High Level Constitutive Expression of Luciferase Reporter by lsd90 Promoter in Fission Yeast

Because of a large number of molecular similarities with higher eukaryotes, the fission yeast Schizosaccharomyces pombe has been considered a potentially ideal host for expressing human proteins having therapeutic and pharmaceutical applications. However, efforts in this direction are hampered by lack of a strong promoter. Here, we report the isolation and characterization of a strong, constitutive promoter from S. pombe. A new expression vector was constructed by cloning the putative promoter region of the lsd90 gene (earlier reported to be strongly induced by heat stress) into a previously reported high copy number vector pJH5, which contained an ARS element corresponding to the mat2P flanking region and a truncated URA3m selectable marker. The resulting vector was used to study and compare the level of expression of the luciferase reporter with that achieved with the known vectors containing regulatable promoter nmt1 and the strong constitutive promoter adh1 in S. pombe and the methanol-inducible AOX1 promoter in Pichia pastoris. Following growth in standard media the new vector containing the putative lsd90 promoter provided constitutive expression of luciferase, at a level, which was 19-, 39- and 10-fold higher than that achieved with nmt1, adh1 and AOX1 promoters, respectively. These results indicate a great potential of the new lsd90 promoter-based vector for commercial scale expression of therapeutic proteins in S. pombe.


Introduction
Among various yeast species, the methylotrophic yeast Pichia pastoris has emerged as a useful expression host for commercial scale production of therapeutic proteins. Reasons behind popularity of the Pichia system are availability of AOX1 promoter, a strong methanol-inducible promoter of alcohol oxidase I and ease of growth to very high cell density in an inexpensive, non-complex and chemically defined medium [1][2][3][4]. However, despite these advantages, there are serious issues regarding safety aspects of Pichia expression system, which uses inflammable methanol for induction; methanol is added after every 24 hrs of growth phase and maximum expression is achieved only after 4-6 days of induction. Large-scale fermentation using considerable amounts of methanol necessitates suitable precautions and special fermenter design to ensure safety. In addition, disposal of toxic waste containing methanol is of grave environmental concern.
Among other well-characterized yeast species, the fission yeast Schizosaccharomyces pombe has been recognized as an excellent model system for understanding biological phenomena at the cellular and molecular level [5,6]. In comparison with the budding yeast Saccharomyces cerevisiae, S. pombe shares greater level of similarity with higher eukaryotes. For example, the conservation of splicing machinery which allow human intron-containing genes to be spliced in S. pombe, the complexity of replication origins and centromeric regions, gene regulatory mechanisms, existence of similar components of RNAi and heterochromatin mechineries, presence of intact Golgi apparatus, etc. [5,6]. Because of these similarities, S. pombe is considered as an attractive host for the production of proteins of eukayotic origin [5,6]. Expression vectors for high-level expression in S. pombe have been developed and many foreign proteins successfully expressed [5,6]. However, S. pombe has lagged behind the P. pastoris expression system mainly because of lack of strong promoters. Among them, nmt1, the strongest known regulatable promoter, is repressed by thiamine and has been widely used for heterologous gene expression [5,7]. To further achieve graded levels of expression, two additional weakened versions of the nmt1 promoter, denoted as nmt41 and nmt81, provide medium and low levels of expression, respectively [8]. Use of mammalian viral promoters has also been reported in S. pombe [9]. However, these promoters are not suitable for commercial scale expression because they do not achieve expression levels comparable to P. pastoris and are not userfriendly. In our experience, nmt1 promoter yielded only moderate level of expression of a therapeutic protein, indicating inadequacy of nmt1 for commercial scale expression of proteins [10]. Several other promoters have been used for regulated expression of heterologous genes in S. pombe but the expression level was even lower than nmt1 promoter [11][12][13][14]. adh1, the strongest known constitutive promoter can provide expression levels up to 0.5-2.5% of total protein [15,16]. However, the maximum bgalactosidase activity achieved under the adh1 promoter was almost half of that achieved under the nmt1 promoter [17]. Moreover, constitutive expression of the heterologous gene may sometimes be toxic to the host cell, although it depends on the nature of the protein to be expressed. Thus, isolation of new stronger promoters in S. pombe that can be employed at industrial scale with minimum effort and inexpensive media is an area requiring greater systematic and sustained effort.
Luciferase has often been used as a reporter gene in bacterial and mammalian cells. There are very few reports describing the use of luciferase as a reporter in S. cerevisiae [18,19]. Firefly luciferase (Fluc) has been used to study transcriptional and translational fidelity in yeast [20]. Recently, Fluc protein was used as a marker to evaluate the efficacy of the vectors with high copy number and mitotic stability for high-level expression of heterologous proteins in Hansenula polymorpha [21]. The Fluc reporter has also been used as a biosensor for screening compounds toxic to eukaryotes [22]. Fluc has proved highly effective as a reporter gene, since the luciferase assays are extremely sensitive, rapid, reproducible and easy to perform [19].
Earlier, we constructed a new high copy vector pJH5 containing the mat2P-linked ARS element from S. pombe and a truncated URA3m gene from S. cerevisiae as a selectable marker, which occurs at very high copy number, ,200 copies/cell [23]. In this study, we used this vector as backbone for cloning the new promoter. We initiated screening of new promoters based on the published microarray data [24] and measured the activity of putative promoter elements using luciferase as the reporter gene. The resulting vector provides constitutive expression of luciferase reporter which is considerably higher than not only the expression achievable with the vectors containing nmt1 and adh1 promoters in S. pombe but also the AOX1 promoter based vector in P. pastoris.

