A Simplified Strategy for Nanobody Production and Use Based on Functional GST-Nanobody Fusion Proteins

TL;DR

This paper introduces a simplified method for producing and using nanobodies by fusing them with GST, enabling efficient purification and detection without requiring complex modifications.

Contribution

A novel GST-nanobody fusion strategy is presented for cytoplasmic expression in E. coli, enabling functional nanobody production and detection.

Findings

GST-nanobody fusions were produced at ~8–12 mg/L in E. coli and retained nanomolar binding affinity to their antigens.

The GST tag enabled indirect immunofluorescence and pulldown assays without tag removal.

The method complements existing nanobody designs and provides a low-cost alternative for functional reagent production.

Abstract

Nanobodies (VHHs or single-domain antibodies) are powerful affinity reagents, but their routine use is often limited by production constraints and by the lack of a conserved Fc region for secondary detection. We describe a simplified strategy in which functional GST–nanobody fusion proteins are expressed directly in the cytoplasm of Escherichia coli OrigamiTM 2 (DE3), a strain that supports disulfide bond formation through trxB/gor mutations. Using well-characterized nanobodies against GFP (Lag2) and mCherry (C11), we designed N-terminal GST fusions and confirmed by AlphaFold3-based modeling that both constructs preserve the GST fold and the VHH (Variable domain of the Heavy-chain antibody of Heavy-chain-only antibodies) β-sandwich with defined CDR loops and a predicted intradomain disulfide bond. Following IPTG induction and purification by glutathione affinity and size-exclusion chromatography, we obtained soluble GST-nb-GFP and GST-nb-mCherry at ~8–12 mg/L. Isothermal titration calorimetry showed nanomolar binding to their antigens (Kd ~123 nM for GFP and ~199 nM for mCherry). Consistent with conformational epitope recognition, GST-nanobodies were reactive in native-state dot blots but not in denaturing Western blots under the conditions tested. The GST moiety enabled indirect immunofluorescence via anti-GST antibodies, yielding specific labeling of GFP- or mCherry-tagged TGN38 in HeLa and H4 cells. Finally, we demonstrate “GST-nanobody pulldown” as a robust method for affinity capture from cell lysates. Together, this platform provides a low-cost, versatile route to functional nanobody reagents without requiring tag removal, and complements other nanobody designs (e.g., VHH-Fc fusions) in an application-dependent manner.

Figures

Full text

1. Introduction

Single-domain antibodies derived from camelid heavy-chain antibodies, commonly referred as nanobodies or VHHs (Variable domain of the Heavy-chain antibody of Heavy-chain-only antibodies), have emerged as powerful molecular tools in biological research and biotechnology. Because nanobodies consist of a single immunoglobulin variable domain, they combine high antigen specificity with exceptional stability, solubility, and ease of genetic manipulation, enabling their expression in heterologous systems and fusion to a wide variety of functional modules such as fluorescent proteins, enzymes, or affinity tags [1,2,3]. These properties make nanobodies particularly attractive alternatives to conventional IgG antibodies, which require the coordinated expression and assembly of heavy and light chains and typically depend on eukaryotic secretion pathways for proper folding and glycosylation. In addition to their favorable biochemical properties, nanobodies frequently recognize conformational epitopes that reflect the native three-dimensional structure of their target proteins [4]. This feature makes them especially suitable for applications such as immunofluorescence microscopy, immunoprecipitation, flow cytometry, and native-state blotting. Still, it also means that many nanobodies fail to recognize their targets after denaturation in SDS-PAGE and Western blotting. Consequently, nanobody-based detection strategies are optimally deployed in assays that preserve protein structure, including dot blotting, affinity-capture, and microscopy of fixed or living cells [2,5,6,7].

Despite these advantages, the routine adoption of nanobodies in research laboratories remains limited by practical constraints in their production and detection. Most recombinant nanobodies are expressed in the periplasm of Escherichia coli or secreted from eukaryotic cells to allow correct disulfide-bond formation and folding [8,9]. Although effective, these approaches introduce additional complexity and frequently require signal peptides, specialized strains, or low yields. Moreover, nanobodies do not contain a conserved constant region analogous to the Fc domain of IgG molecules, precluding the use of generic secondary antibodies for signal amplification in microscopy and blotting assays. To circumvent this limitation, nanobodies are often fused to Fc domains or reporter modules, which can be advantageous (e.g., bivalency and Protein A/G compatibility) but may add cloning and production complexity depending on the intended use [10,11,12,13]. An attractive alternative is to fuse nanobodies to Schistosoma japonicum glutathione S-transferase (GST), a widely used affinity and solubility tag that enables high-level cytosolic expression in bacteria and robust purification on glutathione matrices [14,15]. GST also provides a large, well-folded domain that can be recognized by commercially available anti-GST antibodies, thereby offering a universal handle for detection analogous to secondary antibodies in classical immunofluorescence and immunoblotting. Importantly, GST-VHH and VHH-Fc formats are complementary rather than universally superior, and the choice depends on the experimental context. Other soluble fusion-tag strategies (e.g., MBP- or SUMO-fused VHHs) have also been used to improve bacterial production of nanobodies [16], underscoring that tag choice can differentially impact yield, solubility, and antigen-binding performance and is often application-dependent [17]. However, whether nanobodies remain properly folded, functional, and quantitatively useful when expressed as cytosolic GST fusions has not been systematically addressed.

The formation of the conserved disulfide bond within VHH domains is a critical determinant of nanobody stability and binding activity [18]. In standard E. coli strains such as BL21(DE3), the cytoplasm is strongly reducing due to the activity of thioredoxin reductase (trxB) and glutathione reductase (gor), which prevents efficient disulfide-bond formation. In contrast, engineered strains such as Origami^TM^ 2 (DE3), which carry trxB and gor mutations, exhibit a more oxidizing cytoplasmic environment that supports the formation of structural disulfide bonds in recombinant proteins expressed in the cytosol [19,20,21]. These strains have been successfully used to express antibody fragments and other disulfide-bonded proteins without requiring periplasmic targeting.

