All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, United States). Water was desalted and purified using a milli-Q^® (Merck, Darmstadt, Germany) system. This study used Haloferax volcanii H1895 and the vector pTA1992 (Haque et al., 2019).

The purification-tag-fusion-sfGFP protein synthetic gene library (GenScript Biotech Corporation, HK), together with the mCherry and ADH constructs were codon-optimized using the java codon adaptation online tool Jcat for Halobacterium sp. (strain NRC-1/ATCC 700922/JCM 11081) (Grote et al., 2005) and cloned into the pTA1921 vector via NdeI and BamHI restriction enzymes (Supplementary Figure S1 and Supplementary Table S1).

The construct containing vectors was transformed into the Haloferax volcanii H1895 using the PEG/EDTA method (Dyall-Smith, 2008). H. volcanii and derivatives were cultured in the Hv-YPC medium at 45°C with shaking for 24–36 h, as previously described (Strillinger et al., 2016; Grötzinger et al., 2018; Akal et al., 2019; Karan et al., 2020; Alshehri et al., 2022). For solid media, 2% (w/v) agar was added. Stock cultures were maintained in glycerol at −80°C. For short-term use, cultures were maintained on stock plates.

The expression level of the sfGFP or mCherry constructs was evaluated by measuring the fluorescence (ex 485 nm/em 507 nm) or mCherry (ex 587 nm/em 610 nm) of 24 h grown cell culture/OD₆₀₀ with three different colonies. The expression level of the ADH constructs was evaluated by determining the final protein concentration after purification using NanoDrop absorption at 280 nm. The results were transformed into relative expression, with the highest fluorescence (for sfGFP N-ter Strep-tag, mCherry C-ter 8xHis-tag, and ADH N-ter 8xHis-tag) set to 100%.

The solubility of sfGFP and mCherry was evaluated by comparing the fluorescence after sonication, before and after centrifugation. For ADH, which lacks fluorescence, solubility was evaluated using SDS-PAGE and Western blotting to compare the soluble fraction against the pellet.

One liter of cell culture was grown until an OD₆₀₀ of 1.0–1.5. Cells were harvested by centrifugation (4,000 × g, 4°C, 45 min) in a Legend XFR Centrifuge (Thermo Scientific, Waltham, United States) and disrupted in binding buffer (50 mM sodium phosphate buffer, pH 7.4 containing 200 mM or 2 M NaCl) containing Pierce^™ Protease Inhibitor EDTA-free tablet (Thermo Scientific) using a sonicator (Model Q500, QSONICA, Newtown, CT, United States) with a 1.9 cm probe. Cell debris was removed by centrifugation (24,000 × g, 4°C, 45 min) in a Multifuge X1R Centrifuge (Thermo Scientific).

All constructs were purified using the ÄKTAprime plus chromatography system (GE Healthcare Life Sciences, Piscataway, NJ, United States) according to the purification tag in the binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM or 2 M NaCl).

The polyHis and SUMO fused constructs were purified on a 1 mL HiTrap Ni²⁺ chelating column (GE Healthcare Life Sciences, Piscataway, NJ, United States). The supernatant was loaded at a 1.0 mL/min flow rate onto the column pre-equilibrated with a binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM, or 2 M NaCl) containing 20 mM imidazole. After washing the column with binding buffer, the protein was eluted in one step with His-elution buffer (binding buffer containing 250 mM imidazole).

The Strep-tag^®II and Twin-Strep-tag^® tag fused constructs were purified on a 1 mL Strep-Tactin^®XT 4Flow (GenScript Biotech Corporation, HK). The supernatant was loaded at a 1.0 mL/min flow rate onto the column pre-equilibrated with binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM, or 2 M NaCl). After washing the column with binding buffer, the protein was eluted in one step with a strep-elution buffer (binding buffer containing 50 mM biotin).

