Illuminating T cell-dendritic cell interactions in vivo by FlAsHing antigens

Delineating the complex network of interactions between antigen-specific T cells and antigen presenting cells (APCs) is crucial for effective precision therapies against cancer, chronic infections, and autoimmunity. However, the existing arsenal for examining antigen-specific T cell interactions is restricted to a select few antigen-T cell receptor pairs, with limited in situ utility. This lack of versatility is largely due to the disruptive effects of reagents on the immune synapse, which hinder real-time monitoring of antigen-specific interactions. To address this limitation, we have developed a novel and versatile immune monitoring strategy by adding a short cysteine-rich tag to antigenic peptides that emits fluorescence upon binding to thiol-reactive biarsenical hairpin compounds. Our findings demonstrate the specificity and durability of the novel antigen-targeting probes during dynamic immune monitoring in vitro and in vivo. This strategy opens new avenues for biological validation of T-cell receptors with newly identified epitopes by revealing the behavior of previously unrecognized antigen-receptor pairs, expanding our understanding of T cell responses.


Introduction
The interactions between T cells and APCs form the cornerstone of adaptive immunity.Within the intricate milieu of lymphoid tissue, T cell priming stands out as a discrete and precise process, occurring amidst the multitude of dynamic cellular interactions.Multimerized peptide-major histocompatibility complex (pMHC) labeling has been the gold standard tool for capturing antigen-speci c T cells from a polyclonal repertoire [1].However, such multimers, most commonly tetramers, have limitations in monitoring dynamic interactions of antigen-speci c T cells in real-time due to their blocking effect on the immune synapse.Likewise, currently available T cell receptor like (TCRL), also known as peptide-ingroove (PIG), antibodies that detect pMHC complexes are known to disrupt the immune synapse, making them unsuitable for real-time monitoring of antigen-speci c T cell interactions [2,3].Although alternative approaches for labeling antigenic peptides with conventional tags such as uorescent proteins, uorophores, and biotin-streptavidin have been used, these tags have limitations such as size constraints, membrane impermeability, and photophysically unstable nature, which restrict their use in applications for dynamic monitoring [4][5][6][7].
To overcome these drawbacks, we have developed a highly speci c and minimally disruptive labeling strategy for studying antigen-speci c T cell-APC interactions.Our approach involves introducing a sixamino acid tetracysteine tag (CCPGCC) on antigenic peptides and inducing uorescence using membrane permeable thiol-reactive Arsenical Hairpin (AsH) probes [5,8].We have engineered the OVA (323- 339) peptide by placing the tetracysteine tag at either the amino or the carboxy terminus, away from the TCR and MHC binding sites, to avoid disrupting T cell-APC synapse in vitro and in vivo [9].Owing to these modi cations, our labeling strategy preserves the T cell priming ability and allows for the detection of the uorescent signal on the surface of APCs in a MHCII-dependent manner.Using ow cytometry and live microscopy, we found that the labeled peptide is readily taken up by the cognate CD4 + T cells with precise speci city in vivo, as previously shown for pMHC complexes taken up from the immune synapse via processes such as trogocytosis and transendocytosis [10][11][12][13][14]. Furthermore, we have created new variants of the tetracysteine-tagged OVA (323-339) peptide with altered TCR-binding sites and have shown that the AsH probe labeling can track the TCR mediated uptake of nuanced pMHC complexes.This indicates that our labeling strategy is highly robust and forges a novel path for investigating immune responses against altered antigens such as those derived from self, microbiota and tumor microenvironment.Overall, AsH probe mediated immune monitoring offers signi cant advantages over existing methods, making it a versatile tool for in vitro and in vivo applications including live imaging of the T cell synapse within intact tissue architecture.This novel approach will further our understanding of T cell-mediated immune responses by also complementing tetramers for sorting antigen-speci c T cells with precision and extending it to newly discovered epitopes.

Results
Signal intensity is determined by the sequence and placement of tetracysteine tag We rst generated OVA 323-339 variants by placing a tetracysteine tag "CCPGCC" at amino or carboxy termini, either consecutively or via a exible ε-amino-caproic acid (ε-ACA) linker to prevent endosomallysosomal enzymatic digestion to the tag.This yielded four variant peptides named as C-(carboxy terminus without linker), N-(amino terminus without linker), NACA-(amino terminus with ε-ACA) and CACA-(carboxy terminus with ε-ACA) (Fig 1a).We used these variants to load mature C57BL/6 CD11c + splenic DCs from C57BL/6 mice and induced uorescence by treating cells with Fluorescein-AsH (FlAsH) probe.Flow cytometry analysis showed that CACA-peptide combination but not the other tested variants induced a distinguished uorescent signal (Fig 1b).Cysteine is naturally oxidation prone and this may reduce FlAsH binding to tetracysteine tag.We addressed this problem by pre-treating OVA CACA with tris-2carboxyethylphosphine (TCEP), an irreversible reducing agent that blocks cystine disul de bridge formation and elicited a consistent FlAsH signal (Fig. 1c).Moreover, we used another AsH probe excitable by 561 nm laser, called ReAsH, to occupy non-speci c cysteine residues that would otherwise attract FlAsH [15].Pre-incubation of DCs with ReAsH improved signal to noise ratio signi cantly (Fig 1d).We have created additional OVA variants to enhance signal intensity.These variants include one with two tetracysteine residues and another with a longer tetracysteine tag that withstands the British anti-Lewisite (BAL) wash, a crucial step to reduce background labeling by removing excess AsH probe [16].However, new variants provided FlAsH intensity comparable to the original CACA variant, indicating no additional bene t for longer tags (Fig 1e).
