Zirex: a Novel Zinc-Regulated Expression System for Lactococcus lactis

Abstract

Here, we report a new zinc-inducible expression system for Lactococcus lactis, called Zirex, consisting of the pneumococcal repressor SczA and P_czcD. P_czcD tightly regulates the expression of green fluorescent protein in L. lactis. We show the applicability of Zirex together with the nisin-controlled expression system, enabling simultaneous but independent regulation of different genes.

TEXT

Lactococcus lactis is a Gram-positive bacterium that has been intensively engineered for the production of heterologous proteins (1, 2). In addition, it is an organism generally recognized as safe (GRAS). To date, several promoters originally from Lactococcus, regulated by inducers or environmental factors, have been documented, including the dnaJ promoter, induced by heat shock (3); the PA170 promoter, which can be upregulated at a low pH during the transition to stationary phase (4); the prtP promoter, which is regulated by the peptide concentration in the medium (5); and the P_Zn zitR promoter, which responds to divalent cation starvation (6). The promoter of nisin, P_nisA, is the most widely used promoter for inducible protein expression in L. lactis (1, 7) and other Gram-positive bacteria (8). The expression from the P_nisA promoter is regulated by the two-component regulatory system NisRK, which is triggered by nisin. For the other promoters mentioned above, there are still some drawbacks, such as relatively low induction levels or high background level at the uninduced stage, which may complicate efforts to tightly control the expression or coexpression of one or two different proteins in the same cell (9). The aims of the present work were to develop a novel tightly controlled promoter for L. lactis and to investigate if such an inducible promoter system could be coupled to the P_nisA promoter to create a dual-promoter-regulated production system for different proteins. First, we searched in the genome of L. lactis MG1363 (NCBI reference sequence NC_009004.1) for proteins putatively involved in cation transport that may be regulated by the presence of cations. A putative promoter, namely, P_Zn3, preceding the translation of a cationic ion efflux protein (NCBI reference sequence YP_001032214.1) in L. lactis was further investigated (see below). Additionally, we explored the genome of other related Gram-positive bacteria for cation-regulated promoters. In the case of Streptococcus pneumoniae, a zinc-inducible promoter was previously described by Kloosterman et al. (10) and Eberhardt et al. (11). sczA and czcD are transcribed divergently (Fig. 1A). The promoter of czcD gene is regulated by SczA. SczA binding to the motif 2 sequence located downstream of the −10 sequence of P_czcD blocks transcription of czcD in the absence of zinc. After the addition of zinc, SczA will move to motif 1 unblocking the transcription (10).

Fig 1.

(A) Nucleic acid sequence of P_czcD. Motif1 (in boxes) and Motif2 (red letters) are the binding sites for the SczA regulator. The start codon of czcD is indicated in italics. The ribosome binding site (RBS) of P_czcD is shown in boldface. −35 and −10 sequences are underlined (10). (B) Expression systems used in this study. Hairpins represent terminators.

Primers czcD-f and czcD-r (Table 1) were designed to amplify the regulator protein SczA and the P_czcD region from the S. pneumoniae D39 (12) genome, including the restriction sites KpnI and NcoI, respectively. The gene coding for green fluorescent protein (GFP) with its own terminator was amplified from pJWV102_gfp (a kind gift from J. W. Veening) by PCR with primers gfp-f and gfp-r. A BglII site was added on the 5′ end of primer gfp-r. czcD-r and gfp-f were designed to be reverse complementary by overlapping the 5′ ends of each other, and an NcoI site was inserted in both primers. The fragment SczA-P_czcD-GFP was generated by spliced overlap extension PCR with primers czcD-f and gfp-r using the mixture of SczA-P_czcD and GFP-specifying amplicons as the templates (13). After digestion with KpnI and BglII, SczA-P_czcD-GFP was cloned into pNZ8048 (7), digested with the same enzymes to create the plasmid pCZG (Fig. 1B). pZn3G was constructed based on pCZG, in which the gfp gene was controlled by P_Zn3 (Fig. 1B). Unfortunately, the low production level obtained after induction with zinc and the leakage in the noninduced state made P_Zn3 an unsuitable candidate for further characterization (data not shown). pNZ8048G was created by cloning gfp amplified with the primers gfp-f and gfp-r2 (Table 1) in the NcoI-HindIII sites of pNZ8048. In pNZ8048G, GFP expression is under the control of P_nisA (Fig. 1B).