The FLAG-tag and 3X FLAG-tag fused constructs were purified on a 1 mL Anti-DYKDDDDK G1 Affinity Resin column (GenScript Biotech Corporation, HK). The supernatant was loaded at a 1.0 mL/min flow rate onto the column pre-equilibrated with binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM, or 2 M NaCl). After washing the column with binding buffer, the protein was eluted in a one-step FLAG-elution buffer (100 mM Tris/HCl buffer, pH 12.0, 200 mM, or 2 M NaCl). To purify the FLAG/3xFLAG-tag fused constructs, MonoRab^™ Anti-DYKDDDDK Magnetic Beads (Cat No. L00835, GenScript) were tested as well. The cell lysate was added to the pre-washed beads with bead binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM or 2 M NaCl) and incubated on a shaker at 4°C for 2 h. After magnetic separation, the supernatant was removed, and the magnetic beads were washed with the bead binding buffer three times. Finally, the DYKDDDDK fused construct bound to the beads was eluted by bead elution buffer (100 mM Tris/HCl buffer, pH 12.0).

The C-tag fused constructs were purified on a 1 mL CaptureSelect^™ C-tagXL column (Thermo Scientific, Waltham, United States). The supernatant was loaded at a 1.0 mL/min flow rate onto the column pre-equilibrated with binding buffer (50 mM sodium phosphate, pH 7.4, 200 mM, or 2 M NaCl). The protein was eluted in one step with a C-tag-elution buffer (50 mM sodium phosphate, pH 7.4, 2 M MgCl₂).

The purified fractions were combined, further purified, and concentrated with Amicon^® Ultra-4 Centrifugal Filter Units, 10 kDa. The protein was then dialyzed against dialysis buffer (20 mM sodium phosphate, pH 7.4, 200 mM, or 2 M NaCl). Protein concentration was determined using NanoDrop absorption at 280 nm.

Note: The purification procedures of each construct were not optimized (e.g., elution gradient, wash steps) to compare better the purity of sfGFP, respectively, mCherry and ADH, constructs for each purification tag.

For fusion constructs containing cleavable tags, the cleavage of the purification tags was performed overnight at 4°C in dialysis buffer (50 mM Tris-HCl, 1 mM DTT, pH 8.0) containing 200 mM NaCl or 2 M NaCl, with the respective SUMO-Protease, containing a His-tag, (ratio of 1:10 w/w) and TEV-Protease, containing a His-tag, (200 U for 1 mg of fusion protein), and then reapplied to the respective affinity column to remove the cleaved purification tag. The SUMO- and TEV-Protease were removed by the HiTrap Ni²⁺ chelating column (GE Healthcare Life Sciences, Piscataway, NJ, United States).

The FLAG tag contains an internal enterokinase (EK) cleavage site, which allows the removal of the affinity tag after purification. The eluted fractions were incubated with Enterokinase, containing a His-tag (Z03376-1, Genscript, 200 U for 1 mg of fusion protein) to remove the FLAG-tag. The Enterokinase was removed by the HiTrap Ni²⁺ chelating column (GE Healthcare Life Sciences, Piscataway, NJ, United States).

Western blotting analysis was performed as previously described (DasSarma et al., 2013; Andar et al., 2017; Karan et al., 2020). The SDS-PAGE analysis was performed using Novex^® Tris-glycine gels (4–20%, Invitrogen, Carlsbad, CA, United States). Proteins were then transferred to 0.45 μm nitrocellulose membranes (Millipore Corp., Boston, MA). Membranes were blocked for 15 min in Pierce Fast Blocking Buffer (Thermo Fisher Scientific) and then incubated in Anti-GFP Polyclonal Antibody (Thermo Fisher Scientific) at 1:2000 dilution, followed by three washing steps with TBST, then incubated with Goat anti-Rabbit IgG (H + L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 488 (Invitrogen, Catalog # A-11008).

The purity of the proteins was evaluated using SDS-PAGE stained with Coomassie brilliant blue and quantified with ImageJ software¹ (Schneider et al., 2012). Purity was specified as a percentage of the total protein content, with the intensity of the target protein band compared to the total protein intensity on each lane of the gel.

The identification of corresponding peptides was performed by LC-MS/MS analysis. 10 μg of the sample was digested with trypsin using the FASP protocol (Wiśniewski et al., 2009). Peptides were measured using an LTQ-Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, United States) and analyzed by MASCOT v2.3 (Matrix Sciences Ltd., United Kingdom).