2. Materials and Methods

2.1. Protein Structure Modeling and Analysis

Protein structure models of Schistosoma japonicum glutathione S-transferase (GST)-nanobodies were generated using Protenix, an advanced AI-based open-source reproduction of AlphaFold3 (https://protenix-server.com/model/prediction; URL accessed on 2, 8, 9 and 27 January 2026) [22]. Structural images and root-mean-square deviation (RMSD) values were obtained using PyMOL Molecular Graphics System v3.0.4 (Schrödinger, LLC, New York, NY, USA). RMSD values were calculated by superposition of amino acids 2–215 of GST in the isolated GST structure and in the GST moiety of GST-nanobody fusion proteins. V(D)J sequence analysis of nanobodies was performed using IMGT/V-QUEST v3.7.1 (https://www.imgt.org/IMGT_vquest/input; URL accessed on 27 November 2025) [23,24].

2.2. Reagents and Antibodies

Unless otherwise stated, all chemicals and reagents were from Sigma-Aldrich. Rabbit antisera against GFP and mCherry were kindly provided by Ramanujan Hegde (MRC Laboratory of Molecular Biology, Cambridge, UK). We used a goat anti-GST antibody (Cytiva) and a rabbit anti-Giantin antibody (cat # NBP2-22321, Novus Biologicals, Centennial, CO, USA). The following secondary antibodies were obtained from ThermoFisher Scientific (Waltham, MA, USA): horseradish peroxidase (HRP)-conjugated donkey anti-goat IgG, Alexa Fluor 488-conjugated donkey anti-sheep IgG, Alexa Fluor 594-conjugated donkey anti-sheep IgG, and Alexa Fluor 647-conjugated donkey anti-rabbit IgG. The Alexa Fluor 488- and Alexa Fluor 594-conjugated donkey anti-sheep IgG secondary antibodies were used to detect goat anti-GST antibodies, taking advantage of documented cross-reactivity between sheep and goat IgG [25]. This cross-reactivity is also indicated in the manufacturer’s validation information for Alexa Fluor 594-conjugated donkey anti-sheep IgG (ThermoFisher Scientific, cat # A-11016). The fluorescent nuclear stain 4’,6-diamidino-2-phenylindole (DAPI) was also from ThermoFisher Scientific.

2.3. Recombinant DNA Procedures

To generate constructs for expression of fluorescent proteins in Escherichia coli B834(DE3), the nucleotide sequences of full-length GFP and mCherry were obtained by PCR amplification from pEGFP-N1 (Clontech, Mountain View, CA, USA) or pmCherry-N1 (Takara, Kyoto, Japan), respectively, and cloned in-frame into the EcoRI and SalI sites of the pGST-Parallel-1 vector [26]. The oligonucleotides used for PCR amplification of the GFP coding sequence were 5′-GGGGGAATTCATGGTGAGCAAGGGCGAGGAGCTGTTC-3′ (forward) and 5′-CCCCGTCGACTTACTTGTACAGCTCGTCCATGCCGAG-3′ (reverse), and those used for amplification of the mCherry coding sequence were 5′-GGGGGAATTCATGGTGAGCAAGGGCGAGGAGGATAAC-3′ (forward) and 5′-CCCCGTCGACCTACTTGTACAGCTCGTCCATGCCGCC-3′ (reverse) (Eurofins Scientific, Fremont, CA, USA). To generate a construct for expression of an N-terminally GST-tagged anti-GFP nanobody (GST-nb-GFP) in E. coli Origami^TM^ 2 (DE3), the coding sequence of the Lag2 anti-GFP nanobody [27] was assembled into pGST-Parallel-1 downstream of, and in-frame with, the GST-coding region. The Lag2 coding sequence was amplified by PCR using primers 5′-GCCATGGATCCGGAATTCATGGCGCAGGTGCAGC-3′ (forward) and 5′-TAGTTGAGCTCGTCGACTACTGTAACTTGTGTACCCTGG-3′ (reverse) (Integrated DNA Technologies), using an DNA-synthesized gBlock (Integrated DNA Technologies, Coralville, IA, USA) containing the coding region of NbLag2 [27] as template. The pGST-Parallel-1 backbone was amplified by using primers 5′-GAATTCCGGATCCATGGCGCCCTG-3′ (forward) and 5′-GTCGACGAGCTCAACTAGTGCGGC-3′ (reverse) (Integrated DNA Technologies). Primer overhangs were designed to generate overlapping homology between the Lag2 insert and the linearized vector backbone for seamless Gibson assembly while maintaining the correct GST-Lag2 reading frame [28]. PCR products were generated using the DeCodi-Fi^TM^ High-Fidelity PCR Kit (Kura Biotech, Puerto Varas, Chile) and assembled using Gibson Assembly^®^ Master Mix (New England Biolabs, Ipswich, MA, USA). Assembly reactions were transformed into chemically competent E. coli TOP10 cells. To generate an N-terminally GST-tagged anti-mCherry nanobody construct (GST-nb-mCherry) for the expression in E. coli Origami^TM^ 2 (DE3), the full-length C11 anti-mCherry nanobody coding sequence was obtained by PCR amplification from pHEN2-C11 [29] and cloned in-frame into the EcoRI and SalI sites of pGST-Parallel-1 using primers 5′-GGGGGAATTCATGGCGGAAGTGCAGCTGCAGGCTTCC-3′ (forward) and 5′-CCCCGTCGACCTAGCTACTCACAGTTACCTGCGTCCC-3′ (reverse) (Eurofins Scientific). All recombinant constructs were verified by Sanger sequencing using the AUSTRAL-omics core facility (Universidad Austral de Chile, Valdivia, Chile; https://www.australomics.cl/, accessed on 27 November 2025). Mammalian expression constructs encoding GFP-TGN38 or mCherry-TGN38, in which amino acids 304–357 of rat TGN38 were fused to the C-terminus of GFP or mCherry, respectively, were a generous gift from Jennifer Lippincott-Schwartz (Janelia Research Campus, HHMI, Ashburn, VA, USA).

