Standard formats for genetic tools have been proposed for different organisms to facilitate the implementation of synthetic circuits and the distribution of material between different laboratories [18,21,22,23]. As a starting point for our modular archaeal vectors, we inspired our design in the SEVA format (Standard European Vector Architecture), which was used to construct modular vectors for a broad range of gram-negative bacteria [18]. The design used here is represented in Fig 1. As shown in the schema, the pHsal vectors have four modules, each one with specific functionalities. The first module allows the replication of the vectors in the E. coli host to facilitate genetic manipulation of the plasmids: it is formed by the multi-copy ColE1 replication origin and the bla gene for ampicillin resistance. Both segments are taken from pUC19 vector [24]. The second element is a replication origin that confers autonomous maintenance in the archaeal host. We used the minimal origin from pGRB1 plasmid [25], which is widely used in existing vectors for H. salinarum [2]. The third element is a resistance marker that allows the selection of H. salinarum strains harboring the plasmids. We used the mevinolin resistance marker mev^R, which has been successfully used in H. salinarum [2,7]. Finally, the fourth module is the Cargo, which represents an extensive MCS flanked by universal M13 primers. This architecture facilitates cloning procedures and further confirmation steps, such as digestions and sequencing. As shown in Fig 1, each functional module is flanked by a unique restriction site that can be used to replace these elements by alternative variants [18]: the cargo module is flanked by PacI and SpeI restriction sites, the resistance marker by SpeI and AscI, the archaeal replication origin by PacI and BglII and the E. coli’s specific part by BglII and AscI.

Since unique restriction sites should be exclusively used to replace the functional modules, any fragment used to construct the vectors should be first sequence-edited to eliminate these sites, as well as those restricted to the MCS region. The modular design used in pHsal series also includes a specific architecture for other functional elements such as universal primers and insertion of expression and reporter systems (Fig 2). In this case, universal primers (F24 and R24) are placed between the flanking restriction sites (PacI or SpeI) and the nearest restriction sites in the MCS (Fig 2A). Similarly, any expression element (either a promoter or a promoter plus its regulator) can be inserted between PacI and AvrII sites in the MCS (Fig 2B), while reporter systems are introduced at the end of this region between HindIII and SpeI restriction sites (Fig 2C). The next sections describe the construction of functional modules and the implementation of new functional elements such as promoters and reporter systems for H. salinarum.

Based on the modular design presented above, we constructed a series of minimal, modular vectors for H. salinarum. The maps of the two main vectors are presented in Fig 3. In the pHsal series, pHsal-C is a 5.3 kb cloning vector that has a mevinolin resistance marker and a pGRB1 replication origin for autonomous maintenance in H. salinarum (Fig 3A). This vector is devoid of any expression system. and can be used with user-defined promoters and genes. Two variants of pHsal-C were constructed for different applications: pHsal-E and pHsal-GFP. The vector pHsal-E harbors an expression system based on a variant of the strong fer2 promoter (see below) that is cloned between the PacI and AvrII sites in the MCS (S1A Fig at the Supporting Information file 1). This vector allows high-level, constitutive expression of heterologous genes in H. salinarum. The pHsal-GFP vector is a variant of pHsal-C that harbors a GFP reporter gene between HindIII and SpeI sites and can be used to quantify promoter activities in H. salinarum (S1B Fig).

While the three vectors described above allow autonomous maintenance of the genes of interest in H. salinarum, some applications require the modification of chromosomal sequences to generate stable and permanent genotypes. Construction of knockouts strains, tagging of proteins with an epitope or replacement of a wild type gene by a mutant variant [2,7,13] require the use of a suicide vector with a marker for counter selection, such as the ura/pyr strategy [26]. We have thus constructed a vector named pHsal-S, which is devoid of a replication origin for H. salinarum and has both the mev^R resistance marker for positive selection and a sequence-optimized ura3 (VNG1673G) gene for counter selection with 5-FOA [26]. It is important to highlight that the pHsal-S vector not only has a modular architecture and an extended MCS, but is also about 1.6 kb shorter than the currently available vector for knockout generation and protein tagging [13]. This size reduction is possible since only minimal functional sequences were used, facilitating the manipulation of the vector in vitro. All these new modular vectors, along with their functional components were tested in vivo in H. salinarum and were found to be functional in this organism, as described in the next sections. Taken together, these new vectors provide a set of valuable genetic tools for engineering H. salinarum that could significantly aid functional studies in this model organism.