In the study, H. volcanii strain H1895 as the host, in combination with the pTA1992 plasmid that features the exceptionally strong constitutive p.syn promoter (Haque et al., 2019, 2020) was employed. Next, a versatile purification-tag-fusion-protein library using the sfGFP protein as a reporter (Pédelacq et al., 2006), was designed and constructed. Within this library, a range of purification tags, including 6xHis, 8xHis, Strep-tag^®II, Twin-Strep-tag^®, FLAG, and 3x-FLAG, strategically placed at both the N-terminus and C-terminus of sfGFP (Figure 1) were introduced.

His-tags are the most used and cost-effective affinity tags to purify recombinant proteins using an immobilized metal affinity chromatography (IMAC) column (Porath et al., 1975). Notably, protein activity is rarely affected by polyhistidine affinity tags because of their relatively small size and charge (Block et al., 2009). In the context of H. volcanii strains where naturally histidine-rich proteins PitA and Cdc48d have been replaced with orthologous genes containing fewer histidine residue, the 6xHis-tag has been widely adopted for recombinant protein purification (Grötzinger et al., 2018; Haque et al., 2019, 2020; Karan et al., 2020). The Small Ubiquitin-like Modifier (SUMO)-tag, which incorporated either a 6xHis-tag or an 8xHis-tag for expression and purification (Butt et al., 2005) was used for the library design. Strep-tag^®II, an 8-amino acid peptide tag (WSHPQFEK), exhibits a strong affinity for Strep-Tactin^® (or Strep-Tactin^®XT), a specially engineered streptavidin under physiological buffer conditions (Skerra and Schmidt, 1999; Schmidt et al., 2013; Yeliseev et al., 2017). This interaction allows for the rapid one-step purification of nearly any recombinant protein at high purity and functionality. Moreover, the Strep-tag^®II can be found as a repetitive motif, generating a Twin-Strep-tag^®. The Twin-Strep-tag^®, characterized by its relatively higher affinity for Strep-Tactin^® matrices, offers enhanced purification capabilities for recombinant proteins.

The FLAG-tag is a small, highly soluble tag consisting of a hydrophilic octapeptide epitope (DYKDDDDK) with an enterokinase-cleavage site (DDDDK). This tag’s hydrophilic nature positions it primarily on the surface of the recombinant protein, increasing accessibility to resins and minimizing potential adverse effects on the fusion protein (Hopp et al., 1988; Schmidt et al., 2012). Similarly, the 3xFLAG-tag is a robust epitope tag (DYKDHDG-DYKDHDI-DYKDDDDK) was fused with sfGFP at the N- and C-termini.

A new class of purification tags, the four-peptide (E-P-E-A) short C-tag, which exhibits a strong affinity for a camelid antibody fragment when expressed at the C-terminus of a protein was also explored (Djender et al., 2014).

We also scrutinized the effect of positioning the TEV cleavage site at the N-terminus and C-terminus of the sfGFP fusion constructs (Supplementary Table S1 and Supplementary Figure S1). A total of 26 distinct sfGFP constructs were studied, each evaluated for expression through fluorescence measurements (Supplementary Figure S2), further corroborated by Western blot analysis (Supplementary Figure S3). The different sfGFP constructs were compared for their relative expression level, solubility, and purification profile (Supplementary Figure S4). Tryptic digest analysis was additionally performed to confirm the identity of the fusion proteins (Supplementary Figure S5). The focus was primarily on purification at low salt (200 mM NaCl) and high salt concentrations (2 M NaCl). This comprehensive evaluation aimed to determine the most suitable purification tag for the native purification of extremophilic halophilic proteins, capitalizing on the suitability of H. volcanii as the expression host.

The initial systematic screening library was expanded to examine the applicability of the sfGFP results to the red fluorescent protein mCherry (Shaner et al., 2004) and a previously fully characterized halophilic alcohol dehydrogenase (ADH) from the deep-sea brine pool of the Red Sea (Grötzinger et al., 2018) (Supplementary Figure S7). These investigations revealed the significant impact of different tags, tag lengths, and positions on protein expression and purification. This resulted in a total of 37 screened constructs.