2.4. Expression and Purification of Recombinant Proteins

Recombinant nanobodies and fluorescent proteins were expressed from the pGST-Parallel-1 GST-fusion vector, which encodes a standard TEV protease cleavage site in the linker region commonly used for optimal tag removal. This TEV-containing linker was retained without further engineering and is described here for completeness. Proteins were expressed and purified as described previously [30], with minor modifications. Expression of GST-nanobodies in E. coli Origami^TM^ 2 (DE3) (Novagen, Madison, WI, USA) or GST-tagged fluorescent proteins in E. coli B834(DE3) (Novagen) was performed in lysogeny broth (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl) after induction with 0.2 mM isopropyl-β-D-1-thiogalactopyranoside (IPTG) for 40 h at 20 °C. Bacteria were harvested by centrifugation (4500× g for 30 min at 4 °C), resuspended in homogenization buffer (50 mM Tris-HCl, 500 mM NaCl, 5 mM EDTA, 1 mM dithiothreitol (DTT), and 2 mM phenylmethylsulfonyl fluoride, pH 8.0), and lysed by sonication at 4 °C. Lysates were centrifuged at 14,600× g for 1.5 h at 4 °C, and the supernatant was filtered through a 0.45 μm SFCA filter (ThermoFisher Scientific) and applied to Glutathione Sepharose^TM^ 4B (GE Healthcare, Chicago, IL, USA) at 4 °C in affinity buffer (50 mM Tris-HCl, 500 mM NaCl, 5 mM EDTA, 1 mM DTT, pH 8.0). After elution with 20 mM reduced glutathione in elution buffer (50 mM Tris-HCl, 150 mM NaCl, 5 mM EDTA, 1 mM DTT, pH 8.0) and concentration in Vivaspin^®^ 20 centrifugal concentrators (Cytiva, Marlborough, MA, USA), GST-tagged proteins were further purified by size-exclusion chromatography (SEC) on a Superdex 200 pg column (GE Healthcare) equilibrated in S200 buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM DTT, pH 8.0). DTT (1 mM) was included in lysis and purification buffers to minimize nonspecific oxidation during extraction and chromatography (e.g., oxidative aggregation and/or unintended intermolecular disulfide formation) [31], which should not affect nanobody intramolecular disulfide-bond formed during expression of GST-nanobodies. SEC elution profiles were used to estimate the apparent oligomeric state of GST-nanobodies based on their hydrodynamic behavior (Stokes radius), using globular protein standards (ovalbumin, 30.5 Å; albumin, 35.5 Å; aldolase, 48.1 Å; ferritin, 61.0 Å; Cytiva). Aliquots of purified GST-nanobodies were stored at −80 °C until use. GST-tagged fluorescent proteins were subsequently incubated overnight at 4 °C with 1% (w/w) in-house-produced His_6_-tagged TEV protease. After verification of TEV cleavage by protein electrophoresis, proteins were concentrated in Vivaspin^®^ 20 centrifugal concentrators and subjected to size-exclusion chromatography on a Superdex 200 pg column equilibrated in S200 buffer. To remove cleaved GST, pooled fractions containing untagged fluorescent proteins were applied to Glutathione Sepharose^TM^ 4B equilibrated in S200 buffer. To remove His_6_-tagged TEV protease, the eluate was supplemented with 20 mM imidazole and 7 mM MgCl_2_ and applied to Ni Sepharose^TM^ High Performance (Cytiva) equilibrated in Ni buffer (50 mM Tris-HCl, 150 mM NaCl, 20 mM imidazole, pH 8.0). The final eluate was supplemented with 1 mM DTT, and aliquots of purified GFP and mCherry were stored at −80 °C until use.

2.5. Protein Electrophoresis

Proteins were analyzed by electrophoresis in polyacrylamide gels under denaturing conditions (SDS-PAGE) as described previously [30], with minor modifications. Briefly, protein samples were denatured for 5 min at 95 °C in sample buffer (50 mM Tris-HCl, 2% SDS, 10% glycerol, 5% β-mercaptoethanol, 0.05% bromophenol blue, pH 6.8) and separated on 10% polyacrylamide gels (29:1 acrylamide:bis-acrylamide) at 50 mA for 45 min at room temperature using running buffer (25 mM Tris, 192 mM glycine, 0.1% SDS). For these gels, AccuRuller RGB Plus molecular mass marker (MaestroGen Inc., Hsinchu City, Taiwan) was used. Alternatively, proteins were denatured for 5 min at 95 °C in NuPAGE^TM^ LDS sample buffer (ThermoFisher Scientific) and separated on 4–12% NuPAGE^TM^ Bis-Tris gradient gels (ThermoFisher Scientific) at 110 mA for 45 min at room temperature using NuPAGE^TM^ running buffer (ThermoFisher Scientific). For these gels, SeeBlue^TM^ Plus2 molecular mass marker (ThermoFisher Scientific) was used. Proteins were visualized by staining with SimplyBlue^TM^ SafeStain (ThermoFisher Scientific). Stained gels were scanned, and images were processed using ImageJ software [32] v1.52q for figure preparation.

