SLAPTAG binds rapidly and efficiently to the Bio-matrix (BM)
As described in Materials and Methods the SLAPTAG was fused in-frame to the carboxy-terminus of the green fluorescent protein (GFP) to generate the chimeric protein (H6)-GFP-SLAPTAG adapted here as a SLAPTAG reporter protein. As shown in Fig. 1A the structure of the chimeric protein (H6)-GFP-SLAPTAG was predicted based on the alphaFold-2 method 16. The beta-barrel corresponding to GFP and the two globular domains corresponding to SLAPTAG were predicted with high performance and as expected, linkers and 6xHis-tag, which are flexible regions, showed low prediction values.
As a working chromatographic matrix, a culture of Bacillus subtilis natto was processed as described in the Methods section to generate a bacterial-derived affinity chromatography matrix, named here Bio-Matrix (BM). Interestingly, although B. subtilis has no S-layer, we have demonstrated previously that this bacterium can be externally covered with SLAP-tagged proteins to generate a recombinant S-layer on its bacterial cell wall, a process that we called decoration 15. Decoration of B. subtilis with SLAP-tagged proteins is possible because teichoic acid and lipoteichoic acid (which are the molecules responsible for the SLAPTAG association) have the same chemical composition as Lactobacillus acidophilus. Interestingly, in B. subtilis and L. acidophilus teichoic acid and lipoteichoic acid are distributed homogeneously on their cell wall.
To characterize the binding properties of the SLAPTAG, the reporter protein (H6)-GFP-SLAPTAG, was recombinantly expressed in Escherichia coli, purified by affinity chromatography mediated by its Hisx6 tag (H6), and further incubated with BM under a variety of conditions. Initially, as shown in Fig. 1B, the optimal binding temperature of (H6)-GFP-SLAPTAG to the BM was evaluated, finding that the best binding efficiency was reached when the incubation was performed at 4oC. To get insights into the SLAPTAG association dynamics, (H6)-GFP-SLAPTAG was incubated for 5, 30, and 60 minutes with BM and as shown in the Fig. 1C, the maximal binding of (H6)-GFP-SLAPTAG to BM was reached very rapidly (already at 5 minutes), showing no significant increments in its association at longer time points. These results indicated that the SLAPTAG binds very rapidly and with an apparent high affinity to the BM. To quantify the affinity of the interaction between the SLAPTAG with the BM, an adsorption isotherm was performed to determine the equilibrium apparent dissociation constant as described in the Methods section. As shown in Fig. 1D the apparent KD was estimated as 4.7 uM. Interestingly, this dissociation constant value was in the same order as other KD described for different microbial S-layer proteins with their respective bacterial walls 17–19. In addition, the maximum adsorption capacity (Bmax) of BM was estimated as 1.152 mmol of (H6)-GFP-SLAPTAG or 53.9 mg of protein per milliliter of BM or 1.03 g of SLAPTAG protein/g of dry BM.
Elution of (H6)-GFP-SLAPTAG protein can be performed with different buffers
As it was mentioned, the SLAPTAG contains the protein region responsible for the association of the protein SlpA to the L. acidophilus cell wall. As reported, L. acidophilus SlpA has the natural ability to self-assemble on the bacterial surface to generate a proteinaceous layer named S-layer 20, a highly ordered wall structure that functions as a protective barrier against bacteriophages, resistance to low pH and proteases, and bacterial adhesion. It is well characterized that remotion of the S-layer can be performed efficiently by the addition of chaotropic agents like LiCl, a compound that disrupts hydrogen bonds leading to a partial denaturation of proteins and the consequent detachment of the S-layer 20. Thus, initially, to study elution conditions for (H6)-GFP-SLAPTAG, LiCl solution was selected as the positive control for elution and as shown in Fig. 2A detachment of BM-associated (H6)-GFP-SLAPTAG was achieved with high efficiency with a 2M LiCl solution at room temperature, in a short lapse of five minutes (Fig. 2B). As it was mentioned above, lithium solutions have deleterious effects on proteins, which is an undesired effect for protein purification, especially for enzyme purification. Therefore, different milder alternatives of buffers were explored to elute SLAPTAG-tagged proteins, conditions that can preserve the protein structure and therefore their function. As shown in Fig. 2C, a series of sugars (monosaccharide and disaccharides) described previously 21 were tested for the elution of the reporter SLAP observing a partial elution efficiency when compared with the LiCl elution control (Fig. 2C). Next, the cationic nature of the SLAPTAG at neutral pH was considered. Taking into consideration the importance of this positive charge in the interaction of SLAPTAG with the membrane-bound compound teichoic acid, a set of cationic compounds was evaluated for the elution. As shown in Fig. 2C, 0.3M of CTAB was able to elute (H6)-GFP-SLAPTAG with similar efficiency as the LiCl solutions. Although CTAB was very efficient in this step, we found that this compound was difficult to remove downstream from the elution fraction. Then, considering that SLAPTAG has a theoretical isoelectric point (pI) value of 9.92, we explored if the modification of pH can be adapted as an elution method. As shown in Fig. 2C, when the pH of the bicarbonate buffer was close to the theoretical SLAPTAG isoelectric point (pI), (H6)-GFP-SLAPTAG was efficiently eluted from BM. In addition, after establishing the bicarbonate buffer at pH 10 as an elution buffer, we explore if the addition of NaCl can improve the elution process. As shown in Fig. 2C, the addition of 200 mM of NaCl was able to maximize the yield of recovery of (H6)-GFP-SLAPTAG in the eluate.