Table 1.

Primers used in this study

Primer	Sequencea	Restriction site(s)
czcD-f	CGGGGTACCGGATCCTGCAGGCAGATATAGTTGATAATCAAGG	KpnI, SbfI
czcD-r	CAGCTCTTCTCCTTTTCCCATGGTTCTCATTCCTTTGTTATAATAG	NcoI
gfp-f	CTATTATAACAAAGGAATGAGAACCATGGGAAAAGGAGAAGAGCTG	NcoI
gfp-r	GGAAGATCTATTAATCGAAATACGGGCAGAC	BglII
gfp-r2	CCCAAGCTTCGAAATACGGGCAGAC	HindIII
mCherry-f	CGGGGTACCTCCTGGTTGCAAATTTTG	KpnI
mCherry-r	CGTACTCACGTGCTGCAAGGCGATTAAGTTG	PmlI
sczA-czcD-f	ATCAAGATCTAGAAATAAGACAACTGAAGCTTTAC	BglII
sczA-czcD-r	AGATCCATGGTTCTCATTCCTTTGTTATAATAG	NcoI
pIL-f	ATCAAGATCTACAGCAAAGAATGGCGGAAACG	BglII
pIL-r	AATCGATAAGCTTGGCTGCAGGTC

^aRestriction sites engineered in the primers are underlined.

The expression assays were carried out in L. lactis NZ9000 (9), which was transformed with pCZG (containing SczA-P_czcD-GFP) according to Holo and Nes (14). All of the expression assays were conducted at 30°C in a chemically defined medium for prolonged cultivation (CDMPC) without ZnSO₄ supplemented with 10 μg/ml chloramphenicol (B. Teusink, F. Santos, O. P. Kuipers, C. E. Price, J. Kok, and D. Molenaar, unpublished data). Each assay was repeated in triplicate in a 96-well microtiter plate and monitored with an Infinite 200 Pro microplate spectrophotometer (Tecan Group, Ltd., Mannedorf, Switzerland).

First, we investigated the optimal induction moment. For this purpose, ZnSO₄ was added after 0 h, 2 h, or 4 h of growth at a final concentration of 0.5 mM. The cell growth was monitored measuring the optical density at 600 nm (OD₆₀₀), and the signal of GFP was measured using an excitation wavelength of 485 nm and an emission wavelength of 535 nm (15). NZ9000(pNZ8048) was used as a negative control. We observed that the earlier we induced, the stronger the displayed signal was, with no induction observed in the stationary phase (Fig. 2). It should be noted that 0.5 mM ZnSO₄ showed comparatively low toxicity when it was introduced before inoculation (0 h) of the strains with (pCZG) or without (pNZ8048) GFP. The addition after 2 h of growth does not cause a visible reduction in growth. To assess if the slower growth when the cells were induced at 0 h was caused by the Zn salt used in the study, we also studied the growth tendencies and fluorescent signals of cells induced with the same amount of Zn²⁺ using ZnCl₂. The growth curves and the expression profiles were almost the same as those after induction with ZnSO₄ (data not shown). These results indicate that different Zn²⁺ sources do not affect the toxicity or potency of the induction.

Fig 2.

Growth (dotted lines) and GFP expression level (solid lines) after induction with 0.5 mM zinc at different time points. NZ9000(pCZG) induced with ZnSO₄ at a final concentration of 0.5 mM at 0 h (□), 2 h (△), and 4 h (○). Two control strains were run in parallel: NZ9000(pNZ8048) uninduced (◇) and induced with 0.5 mM ZnSO₄ at 0 h (×). The values represent the means from three independent measurements.

