The Escherichia coli lac operon promoter is widely used as a tool to control recombinant protein production in bacteria. Here, we give a brief review of how it functions, how it is regulated, and how, based on this knowledge, a suite of lac promoter derivatives has been developed to give a controlled expression that is suitable for diverse biotechnology applications.
Most of the recombinant protein production (RPP) systems used for expressing proteins in bacteria were constructed in the last century [1–4]. High-level RPP provided by these systems enables the synthesis and purification of large amounts of soluble recombinant protein. However, the expression of difficult protein targets (e.g. membrane proteins or proteins secreted out of the cytoplasm), using these RPP systems, may be too high for cells to cope and adequately fold protein, resulting in substantial target degradation or the production of insoluble aggregates (i.e. inclusion bodies) [5–8]. Many ‘tricks of the trade' can be employed to slow down RPP expression and increase the level of soluble product, e.g. lowering the growth temperature, decreasing the inducer concentration or using a weaker promoter [6–9]. Although such tinkering can be very successful, determining the correct combination of refinements can be time-consuming. This can also be very ‘hit-and-miss', being dependent on the particular target protein in question [6,9], and for some induction regimes, for example when using low inducer concentrations, only a proportion of the cells in a culture may in fact express recombinant protein [10–12].
The Escherichia coli lac operon promoter was one of the first bacterial promoters to be adopted by biotechnologists for RPP, and it is still used today, especially when E. coli is used as the host . Here, we give a brief update of our current understanding of transcript initiation in bacteria, emphasising special features of the lac promoter and its regulation. We then review how, based on this information, many lac promoter derivatives have now been engineered in order to facilitate controlled RPP expression and avoid the problems that are concomitant with high-level overexpression.
Transcript initiation and regulation at the E. coli lac operon promoter
Transcript initiation in bacteria takes place when the multisubunit DNA-dependent RNA polymerase (RNAP) holoenzyme interacts with a DNA promoter sequence (Figure 1A). In brief, the RNAP first interacts with double-stranded DNA to form a ‘closed complex’ in which determinants in different RNAP subunits interact with different promoter sequence elements (Figure 1B) . Thus, a determinant in the RNAP α-subunit C-terminal domain interacts with the promoter UP element, a determinant in Domain 4 of the RNAP σ-subunit interacts with the promoter −35 element, and a determinant in Domain 3 of the RNAP σ-subunit interacts with the extended −10 element (Figure 1B). Following this, Domain 2 of the RNAP σ-subunit drives the local unwinding of 13–15 base pairs (bp) of promoter DNA to form the ‘open complex’, in which the single-stranded DNA template strand is positioned in the active site of the RNAP, such that initiation of DNA-templated RNA synthesis can take place (Figure 1C) . Formation of the initiation-competent ‘open complex’ from the ‘closed complex’ is driven by further specific interactions between other RNAP determinants and promoter sequences [15–17]. Thus, determinants in Domain 2 of the RNAP σ-subunit interact with single-stranded bases of the promoter −10 element and with the promoter discriminator element. These interactions involve only the non-template strand of the locally unwound segment of promoter DNA, thereby permitting the single-stranded template strand to access the RNAP active site. The exact position of the template strand in the active site, and the location of the downstream junction between single-stranded and double-stranded DNA, is set by other contacts involving amino-acid side-chains of the RNAP β- and β′-subunits (which interact with the promoter Core-recognition element, CRE; Figure 1C) .
Transcript initiation at bacterial promoters and regulation at the lac promoter.