2.6. Isothermal Titration Calorimetry

Isothermal titration calorimetry (ITC) was performed as described previously [33]. The buffer of purified recombinant GFP, GST-nb-GFP, mCherry, and GST-nb-mCherry was exchanged into ITC buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM DTT, pH 7.4) using 7K MWCO Zeba^TM^ spin desalting columns (ThermoFisher Scientific). ITC measurements were carried out at 25 °C using a MicroCal PEAQ-ITC instrument (Malvern Panalytical, Worcestershire, United Kingdom). Typically, the sample cell contained ~0.3 mL of GFP or mCherry (10–30 μM), and GST-nb-GFP or GST-nb-mCherry (100–300 μM) was titrated in 18 injections of 2 μL each following an initial injection of 0.4 μL. Titration curves were analyzed using MicroCal PEAQ-ITC Analysis Software v1.21 (Malvern Panalytical). Binding constants and stoichiometry were obtained by fitting the data to a one-site binding model.

2.7. Dot Blot

For dot blot analysis, serial dilutions of purified recombinant GFP or mCherry were prepared in dot blot buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM DTT, pH 8.0). Aliquots of each protein dilution (3 μL) were spotted onto nitrocellulose membranes. After air-drying, membranes were blocked for 1 h at room temperature in phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na_2_HPO_4_, 1.8 mM KH_2_PO_4_, pH 7.4) supplemented with 5% (w/v) non-fat dry milk. Membranes were then incubated for 1 h at room temperature with GST-nb-GFP or GST-nb-mCherry (100 μg/mL) diluted in blocking buffer. After four washes of 3 min each in PBS containing 0.3% Tween 20, membranes were incubated with goat anti-GST antibody (1:1000 in blocking buffer). Membranes were washed four times for 3 min in PBS containing 0.3% Tween 20 and then incubated with HRP-conjugated donkey anti-goat IgG (1:1000 in blocking buffer). As a negative control, membranes were incubated in parallel with purified GST alone at the same concentration as the GST-nanobody fusions and processed identically. Chemiluminescent detection was performed using SuperSignal^TM^ West Dura (ThermoFisher Scientific), and images were acquired with the ChemiScope 6200Touch Chemiluminescence Imaging System (Clinx Science Instruments, Shanghai, China).

2.8. Cell Culture, Transfection, and Preparation of Protein Extracts

HeLa human adenocarcinoma-like cells and H4 human neuroglioma cells were obtained from the American Type Culture Collection and maintained in Dulbecco’s modified Eagle’s medium (DMEM; ThermoFisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Cytiva), 100 U/mL penicillin, 100 μg/mL streptomycin (ThermoFisher Scientific), and 5 μg/mL plasmocin (InvivoGen, San Diego, CA, USA), at 37 °C in a humidified incubator with 5% CO_2_. For transient transfections, cells were seeded onto glass coverslips in 24-well plates and transfected at approximately 60% confluence using Lipofectamine 2000 (ThermoFisher Scientific), according to the manufacturer’s instructions. For preparation of protein extracts, transfected or non-transfected cultured cells were washed with cold PBS and lysed on ice in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1 mM DTT, 1% Triton X-100, pH 7.4) supplemented with a protease inhibitor cocktail (cat # P8340, Sigma-Aldrich, Saint Louis, MO, USA). Cell lysates were clarified by centrifugation at 16,000× g for 20 min at 4 °C, and protein concentrations in the supernatants were determined using a protein assay (Bio-Rad Laboratories, Hercules, CA, USA). Aliquots were stored at −80 °C until use.

2.9. GST-Nanobody Fluorescence Microscopy and Colocalization Analysis

For fluorescence microscopy, transfected cells were processed as described previously [34], with minor modifications. Cells grown on glass coverslips were washed with PBS supplemented with 0.1 mM CaCl_2_ and 1 mM MgCl_2_ (PBS-CM) and fixed with 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA, USA) in PBS-CM for 1 h at room temperature. After washing in PBS-CM, cells were permeabilized with 0.3% Triton X-100 in PBS-CM for 15 min at room temperature. Coverslips were then washed in PBS-CM and incubated with 0.2% (w/v) Type A porcine skin gelatin in PBS-CM (PBS-CM-G) for 10 min at room temperature. Cells were incubated with GST-nanobodies (75 μg/mL in PBS-CM-G) for 30 min at 37 °C or overnight at 4 °C. After washing in PBS, cells were co-incubated with goat anti-GST antibody (1:300 dilution) and rabbit anti-Giantin antibody (1:300 dilution) in PBS-CM-G for 30 min at 37 °C. Following PBS washes, cells were incubated with Alexa Fluor 647-conjugated donkey anti-rabbit IgG (1:750) and either Alexa Fluor 488-conjugated or Alexa Fluor 594-conjugated donkey anti-sheep IgG (1:750 dilution) in PBS-CM-G for 30 min at 37 °C. Negative controls included staining of transfected cells with GST alone and staining of non-transfected cells with GST-nanobodies. After washing in PBS, coverslips were mounted on glass slides using Fluoromount-G mounting medium (Electron Microscopy Sciences). Fluorescence 12-bit depth images were acquired with an AxioObserver.D1 microscope (Carl Zeiss, Oberkochen, Germany) equipped with a Plan-Apochromat 63× oil-immersion objective (NA 1.4) and an AxioCam MRm digital camera controlled by AxioVision 4.9 software (Carl Zeiss). Alternatively, fluorescence images were acquired using a Zeiss LSM 990 confocal microscope (Carl Zeiss) equipped with a Plan-Apochromat 63× oil-immersion objective (NA 1.40). Confocal images were acquired in point-scanning mode using spectral detection with GaAsP-PMT detectors (Carl Zeiss). Laser power for the 488, 543, and 639 nm lines was adjusted to the minimum required to avoid signal saturation and photobleaching (0.07%, 2.0%, and 0.9%, respectively). Fluorescence emission was collected in the following spectral ranges: 481–552 nm for EGFP, 543–623 nm for mCherry or Alexa Fluor 594, and 667–693 nm for Alexa Fluor 647. The pinhole was set to 1.0 AU, image resolution was 512 × 398 pixels, and images were acquired with a 16-bit depth, using unidirectional scanning. Images were processed for figure preparation using ImageJ software v1.52q [32] or Adobe Photoshop CS3 software (Adobe Systems, San Jose, CA, USA). Co-localization between the GST-nanobody signal or the signal of GST-alone (anti-GST channel) and GFP-TGN38 or mCherry-TGN38 was quantified on background-subtracted images using Pearson’s correlation coefficient. Analyses were performed in ImageJ v1.53a [32] using the JACoP plugin [35] on regions of interest encompassing the perinuclear Golgi area, with identical thresholding and analysis settings applied across all conditions. For each condition, 20 cells from 3 independent experiments were quantified. Pixel-intensity scatter plots (cytofluorograms) were generated for figure preparation.

