Live-cell Imaging of Endocytic Transport using Functionalized Nanobodies in Cultured Cells

Endocytosis of receptors and other transmembrane (TM) proteins from the plasma membrane to various intracellular compartments is important for the maintenance of membrane homeostasis^1,2. Following internalization by clathrin-dependent or -independent endocytosis, protein cargoes first populate early endosomes from where they are further sorted either along the endo-lysosomal system, recycled to the plasma membrane, or retrogradely shipped to the trans-Golgi network (TGN)^3,4,5. Recycling from endosomes and/or the cell surface to the TGN is part of the functional cycle of a number of transmembrane cargo receptors, such as the cation-dependent and cation-independent mannose-6-phosphate receptors (CDMPR and CIMPR), delivering newly synthesized lysosomal hydrolases from the TGN to late endosomes and lysosomes^6,7,8, Wntless (WLS) transports Wnt ligands to the cell surface^9,10,11,12, or TGN46 that escorts soluble secretory cargo proteins, including pancreatic upregulated factor (PAUF) and other CARTS cargo, from the TGN^13,14,15,16. While all these cargo receptors commute through the TGN for the collection of new client proteins, the transferrin receptor (TfR) that internalizes iron-bound transferrin is excluded from the Golgi return^{17,18,19,20,21}.

To investigate endocytic and retrograde traffic, we previously established a nanobody-based toolkit designed to label and trace cargo proteins from the cell surface to intracellular compartments¹⁷. Nanobodies constitute a novel class of protein binders derived from homodimeric heavy-chain-only antibodies (hcAbs) naturally found in camelids and cartilaginous fishes^22,23. They represent the variable heavy-chain domain (VHH) of hcAbs and offer numerous advantages over conventional antibodies (e.g., IgGs): they are monomeric, small (~15 kDa), highly soluble, lack disulfide bonds, can be bacterially expressed, and are amenable to selection for high-affinity binding²⁴. To enhance the versatility and broad applicability of this nanobody toolkit, we employed functionalized anti-GFP nanobodies to surface-label and track proteins tagged with GFP tags on their extracellular or lumenal domains. Through recombinant conjugation of nanobodies with mCherry, ascorbate peroxidase 2 (APEX2), or tyrosine sulfation sequences, the retrograde trafficking of bona fide transmembrane cargo proteins can be analyzed by fixed and live-cell fluorescence microscopy, electron microscopy, or biochemical assays using autoradiography. Given that tyrosine sulfation, catalyzed by tyrosylprotein sulfotransferases TPST1 and TPST2, is a posttranslational modification confined to the trans-Golgi/TGN, this strategy enables direct assessment of transport dynamics and kinetics from the cell surface to the Golgi compartment^25,26,27. More recently, we have expanded this repertoire of functionalized protein binders to include derivatized anti-mCherry nanobodies. These reagents have proven instrumental in dissecting machinery-dependent transport pathways of MPRs, particularly of the CDMPR^17,28, to the TGN through biochemical analyses.

While the previous study mainly focused on the application of nanobodies modified with tyrosine sulfation sequences to biochemically assess TGN arrival by autoradiography¹⁹, this methods article describes the generation of fluorescently labeled, functionalized nanobodies (VHH-mCherry) and reporter cell lines for live-cell imaging of endocytic trafficking. In combination with an additional Golgi-resident fluorescent reporter protein, this protocol can be further adapted to serve as a robust workflow for quantitatively analyzing live-cell imaging of endosome-to-TGN transport of selected cargo proteins.

Before applying this protocol, researchers should ensure that the target protein is expressed as a GFP-tagged fusion construct, typically generated by standard molecular cloning approaches. Functionalized nanobodies can be readily produced in bacteria at yields sufficient for surface labeling experiments, with working concentrations in the low nanomolar range being suitable for most live-cell imaging assays. Users should also consider that transport kinetics may vary between cargo proteins, requiring adjustment of labeling duration or imaging frequency for optimal visualization of endocytic or endosome-to-TGN trafficking.

1. Bacterial transformation with VHH-mCherry

NOTE: This protocol has been optimized for the expression, purification, and analysis of functionalized anti-GFP nanobodies as previously described¹⁷. The following steps are adapted for VHH-mCherry and are consistent with the earlier methodological report¹⁹.

1. Thaw chemocompetent bacteria (50-100 µL) suitable for protein expression (e.g., Escherichia coli Rosetta BL21 (DE3) cells) by placing them on ice.
   NOTE: Prepare chemocompetent bacterial cells following standard laboratory protocols. Commercially available BL21 (DE3) cells from NEB can be employed for this.
2. Add 50 ng of a plasmid encoding the functionalized VHH-mCherry construct (Addgene #109421). To ensure efficient site-specific biotinylation of the nanobody reporter during bacterial expression, co-transform with a threefold excess (150 ng) of a plasmid encoding the bacterial biotin ligase BirA (Addgene #109424). Gently flick the tube 4x-5x to mix cells and plasmid DNA.
   NOTE: Co-transformation with the BirA-encoding plasmid is not necessary if biotinylation by the biotin acceptor peptide (BAP) is not required.
3. Incubate the mixture on ice for 30 min. Avoid agitation during this step.
4. Heat-shock bacterial cells by placing them for exactly 20 s at 42 °C in a water bath or a heating block.
5. Add 1 mL of room-temperature (RT) Luria broth (LB) medium and incubate the transformed bacteria in a thermoshaker (250-500 rpm) for 1 h at 37 °C to allow phenotypic expression of antibiotic resistance genes.
   NOTE: To prepare 1 L LB medium, add 5 g yeast extract, 10 g tryptone, 10 g NaCl and make up the volume with water, and sterilize by autoclaving.
6. Pellet the bacteria by centrifugation at 11,000 x g for 1 min and resuspend the pellet in 100 µL of fresh LB medium. Mix thoroughly by pipetting.
7. Plate the suspended bacteria onto pre-warmed LB plates containing the appropriate antibiotics (e.g., with 50 µg/mL kanamycin; if co-transformed with BirA, also include 50 µg/mL carbenicillin, see step 1.2).
8. Incubate the antibiotic-supplemented LB plates upside-down at 37 °C for 13-15 h.
   NOTE: The protocol can be paused here. Plates with grown colonies can be stored at 4 °C and sealed with permeable film to prevent drying.