The activity of any bacterial promoter is set by the formation of the ‘closed complex’ and the ensuing isomerisation to, and escape from, the ‘open complex’, as the RNAP copies the template strand to elongate its transcript. In the case of the E. coli lac operon promoter, defects in the UP element, −35 element and extended −10 element hinder ‘closed complex’ formation but this is remedied by binding of an activatory factor, the cyclic AMP receptor protein (CRP) to a target sequence centred between bp 61 and 62 (referred to as position −61.5) upstream of the transcript start (denoted as +1; Figure 1D) [18,19]. A second regulator protein, the lactose operon repressor protein (LacI) binds to a high-affinity target sequence, known as operator O1, centred at position +11 (Figure 1D) [19,20]. LacI binding to its target effectively shuts down lac promoter activity, but repression can be broken by the presence of allolactose (a breakdown product of lactose) or by the sugar analogue IPTG (isopropyl β-d-1-thiogalactopyranoside), which both bind to the LacI repressor and cause it to release operator DNA . LacI-dependent repression of the lac promoter is supported by LacI binding to two secondary weaker operators, O3, located at position −82 (Figure 1D), and far downstream, O2 [19–23]. Here, we review how the starting lac promoter has been engineered to make it fit for different biotechnology purposes. We focus on constructions that release the requirement for CRP, base changes in key promoter elements, and the exploitation of the O3 operator.
Activator-independent lac promoter derivatives
Figure 2A illustrates the organisation of the E. coli lac operon promoter, showing key promoter elements, the DNA site for CRP, and the location of LacI-binding operators, O1 and O3. Figure 2B shows the base sequence of a typical DNA fragment carrying the lac promoter that might be used in any biotechnology application. The fragment, which carries a useful restriction site at each end, is denoted lac O3O1. Potentially, regulation by CRP might be exploitable but, since CRP activity is difficult to control by external cues, many biotechnology applications that use the lac promoter have sought to eliminate CRP effects and focussed on regulation by LacI. One way to do this is by the use of the lacUV5 mutant promoter. This mutant promoter carries a 2 bp change in the promoter −10 hexamer element that creates a consensus −10 promoter element (Figure 2B) [19,24]. The alternative is to replace lac promoter upstream DNA sequences and this has been done in the tac promoter, which is a chimeric fusion between the upstream elements of the E. coli trp promoter and the downstream elements of the lacUV5 promoter [1,2]. Figure 2A illustrates both the lacUV5 and tac promoters, and Figure 2B shows the base sequence of DNA fragments carrying these promoters. Note that the tac promoter carries consensus −35 and −10 promoters elements and a single operator for LacI, O1. For E. coli promoters that are dependent on the housekeeping sigma factor σ70 the optimal spacing between the −35 and −10 promoter elements is 17 bp and deviation from this leads to a reduction in promoter activity [25,26]. Thus, it worth noting that both the tac and lac promoters are in fact suboptimal, having spacers of 16 and 18 bp, respectively (Figure 2B).
Organisation of lac, tac and lacUV5 promoter constructs.
Figure 3A–C illustrates the results of simple assays to compare the activities of the tac, lacUV5, and lac promoters. To perform these assays, the tac O1 fragment, the lacUV5 O3O1 fragment and the lac O3O1 fragment were each cloned into a plasmid expression vector (pRW50) such that the promoters controlled transcription of the gene (lacZ) encoding β-galactosidase . Recombinant plasmids were then transformed into a Δlac E. coli host strain and β-galactosidase expression was measured. As expected the hierarchy of promoter activity was tac > lacUV5 > lac (Figure 3).
Expression levels of engineered ptac and plac promoter derivatives.
Modulation of promoter activities using upstream lac operator sequences
A key feature of LacI-dependent repression of the E. coli lac promoter is the contribution of the auxiliary upstream operator, O3, and this has been exploited to tailor promoter activity levels. Data in Figure 3A illustrate how the introduction of certain lac operator sequences (i.e. O1 or the high-affinity ‘ideal' lac operator Oid, see Figure 2A [21,28,29]) at position −82 of the tac O1 promoter fragment reduced the high activity of the tac promoter. Thus, expression from the starting tac O1 promoter fragment and each of the tac O3O1, tac O1O1, and tac OidO1 derivatives was induced by IPTG but the introduction of the O1 and Oid operator sequences decreased IPTG-induced expression levels. Similarly, data in Figure 3B,C illustrate how the introduction of higher affinity lac operators into the upstream region of the lacUV5 promoter (i.e. in the lacUV5 O1O1 and lacUV5 OidO1 promoter fragments) or the lac promoter (i.e. in the lac O1O1 and lac OidO1 promoter fragments) decreased promoter activity.