In order to assess the optimal concentration of ZnSO₄ for the induction, NZ9000(pCZG) was induced at an OD₆₀₀ of 0.06 (the middle of the exponential phase) with a final concentration of 0, 0.1, 0.3, 0.5, 0.7, or 1.0 mM ZnSO₄. NZ9000(pNZ8048G) grown in CDMPC was induced with 5 ng/ml of nisin at an OD₆₀₀ of 0.06 as a reference. The growth rate was not affected by the addition of ZnSO₄, and the GFP signal produced by NZ9000(pNZ8048G) or NZ9000(pCZG) reached the highest level after 2 h or 2.5 h of induction, respectively (Fig. 3A). The highest intensity of GFP produced using the zinc-inducible system was almost 80% of that produced with the nisin-inducible system. Moreover, the GFP signal in the induced cells increased nearly proportionally with the ZnSO₄ concentration in the range between 0 and 0.3 mM (Fig. 3B). Furthermore, almost no fluorescent signal was detected under uninduced conditions, which demonstrated that this pneumococcal system is also effectively repressed in the absence of zinc in L. lactis.

Fig 3.

(A) Growth (dotted lines) and GFP intensity (solid lines) after induction with different zinc concentrations. L. lactis NZ9000(pCZG) was induced at an OD₆₀₀ of 0.06 with different concentrations of ZnSO₄: 0 (◇), 0.1 mM (□), 0.7 mM (△), or 1.0 mM (○). A control, NZ9000(pNZ8048G) (×), induced with 5 ng/ml of nisin was also used to compare the production levels of P_czcD and P_nisA. The arrow indicates the time point for induction. These values represent the means from three independent measurements. (B) Dose-response curve of GFP expression of L. lactis NZ9000(pCZG). Fluorescent signal is shown as specific units per OD₆₀₀. The fluorescent signal changed less sensitively at the concentration of ZnSO₄ above 0.3 mM. The standard errors are less than 15% for each value.

pCZGM was constructed to observe the effect of the induction with nisin and zinc at the same time. In pCZGM, P_nisA controls the expression of mCherry, whereas P_czcD controls the expression of GFP. To construct this vector, a fragment encompassing from P_nisA to the terminator of mCherry was amplified from pHK35C (a generous gift from H. Karsens) using the primers mCherry-f and mCherry-r, containing at their 5′ end a KpnI site and a PmlI site, respectively. After digestion with KpnI and PmlI, the fragment was inserted into pCZG cut with the same enzymes, resulting in pCZGM (Fig. 1B). The signal of mCherry was measured using an excitation wavelength of 590 nm and an emission wavelength of 620 nm, and GFP was measured as mentioned above. Cultures were induced with 0.7 mM ZnSO₄ at an OD₆₀₀ of 0.06 and with 5 ng/ml nisin 1 h later. Uninduced controls lacking either nisin or zinc were run in parallel. In Table 2, we can observe the expression level of GFP or mCherry achieved after 2.5 h of the induction with ZnSO₄ or nisin. These data show that simultaneous overexpressions of mCherry and GFP in this system cause around 23% and 11% reduction of the two fluorescent signals, respectively.

Table 2.

Simultaneous production of GFP and mCherry

ZnSO4 as inducer (0.7 mM)a	Nisin as inducer (5 ng/ml 1 h later)a	Fluorescent intensity (AU)b
		GFP (2.5 h)	mCherry (2.5 h)
+	−	181.00 ± 8.01	0.67 ± 4.19
+	+	138.00 ± 2.49	544.33 ± 17.68
−	+	5.33 ± 3.00	612.00 ± 39.04

^a+, inducer present; −, inducer absent.

^bAU, arbitrary units.

Based on the results described above, we created pCZ-Cm for general use as a chloramphenicol-resistant expression vector for L. lactis. In this vector, the multiple-cloning site (MCS) of pNZ8048 was fused behind P_czcD. For this purpose, the region SczA-P_czcD was amplified from pCZG with the primers sczA-czcD-f and sczA-czcD-r (Table 1). After digestion with BglII and NcoI, the fragment SczA-P_czcD was inserted into pNZ8048 digested with the same restriction enzymes, rendering pCZ-Cm (Fig. 1B). An additional expression vector, termed pILZ-Em, containing this zinc-inducible expression system with the same MCS and erythromycin resistance was also constructed from the plasmid pIL253 (Fig. 1B) (16). A BglII site was inserted into pIL253 by round PCR with the primers pIL-f and pIL-r (Table 1) in order to insert the BglII-SacI region from pCZ-Cm.