2. Bacterial liquid culture and induction of VHH-mCherry expression

1. Select a single bacterial colony from the transformation plate and inoculate it into an Erlenmeyer flask containing 20 mL of LB medium supplemented with antibiotics. Incubate the culture in a shaking incubator for 13-15 h at 37 °C (see also step 1.7 regarding selection antibiotics).
2. The following day, dilute the overnight 20 mL bacterial culture into a fresh flask containing 1 L of LB medium with the corresponding selection antibiotics.
3. Continue to incubate at 37 °C until the culture reaches an optical density at 600 nm (OD₆₀₀) of 0.6-0.7.
   NOTE: Allow the culture to cool down to RT or to 16 °C prior to induction of protein expression.
4. Induce expression of VHH-mCherry (and BirA, if co-expressed) by adding 1 mL of 1 M isopropyl-β-D-thiogalactopyranosid (IPTG) to achieve a final concentration of 1 mM of the inducer (1:1000 dilution). If co-transformation with the BirA expression plasmid was performed, also supplement the culture with 10 mL of a 20 mM D-biotin stock solution, yielding a final concentration of 200 µM D-biotin in the growth medium. This facilitates in vivo biotinylation of the biotin acceptor (BAP) epitope present in the VHH-mCherry construct.
   NOTE: Prepare the D-biotin stock in ddH₂O and solubilize it by gradual addition of 500 mM NaH₂PO₄.
5. Incubate the IPTG-induced 1 L culture for 13-15 h at 16 °C to promote optimal protein folding and yield.
   NOTE: Expression conditions for newly designed nanobody constructs with other fluorescent tags may require empirical optimization by the investigator.
6. Transfer the culture into a 1 L centrifugation bottle and harvest cells by centrifugation at 5,000 x g at 4 °C for 45 min. Discard the supernatant into the appropriate liquid biowaste and proceed with purification steps.
   NOTE: The protocol can be paused at this point by storing the bacterial pellet at -80 °C. The VHH-mCherry pellet typically appears pink, indicating proper folding of the fluorescent fusion protein.

3. IMAC-based purification of VHH-mCherry

1. If necessary, thaw the frozen bacterial pellet on ice (see also step 2.6).
2. Add 30 mL of ice-cold binding buffer (20 mM imidazole in 1x PBS) to the bacterial cell pellet and resuspend thoroughly by pipetting up and down. Transfer the suspension into a labeled 50 mL centrifuge tube.
3. Supplement the resuspended cells with 200 µg/mL lysozyme, 20 µg/mL DNase I, 1 mM MgCl_2, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Incubate the mixture for 10 min at RT, followed by 1 h at 4 °C on an end-over-end rotator.
   NOTE: PMSF is toxic. Prepare and handle stock solutions in a chemical fume hood wearing gloves, a lab coat, and eye protection; dispose of PMSF-containing waste in halogenated organic waste containers. For experiments, a 0.1 M stock solution was used and further diluted.
4. Mechanically disrupt the bacterial cells using a tip sonicator inserted directly into the suspension. Apply constant three 1 min sonication pulses, allowing a 1 min cooling interval between each pulse with specified sonicator settings (Probe tip size: 6 mm, solid tip; Amplitude 40%; Duty cycle (pulse mode): 1 s ON/1 s OFF)
5. Clarify the lysate by centrifugation at 15,000 x g at 4 °C for 45 min to pellet bacterial cell debris and intact bacteria.
   NOTE: The lysate can be transferred to a centrifuge bottle or aliquoted into 5 mL tubes for benchtop centrifugation.
6. Carefully transfer the resulting supernatant to a new 50 mL tube and discard the pellet into the appropriate biowaste.
7. Keep the cleared lysate on ice while preparing for immobilized metal affinity chromatography (IMAC). For isolation of histidine-tagged nanobodies, utilize pre-packed, single-use His-tag purification columns optimized for gravity-flow operation.
8. Secure Ni-NTA (nickel-nitrilotriacetic acid) columns onto a metal stand or an appropriate column holder.
   1. Drain the storage buffer from the columns and equilibrate with 10 mL of binding buffer (20 mM imidazole in 1x PBS).
   2. Allow the buffer to pass through by gravity; discard the flow-through as biowaste.
9. Gradually apply the cleared bacterial lysate (~30 mL) to the column. Let it flow through by gravity and discard the flow-through.
10. Wash the column with two consecutive 10 mL volumes of binding buffer (20 mM imidazole in 1x PBS).
11. Elute the bound nanobodies with 2 mL of elution buffer (500 mM imidazole in 1x PBS) into a 2 mL microcentrifuge tube.
12. Perform a buffer exchange.
    1. Equilibrate a desalting column placed in a 50 mL tube adapter by rinsing 5x with 5 mL of 1x PBS.
    2. Allow the buffer to fully enter the packed resin; discard the flowthrough. After the final PBS wash, centrifuge the column at 1,000 x g for 2 min.
    3. Discard the flowthrough. Place the column with the adapter onto a new 50 mL tube. Load 2 mL of eluted functionalized nanobody (from step 3.11) onto the PBS-equilibrated desalting column and spin at 1,000 x g for 2 min and collect the eluate.
       NOTE: Dialysis may alternatively be used for buffer exchange.
13. Quantify the concentration of the purified VHH-mCherry protein using a bicinchoninic acid (BCA) or Bradford assay according to the manufacturer's protocols.
    NOTE: For efficient nanobody uptake in live-cell imaging applications, a final stock concentration of 5-10 mg/mL is recommended.