The corresponding overview of the expression and purification of the sfGFP constructs is presented in the order of the different studied affinity purification tag systems.

The experimental design explored the efficacy of both a 6xHis-tag and an 8xHis-tag at the N- and C-termini of sfGFP, both with and without a TEV cleavage site. Furthermore, the N-terminal solubility-enhancing and cleavable Small Ubiquitin-like Modifier (SUMO)-tag, which incorporated either a 6xHis-tag or an 8xHis-tag for expression and purification (Butt et al., 2005), was used. All the constructs underwent purification and cleavage, where applicable, under varying salt concentrations, ranging from low (200 mM NaCl) to high (2 M NaCl) in a 50 mM sodium phosphate buffer, pH 7.4. The expression level, solubility, purity, and final yield of the fusion sfGFP constructs are presented in Table 1 and Figure 1.

The expression of sfGFP is significantly influenced by the choice of tag and its location (Yilmaz and Arslanyolu, 2015). Notably, the highest level of sfGFP expression was achieved with the N-terminal 8xHis-tag (Table 1). Surprisingly, the introduction of the N-terminal 8xHis-SUMO-tag did not significantly enhance expression yield, and the addition of a cleavable TEV cleavage site had a noticeable impact on expression levels. Conversely, sfGFP expression with C-terminal His-tags demonstrated significantly lower levels, up to one-fifth, compared to the N-terminal 8xHis-tag. While the expression level of sfGFP varied regarding the position and length of the tag, it’s noteworthy that these variations did not substantially impact solubility. All constructs exhibited consistently high solubility levels ranging from 92 to 100%, irrespective of salt concentration. The purity of the sfGFP fusion proteins displayed notable variation concerning His-tag length, position, and salt concentration in the purification buffer. The 8xHis-tag consistently outperformed all other His-tag constructs, resulting in high purity and yield in high and low salt buffers. This outcome aligns with expectations, as the increased number of histidine residues in the tag leads to more efficient binding on IMAC columns. Furthermore, it allows for more rigorous washes, enhancing the overall purity of the protein.

A comparative assessment of the solubility-enhancing and cleavable SUMO-tag versus the TEV cleavage site yielded valuable insights into the performance of these constructs. The SUMO-tag exhibited robust performance, outperforming the cleavable TEV cleavage site. The SUMO-tag was pivotal in enhancing sfGFP expression and ensuring high purity, irrespective of the buffering composition. sfGFP constructs with the TEV cleavage site also displayed strong expression, achieving even higher purity levels. However, the TEV cleavage site did not cleave efficiently at high salt concentrations (Figure 2). This limitation resulted in approximately 40–50% final yields compared to the uncleaved constructs. Of note, the sfGFP construct featuring a C-terminal 6xHis-tag with a TEV cleavage site exhibited the lowest production level among all the tested His-tagged constructs.

Overall, the N-terminal 8xHis-tag proved the best performer, offering a simple, rapid, and highly efficient one-step purification method for sfGFP. This approach employed cost-effective materials and delivered approximately double the final yield compared to the 6xHis-tag and nearly five times the yield compared to the C-terminal 6xHis- or 8xHis-tag constructs (Table 1).

The 6xHis-tag remains the most favored purification tag for H. volcanii due to its compatibility with high-salt media (Haque et al., 2019). However, the results, which contradict earlier studies on N-terminal His-tagged GFP that indicated adverse effects on protein production, emphasize the ongoing challenge of selecting the optimal position and length for the His-tag within various expression systems (Nguyen et al., 2014; Phan et al., 2015; Tian et al., 2019; Le et al., 2022). This study underscores the importance of making informed choices regarding His-tag design to maximize expression levels and purification efficiency.

Next, the sfGFP constructs fused with Strep-tag^®II and Twin-Strep-tag^® were investigated.