As mentioned above, pHsal-E vector allows the expression of heterologous fragments in H. salinarum from a strong, constitutive fer2 promoter [27] that was sequence edited to eliminate useful restriction sites. Fig 4A shows a schematic representation of the fer2 promoter region (Pfer2), where TATA box, BRE and PPE regions [28] are represented. In the Pfer2 wild-type sequence, the region 150 bp upstream of BRE encompasses the Upstream Activation Region (UAS), where we identified three restriction sites for the enzymes NcoI, SphI and SmaI. We used mutagenic PCR to change four bases in the wild type sequence, eliminating these restriction sites. The mutated UAS sequence was assembled by PCR and cloned into a GFP reporter vector, generating the promoter Pzero (Fig 4A). To check the functionality of the sequence-edited promoter, we introduced reporter plasmids with the wild type Pfer2 or with Pzero into H. salinarum. As a control, a plasmid devoid of any promoter was also introduced to determine the basal activity of the system. Recombinant H. salinarum strains harboring the plasmids of interest were grown in CM media and the promoter activities were assayed at mid (16h) and late (24h) exponential phase. As shown in Fig 4B, the sequence-optimized Pzero displayed a promoter activity very similar to the wild type Pfer2, showing that the mutations introduced into the promoter did not affect significantly its activity. Thus, the data provided here show the construction of sequence optimized, strong expression system for high-level expression of heterologous genes in H. salinarum.

To demonstrate the efficacy of the new tools constructed here for genome editing, we used the pHsal-S vector for the generation of H. salinarum strains harboring chromosomal tags in genes of interest. As a proof of concept, we targeted the RNA chaperone Lsm encoded by the gene VNG1496G. We selected the FLAG-tag since it is a widely used and short (8 aa) tag peptide [29], with reduced chances of affecting the structure of the target protein. For the construction of the tagged strain, flanking regions of 500 bp upstream and downstream of the stop codon of the VNG1496G gene were PCR amplified and assembled into a 1.0 kb fragment by recombinant PCR. In this procedure, the homology primers of the recombinant PCR harbor the nucleotide sequence coding the FLAG-tag, allowing the insertion of the tag at the 3’-end of the gene. The final recombinant fragment was then cloned into pHsal-S vector and the resulting plasmid was used to transform H. salinarum. Recombinant strains with the plasmid integrated into the chromosome were selected by plating the transformation in selective media supplemented with mevinolin. After the appearance of colonies in the selective media, a single one was picked and inoculated into liquid media without selective pressure to allow the second event of recombination. After saturation of the liquid culture, cells were plated in solid media with 5-FOA to counter-select non-recombinant strains [13]. Colonies able to grow under these conditions were checked by PCR to verify if the tag was correctly inserted into the chromosome or if the strains were able to revert to the wild-type genotype (Fig 5). As shown in Fig 5B, we detected 7 out of 10 colonies that were positive for the correct incorporation of the FLAG-tag. This result highlights the applicability of the new suicide vector pHsal-S for the generation of stable genotypes in H. salinarum.

Halobacterium salinarum NRC-1 cells were grown in enriched complex media (CM) consisting of 25% NaCl, 2% MgSO₄.7H₂O, 0.2% KCl, 0.3% Na-citrate and 1% peptone, at 37°C under light and constant agitation of 225 rpm [31]. When required, the media was supplemented with 20 μg/mL of mevinolin (A.G. Scientific, San Diego, CA) or 5-fluoroorotic acid (5-FOA, 300 μg/mL). Cultivation and transformation of H. salinarum were performed according to standard protocols [31]. Escherichia coli DH5α cells were grown at 37°C with air shaking at 225 rpm in Luria-Bertani media (LB; 1% triptone, 0.5% yeast extract, 0.5% Na-chloride). When required, the media was supplemented with 100 μg/mL of carbenicilin (Sigma) to ensure plasmid retention. Synthetic DNA sequences were obtained from GeneArt (Life Technologies). All DNA manipulation techniques were performed according to standard protocols.