4. Validation of VHH-mCherry expression and purity (Coomassie staining)

1. Prepare a 10% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) according to established laboratory protocols.
   NOTE: As an alternative, commercially available precast gradient gels from Bio-Rad can be employed for convenience and reproducibility.
2. Aliquot 20 µg of purified VHH-mCherry in a 1.5 mL microcentrifuge tube and denature by boiling in sample buffer at 95 °C for 5 min.
3. Load the denatured nanobody sample onto the SDS-polyacrylamide gel and perform electrophoresis following standard PAGE procedures until the tracking dye (e.g., bromophenol blue) reaches the bottom of the resolving gel.
4. Proceed with Coomassie staining and destaining of the gel.
   1. Carefully remove the gel from the cassette and immerse it in Coomassie staining solution (5% of a 10 g/L Coomassie Brilliant Blue stock in 10% acetic acid and 45% methanol in ddH₂O) for 20-30 min at RT on a gently shaking platform. Ensure that the gel is fully submerged in the staining solution.
   2. Destain the gel using a destaining solution (7.5% acetic acid and 15% methanol in ddH₂O), performing 2-3 sequential washes of 1 h each at RT. For optimal background reduction, leave the gel for 13-15 h in fresh destaining solution.
      NOTE: Excess dye can be effectively absorbed by placing household paper towels around the gel during the destaining process. Staining/destaining solutions contain flammable methanol and irritant acetic acid. Use in a fume hood, wear gloves and goggles, and collect spent solutions in designated flammable-liquid waste containers.
5. Visualize the gel using a gel documentation system or camera of choice.
   NOTE: For additional confirmation of nanobody expression, immunoblotting using epitope-specific antibodies can be performed (see also Figure 1C).

5. Generation of GFP-tagged reporter HeLa cell lines using retroviral transduction

NOTE: Various GFP reporter HeLa α Kyoto cell lines (here simply called HeLa or HeLa α) have been previously generated¹⁷. For the purpose of demonstrating the live-cell imaging assay, GFP-tagged CDMPR or TfR, as representative cargo proteins, are employed. While TM proteins with type I topology are GFP-tagged at the extreme N-terminus, TM proteins with type II topology require C-terminal GFP fusion.

1. The genes encoding for EGFP-CDMPR (Addgene #182642) or TfR-EGFP (#243763) have been previously subcloned into the retroviral vector pQCXIP using standard restriction enzyme cloning techniques¹⁷. Following transformation into chemocompetent E. coli, prepare high-purity plasmid DNA using midi prep kits (e.g., from Macherey-Nagel).
   NOTE: Chemocompetent bacterial cells should be prepared following standard laboratory protocols. NEB 5-alpha Competent E. coli from NEB can be used for cloning and plasmid purification.
2. To generate stable GFP reporter HeLa cell lines, retroviral particles are produced by transfection of the packaging cell line Phoenix ampho. All cell culture work is conducted under BSL2 laminar flow conditions.
   1. The day prior to transfection, seed 2.0-3.0 x 10⁶ Phoenix ampho cells into a 10 cm dish in 10 mL of medium to achieve 60%-80% confluency by the following day. Phoenix ampho are maintained in DMEM high-glucose GlutaMAX-I with 10% fetal bovine serum (FBS), 100 units/mL penicillin/streptomycin, and 1 mM sodium pyruvate.
   2. On the day of transfection, replace the medium with 10 mL of fresh complete medium.
   3. In a 1.5 mL microcentrifuge tube, combine 1 mL of serum- and supplement-free DMEM high-glucose GlutaMAX-I with 10 µg of either pQCXIP-EGFP-CDMPR or pQCXIP-TfR-EGFP and 30 µL of FuGENE HD transfection reagent. Mix thoroughly by pipetting.
   4. Incubate the mixture at room temperature for 15 min to allow formation of DNA-reagent complexes.
   5. Add the entire transfection mixture dropwise to the Phoenix ampho culture. Swirl the plate gently to distribute the complexes evenly.
   6. Incubate transfected Phoenix ampho cells for 13-15 h at 37 °C with 5% CO₂.
      NOTE: Alternative packaging cell lines may be used, provided they are compatible with the Retro-X Q vector system.
3. The following day, discard the transfection medium and replace it with 7 mL of fresh complete medium to enhance viral production.
4. At 48 h and optionally 72 h post-transfection, collect the culture supernatant (~7 mL) containing viral particles using a pipette aid. Monitor GFP expression in transfected Phoenix ampho cells using a fluorescence microscope or imaging system (e.g., Bio-Rad ZOE).
   NOTE: Again, retroviral manipulations require BSL-2 containment. Perform all work in a certified biosafety cabinet, wear gloves, a lab coat, and eye protection, and decontaminate surfaces with 1% hexaquart followed by 70% ethanol before disposing of waste via autoclave.
5. Filter the collected viral supernatant through a 0.45 µm-sized filter. The filtered supernatant can be used immediately (see step 5.7) or stored at 4 °C for short-term use or at -80 °C for long-term storage. Note that viral titers may decline upon freezing.
6. To establish stable GFP reporter HeLa cells, proceed with retroviral transduction using the filtered viral supernatant. All procedures must be carried out under BSL2 conditions.
   1. One day prior to transduction, seed approximately 2.0-3.0 x 10⁶ HeLa α cells in a 10 cm dish with 10 mL of medium to reach 60%-80% confluency by the following day. Culture HeLa α cells are cultured in DMEM high-glucose GlutaMAX-I supplemented with 10% FBS and 100 U/mL penicillin/streptomycin.
   2. Mix filtered viral supernatant with 15 µg/mL polybrene (hexadimethrine bromide) to enhance the infection efficiency.
      NOTE: Polybrene is an irritant - during stock preparation, handle with gloves and eye protection, avoid aerosol formation, and dispose of Polybrene-containing solutions as hazardous chemical waste.
   3. The next day, replace the standard growth medium with undiluted viral supernatant containing Polybrene, ensuring the entire surface of the dish is covered.
   4. Incubate HeLa α cells under transduction for 13-15 h at 37 °C with 5% CO₂.
      NOTE: To determine viral titer, serial dilution and infection assays on HeLa α cells can be performed. Titering in this case is optional, as retroviruses infect only dividing cells.
7. The day after transduction, discard the viral medium into BSL2 waste and replace it with 10 mL of fresh complete HeLa α medium. Incubate for 13-15 h at 37 °C with 5% CO₂.
8. The following day, initiate selection by replacing the medium with complete medium containing 1.5 µg/mL puromycin.
9. Maintain selection pressure for 2-3 weeks, passaging cells as needed. Prior to transferring genetically modified cells to a BSL1 environment, ensure that they are free of replication-competent viral particles.
10. To obtain a homogeneous population, either use clonal isolation using cloning rings or cylinders, or fluorescence-activated cell sorting (FACS). FACS can be performed on transduced HeLa α cells using FACSAria III or Fusion (BD Biosciences) to enrich for a population with uniform GFP expression based on fluorescence intensity.
    NOTE: Expression of the GFP reporter in HeLa α cells can be further validated by fixed-cell fluorescence microscopy or immunoblotting with anti-GFP antibodies (see also Figure 2C).