Construction of expression vectors
To search for new stronger and regulatable promoters, we screened the published DNA microarray data of global transcriptional responses of S. pombe to various environmental stresses [24]. The published data showed that genes SPAC1F8.02C and SPBC24C6.09C showed maximum induction of RNA levels (260and 108-fold, respectively) when exposed to 0.5 mM H 2 O 2 , while lsd90/SPBC16E9.16C showed maximum induction (117-fold) upon shifting the culture from 30uC to 39uC. Accordingly, approximately 1.0 to 1.5 kb upstream regions of these genes were PCR amplified as putative promoters and cloned into the vector pJH5 [23] to obtain new expression vectors pJH6a, pJH6b (not shown) and pJH6c (Figure 1), respectively. To assess the promoter efficiency, Fluc reporter gene was cloned downstream of these promoters (Figure 1). For a comparative study with existing expression systems, Fluc gene was also cloned downstream of adh1 (vector pART1) and nmt1 promoter (vector pREP3X) of S. pombe and AOX1 promoter (vector pPIC3.5) of P. pastoris ( Figure 1).
Strong, constitutive expression of Fluc-gene by putative promoter region of lsd90 gene. The putative expression vectors were transformed into the S. pombe strain SPJ25. Transformants were grown in synthetic medium and expression of the luciferase reporter gene in response to various stresses [24] was studied. However, contrary to expectation based on the published microarray data [24] very low Fluc expression was observed in cells expressing Fluc gene under the control of the Psp1 and Psp2 promoters (1.7610 218 moles/100 ng and 2.3610 218 moles/100 ng, respectively). Subjecting the cells to oxidative stress also showed no effect (data not shown). The discrepancy of these results from the published microarray data is surprising. This may be due to lack of some further upstream regulatory sequences or presence of negative regulatory sequences inhibiting the induction of the existing promoter elements in the putative promoter region in response to oxidative stress. Alternatively, growth conditions like media composition may account for the difference from the published results [24].
In contrast, the vector pH6c-Fluc containing the lsd90 promoter (Plsd90) provided considerably higher level of Fluc expression, which showed increase with cell density, reaching the highest level of 2.4610 214 moles [(,2.4 ng)/100 ng protein] during early stationary phase (48 hrs). Thereafter, expression level remained almost constant, with a slight increase after 64 hrs [2.5610 214 moles (,2.5 ng)/100 ng protein], which amounted to 2.5% of total cellular protein ( Figure 2A). Furthermore, contrary to the published microarray data [24] exposure of cells to heat stress did not cause a further increase in the level of expression of Fluc. Thus, the putative lsd90 promoter fragment appears to elicit constitutive expression of luciferase (not shown).
It is possible that constitutive and high level of heterologous gene expression may exert a metabolic load thus leading to slow growth of host cells compared to cells with vector alone. Surprisingly, cells expressing Fluc gene under lsd90 promoter showed a faster growth rate and achieved higher cell density than those containing the control vector ( Figure 2B).
The increase in OD 600 may also occur indirectly; for example, it may occur due to an increase in cell size. Therefore, we counted the cell density by counting the number of cells/ml by haemocytometry. The results closely parallel the data shown in Figure 2B, indicating a faster growth rate and higher cell density in case of cells harboring the recombinant vector pJH6c-Fluc in comparison with those harboring the control vector pJH6c ( Figure 2C).