2.10. GST-Nanobody Pulldown and Western Blotting

For GST-nanobody pulldown assays, Glutathione Sepharose^TM^ 4B beads (40 μL) were equilibrated with 1 mL of cold pulldown buffer (PBS, 0.3% Tween 20, 1 mM DTT) supplemented with a protease inhibitor cocktail (cat # P8340, Sigma-Aldrich). After removal of the supernatant, increasing amounts of GST-nanobody (250–2000 ng) were added together with ice-cold pulldown buffer containing protease inhibitors to a final volume of 1 mL. As a negative control, GST alone (2000 ng) was incubated with glutathione beads and processed in parallel. Samples were incubated by end-over-end rotation for 3 h at 4 °C, and beads were washed four times with 1 mL of ice-cold pulldown buffer. After removal of the final wash, aliquots of cell lysates (0.5 mg of total protein) were added together with 1 μg of the corresponding fluorescent protein (GFP or mCherry) and ice-cold pulldown buffer containing protease inhibitors to a final volume of 1 mL. Samples were incubated by end-over-end rotation overnight at 4 °C, followed by four washes with 1 mL of ice-cold pulldown buffer. Bound proteins were eluted in sample buffer (50 mM Tris-HCl, 2% SDS, 10% glycerol, 5% β-mercaptoethanol, 0.05% bromophenol blue, pH 6.8) or in NuPAGE^TM^ LDS sample buffer and analyzed by SDS-PAGE as described above. Proteins were transferred to Amersham^TM^ Protran^®^ 0.2 μm nitrocellulose membranes (Cytiva) at 250 mA for 90 min at 4 °C. Membranes were blocked for 1 h at room temperature in PBS containing 5% (w/v) non-fat dry milk and then incubated overnight at 4 °C with either rabbit anti-GFP antibody (1:2000 dilution) or rabbit anti-mCherry antibody (1:1000 dilution). After four washes of 3 min in PBS containing 0.2% Tween 20, membranes were incubated for 1 h at room temperature with HRP-conjugated goat anti-rabbit IgG (1:5000). After four washes of 3 min in PBS containing 0.3% Tween 20, chemiluminescent detection was performed using SuperSignal^TM^ West Dura (ThermoFisher Scientific), and images were acquired using the ChemiScope 6200Touch Chemiluminescence Imaging System.

3. Results

3.1. GST-Nanobodies Can Be Produced with a Simple Procedure

We sought to determine whether a simple GST-nanobody design enables inducible cytoplasmic expression in suitable Escherichia coli strains. For recombinant construction, we used the vector pGST-Parallel-1, a modified version of the widely used pGEX4T1 vector [26]. The final constructs encoded fusion proteins consisting of GST at the N-terminus, followed by a peptide linker and the nanobody sequence at the C-terminus. The linker comprised the spacer encoded downstream of GST (Ile-Asp-Thr-Thr), followed by the TEV protease cleavage site (Glu-Asn-Leu-Tyr-Phe-Gln-Gly) and residual polylinker-encoded residues (Ala-Met-Asp-Pro-Glu-Phe) upstream of the nanobody coding sequence (Figure 1A). This TEV-containing linker architecture is widely used in GST fusion constructs and is included here to document the exact pGST-Parallel-1-encoded fusion design, not to imply a possible advantage of the TEV site.

Because nanobody function depends on proper folding of the VHH domain, we evaluated AlphaFold3 predictions of the designed GST-nanobody fusion proteins. In both GST-nb-GFP and GST-nb-mCherry, the GST moiety was predicted to adopt a structure highly similar to native Schistosoma japonicum GST (Figure 1B,C). Indeed, superposition of Cα coordinates yielded RMSD values of 0.458 Å for GST-nb-GFP and 0.454 Å for GST-nb-mCherry relative to GST (PDB: 1UA5), indicating nearly identical folding [36,37]. The nanobody moieties in both fusion proteins were predicted to adopt the expected nine-β-strands β-sandwich fold characteristic of VHH domains, with well-defined complementarity-determining regions (CDR1–CDR3; Figure 1B,C). In addition, disulfide bonds were predicted between C259 and C333 in nb-GFP and between C259 and C334 in nb-mCherry (Figure 1B,C), consistent with a conserved structural disulfide present in camelid VHH domains [18]. GST can be found in a monomer-dimer equilibrium [38], suggesting that GST-nanobody fusions may likewise occur as dimeric species. Thus, we generated AlphaFold3 predictions of dimeric GST-nanobodies and found that these models preserve the expected GST fold and the characteristic VHH domain architecture (Supplementary Figure S1). Importantly, AlphaFold3 provides a static structural prediction and does not capture the conformational ensemble of the GST-linker-VHH fusion in solution. Because GST and the VHH are connected by a flexible linker (spacer plus TEV/polylinker-derived segment), their relative orientation is expected to be dynamic, and any apparent proximity between GST and CDR loops in the models is unlikely to reflect persistent CDR occlusion. Thus, these in silico analyses suggested that the GST-nanobody fusion proteins should be structurally competent for recombinant production.