6. Uptake of VHH-mCherry by cultured cells for live-cell imaging

NOTE: All cell culture procedures are carried out under sterile conditions within a laminar flow hood prior to live-cell microscopy.

1. 6.1Under sterile laminar flow, seed approximately 50,000 to 110,000 HeLa cells stably expressing GFP-tagged CDMPR or TfR reporter constructs into each well of a µ-slide 4-well ibiTreat chamber slide. Use 700 µL of complete culture medium containing antibiotics (DMEM high-glucose, phenol red-free), supplemented with 10% FBS, 100 U/mL penicillin/streptomycin, 2 mM l-glutamine, and 1.5 µg/mL puromycin.
2. Incubate the cells for 13-15 h at 37 °C in a humidified incubator with 5% CO₂ to allow for proper adhesion and proliferation. Cells should reach approximately 80% confluency by the following day.
3. On the day of imaging, just prior to the experiment, gently replace the medium in each well with 350 µL of fresh phenol red-free complete medium.
4. In parallel, prepare a VHH-mCherry working solution at a final concentration of 2.5 µg/mL (approximately 50 nM) in pre-warmed, phenol red-free medium. Maintain the solution at 37 °C until use.
5. After initializing the automated live-cell imaging system (e.g., FEI MORE or Nikon Ti2 X-Light V3), pre-heat the microscope's incubation chamber to 37 °C with 5% CO₂ before transferring the ibidi microscopy chamber onto the microscope and configuring the acquisition settings as follows.
   1. Choose the U PlanS Apo 100x NA 1.4 and add the ibidi microscopy chamber on the microscope stage.
   2. Set up two channels with the appropriate excitation wavelength and single band pass emission filters for EGFP (ex: 470/20 nm, em: 517/20 nm) and mCherry (ex: 550/15 nm, em: 590/20 nm).
   3. Choose a short exposure time and low illumination intensity to avoid photobleaching (e.g., 50 ms and 5% power for EGFP, 160 ms and 25% for mCherry). It is important that the chosen values are kept constant for all experiments.
   4. Per condition (=well), store the coordinates of 10 different fields of view with 3-4 cells in focus in a point list.
   5. Image every position in this point list once with a single snapshot. These are the reference images before VHH-mCherry addition.
   6. To initiate endocytic uptake, gently add 350 µL of the pre-warmed VHH-mCherry solution to each well, ensuring minimal disturbance to the cell monolayer. Continue incubation at 37 °C with 5% CO₂.
   7. Image every point in the point list repeatedly in a time-lapse experiment for 100 frames with a 36 s delay between frames at 37 °C with 5% CO₂.
   8. To decrease photobleaching, order the acquisition of the channels to image first mCherry, then EGFP.
   9. To increase throughput, order the acquisition to first acquire all positions sequentially and repeat this for each time point.