Comparison of new expression vector with existing systems of S. pombe
To compare the suitability of the new lsd90 promoter containing vector as an alternative expression vector, kinetics of expression of luciferase under the control of adh1 and nmt1 promoters was also studied. Luciferase expression under adh1 promoter (in vector pART1) showed an increase up to log phase of growth with maximum level of 6.4610 216 moles [(,0.064 ng)/100 ng protein] after 32 hrs of growth. A slight decline (5.1610 216 moles/100 ng protein) was observed at 96 hrs of growth, probably due to decrease in the growth rate upon nutrient depletion ( Figure 3A). Importantly, the maximum level of luciferase expression achieved with vector pART1-Fluc under the adh1 promoter was ,39-fold lower as compared to that achieved with the vector pJH6c-Fluc under lsd90 promoter ( Figure 2A; Table 1).
Similar, when expressed under the control of the nmt1 promoter (vector pREP3X), the maximum level of luciferase expression of 1.3610 215 moles [(,0.13 ng)/100 ng protein] was recorded after 18 hrs of induction ( Figure 4A). This level was almost 19-fold lower as compared to maximum expression level with vector pJH6c-Fluc under lsd90 promoter ( Figure 2A; Table 1). The growth rate of culture expressing luciferase under the adh1 promoter ( Figure 3B) and the nmt1 promoter ( Figure 4B) was almost similar to that of culture harbouring control vectors pART1 and pREP3X, respectively, indicating no deleterious effect of luciferase expression on host cells' growth ( Figure 4B).
Kinetics of luciferase expression under AOX1 promoter in P. pastoris. For comparison, we also studied the kinetics of expression of luciferase under control of the methanol-inducible AOX1 promoter in the vector pPIC3.5-Fluc in P. pastoris. A continuous increase of luciferase activity was recorded, reaching maximum level of 2.5610 215 moles [(,0.25 ng)/100 ng protein] after 5 days of induction. The luciferase activity decreased to almost half of the maximum level after 6 days of induction [1.2610 215 moles; ,0.12 ng/100 ng protein] ( Figure 5). In comparison, the maximum level of luciferase expression was obtained with lsd90 promoter after 48-64 hrs (Figure 2A). This level was at least 10-fold higher than that the maximum level of luciferase expressed achieved under the control of AOX1 promoter after 120 hrs of induction in P. pastoris ( Table 1).
Measurement of Luciferase Mrna. The level of expression of luciferase mRNA was measured by RT-qPCR and normalized against internal control housekeeping gene act1 to determined the value of DC T for each Fluc transcript at selected time points. Relative expression fold change of transcripts with respect to pJH6c-Fluc was quantified by DDC T method using 2 2DDCT calculation [25]. As shown in Fig. 6, consistent with level of protein expression, highest level of Fluc transcript was obtained with the vector pJH6c-Fluc (32 hrs), followed by the Pichia vector pPIC3.5-Fluc (96 hrs), and pombe vectors pART1-Fluc (48 hrs) and pREP3-Fluc (16 hrs). Surprisingly, in contrast with the relative levels of Fluc protein, the levels of Fluc mRNAs were relatively   Table 1). Furthermore, while Fluc protein expression with pREP3-Fluc was about twice as high as compared to that obtained with pART1-Fluc (Table 1), the mRNA expression level by these vectors are somewhat similar. The cause of this discrepancy is not known.