Given the requirement for disulfide bond formation within the nanobody domain, we expressed the GST-nanobody fusions in E. coli Origami^TM^ 2 (DE3), a K-12 derivative strain carrying trxB and gor mutations that create a more oxidizing cytoplasmic environment favorable for disulfide bond formation [19,20,21]. Although expression levels in Origami strains are typically lower than BL21 derivatives, we obtained substantial amounts of soluble GST-nanobody fusions, with yields of approximately 8 mg/L for GST-nb-GFP and 12 mg/L for GST-nb-mCherry (Figure 2A). In our hands, expression of the same GST-nanobody constructs in reducing cytosolic strains such as BL21(DE3) and B834(DE3) yielded ~10% of the soluble protein obtained in Origami^TM^ 2 (DE3) using the same workflow. Importantly, during purification, SEC analyses indicated that both GST-nanobodies eluted predominantly at an apparent molecular size consistent with a dimeric species (Supplementary Figures S2 and S3), in agreement with the possibility of GST-driven dimerization.

For antigen production, recombinant GFP and mCherry were expressed in E. coli B834(DE3), a methionine-auxotrophic derivative of BL21(DE3) that supports high-level expression in a reducing cytoplasm. As expected, this strain yielded large amounts of soluble fluorescent proteins following IPTG induction and purification (Figure 2B).

3.2. GST-Nanobodies Bind Their Antigens with Nanomolar Affinity

To determine whether GST-nanobodies retained their binding capacity toward their cognate antigens, i.e., GFP and mCherry, we performed isothermal titration calorimetry (ITC). Both GST-nb-GFP and GST-nb-mCherry bound their respective targets with high affinity. GST-nb-GFP bound GFP with a dissociation constant (K_d_) of 123 ± 39.5 nM (Figure 3A), whereas GST-nb-mCherry bound mCherry with a K_d_ of 199 ± 31.5 nM (Figure 3B). These results demonstrate that fusion to GST does not compromise antigen binding and that GST-nanobodies are suitable for downstream biochemical and cell biological applications.

3.3. GST-Nanobodies Preferentially Recognize Native Antigens

We next examined whether GST-nanobodies were suitable for protein blotting applications. We first tested them in dot blot assays, which preserve antigens in a more native-like conformation [39]. Both GST-nb-GFP and GST-nb-mCherry exhibited antigen concentration-dependent reactivity toward their respective targets (Figure 4). Under the same conditions, incubation with GST alone produced no detectable signal (Figure 4), supporting the specificity of the GST-nanobody fusions. In contrast, neither GST-nanobody showed detectable binding in Western blot assays following SDS–PAGE under the denaturing conditions tested (samples boiled before loading; Supplementary Figures S4 and S5). This result is not unexpected, as many nanobodies preferentially recognize conformational epitopes that are preserved in native proteins but disrupted upon denaturation [4]. Collectively, these results indicate that GST-nanobodies are well-suited for applications that maintain antigen structure, such as dot blotting, native-state affinity capture, and fluorescence microscopy.

3.4. GST-Nanobodies Are Suitable for Fluorescence Microscopy

We next examined whether our GST-nanobodies could be used for fluorescence microscopy. One limitation for the use of conventional nanobodies in indirect immunofluorescence is that they lack a conserved constant region that can be recognized by a generic secondary reagent, such as a secondary antibody. We therefore reasoned that the GST moiety could serve as a universal detection handle, allowing for the visualization of nanobody binding through anti-GST antibodies, followed by fluorescent secondary antibodies. To test this possibility, HeLa and H4 cells were transfected with constructs encoding GFP-TGN38 or mCherry-TGN38. These fusion proteins contain GFP or mCherry linked to amino acids 304–357 of the rat Golgi protein TGN38, which include the transmembrane and cytoplasmic domains sufficient for Golgi localization [40]. As expected, both GFP-TGN38 and mCherry-TGN38 localized to the Golgi apparatus in HeLa and H4 cells, as confirmed by colocalization with the endogenous Golgi apparatus marker Giantin (Figure 5) [41]. Importantly, GST-nb-GFP and GST-nb-mCherry specifically labeled their corresponding GFP- and mCherry-tagged TGN38 constructs, as indicated by the colocalization of the signal of these constructs with that of the anti-GST signal (Figure 5). To quantify this colocalization, we obtained Pearson’s correlation coefficients (PCC) between the corresponding fluorescent signals in both HeLa and H4 cells. We found that the signal of GST-nb-GFP colocalized extensively with that of GFP-TGN38 in both HeLa (PCC = 0.952 ± 0.020) and H4 cells (PCC = 0.910 ± 0.070; Supplementary Figure S6). Likewise, the signal of GST-nb-mCherry showed extensive colocalization with that of mCherry-TGN38 in both HeLa (PCC = 0.976 ± 0.017) and H4 cells (PCC = 0.952 ± 0.030; Supplementary Figure S6). This analysis confirmed robust colocalization and supports the specificity of the GST-nanobody labeling under the conditions used. In contrast, control conditions (GST alone or incubation of non-transfected cells with GST-nanobodies) showed no detectable staining under the same imaging settings (Supplementary Figures S7 and S8), indicating that GST-nanobodies could be fully compatible with fluorescence microscopy applications.