The results of the sfGFP constructs fused with Strep-tag^®II are summarized in Table 2 and Figure 3. Expression levels of sfGFP varied based on the position of the Strep-tag^®II and Twin-Strep-tag^®. Like the His-tag, the N-terminal Strep-tag^®II yielded the highest sfGFP expression, surpassing the C-terminal Strep-tag^®II or Twin-Strep-tag^® by up to fivefold. Notably, while C-terminal Strep-tag^®II has been utilized for protein production in H. volcanii and proven compatible with high salt concentrations, the study reveals that N-terminal Strep-tag^®II results in approximately five times higher expression, suggesting that C-terminal usage in H. volcanii should be reconsidered (Braun et al., 2019). Intriguingly, Twin-Strep-tag^®, which generally exhibits a higher affinity for Strep-Tactin^® and greater tolerance to salts and detergents in buffers than Strep-tag^®II (Yeliseev et al., 2017), displayed an unexpected reduction in sfGFP expression to about one-third when compared to N-terminal Strep-tag^®II. The introduction of Strep-tags (Strep-tag^®II or Twin-Strep-tag^®) appeared to have no significant effect on the solubility of the fusion proteins, regardless of salt conditions. Furthermore, these tags enabled high-purity purification processes (Table 2 and Figure 2).

Next, experiments to assess the removal of Strep-tags using protease cleavage sites (TEV and SUMO) positioned between the sfGFP protein and the Strep-tags were conducted. Subsequently, a purification step was employed to separate the sfGFP protein from the cleaved Strep-tags. Consequently, the findings indicate that both Strep-tag^®II and Twin-Strep-tag^®, when fused with SUMO or TEV cleavage sites, could be efficiently removed through protease cleavage (SUMO- or TEV-Protease).

Among all the tested tags fused to the N-terminus of sfGFP, Strep-tag^®II consistently yielded the highest levels of sfGFP expression.

Next, the expression and solubility of sfGFP fused with the FLAG-tag and 3xFLAG-tag, in combination with the SUMO-tag or TEV cleavage site (Einhauer and Jungbauer, 2001; Schmidt et al., 2012; Beznosov et al., 2013) were assessed.

Comparable to His and Strep-tagged sfGFP findings, N-terminal FLAG, and 3xFLAG-tags exhibited higher expression levels (50 and 75%, respectively) than C-terminal tags (Table 3). Interestingly, N-terminal 3xFLAG-fused sfGFP demonstrated expression levels 1.5 to 3 times superior to N-terminal FLAG-tag, C-terminal FLAG, or 3xFLAG-tagged sfGFP. Additionally, all constructs displayed high solubility, exceeding 90%, regardless of whether the FLAG or 3xFLAG was positioned at the N-terminus or C-terminus of the sfGFP protein.

The Enterokinase efficiently removed the FLAG-tag even at higher salt concentrations, which led to a relatively lower yield (Figure 4). All FLAG-tagged expressions resulted in highly pure sfGFP with no visible contaminating proteins on SDS-PAGE (Figure 4). To purify FLAG and 3xFLAG-tagged sfGFP from sonicated H. volcanii cell culture, the anti-DYKDDDDK magnetic beads (GenScript) were also tested. Remarkably, we were able to streamline the purification process by eliminating centrifugation or filtration steps and still achieved high purification (data not shown).

While FLAG-tags have previously been employed to purify recombinant proteins in various expression systems, including bacteria, baculovirus, mammalian cells, and yeast, there has been limited information regarding their use in halophilic systems (Einhauer and Jungbauer, 2001; Mishra, 2020). FLAG peptide has been tagged with modified archaellin genes (flaA1, flaA2, and flaB2) in haloarchaea Halobacterium salinarum cells and utilized for the identification of proteins within the archaellum (Beznosov et al., 2013). However, no reports existed on FLAG-tagged protein expression and purification in the halophilic context until this study. These findings suggest that the FLAG and 3xFLAG-tags suit protein expression and purification in halophilic systems. The FLAG-tag carries several advantages over His and Strep tags. Like the 6xHis-, 8xHis-, and Strep-tag^®II, the FLAG-tag comprises a small hydrophilic peptide (8 amino acids), unlikely to impact protein folding or function. Furthermore, it can be easily removed with Enterokinase. However, it is worth noting that Enterokinase, while versatile, is sensitive to high salt concentrations compared to other proteases like the SUMO-Protease, limiting its application in the extremophilic environment (Shahravan et al., 2008).