7. Image analysis of VHH-mCherry endocytic uptake

1. Post-acquisition image pre-processing and data extraction in Fiji.
   1. Load the time-lapse image using the Bio-Formats importer plugin with the split channels option activated.
   2. Correct lateral drift using the MultiStackReg plugin. Use the EGFP channel as a reference and apply the transformation to the mCherry channel as well.
   3. Per cell, draw a region of interest (ROI) with the polygon tool (e.g., around the perinuclear region) and add it to the ROI manager.
   4. For both channels, measure the mean intensity in all ROIs and time points using the Multi measure function of the ROI manager and save the resulting table as text file.
   5. Open the corresponding reference image and repeat the measurement for each ROI.
2. Data analysis to determine endocytic uptake curves
   1. Open the text file and subtract each frame number by 1 so that they start at 0, then multiply each frame number by the delay between frames in s (i.e., 36).
   2. For each ROI, subtract the mean intensity in the VHH-mCherry channel of the reference image from each frame of the corresponding time-lapse image in the VHH-mCherry channel to remove autofluorescence and background.
   3. For each frame and ROI, divide the background-subtracted VHH-mCherry intensity by the EGFP-channels mean intensity to account for intensity fluctuations within the ROI other than VHH-mCherry, e.g., by Golgi movement, cell contraction, z-drift, or illumination instabilities.
   4. Normalize the resulting curve to its own maximum.
   5. Select a range of frames to exclude any artifacts at the beginning (i.e., media addition) or end (i.e., drift, bleaching) of the time-lapse.
   6. Fit the curve data with a single exponential decay function that models a first-order kinetic process:
      
      where I(t) is the normalized intensity at time t, and t₀ accounts for the time delay between VHH-mCherry addition and the start of recording. From the rate constant k, calculate the half-life τ_1/2 of the process as:
      
      ​NOTE: When conducting an experiment with multiple subsequently imaged field of view (FOV) per time point, each corresponding t0 could be slightly different, but it should increment to higher values from one FOV to the next.
   7. Repeat the procedure for all datasets and report the half-life values in box plots.

To investigate retrograde protein trafficking to various intracellular compartments, we have recently established an anti-GFP nanobody-based tool for labeling and tracking recombinant fusion proteins from the cell surface17. Here, we describe the bacterial production of such derivatized nanobodies and demonstrate their utility in monitoring endocytic uptake by live-cell imaging. In combination with a Golgi-resident fluorescent reporter, this protocol offers a robust platform to quantitatively study the retrograde transport of selected cargo proteins to the trans-Golgi network (TGN) in real time.

We have generated a collection of distinct functionalized anti-GFP nanobodies sharing a common modular architecture, using standard molecular cloning techniques17. Although the current study focuses on the fluorescent variant VHH-mCherry, we also include additional derivatized nanobodies to illustrate the versatility of this toolset and highlight their potential for future applications. Our most basic construct, VHH-std (std for standard), comprises the VHH domain, T7 and HA epitopes for antibody-based detection, a C-terminal hexahistidine (His6) tag for purification, and a biotin acceptor peptide (BAP) sequence to enable enzymatic biotinylation and high-affinity streptavidin-based pulldown assays (Figure 1A). From this core design, various nanobody derivatives were developed to facilitate the study of endocytic and retrograde trafficking through biochemical analysis, fixed and live-cell imaging, and electron microscopy.

For live-cell imaging of endocytic transport, we engineered a fluorescent anti-GFP nanobody incorporating a fluorophore with excitation and emission spectra distinct from those of GFP, thereby ensuring optimal spectral separation during fluorescence microscopy. Based on its superior folding properties in IPTG-induced E. coli and its well-characterized photophysical properties, we selected mCherry as the fluorescent tag of choice. Other red-shifted fluorophores, such as monomeric red fluorescent protein (mRFP), would also be suitable alternatives in principle, but mCherry proved particularly robust and effective in this system.

What is the potential of other functionalized nanobodies than VHH-mCherry? To investigate protein trafficking from the plasma membrane to the trans-Golgi network (TGN), we leveraged the compartment-specific localization of TPSTs, which reside exclusively in the TGN and trans-Golgi cisternae. To this end, we modified VHH-std with a tandem tyrosine sulfation motif (2xTS) derived from rat procholecystokinin29, thereby enabling a biochemical readout of cargo arrival at these compartments. While fusion of VHH-std to mCherry allows for direct visualization of retrograde transport by fixed or live-cell imaging, functionalization with peroxidases such as APEX230 permits ultrastructural localization by electron microscopy, targeted cytochemical ablation, or proximity-based biotinylation assays. Moreover, incorporation of a tobacco etch virus (TEV) protease cleavage site into the nanobody scaffold provides a biochemical means to distinguish between internalized and surface-bound nanobodies (Figure 1A). We previously employed the VHH-tev construct to monitor recycling kinetics of nanobody-bound EGFP-CDMPR and TfR-EGFP17. The reappearance of nanobody-tagged receptors at the cell surface could be readily detected by applying recombinant TEV protease extracellularly, resulting in a specific loss of the nanobody's C-terminal epitope cassette. Analogously, the inclusion of a TEV cleavage site within the here presented mCherry-functionalized nanobody VHH-mCherry allows for dynamic assessment of EGFP reporter recycling by live-cell imaging. Functionalization with alternative protein domains - such as additional fluorophores, enzymatic tags, or sequence motifs for posttranslational modification - can be readily accomplished by subcloning desired inserts into the VHH-std backbone by the SpeI and EcoRI restriction sites. All functionalized anti-GFP nanobody constructs used in the original study17 have been deposited with Addgene for public distribution.