3.5. GST-Nanobodies Are Suitable for a Variant of the GST-Pulldown

Because GST pulldown is widely used for the biochemical analysis of protein–protein interactions, we investigated whether GST-nanobodies could serve as affinity reagents in this context. As single-chain molecules, nanobodies are intrinsically more stable and easier to manipulate than conventional multi-chain antibodies, making them attractive tools for specific and robust target capture [2,3]. To test this concept, we developed a modified GST pulldown assay, termed GST-nanobody pulldown, in which GST-nb-GFP or GST-nb-mCherry was used to capture their respective antigens. HeLa cell extracts (500 μg total protein) were supplemented with 1 μg of purified GFP or mCherry and incubated with increasing amounts of GST-nanobody. Both GST-nb-GFP and GST-nb-mCherry specifically and efficiently precipitated their corresponding fluorescent proteins in a concentration-dependent manner (Figure 6A,B; lanes 4–7). In parallel, GST alone did not precipitate GFP or mCherry under the same conditions (Figure 6), supporting the specificity of the capture. These results demonstrate that GST-nanobodies function as effective affinity reagents in pulldown assays and can be used directly without removal of the GST moiety, greatly simplifying their experimental application.

4. Discussion

Nanobodies (VHH single-domain antibodies) have become mainstream affinity reagents because they are small, stable, and genetically amenable, enabling recombinant production and fusion to functional modules [1,2,3]. However, two practical barriers still limit their adoption in many laboratories: routine production workflows often rely on periplasmic targeting to enable disulfide-bond formation, and native nanobodies lack an Fc-like constant region that would permit generic secondary-antibody-based detection [4,8,9]. Our study addresses both limitations by combining cytosolic expression of GST-nanobody fusion proteins in a disulfide-permissive bacterial host with GST-mediated universal detection and affinity capture. We emphasize that this platform is not intended to replace VHH-Fc fusions, which offer distinct advantages in many settings, but rather to provide a convenient option for bacterial production and in vitro workflows where Fc functionality is not required.

A defining structural feature of VHH domains is the conserved intradomain disulfide bond that stabilizes the immunoglobulin fold and contributes to binding capability [18]. In standard expression strains such as BL21(DE3), the strongly reducing cytosol prevents efficient formation of this bond, motivating the widespread use of periplasmic secretion for nanobody production [8,9]. In contrast, redox-engineered strains such as Origami^TM^ 2 (DE3), which carry mutations in thioredoxin reductase and glutathione reductase pathways, create a more oxidizing cytoplasmic environment that supports disulfide-bond formation in cytosol-expressed proteins [19,20,21]. Although some nanobodies can be expressed in standard reducing strains, we observed markedly lower soluble yields for our GST-nanobody constructs in BL21(DE3) and B834(DE3) (~10% relative to Origami^TM^ 2 (DE3)). We did not systematically optimize BL21(DE3) conditions because Origami^TM^ 2 (DE3) reproducibly provided higher yields under our standard workflow and is commercially available. In addition, although our purification buffers contained 1 mM DTT, this was used as a mild antioxidant/anti-aggregation during handling and chromatography [31] and was not intended to promote or prevent intradomain disulfide formation, which occurs during expression in the Origami^TM^ 2 (DE3) cytosol. Thus, our results provide a practical demonstration that this strategy is compatible with functional VHH production: GST-nb-GFP and GST-nb-mCherry were recovered as soluble proteins at milligram-per-liter yields and retained nanomolar affinity for their antigens.

GST fusion has long been used to improve solubility and enable single-step purification on glutathione matrices [14,15]. We also note that TEV-cleavable GST fusion designs are well established, and that the TEV site in our constructs derives from the standard pGST-Parallel-1 linker. In addition, GST-nanobodies have been used previously for affinity capture of GFP fusion proteins, most notably in visible immunoprecipitation (VIP) assays [42]. Building on prior GST-nanobody capture approaches, our work establishes a practical GST-centered workflow for cytosolic production and direct downstream use of functional VHH reagents. First, we explicitly link GST-VHH production to cytosolic disulfide-bond-permissive expression, establishing a rational bacterial platform for producing structurally intact nanobody fusions. Second, we quantify the binding thermodynamics by ITC, confirming that GST fusion and cytosolic folding preserve high-affinity antigen recognition. In this regard, the anti-GFP nanobody Lag2 has been reported to bind GFP with low-nanomolar affinity using methods such as surface plasmon resonance and bead-based binding assays [27]. Because those published values were obtained using different experimental formats than our solution-based ITC measurements, a direct numerical comparison to our ITC binding constant values could be misleading unless performed side-by-side under matched conditions. For the anti-mCherry nanobody C11, a quantitative K_d_ for the untagged nanobody has not, to our knowledge, been reported [29]. Third, we demonstrate that GST-nanobodies can be used directly for immunofluorescence microscopy via anti-GST detection and for a defined biochemical capture workflow that we term the GST-nanobody pulldown.

Other fusion-tag approaches have been successfully applied to VHH production (e.g., MBP and SUMO), and comparative work indicates that tag selection can differentially affect solubility, yield, and antigen-binding activity [43]. For example, it has been directly compared the impact of MBP-, His-, and His-SUMO-tagging of VHHs and found that while MBP improved solubility, the antigen-binding activity of MBP-VHH fusions was lower than His-SUMO and His fusions [16], consistent with potential steric effects. These observations reinforce that no single fusion tag is universally optimal.

In our constructs, GST is fused to the N-terminus of the VHH. This orientation is advantageous because the antigen-binding site of the VHH (often dominated by CDR3 near the C-terminus) is less likely to be sterically obstructed, while GST provides a large, soluble N-terminal module that facilitates expression, purification, and detection. In contrast, a C-terminal GST fusion could, in principle, interfere with CDR3 accessibility or with VHH folding depending on linker length and epitope topology. Consistent with the N-terminal design being well-tolerated, structural superpositions of the GST moiety from our AlphaFold3 models with canonical GST show low RMSD values (0.454–0.458 Å), indicating that the GST fold is preserved in the fusion context. Likewise, low RMSDs for the VHH moieties relative to reference nanobody structures will further support that the GST fusion does not perturb the immunoglobulin β-sandwich. Together, these observations support the notion that N-terminal GST placement is a robust configuration for functional GST-nanobodies, while alternative orientations should be evaluated on a case-by-case basis.

