7, has not been shown, as yet, to possess any different biological activity than L12. A specific enzyme that acetylates L12 to form L7, using acetyl-CoA as the acetyl donor, has been purified from E. coli extracts. L12 is also unique in that it does not contain cysteine, tryptophan, histidine, or tyrosine, is very acidic (pI: 4.85) and has a high content of ordered secondary structure (approximately 50%). The protein is normally found in solution as a dimer and also forms a tight complex with ribosomal protein L10. There are three methionine residues in L12, located in the N-terminal region of the protein, one or more of which are essential for biological activity. Oxidation of the methionines to methionine sulfoxide prevents dimer formation and inactivates the protein. The four copies of L12 are located in the crest region(s) of the 50S ribosomal subunit. There is good evidence that the soluble factors, such as IF-2, EF-Tu, EF-G and RF, interact with L12 on the ribosome during the process of protein synthesis. This interaction is essential for the proper functioning of each of the factors and for GTP hydrolysis associated with the individual partial reactions of protein synthesis. The L12 gene is located on an operon that contains the genes for L10 and beta beta' subunits of RNA polymerase at about 88 min on the bacterial chromosome. DNA-directed in vitro systems have been used to study the unique regulation of the expression of these genes. Autogenous regulation, translational control, and transcription attenuation are regulatory mechanisms that function to control the synthesis of these proteins.

Synthesis and function of ribonucleic acid polymerase and ribosomes in Escherichia coli B/r after a nutritional shift-up

The syntheses of stable ribosomal ribonucleic acid (RNA) and transfer RNA in bacteria depend on the concentration and activity of RNA polymerase and on the fraction of active RNA polymerase synthesizing stable RNA. These parameters were measured in Escherichia coli B/r after a nutritional shift-up from succinate-minimal to glucose-amino acids medium and were found to change in complex patterns during a 1- to 2-h period after the shift-up before reaching a final steady-state level characteristic for the postshift growth medium. The combined effect of these changes was an immediate, one-step increase in the exponential rate of stable RNA synthesis and thus of ribosome synthesis. This suggests that the distribution of transcribing RNA polymerase over ribosomal and nonribosomal genes and the polymerase activity are continuously adjusted during postshift growth to some growth-limiting reaction whose rate increases exponentially. It is proposed that this reaction is the production of amino-acylated transfer RNA and that is exponentially increasing rate results in part from a gradually increasing concentration of aminoacyl transfer RNA synthetases after a shift-up. This idea was tested and is supported by a computer simulation of a nutritional shift-up.

Analysis of a cell cycle model for Escherichia coli

Ribosome and protein synthesis, DNA replication and cell division in Escherichia coli cells are described by a mathematical model that integrates previous descriptions in quantitative terms and proposes a new formalization to relate ribosome net synthesis to cell growth. The model assumes a cell size control of DNA replication and therefore is structurally divided into two subsystems: the first, whose state variables are ribosomes and protein, and the second, which is activated when the protein level reaches a threshold and which is comprised of DNA replication and cell division. The dynamics of the entire system is set only by the first subsystem: the values of its parameters determine whether the cells will be in a resting condition or will grow exponentially and in the latter case the resulting duplication time, while the structure and the parameter values of the second subsystem determine the size and the composition of the cell and the timing of DNA replication during the cycle. Relationships are derived that allow a simple determination of the time of initiation and of termination of DNA replication and the number of chromosome origins involved in any possible cell cycle as well as the macromolecular levels at the beginning of a cycle and on the average in a population of cells in balanced exponential growth.

Autogenous regulation of the synthesis of ribosomal proteins, L10 and L7/12, in Escherichia coli

The in vitro synthesis of Escherichia coli ribosomal proteins, L10 and L7/12, is specifically repressed by the addition of the L10-L7/12 complex, while that of other ribosomal proteins encoded by the neighboring operons is not affected. Thus the expression of the rpoBC operon is controlled by two autorepression systems, one for the two ribosomal proteins and the other for RNA polymerase beta and beta' subunits, both operating probably at the translational level.