Using the protocol described above, all nanobody variants illustrated here were purified to high yield and purity (Figure 1B). Only the mCherry fusion exhibited minor proteolytic clipping of protein domains following purification (Figure 1B, lane 5). The two observed degradation products most likely correspond to the individual VHH and mCherry domains, as inferred from their apparent molecular weights and verified by epitope-specific immunoblot detection (Figure 1C). In the absence of co-expressed BirA, VHH-mCherry typically recovered at yields of approximately 20 mg per preparation. Based on our experience, co-expression of BirA consistently reduces nanobody yield by roughly 1/3rd to 1/2. Biotinylation was fairly complete because the nanobodies from a 1:1 mixture with BSA were fully recovered by streptavidin-agarose (Figure 1D).

To evaluate the suitability of this nanobody toolkit for studying endocytic transport, we generated stable HeLa cell lines expressing EGFP-tagged surface receptor proteins with distinct intracellular trafficking routes. These included TfR, which cycles between the plasma membrane and early (sorting and recycling) endosomes; TGN46, which traffics between the plasma membrane and the TGN by early endosomes; and both MPRs, which shuttle between the TGN, plasma membrane, and both early and late endosomes17. EGFP was fused to the extracellular domain of each receptor - specifically, inserted between the signal peptide and the receptor sequence for CDMPR, CIMPR, and TGN46 - and to the C-terminus of TfR. This design preserved the native cytoplasmic domains, ensuring that all known sorting signals remained intact and the EGFP tag accessible for binding by extracellular anti-GFP nanobodies (Figure 1E). For CIMPR, whose extracellular domain is unusually large, a truncated version was used, consistent with previous studies demonstrating that such truncation preserves normal trafficking behavior31,32.

Stable cell lines were established by retroviral transduction, followed by fluorescence-activated cell sorting (FACS) to isolate homogeneous cell populations with moderate and comparable expression levels. Of note, EGFP-CDMPR consistently appeared as a double band in immunoblots, similar to its endogenous counterpart33, indicative of heterogeneous glycosylation. To assess whether the EGFP fusion proteins recapitulate the steady-state localization and expression patterns of the endogenous proteins, the stably expressing cells were co-cultured with parental HeLa cells and analyzed by confocal fluorescence microscopy (Figure 2B). The EGFP signal faithfully mirrored the distribution of the corresponding endogenous proteins. As expected, CDMPR, CIMPR, and their EGFP-tagged versions localized predominantly to the perinuclear region - reflecting TGN and late endosomal compartments - with additional labeling in peripheral endosomes. Both endogenous and EGFP-tagged TGN46 were found almost exclusively in the perinuclear TGN, while TfR and TfR-EGFP displayed the characteristic early endosome distribution, with prominent peripheral sorting and perinuclear recycling endosomes. Although the antibodies used detected both the endogenous and EGFP-tagged forms (except for CIMPR), the overall staining intensity was not markedly increased in cells expressing the fusion proteins, suggesting that the EGFP constructs were not substantially overexpressed. We included EGFP-TGN46 and EGFP-CIMPR in Figure 2B for direct comparison with EGFP-CDMPR and TfR-EGFP. The protocols used for cell fixation and fluorescence microscopy imaging are outlined in previous studies17,19.

Using VHH-mCherry, endocytic transport of EGFP-tagged reporter proteins can be monitored by live-cell imaging. For this, we used as examples cells expressing EGFP-CDMPR and TfR-EGFP. The cells were imaged over time with an inverted widefield fluorescent microscope upon addition of VHH-mCherry to the medium (Video 1 and Video 2). Still images at various time points are shown in Figure 3. Uptake was quantified by measuring the signal in the mCherry channel, subtracting the autofluorescence background, and normalizing to the EGFP signal to eliminate fluctuations due to the movement of labeled compartments or potential small shifts in the focal plane. The fluorescence of VHH-mCherry in the medium at 25 nM was negligible and did not interfere with the measurements. Analysis of transport of the reporters from the cell surface to their intracellular compartments, to the steady-state distribution, yielded the same kinetic results as the biochemical experiments shown in Figure 2 of the previous publication17, with apparent half-lives of uptake of ~9 min for EGFP-CDMPR and ~4 min for TfR-EGFP, and saturation after ~43 min and ~20 min, respectively. These values were comparable with values obtained from biochemical uptake experiments using immunoblotting assays17. In principle, the kinetics of retrograde transport into subcellular regions of interest, such as the perinuclear region of the highest concentration of MPRs, can also be analyzed. However, the perinuclear region contains not only Golgi/TGN but is also enriched in late endosomes and recycling endosomes. The kinetics of nanobody uptake into the perinuclear region are not sufficiently specific to analyze retrograde transport to a defined organelle. To more reliably assess plasma membrane-to-TGN transport, another fluorescent protein, a TGN-resident transmembrane protein, must be stably co-expressed along with the GFP reporter protein, in a way that there is no spectral overlap of fluorophore properties. Using this extra marker allows the detection of reporter-imported VHH-mCherry, analogous to tyrosine sulfation mediated by TPSTs, as previously documented19.