Importantly, consistent with the known monomer-dimer equilibrium of GST [38], our SEC analyses indicate that the GST-nanobodies predominantly behave as dimers (Supplementary Figures S2 and S3). This property may be advantageous for avidity in some capture settings, but it could also influence apparent binding behavior or steric accessibility depending on the antigen and epitope topology. Therefore, for applications sensitive to stoichiometry or steric constraints, the oligomeric state of individual GST-nanobodies should be considered and can be assessed empirically by SEC. For VHHs with less stable frameworks or for aggregation-prone targets, potential GST-driven oligomerization effects can be minimized empirically by adjusting linker length or rigidity and by optimizing assay conditions (e.g., working at lower protein concentrations and tuning salt, pH, glycerol and mild non-ionic detergents), with SEC used as a routine check of oligomeric state, or other alternative analyses such as dynamic light scattering or size-exclusion chromatography coupled to multi-angle light scattering. Finally, any predicted proximity between GST and CDR loops in the AlphaFold3 models likely reflects the limitations of a static prediction and the flexibility of the interdomain linker, rather than sustained occlusion, as supported by the robust binding and performance of the GST-nanobodies across multiple assays.

The dot blot and Western blot behaviors of our GST-nanobodies are fully consistent with known properties of nanobody epitope recognition. Many nanobodies preferentially bind conformational epitopes that depend on native protein folding and are lost upon SDS denaturation and reduction [4,6]. Accordingly, GST-nb-GFP and GST-nb-mCherry reacted strongly with native antigens in dot blots but failed to recognize denatured proteins after SDS-PAGE. It remains possible that alternative, less-denaturing (semi-native) immunoblot conditions (e.g., omitting the boiling step) could enable detection for specific nanobody-antigen pairs, but this optimization was beyond the scope of the present study. Nevertheless, this behavior reinforces that GST-nanobody reagents are optimally suited for assays that preserve structural integrity, including microscopy, native blotting, and affinity capture, rather than classical Western blotting.

A major practical advantage of the GST fusion format is the introduction of a universal detection epitope. Because nanobodies lack a conserved Fc region, indirect immunofluorescence typically requires direct labeling or fusion to reporter proteins, approaches that increase construct complexity and can complicate signal amplification [4]. By contrast, GST-nanobodies can be detected with standard anti-GST antibodies and fluorescent secondary antibodies, restoring the convenience and amplification capacity of classical immunofluorescence. Using this strategy, we obtained specific labeling of GFP-TGN38 and mCherry-TGN38 at the Golgi apparatus in HeLa and H4 cells, as defined by colocalization with Giantin [41]. Thus, GST-nanobody fusions combine the specificity and recombinant tractability of nanobodies with the practical advantages of antibody-based imaging workflows.

The GST-nanobody pulldown method introduced here further broadens the utility of this platform. Conventional GST-pulldown assays are widely used to study protein–protein interactions by immobilizing a GST-tagged bait on glutathione beads [14,15]. In our adaptation, GST-nanobodies serve as high-specificity capture reagents that immobilize defined antigens from complex lysates, providing an alternative to Protein A/G-based capture or antibody-coupling approaches in workflows where Fc functionality is not required. This approach is conceptually related to earlier GST-nanobody-based capture methods such as VIP [42] but generalizes the strategy to any nanobody target and integrates it into a standard biochemical pulldown workflow. Importantly, GST-alone controls were negative across assays, implying that signals derive from nanobody-antigen recognition rather than GST-driven background.

Several considerations will guide future applications of this system. Not all nanobodies may fold equally well in the bacterial cytosol, even in redox-engineered strains, and expression conditions may need to be optimized for individual VHHs [19,20]. In addition, GST dimerization may introduce avidity effects that enhance capture sensitivity but could complicate quantitative binding analyses. Nevertheless, within these boundaries, our results establish GST-nanobody fusions as versatile, low-cost, and modular reagents that integrate production, detection, and affinity capture into a single recombinant format. In addition, future work will include matched affinity measurements of tagged versus untagged nanobodies under identical conditions to quantitatively assess any fusion-dependent effects on binding.

In summary, by bringing together cytosolic disulfide-permissive expression with GST-mediated purification and detection, this study provides a simplified platform to produce nanobodies for their use in routine microscopy and biochemical assays. This approach lowers the technical barrier to nanobody use and should be readily extensible to a wide range of nanobody specificities and biological applications.

5. Conclusions

In this study, we present a simplified and versatile platform for nanobody production and application based on cytosolic expression of functional GST-nanobody fusion proteins in E. coli Origami^TM^ 2 (DE3). By combining a disulfide-permissive bacterial cytoplasm with an N-terminal GST fusion, we demonstrate that nanobodies can be produced in soluble form, retain nanomolar binding affinity, and be used directly (without tag removal) for native-state detection, fluorescence microscopy, and affinity capture. The GST moiety offers a dual advantage by enabling efficient purification and serving as a universal detection handle compatible with standard secondary antibody-based workflows. Beyond the model nanobodies against GFP and mCherry used here, this strategy should be readily extendable to additional nanobody specificities, alternative fusion orientations, and other redox-engineered expression hosts. Future work may explore systematic comparisons of N- versus C-terminal GST placement, optimization of expression conditions for diverse VHH scaffolds, and integration with proximity labeling, super-resolution imaging, or quantitative interactomics approaches. Together, this platform lowers technical and economic barriers to nanobody utilization, representing a flexible framework for expanding the use of nanobody-based tools in cell biology and biochemical research.