For the construction of the pHsal vector series, some initial vectors were constructed as starting points for assembling the modular vectors (Table 1). First, a ~1.8 kb fragment containing the bla gene and the ori ColE1 was PCR amplified using primers 5-pUC-BglII and 5-pUC-AscI that incorporate BglII and AscI restriction sites in the flanking regions. This fragment was digested and ligated to a 150 bp synthetic DNA fragment (GeneArt) containing the MCS of the system, generating the pRzero vector. Next, the pRzero vector was digested with the enzymes AscI and SpeI and ligated to a 1.2 kb long synthetic sequence of the optimized ura3 gene (named ura3-2.0, GeneArt) digested with the same enzymes, generating the vector pRU. In turn, the mev^R gene was PCR amplified from the plasmid pMTF1025GFP_CHA with primers 5-mev-SpeI and 3-mev-XbaI, digested with the enzymes SpeI and XbaI and cloned into a pRU vector previously digested with SpeI. From the resulting colonies, a recombinant vector that regenerated the single SpeI site at the 5’-end of the mev^R gene was named pHsal-S (Fig 3B), which is a suicide vector for the insertion of chromosomal modifications in H. salinarum. For the generation of the pHsal-S-Lsm::FLAG-tag vector, a 1.0 kb fragment containing the flanking regions of the Lsm gene (VNG1496G) was assembled by recombinant PCR using an upstream fragment amplified using primers 5-Lsm-EcoRI/3-flag and a downstream fragment, obtained with primers 5-flag/Pck2. The two fragments were joined using primers 5-Lsm-EcoRI /Pck2 and cloned into pHsal-S vector in EcoRI and HindIII restriction sites. The resulting recombinant vector was used to transform H. salinarum and generate the tagged version of Lsm as described previously [13].

For the construction of modular vectors able to replicate autonomously in H. salinarum, a ~1.6 kb fragment containing the pGRB1 replication origin [25] was PCR amplified from the pMTF1025GFP_CHA vector and cloned as a BglII/PacI fragment into the pRzero vector, previously digested with the same enzymes. The resulting plasmid was named pRO and used in the next steps. In order to create a sequence-edited version of the mev^R gene, three DNA fragments were PCR amplified using pMTF1025GFP_CHA vector as template with primers 5-mev-SpeI/3-mut1 (fragment 1), 5-mut1/3-mut2 (fragment 2) and 5-mut2/3-mev-AscI (fragment 3). These primers allow the elimination of an AscI and a SalI restriction sites present in the wild type sequence of the gene. Next, the three fragments were used in a PCR reaction with primers 5-mev-SpeI/3-mev-AscI to reconstruct the complete mev^R gene by overlapping PCR. The full 1.8 kb fragment was agarose gel purified and cloned using AscI and SpeI enzymes into the pRO vector, generating pHsal-C (Fig 3A). This vector is able to replicate autonomously in H. salinarum under the selection for mevinolin resistance.

For the construction of pHsal-E, we first generated and validated a sequence-edited version of the fer2 promoter of H. salinarum. For this, two fragments of the Pfer2 sequence were PCR amplified using primers 5-Pfer2-KpnI/3-Pfer2-M and 5-Pfer2-M/3-Pfer2, and assembled into a single fragment in a second round of PCR reaction with primers 5-Pfer2-KpnI/3-Pfer2. This ~100 bp fragment was digested with KpnI and cloned into pMTF1025GFP_CHA vector previously digested with KpnI/SmaI enzymes, generating vector pMTF1025GFP-Pzero. Once Pzero activity was validated in vivo, the synthetic promoter was PCR amplified using primers 5-Pzero-PacI/3-Pzero-AvrII and cloned as a PacI/AvrII fragment into pHsal-C, generating pHsal-E, which is an expression plasmid based on the strong Pzero promoter activity. Finally, for the construction of a GFP-based vector for promoter probing, a 0.8 kb fragment containing the GFP sequence was PCR amplified using pMTF1025GFP_CHA as template and primers 5-GFP-HindIII/3-GFP-SpeI. This fragment was cloned with enzymes HindIII/SpeI into pHsal-C, generating pHsal-GFP. All vector sequences are available at our website (http://labisismi.fmrp.usp.br/index.php/en/resen) and as S1 File. All plasmids are available upon request by e-mail.

To quantify promoter activity, single colonies of H. salinarum strains with different reporter plasmids were inoculated in 5 mL of liquid CM supplemented with 20 μg/mL mevinolin and incubated for 5 days. After pre-growth, cells were diluted to an OD₆₀₀ ~0.05 in fresh media and grown in standard conditions. At two specific time points (16 and 24 hours), samples (1.8 mL) were taken and the OD₆₀₀ measured. Sample cells were centrifuged for 5 min at 13,000 rpm and the supernatant was discarded. Cells were then resuspended in 1.8 mL of GFP assay buffer (10mM Tris-HCl, pH 7.5) and mixed vigorously to allow cell lysis. After lysis, cell debris were removed by centrifugation and samples were analyzed in a RF-5301PC fluorimeter (Shimadzu). GFP assays were performed using wavelengths of 488 nm for excitation and 510 nm for emission. Promoter activities were calculated by normalizing the measured fluorescence by the initial cell density (fluorescence/OD₆₀₀). A control strain of H. salinarum without the reporter plasmid was used to calculate the auto-fluorescence of the cells, and these background values were subtracted from the promoter activities.