Next, the C-tag constructs were examined, and the results detailing the expression level, solubility, purity, and final yield of the fusion sfGFP constructs are outlined in Table 4 and Figure 5. The C-tag-binding resin facilitated the purification of pure proteins in both low-salt and high-salt buffers. Notably, under the 2 M NaCl purification buffer, the protein achieved high purity at 98%, albeit with a lower yield. However, for the 200 mM NaCl purification buffer, the purity of sfGFP could be significantly improved by column washing with 1 M NaCl (Supplementary Figure S6). Moreover, including a TEV cleavage site aided in removing the C-tag, resulting in higher purity, although this came at the expense of yield, particularly in high-salt conditions due to the reduced activity of TEV-Protease in high-salt environments (Parks et al., 1995; Waugh, 2011).

Among all the affinity tags tested, the smallest among them, the C-tag comprising just four amino acids, proved to be a rapid and efficient tool for purification. The C-tag offers several advantages over more established tags such as His, Strep, and FLAG. It is the smallest affinity purification tag with high binding affinity and selectivity. Furthermore, its limited impact on protein folding and functionality due to its small size makes it an attractive choice for applications in protein purification (Hermans et al., 2012; Djender et al., 2014; Jin et al., 2017).

The systematic screening of the different purification tags and tag positions with sfGFP revealed a significant enhancement in protein production, purity, and yield in H. volcanii when an 8xHis-tag or Strep-tag^®II was placed at the N-terminus of sfGFP. This enhancement is potentially linked to favorable folding kinetics of newly synthesized proteins and increased accessibility of these tags during affinity purification, which may facilitate higher yield and purity. To explore whether this trend was unique to the expression system or exclusive to sfGFP, the investigation was extended to include the red fluorescent protein mCherry and a previously fully characterized halophilic alcohol dehydrogenase (ADH) from the deep-sea brine pool of the Red Sea (Shaner et al., 2004; Grötzinger et al., 2018).

The 8xHis-tag and Strep-tag^®II were both positioned at the N- and C-termini to investigate the impact of tag position on mCherry and the ADH (Supplementary Table S2 and Supplementary Figure S1). Surprisingly, we discovered that mCherry demonstrated improved expression levels when the 8xHis-tag, or Strep-tag^®II, was located at the C-terminus of the protein. Conversely, ADH exhibited more efficient expression when the 8xHis-tag or Strep-tag^®II was situated at the N-terminus of the respective proteins (Table 5 and Figure 6). Although these proteins shared similar relative solubility levels, they exhibited divergent expression patterns, emphasizing the nuanced nature of tag effects in different protein contexts.

These findings underscore the critical role of purification tags, their positions, and lengths in determining fusion protein expression, purification, and yields. Moreover, this observation emphasizes the necessity of exploring various fusion strategies to optimize expression and purification efficiencies. We considered applying a dual-affinity-tag approach to create a more general purification tag for H. volcanii.

Assessing different tag combinations, positions, and lengths for a novel protein can be resource-intensive and time-consuming (Bernier et al., 2018). To address this challenge, we applied a dual-affinity-tag approach by combining 8xHis-tag and Twin-Strep-tag^® tags at the N-terminus of sfGFP, mCherry, and ADH (Supplementary Table S3 and Supplementary Figure S1). In the purification process first a HiTrap Ni²⁺ chelating column was used to exploit the 8xHis-tag, followed by a Strep-Tactin^®XT 4Flow column to utilize the Twin-Strep-tag^®, achieving high-purity protein. Surprisingly, this dual-affinity-tag approach consistently outperformed all previously tested tag configurations regarding expression level and purification yields (Table 6 and Figure 7). Moreover, the dual-affinity-tag seems to be a universal solution for protein expression and purification in H. volcanii since all tested proteins were notably expressed in higher amounts and yielded higher overall protein content.

The dual-affinity-tag approach is a method that streamlines the purification of recombinant proteins, yielding homogeneous preparations of the proteins of interest in H. volcanii. This is achieved by enabling two consecutive affinity chromatography steps, efficiently removing contaminating proteins.