Tetracysteine tag does not alter the immunogenicity of peptide Next, we sought to compare the CD4 + T cell priming ability of four tetracysteine-tagged OVA 323- 339 variants and that of native peptide in vitro.To do so, naïve CD4 + T cells were isolated from OVA 323-339 speci c TCR transgenic OTII mice, labelled with a proliferation dye and cocultured with mature CD11c + DCs that were pre-loaded with equimolar amount peptides for three days.T cells served as a negative control due to lack of MHCII expression in mouse T cells [21] (Fig. 4b).It is worth mentioning that streptavidin detection of biotinylated peptide is a viable option for quantifying peptide loading on cells.However, size and membrane impermeable nature of streptavidin make it unsuitable for assays that include live monitoring of the T-DC interaction [22].Next, we addressed whether FlAsH labeling of OVA CACA -pulsed DCs is dependent on the MHC expression.CD11c + DCs were isolated from MHCII single knockout or MHCI and II double knockout (KO) mice, labeled with tracking dyes and mixed 1:1 with WT C57BL/6 DCs.Subsequently, they were incubated with OVA CACA and treated with ReAsH and FlAsH as in Fig. 1 (Fig 4e).To make a fair comparison between WT and KO DCs and minimize auto uorescence, we performed adoptive transfers, aiming to facilitate in vivo elimination of apoptotic and possibly defective KO DCs [23,24].Flow cytometry analysis of the draining lymph node demonstrated that WT DCs, but not MHCII KO DCs, harbored FlAsH.Two-photon microscopy of the lymph node further con rmed these ndings, indicating that FlAsH robustly marks OVA CACA -MHCII complexes To address how FlAsH acquisition from synapse marks antigen-speci c vs. bystander T cells, DCs were labeled with ReAsH and cell membrane dye PKH-26, pulsed with OVA CACA , Eα (52-68) or left unpulsed.Next, they were treated with FlAsH and injected into the recipient mice via footpad.Subsequently, 1:1 mixture of labeled naïve OTII and Eα-speci c T cells were adoptively transferred into via retroorbital sinus i.v. and the popliteal lymph nodes of the recipients were removed 36-42 h after the transfer (Fig. 5a).Flow cytometry analysis showed that OTII and Eα-speci c T cells both acquired PKH-26 from double-pulsed DCs, indicating that they captured DC cell membrane from synapse (Fig 5b).This process is termed as TCRmediated trogocytosis and has previously been described as a feature of the DC-T cell synapse [11].Tetracysteine-tagging of peptides provides a sensitive tool to visualize interactions with altered TCR a nity The lack of adaptable antigen-targeted probes presents a signi cant obstacle in the dynamic visualization of antigen-speci c immunity.To address this gap, we sought to determine robustness of tetracysteine tagging in visualizing nuanced pMHCII variants.We created three variants by introducing alanine mutations at the TCR binding residues 331, 333, and 335 of OVA CACA (Fig 6a).The mutations at 331 and 333, depicted as variants 1 and 2 respectively, were previously reported to abolish the binding of OTII TCR to OVA 323-339 [9].Although the residue 335 is anked between the core OTII TCR binding residues 333 and 336, mutant 335 was shown to elicit normal proliferative response in vitro [9].We addressed how pMHCII uptake by T cells is in uenced by the mutation of TCR binding residues of OVA CACA .DCs were isolated from CD45.1 congenic mice, loaded with variant peptides and cocultured with naïve CD45.2 OTII T cells for three days (Fig 6b-e).In line with earlier observations, variants 1 and 2 interrupted T cell priming while variant 3 led to a proliferative response similar to native peptide (Fig 6cd).TCR mediated uptake of the FlAsH-labeled pMHCII complexes followed the same pattern and only variant 3 complexes produced a comparable signal to that of OVA CACA (Fig 6e).These suggest that AsHprobe labeling can be a sensitive method to track low a nity/ nuanced pMHCII complexes.