Optimized SAC protocol allowed the efficient purification of proteins with similar efficiency to the high-performance Ni 2+ -charged agarose matrix (IMAC).
With all the experimental information obtained, an optimized SAC protocol for the purification of SLAP-tagged proteins was established as shown in Fig. 3A. Remarkably, the entire process of protein purification can be performed in 15 minutes. To have a direct observation of the purification process of (H6)-GFP-SLAPTAG, the BM was fixed and immobilized on coverslips at different steps of the process to be observed by confocal microscopy. As shown in Fig. 3B, before the incubation of BM with (H6)-GFP-SLAPTAG, only a red native autofluorescence can be observed from fixed Bacillus subtilis natto cells present in BM. Remarkably, after the incubation of BM with (H6)-GFP-SLAPTAG, it was possible to detect the adhesion of the reporter protein to the BM forming a recombinant fluorescent S-layer that covers completely the bacterial surface (Fig. 3C). As expected, after the addition of the elution buffer the (H6)-GFP-SLAPTAG was completely removed from the bacterial surface (Fig. 3D). In addition, the whole process of purification can also be monitored by direct observation under UV light (Fig. 3E)
Since SAC was efficient in protein purification, we compared this new technique against an established affinity purification protocol. For this, we took advantage of (H6)-GFP-SLAPTAG also has a His-tag that can be purified by a Metal-Affinity-Chromatography or IMAC. As described in Materials and Methods, a bacterial lysate (H6)-GFP-SLAPTAG was divided into two fractions and both protocols were performed accordingly, confirming that SAC optimized protocol was able to purify (H6)-GFP-SLAPTAG with high efficiency (Table Suppl. 1 and Fig. 3F) and with a similar yield to the one obtained with the commercial IMAC (Fig. 3G).
Proteins of different biological sources, molecular weights, biochemical functions, or expressed by procaryotic or eukaryotic expression systems can be efficiently purified by the SAC.
To evaluate if SLAPTAG and SAC can be adapted as a universal protein purification system, a set of the selected proteins were fused to the SLAPTAG (Fig. 4).
Thus, bacterial proteins like Salmonella flagellin (Fig. 5A), the STEC proteins, EspA (Fig. 5B), the chimeric EIT (Fig. 5C), the B. abortus Omp19 (Fig. 5D), the bacteriophage protein T7-lysozyme (Fig. 5E), the human viral proteins Rhinovirus 3C Protease (Fig. 5E), the mouse Galectin-8 (Fig. 5G) and the commercial molecular tag GST (Fig. 5H), were fused in frame with the SLAPTAG.
As shown in Fig. 5, all the selected proteins were purified efficiently with SAC. In addition to the bacterial expression system, different protein expressions systems (yeast and mammalian cells) were also evaluated (Fig. 6). As shown in Fig. 6A the viral HRV 3C protease fused to the SLAPTAG was expressed and purified from Pichia pastoris supernatant. Also, the SARS-CoV2 SPIKE fused to the SLAPTAG was purified from the supernatant of transfected HEK293 (Fig. 6B). These results confirmed that the SLAPTAG can be widely adopted for affinity chromatography purification.
BM is a reusable chromatography matrix with long-term stability.
As described in Materials and Methods a batch of BM was produced, and several aliquots were frozen at -20oC to study time stability. As shown in Fig. 7A, at different times, a few aliquots were unfrozen and tested for protein purification of the reporter protein (H6)-GFP-SLAPTAG being the BM stable for more than a year (14 months). In addition, BM reusability capacity was evaluated determining that the matrix can be reused five times with no modification of the protein purification yield (Fig. 7B).
Use of the Human Rhinovirus 3C protease fused to the SLAP TAG (SLAPASE) to release the SLAP-tagged proteins from BM.