In order to assess the usefulness of this double inducible system, the structural gene of nisin, nisA, was cloned into plasmid pCZ-Cm under the control of P_czcD and transformed into NZ9800 (17). In this strain, the enzymes responsible for the maturation and modification of nisin are controlled by P_nisA. Comparison of the production of nisin in CDM medium (18, 19) with a constant concentration of nisin and various amounts of ZnSO₄ was performed (Fig. 4). We measured the production of nisin using an activity test against L. lactis NZ9000 (20). The activity assay clearly shows that nisin can be successfully expressed in the system in a tightly regulated fashion when the gene nisA is controlled by P_czcD and the modification enzymes are regulated by nisin (Fig. 4).

Fig 4.

Expression of nisin measured in terms of activity. In the control strain, NZ9800(pNZ-nisA) (20), nisA and the modification enzymes that process nisin are controlled by P_nisA. In NZ9800(pCZ-nisA), nisA was controlled by P_czcD. A constant concentration of nisin (5 ng/ml) was used to induce both strains. In NZ9800(pCZ-nisA), no activity was detected in the absence of Zn²⁺, whereas increasing amounts of Zn²⁺ led to the production of nisin proportionally to the concentration used.

In our study, we introduced the streptococcal promoter P_czcD together with the gene coding for its regulatory protein, SczA, in L. lactis, yielding an effective zinc-regulated expression system, called Zirex. Our results clearly show that this system can effectively control the overexpression of proteins in response to modest and nontoxic zinc additions to the medium in L. lactis. The very low basal expression without inducer suggests that SczA is also expressed in L. lactis and tightly represses the system in the absence of zinc. Notably, L. lactis showed high tolerance to zinc in the millimolar range when induced in the exponential phase (6). Previously, a zinc-repressed expression system (P_Zn zitR promoter) was reported. It was based on the L. lactis zit operon, which encodes an emergency Zn²⁺ uptake ABC transporter (6). The presence of Zn²⁺ can repress the expression of the emergency Zn²⁺ uptake ABC transporter, which partly explains the high tolerance of L. lactis to Zn²⁺. The initiation of the P_Zn zitR promoter is caused by the addition of a chelating agent, which reduces the available zinc in the medium, therefore activating the transcription of the emergency uptake system (6). A drawback of this system is that the induction based on the depletion of Zn²⁺, which is achieved with the addition of EDTA, can hamper the overexpression of proteins or enzymes that require cations. The zinc-inducible system presented here constitutes, to our knowledge, the first zinc-inducible promoter developed for L. lactis. It can be extremely useful for the overproduction of enzymes such as lanthipeptide cyclases, which require Zn²⁺ to be active, or other metalloenzymes. This advantage makes the expression system presented in this paper a suitable candidate for the production of lanthipeptides (21). So far the nisin-inducible expression (NICE) system is the most widely used and potent protein expression system in L. lactis. Compared to the nisin-inducible system, the common drawbacks of other regulated expression systems found in L. lactis are their low expression level and/or high leakage (9). The zinc-inducible system presented here achieves a high expression level comparable to that of nisin (ca. 80%), which is higher than the expression level obtained with the P_Zn zitR promoter (20% of that achieved with nisin) (6). Moreover, we demonstrate that it is possible to combine both inducible promoters for the expression of different proteins at different times during cell growth. This can be a useful tool for the overexpression of proteins or the creation of controlled gene regulatory circuits in L. lactis and expands the toolbox available for this bacterium. Moreover, the plasmid described here could be directly applicable for use in other Gram-positive hosts, as is the case for the NICE system, although this has to be further investigated to assess the specific characteristics.