The power of lac: combinations make anything possible
In addition to controlling promoter activity by upstream-bound LacI, the activity can be modulated by point mutations in different promoter elements. Data in Figure 3D illustrate how the p34G, p14G, p9A, or p8A substitutions (at positions −34, −14, −9, or −8), which make the −35 element, the extended −10 or the −10 element more similar to the respective consensus, can be combined with different operator combinations to produce a suite of IPTG-inducible promoters with a wide range of activities.
To illustrate the use of these promoters in RPP, we selected a subset of eight promoters from the above suite and gave each a promoter activity rating (PAR) value of PAR1 to PAR8 (Figure 4A). Our rationale for this is that, depending on the target protein being expressed, specific IPTG-induced expression levels could be achieved. To examine this, some of the promoter constructs were introduced into the low-copy-number vector pTorA-GFP  and the high-copy-number vector pHAK1 , using standard techniques [32–34]. For pTorA-GFP derivatives, each PAR promoter drives the expression of a torA-GFP-6his fusion (GFP, green fluorescent protein), while, for pHAK1, each construct expresses a torA-hGH-6his fusion (hGH, human growth hormone). Note that the torA signal sequence in each case will direct the recombinant protein to the Tat (twin-arginine translocon) system for periplasmic targeting [7,30,31]. Plasmid constructs were transferred into E. coli BL21 cells and RPP was induced in bacterial cultures by the addition of 1 mM IPTG for 3 h . Normalised total cellular protein from cells, expressing either TorA-GFP-6His or TorA-hGH-6His, were then analysed by Western blotting. Results in Figures 4B,C show that both GFP and hGH were expressed using the PAR promoter constructs. Note that, in some instances, two product bands can be observed, in each case, the species with the higher molecular mass still carries the TorA signal sequence, while the smaller species has been processed and lacks the TorA moiety [7,30,31,35]. Importantly, using this expression system, inducible expression of different target proteins can be set to specific levels when using both low- and high-copy-number vectors.
The expression of TorA-protein fusions can be set to different levels with PAR promoter constructs.
The lac promoter and its derivatives have been widely used to express many recombinant proteins to high levels in E. coli [1,2,7,36] and many currently used vectors have been designed to optimise expression. However, there are situations where expression must be moderated. For example, the secretion of recombinant biopharmaceuticals out of the E. coli cytoplasm into the periplasm is often a preferred industrial strategy, as this minimises downstream processing costs since the target protein can be purified from the periplasmic contents, with minimal cellular and DNA contamination . For this to be successful, RPP needs to be slowed down so that the product is not degraded before it is transported .
In this review, we have sought to show how knowledge of the E. coli lac promoter can be exploited to produce a suite of derivative promoter fragments to cope with any situation. Previous promoter engineering had focused on altering the lac promoter −10 and −35 elements to change the basal promoter expression of constructs [1,38]. Recent advances in our understanding of transcript initiation in bacteria and its regulation now allow lac promoters to be constructed with different operator sequences to alter the induced level of expression. This is possible because, even in the presence of inducer, LacI has some affinity for its operator sequence and, thus, in the induced state the LacI repressor can still remain bound to DNA, as has been observed in single molecule studies .
recombinant protein production
cyclic AMP receptor protein
lactose operon repressor protein
promoter activity rating
green fluorescent protein
human growth hormone
D.F.B., S.J.W.B. and C.R. conceived and designed the research programme. D.F.B., R.E.G. and K.L.R. performed the experiments. D.F.B. wrote the manuscript with input from all authors.
This work was generously supported by an Industrial Biotechnology Catalyst (Innovate UK, BBSRC, EPSRC) [grant BB/M018261/1] to support the translation, development and commercialisation of innovative Industrial Biotechnology processes.
The Authors declare that there are no competing interests associated with the manuscript.