Discussion
Although we selected putative promoter elements based on the high level of induction of mRNA from the linked gene sequences in response to heat and oxidative stress [24], we were surprised to observe that the promoter elements Psp1 and Psp2 failed to elicit significant expression of luciferase, both with and without stress conditions, in S. pombe. On the other hand, lsd90 promoter (Plsd90) elicited high level constitutive expression of luciferase, which was not stimulated further by heat stress, while the lsd90 gene is known to be induced by heat stress [24]. In a recent report, the promoter region of the heat-inducible gene hsp16 has been shown to elicit induction of the GFP reporter in response to heat stress [26]. While a 0.6 kb upstream region of hsp16 gene elicited constitutive expression of the GFP reporter, longer upstream regions of 1.2 and 1.8 kb imparted heat stress inducible expression [26]. In the present study, the lack of temperature regulation of lsd90 promoter is surprising. Possibly, some sequence elements of lsd90 promoter located further upstream to the region used in this study may impart the heat-inducibility. Alternatively, the promoter region  used here may contain elements that respond negatively to heat stress and thus mask the heat stress-dependent induction. It is also possible that the high copy vector may titrate out the negatively acting regulator of the lsd90 promoter in absence of heat stress. All these possibilities will be tested in future studies.
Surprisingly, cells expressing luciferase under the control of the lsd90 promoter grew at faster rate and achieved higher OD 600 as  well as cell density than cells containing vector alone. It is possible that host cells having gene insertion in the backbone vector pJH5 [23] may impart some growth advantage. Although the exact mechanism of this characteristic is not known, it would serve as an advantageous feature for heterologous gene expression.
Here, it is pertinent to evaluate the contribution of different elements of the vector pJH6c towards the overall yield of proteins. The vector pJH5, containing the truncated S. cerevisiae URA3m as a selectable marker and the mat2P RF as ARS element, as a part of the backbone of the vector pJH6c, occurs at a higher copy number (,200 copies/cell), and has greater mitotic stability (,1.5 fold) as well as greater loss rate (,1.25 fold) than the vector pJH2 [23], which is similar to pART1 and pREP3 (having LEU2 as a selectable marker and replication origin ars1(23; data not shown).
The latter two vectors are, however, present at lower copy number of 50 copies/cell [23]. While a higher plasmid copy number does not necessarily cause proportionate increase in the level of expression, even assuming a linear correlation between the copy number and expression level, a cumulative contribution of plasmid copy number, plasmid stability and loss rate (461.5/1.25) yields a theoretically ,4.8 fold advantage of the backbone of the vector pJH5 over pART1 and pREP3X for protein expression level. Even after accounting for this contributory factor, the lsd90 promoter provides at least ,4-and 8-fold higher expression level per plasmid copy than that provided by vectors containing adh1 and nmt1 promoters of S. pombe, respectively.
In conclusion, this study has led to the construction of a potentially powerful expression vector harbouring a strong constitutive lsd90 promoter region in S. pombe, which can yield expression level 19-and 39-fold higher than known strong promoters of S. pombe, namely nmt1 and adh1, respectively (Table 1). Most importantly, under shake flask conditions, it yields nearly 10-fold higher expression of luciferase after 48 hrs than that achieved with the AOX1 promoter in P. pastoris after 120 hrs of culture (Table 1). After accounting for the time factor of 2.5-fold faster expression, the new expression vector shows 25-fold greater productivity of the lsd90 promoter based vector in S. pombe as compared to the AOX1 promoter in P. pastoris.
Based on its high level of expression, the lsd90 promoter based vector pJH6c has great potential for commercial application. Further development of the system will involve optimization of secretion of protein by inclusion of secretory signals in the vector [27], use of protease deficient mutants [28] and addition of inexpensive chemicals like dextran sulphate to the culture medium [29]. Use of simple medium and expression regime further obviate the need for methanol or other inducers and the associated special fermenter design, also making the process more environmentfriendly. Thus, the new expression vector could prove to be a viable, cost-effective and both user and environment-friendly alternative to the expression under the control of the AOX1 promoter in P. pastoris for commercial scale expression of heterologous proteins of therapeutic and industrial interest in S. pombe.

Cloning of the promoters of S. pombe
For cloning the promoters of SPAC1F8.02C, SPBC24C6.09C and lsd90 genes of S. pombe, PCR primers ( Table 2) were designed to amplify 59-upstream region of about 1.5 Kb, 1.2 Kb and 1 Kb, respectively. Primers used for amplification of the promoters were as follows: for Psp1, SP1 and ASP1; for Psp2, SP2 and ASP2; for Plsd90, SP3 and ASP3. The 59-and the 39-primers contain Sph1 and NdeI restriction sites, respectively. The PCR products were digested with these restriction endonucleases and inserted into the same sites in the vector pJH5 [23]. The resulting vectors were named as pJH6a, pJH6b and pJH6c, respectively.