Figure 1: Design and production of derivatized nanobodies for tracking EGFP-tagged cell surface proteins. (A) Schematic overview of the derivatized nanobodies. The standard nanobody construct comprises a GFP-specific VHH domain, T7 and HA epitope tags, a biotin acceptor peptide (BAP), and a C-terminal hexahistidine (His6) tag for purification. Additional nanobody variants include modifications with tandem tyrosine sulfation sites (2xTS), the engineered peroxidase APEX2, or the fluorescent protein mCherry. Scale bar indicates amino acid (aa) length. (B) Bacterially expressed and affinity-purified nanobodies (20-50 µg) were analyzed by gradient SDS-PAGE and visualized by Coomassie staining. Molecular weight standards (in kDa) are indicated on the left. Minor proteolytic clipping was observed only for VHH-mCherry, likely occurring between the VHH and mCherry domains. (C) Immunoblot analysis of nanobody preparations (10 ng) using antibodies directed against T7, HA, or His6 epitopes, or detected by streptavidin-HRP (SA-HRP) for biotinylation assessment. (D) The extent of nanobody biotinylation was evaluated by incubating nanobodies at a 1:1 ratio with BSA, followed by streptavidin-agarose pulldown, pelleting, and washing of the beads. Equal volumes of the supernatant (S) and the bead-bound material (B) were subsequently analyzed by SDS-PAGE and Coomassie staining. Quantitative recovery of the nanobody in the bead-bound fraction indicates complete biotinylation. The presence of both VHH and mCherry fragments in the bound fraction suggests partial degradation, likely occurring during sample preparation for SDS-PAGE analysis. The white line between lanes 2 and 3 indicates the deletion of two unrelated lanes. This figure has been modified from17. Please click here to view a larger version of this figure.

Figure 2: Expression and intracellular localization of fluorescently labeled EGFP-tagged cargo receptors. (A) Schematic representation of the EGFP fusion constructs. Sequences derived from secretory transmembrane proteins are depicted in black, with N-terminal signal peptides and internal transmembrane domains highlighted in yellow. The EGFP moiety is illustrated in green. Full-length constructs were generated for CDMPR, TfR, and TGN46, while a well-characterized truncated variant was used for CIMPR to preserve normal trafficking behavior. Scale bar indicates amino acid (aa) length. EGFP-CDMPR and TfR-EGFP have been described previously17. (B) To assess subcellular localization, HeLa cells stably expressing EGFP fusion proteins were co-cultured with parental HeLa cells and analyzed by fluorescence microscopy. Scale bar represents 10 µm. (C) HeLa cells stably expressing EGFP-tagged reporter proteins were lysed, and protein extracts were subjected to SDS-PAGE followed by immunoblotting using antibodies against GFP and actin. Molecular weight markers (in kDa) are indicated on the right. This figure has been modified from17. Please click here to view a larger version of this figure.

Figure 3: Live-cell imaging of nanobody endocytic uptake kinetics mediated by EGFP-CDMPR and TfR-EGFP. Live-cell imaging was performed following the addition of 25 nM VHH-mCherry to HeLa cells stably expressing (A) EGFP-CDMPR or (B) TfR-EGFP in phenol red-free complete medium at 37 °C. Cells were imaged in the GFP and mCherry channels using a semi-automated widefield fluorescence microscope at 36 s intervals. Representative merged still images are shown, accompanied by enlarged views of the perinuclear region (magnification: 2.2x) in individual channels below. (Scale bars, 10 µm.) See also Video 1 and Video 2. Quantitative analysis of nanobody uptake kinetics in (C) EGFP-CDMPR and (D) TfR-EGFP was performed using data from three independent experiments, each capturing approximately 40 individual cells. To account for organelle movement and focal plane variability, the mCherry fluorescence intensity was normalized (norm) to the GFP signal and plotted as the mean ± SD across all cells from the three experiments. The average maximal uptake signal was normalized to 1. Uptake kinetics for each cell were individually fitted using a first-order kinetic model. The plotted lines represent the averaged curves derived from the mean of the corresponding rate constants. This figure has been modified from17. Please click here to view a larger version of this figure.

Video 1: Live-cell imaging of mCherry-nanobody uptake by EGFP-CDMPR. HeLa cells stably expressing EGFP-CDMPR were incubated at 37 °C in complete medium containing 25 nM VHH-mCherry and imaged in both the EGFP and mCherry fluorescence channels at 36 s intervals. The movie was rendered at a playback rate of 5 frames/s, corresponding to a real-time compression of 3 min/s. This movie has been modified from17. Please click here to download this Video.

Video 2: Live-cell imaging of mCherry-nanobody uptake by TfR-EGFP. HeLa cells stably expressing TfR-EGFP were incubated at 37 °C in complete medium supplemented with 25 nM VHH-mCherry and imaged for EGFP and mCherry fluorescence at 36 s intervals. The movie was rendered at 5 frames/s, representing a time-lapse rate of 3 min/s. This movie has been modified from17. Please click here to download this Video.

Nanobodies represent a rapidly emerging class of protein-binding scaffolds that offer numerous advantages over conventional antibodies: they are small, highly stable, monomeric, non-crosslinking, lack disulfide bonds, and can be selected for high-affinity binding²⁴. In our previous study, we developed and applied functionalized nanobodies produced by bacterial expression to label surface-exposed proteins and monitor their intracellular trafficking to various compartments, with a particular focus on retrograde transport to the trans-Golgi network (TGN)¹⁷. We utilized an anti-GFP nanobody to ensure versatility, enabling the targeting of fusion proteins tagged with extracellular GFP, YFP, or mCerulean -- fluorescent tags that are widely available and often already characterized. Functionalization of the anti-GFP nanobody with mCherry, APEX2, or tyrosine sulfation motifs allowed us to visualize cargo uptake by fixed and live-cell imaging, localize retrograde transport compartments at the ultrastructural level, perform targeted compartment ablation, and quantitatively assess transport kinetics to the TGN. One limitation of this approach is the prerequisite of genetic modification of the target cell lines -- either by stable overexpression or endogenous tagging -- before functionalized anti-GFP nanobodies can be effectively employed. However, this strategy can be readily adapted to the growing repertoire of nanobodies directed against untagged, endogenous target proteins generated through animal immunization or in vitro selection from synthetic VHH libraries using ribosome or phage display technologies³⁴. In addition, nanobodies targeting commonly used epitope tags, such as mCherry, V5, or ALFA, have been developed in recent years^35,36,37. Notably, we recently utilized a sulfation-competent anti-mCherry nanobody to investigate retrograde trafficking of CDMPR in GGA2-depleted HeLa cells through rapid functional inactivation²⁸. Small epitope tags offer the advantage of preserving the native conformation and function of the target protein, as they are less likely to interfere with critical structural domains. Nanobodies recognizing these compact epitopes (e.g., ALFA, V5) can be seamlessly integrated into the functional nanobody toolkit to extend its utility in studying endocytic and retrograde trafficking mechanisms.