As shown in Figs. 4 and 5 the gene sequence of the viral protease HRV-3C was fused to the SLAPTAG (named SLAPASE), recombinantly expressed, and purified (Fig. 8B and 8D, lane SLAPase) to evaluate its activity. A reporter protein for SLAPASE activity was generated (GFP-LEVLFQGP-SLAPTAG) (Table 1) and a protocol for tag-remotion by SLAPase was set as described in the Methods section. As shown in Fig. 8A the GFP-LEVLFQGP-SLAPTAG was expressed (Fig. 8B and Fig. 8C, lane Input or IN) and mixed with the BM for 5 minutes allowing the binding process. After binding, a purified SLAPASE was added to the mix and incubated for 60 minutes releasing a tag-less GFP (Fig. 8B and Fig. 8C, Fig. 8E, lanes flow-through or FT) by proteolysis. Cut SLAPTAG is not observed in the flow-through indicating that was retained by the BM along with the SLAPase. To confirm these steps, post-proteolysis BM was incubated with LiCl solution to remove any residual bound SLAP-tagged protein. As shown in Fig. 8B and 8D (lane Elution or E), the LiCl solution released the SLAPase and the cut SLAPTAG.
Table 1
Strain | | |
E. coli DH5α | F − φ80lacZΔM15 Δ(lacZYA-argF) U169 recA1 endA1 hsdR17(rK−, mK+) phoA supE44λ- thi-1 gyrA96 relA1 | Invitrogen |
E. coli BL21 Codon plus | [ompT hsdS(rB– mB–) dcm + Tcr gal λ (DE3) endA Hte Cmr] | Stratagen |
B. subtilis | Wild type strain var. natto | ATCC 15245 |
Pichia pastoris | P. pastoris strain GS115 | BioRad, USA |
Plasmid | | |
pET28-eGFP-SlpA | A vector containing the GFP gene from Aequorea victoria fused to SLAPTAG | 21 |
pET22-eGFP-SlpA | A vector containing the GFP gene from Aequorea victoria fused to LEVLFQGP sequence and SLAPTAG | This study |
pGEX-SlpA | A vector containing GST form Shitosoma japonicum fused to SLAPTAG | 15 |
pLC3-EITH7-SlpA | A vector containing EspA-Intimin-Tir fused genes from STEC E. coli fused to SLAPTAG | 15 |
pLC3-Omp19-SlpA | A vector containing the Omp19 gene from Brucella abortus fused to SLAPTAG | 15 |
pLC3-FliC-SlpA | A vector containing the FliC gene from Salmonella enterica fused to SLAPTAG | 15 |
pLC3-Gal8-SlpA | A vector containing the Gal8 gene from Mus musculus fused to SLAPTAG | 15 |
pET28-Lys-SlpA | A vector containing the Lysozyme gene from Gallus gallus fused to SLAPTAG | This study |
pMAL-c5X-HRV3c-SlpA | A vector containing human rhinovirus (HRV) type 14 3C protease gene fused to SLAPTAG (SLAPase) | This study |
pGEX-2T-EspA-SlpA | A vector containing the EspA gene from STEC E. coli fused to SLAPTAG | This study |
pPICZalphaB-HRV3c-SlpA | A vector containing human rhinovirus (HRV) type 14 3C protease gene fused to SLAPTAG (SLAPase) | This study |
Addgene #154754 | A vector containing the sequence of SARS CoV2 Spike ectodomain Hexa-pro | 27 |
pSpike-SLAPTAG | Plasmid Addgene #154754 fused with the SLAPTAG | This study |
pLysS | Vector for expression of T7 lysozyme. | Novagene |
Adaptation of SAC to magnetic affinity chromatography.
Since we confirmed the efficiency and universality of SAC for the purification of recombinant proteins, we explored a magnetic affinity chromatography SAC alternative (Fig. 9A). As shown in Fig. 9B, 9C and Supplementary Video-1 the (H6)-GFP-SLAPTAG was purified in a few and easy steps. As shown in the SDS-PAGE (Fig. 9B) the adapted SAC protocol for magnetic chromatography (BMmag) was able to purify the reporter protein (H6)-GFP-SLAPTAG very efficiently (Fig. 9B, lane Elution or E). Interestingly, this protocol can be adapted to commercial devices that use magnetic force for protein or DNA purification (like King Fisher Flex from ThermoFisher). In Fig. 9C the entire process of purification was monitored by direct observation under UV light (Supp. video 1 ). Noteworthy, the binding of the BMmag to GFP quenched the fluorescence of this protein, an effect that was described for transition metal binding to GFP or binding of iron cations to fluorescent proteins (Fig. 9C, tube IN + BMmag) 22,23.