Transformation and selection in S. pombe
The Fluc containing vectors (pJH6a-Fluc, pJH6b-Fluc, pJH6c-Fluc, pART1-Fluc and pREP3X-Fluc) described above were transformed into S. pombe wild type strain SPJ25 having leu1-32 and ura4D18 mutations (genotype: mat1Msmt0, leu1-32, ura4D18, his2, ade6-210), by lithium acetate method [30]. Transformants were confirmed for presence of Fluc-gene insert by colony PCR using gene-specific primers.  Transformation and selection in P. pastoris P. pastoris transformation and selection was done according to instruction manual (Pichia Expression Kit, Invitrogen, USA). The Fluc-expression vector pPIC3.5-Fluc described above and the control vector pPIC3.5 were linearized with SalI and transformed into P. pastoris host strain GS115 (His 2 Mut + phenotype) by lithium chloride method. Transformants were selected on minimal plates lacking histidine and screened for Mut + phenotype. Pichia transformants were confirmed by colony PCR using gene-specific primers.
Expression of Fluc gene in S. pombe S. pombe strains harboring recombinant expression vectors were inoculated in selective pombe minimal media (PMA ura 2 or leu 2 ) and cultures were grown for 16-18 hrs at 30uC and 200 rpm. In case of the vector containing the nmt1 promoter, thiamine was added at 25 mM final concentration to repress the promoter during growth phase. This culture was inoculated into fresh selective minimal media to a final OD 600 of 0.1 and grown at 30uC and 200 rpm. For protein extraction and growth kinetics, samples from the culture were harvested at the indicated time points. In case of the nmt1 promoter, secondary culture was grown up to mid-log phase (OD 600 0.4-0.6) in PMA leu 2 medium containing 25 mM thiamine. This culture was centrifuged and washed thrice with the same media without thiamine at room temperature. Cell pellet was resuspended in fresh PMA-leu media and grown further at 30uC and 200 rpm. Samples were harvested at specific time intervals to study growth kinetics and expression of luciferase.
All measurements of each time point were carried out in triplicate.

Expression of Fluc gene in P. pastoris
Optimization and scale-up of recombinant protein expression in Pichia was done according to instruction manual (Pichia Expression Kit, Invitrogen, USA). A single colony of the recombinant GS115 strain of P. pastoris harboring the Fluc gene, cloned under AOX1 promoter was inoculated into 25 ml of BMG (Buffered Minimal Glycerol) medium in a 250 ml baffled flask and grown at 30uC in a shaker-incubator at 250 rpm until the culture reached log phase with OD 600 of 2-6 (,16-18 hrs). The culture was harvested by centrifugation at 30006g for 5 minutes at RT. The supernatant was discarded and cell pellet was resuspended in 100 ml BMM (Buffered Minimal Methanol) for induction in a 1L baffled flask and grown further at 30uC with shaking at 250 rpm. To this culture, 100% methanol was added after 24 hrs intervals to give a final concentration of 1% to maintain the induction. 1 ml of culture was harvested after 24 hrs intervals for protein extraction.
All measurements of each time point were carried out in triplicate.

Protein extraction from yeast strains
Cell extracts of S. pombe cultures were prepared as described previously [10]. Cell extracts of P. pastoris were prepared as described in Kit manual (Pichia Expression Kit, Invitrogen, USA). In case of P. pastoris, cells lysis was done as described for S. pombe, except that less number of cycles (5)(6) in bead beater (Biospec, USA) were sufficient for cell breakage.

Luminescence assay
The level of expression of luciferase was determined with Luciferase Assay System (Promega Corp., Madison, WI). All reagents were prepared as described by the manufacturer. We directly used 100 ng of cell extract for luciferase assay in place of cell lysate prepared with passive lysis buffer. All measurements were performed with GloMax TM 20/20 luminometer. Protein concentration of luciferase (moles/reaction) was determined against the calibrated standard curve, plotted between known amounts of purified recombinant luciferase (Promega Corp., Madison, WI) in moles and the respective luminescence values in Relative Light Units (RLU).

Real-Time quantitative polymerase chain reaction (RT-qPCR)
Two-step RT-qPCR was performed to assess the relative expression of transcripts. Total RNA of cells for different time points was isolated using phenol/chloroform extraction method [33] and treated with DNase1. cDNA was synthesized from the total RNA using the cDNA Reverse Transcription Kit (Thermo scientific) in accordance with manufacturer's instructions. Maxima SYBR Green/Fluorescein qPCR Master Mix (Thermo scientific), specific forward and reverse primers of target (luciferase) and internal control (act1) genes were used in PCR reactions and RT-qPCR was performed in Mastercycler ep realplex Real-time PCR System. The parameters of thermocycling consisted of an initial denaturation at 95uC for 3 minute, followed by 40 cycles of denaturation (95uC/1 minute), annealing (55uC/30 second), and extension (72uC/1 minute), melting-curve analysis was carried out starting from the initial temperature 50uC to 95uC, with gradual increase of 0.5uC/15 second. The generated C T values of target gene were normalized to the C T value of act1 gene and the relative fold expression changes were estimated by DDC T method [25]. RT-qPCR experiments were performed in triplicates and on three different days.