In the present study, we demonstrate the application of VHH-mCherry in live-cell imaging experiments. Using EGFP-CDMPR and TfR-EGFP as reporter proteins, we were able to visualize receptor-specific endocytic uptake and their progression toward steady-state distribution. Quantitative analysis of the trafficking of these reporters from the plasma membrane to their respective intracellular compartments revealed kinetics consistent with those obtained through biochemical assays presented in Figure 2 of a previous study¹⁷. We found that the apparent half-life was considerably longer for EGFP-CDMPR (~9 min) than for TfR-EGFP (~4 min), with the signals becoming saturated after about 43 min and 20 min, respectively. These quantitative findings align well with previously published data on MPR and TfR trafficking dynamics^38,39,40. Notably, the VHH-mCherry construct has garnered significant attention from the research community, as evidenced by numerous Addgene distribution requests, and has already been successfully implemented in several independent studies^{28,41,42,43,44}.

Depending on the microscope configuration or experimental requirements, mCherry-tagged anti-GFP nanobodies may not always be optimal. In such cases, alternative fluorophores or fluorescent protein variants may provide more suitable spectral properties when used in combination with GFP reporter cell lines. To address this, we established a library of functionalized nanobodies incorporating a diverse range of fluorescent proteins, including mTagBFP2, mNeptune2.5, mCardinal2, TagRFP657, and iRFP670 (243764-68). These additional nanobody constructs will be made available through Addgene. The use of spectrally distinct fluorophores enables clear separation of excitation and emission signals, thereby facilitating optimal live, real-time imaging of up to three fluorescence channels with minimal spectral overlap.

While monitoring endocytic uptake to steady state may appear straightforward, tracking retrograde transport of proteins to the TGN presents a greater experimental challenge. In addition to VHH-mCherry and a GFP-tagged reporter protein, an additional fluorescent or dye-conjugatable TGN-resident marker is required for visualization and quantification. However, certain Golgi-resident proteins exhibit partial leakiness to the cell surface, which can be further exacerbated by perturbations of intracellular trafficking pathways, such as those introduced by experimental manipulation of transport machinery components. TGN46, for instance, is commonly used as a Golgi-resident marker despite its functional role as a cargo receptor mediating TGN-to-cell surface transport of certain secretory cargoes^16,45. Similarly, glycosyltransferases such as β-1,4-galactosyltransferase have also been detected at the plasma membrane⁴¹. Although these proteins can cycle to the surface, their predominant steady-state localization remains within the TGN. Tagging such markers with a spectrally distinct fluorescent protein or a dye-compatible tag (e.g., SNAP, Halo, etc.) and stably co-expressing them with a GFP-tagged transmembrane reporter would enable quantitative live-cell imaging of TGN arrival. This could be measured analogously to the workflow described above, with quantification based on the accumulation of mCherry signal within the fluorescence channel corresponding to the Golgi-resident fusion protein. Current strategies for investigating retrograde trafficking in fixed or live cells often rely on chimeric receptor constructs in combination with IgG antibodies (e.g., CD8)^46,47. However, the inherent risk of cross-linking due to the bivalency of conventional IgGs during surface labeling can significantly alter the intracellular fate of the labeled protein, frequently redirecting it toward lysosomal degradation after endocytosis^48,49. In contrast, nanobodies, due to their monovalent nature and 1:1-binding stoichiometry, do not induce crosslinking, thereby preserving the physiological trafficking routes of the target protein. This makes nanobody-based approaches particularly valuable for studying authentic retrograde transport dynamics without introducing artificial routing artifacts.

Optimal performance of this protocol for live-cell imaging requires careful titration of nanobody concentration, as excessive amounts lead to elevated background fluorescence from unbound nanobody in the medium. Furthermore, differences in expression levels of GFP reporter cell lines must be accounted for during titration to ensure accurate quantitative comparison. For nanobody production, low-temperature induction in E. coli is critical to promote proper folding of the mCherry domain. Higher induction temperatures typically result in misfolded or degraded protein and are, therefore, not recommended.

The success of this protocol critically depends on three key aspects: first, the high-yield expression and purity of functional nanobody constructs to ensure consistent labeling and minimal background; second, the use of well-characterized cell lines stably expressing GFP-tagged reporter proteins with cell surface localization to recapitulate physiological trafficking pathways; and third, optimal handling and configuration of the fluorescence microscopy setup, including imaging parameters and spectral separation, to enable accurate quantitative and qualitative analysis of cargo transport dynamics.

The method and protocol presented here, centered on the use of mCherry-tagged anti-GFP nanobodies, enable both quantitative and qualitative tracking of cargo proteins from the cell surface to endocytic compartments in cultured mammalian cells. When co-expressed with a fluorescently labeled Golgi-resident marker, this approach further allows the analysis of retrograde transport pathways. As such, this tool provides a powerful strategy to monitor the intracellular trafficking of cell surface transmembrane proteins and receptors that were previously inaccessible to labeling due to the absence of specific ligands or compatible monomeric binding reagents.