Discussion
Dynamic monitoring of antigen-speci c T cell responses in vivo poses challenges due to the limited availability of reagents and the size constraints of the immune synapse.Here, we address this issue by presenting a novel strategy for visualizing T cell-APC interactions via tetracysteine tagging of antigenic peptides.While tetracysteine tagging has been previously applied in cell biology and virology, these applications have been mostly limited to cell lines and in vitro studies in immunology unrelated experimental settings [25][26][27][28].On the other hand, our technique labels antigenic peptides and tracks pMHCII complexes in primary APCs during their interactions with T cells both in vitro and in vivo.Our approach is superior as compared to other strategies for monitoring T cell-DC interactions, in the way it robustly targets speci c pMHCII complexes in real time in vivo.There is no need for special organisms or genetic modi cations, and the technique can easily be expanded to new antigens and T cell-APC pairs from both mice and humans.This strategy can, for instance, be integrated into validation efforts for antigen discovery and utilized in the investigation of the synapses of newly deorphanized human T cells, a largely unexplored area where tools are scarce [29][30][31][32].Given that is a two-part system (tag and FlAsH), it also carries the added capability of being inducible at different time points to allow for increased exibility in interrogating pMHCII-TCR interactions over time.On the other hand, this introduces a high background due to the non-speci c binding of FlAsH to endogenous cysteines.Thus, we implemented a ReAsH pretreatment strategy, successfully improving the signal-to-noise pro le for FlAsH.It's worth mentioning our unpublished attempts to load dendritic cells (DCs) with preconjugated peptide and FlAsH to mitigate the background.Unfortunately, this endeavor led to almost a complete loss of the FlAsH signal and has not been pursued further.Therefore, exploring new peptide probes chemically conjugated to small organic dyes emerges as a potential avenue for future investigation.
Most important of all, our experiments demonstrate that tetracysteine-tagged peptides can be transferred from DCs to T cells via synaptic processes, such as antigen-speci c pMHCII trogocytosis/transendocytosis, marking acceptor T cells for several days.Although the turnover of pMHCII in acceptor T cells and its temporal relationship to the TCR downstream events require further elucidation, our visualization of tetracysteine-tagged pMHCII can complement tetramers in detecting antigen-speci c T cells in vivo.Additionally, we show that tetracysteine tagging of pMHCII does not broadly alter or disturb T cell-DC synapse by functional read-outs.Although the ratio of peptide-FlAsH conjugates to free peptide in DCs would be needed to support this claim, our approach offers an advantage over tetramers in that it is suitable for real-time monitoring of the immune synapse.Finally, it is worth noting that tetracysteine tags can be visualized as electron-dense particles by Transmission Electron Microscopy [33,34], providing potential for the detailed investigation of T-DC synapse and subcellular localization of pMHCII complexes.Although TEM application was out of scope for this study, it offers a promising avenue for future research.In conclusion, the tetracysteine labeling approach described here provides a powerful and unique tool to advance our understanding of T cell-APC interactions, offering new opportunities to investigate the immune response and develop improved immunotherapies.
Cell cultures were maintained in sterile complete Roswell Park Memorial Institute (RPMI) media.The RPMI 1640 medium was supplemented with 10% heat-inactivated fetal bovine serum, 50 U/ml penicillin, 50 µM streptomycin, 1 mM sodium pyruvate, 2 mM L-glutamine, 0.1 mM non-essential amino acids, 50 µM 2-mercaptoethanol, and 10 mM HEPES (Thermo Fisher Scienti c).For magnetic separations, cells were maintained in ltered and degassed magnetic-activated cell sorting (MACS) buffer.The MACS buffer consisted of PBS (Lonza) supplemented with 0.5% BSA (Sigma-Aldrich) and 2 mM EDTA (Sigma-Aldrich).For splenic dendritic cell (DC) isolation, Liberase TM (Blendzyme) and DNase I were purchased from Roche.50 mg Liberase was dissolved in complete RPMI, resulting in a nal concentration of 5 mg/mL.The solution was kept on ice, stirred gently every 5 minutes for 30 minutes and aliquoted into 1 mL portions.10 mg/mL DNAse stock solution was prepared with 0.15 M NaCl and divided into 500 µL aliquots.Both enzymes were stored at -20ºC.
For AsH probe labeling, peptides were pre-treated with fresh Tris Carboxy Ethyl Phosphene (TCEP, Sigma-Aldrich) to protect tetracysteine residues from oxidation.10 mM ReAsH stock solution was prepared by adding 45 µL sterile cell-culture grade DMSO (Sigma-Aldrich) to 250 µg solid ReAsH (Cayman Chemical) and 10 mM FlAsH solution was prepared by adding 150uL DMSO to 1 mg FlAsH powder (Cayman Chemical).British Anti-Lewisite (BAL, Alfa Aesar) and HBSS with Ca ++ and Mg ++ (Thermo Fisher Scienti c) was used to wash excess FlAsH off the cells.BAL is resuspended in sterile degassed water to obtain 8 mM stock solution.FlAsH, ReAsH and BAL are highly sensitive to oxidation, therefore their activity may decline in subsequent uses.This may be averted, at least in part, by quickly purging stock solution containers with 4M Argon gas (Sigma-Aldrich, 501247), placing container in a 50mL tube, purging again with Argon before storing FlAsH and ReAsH at -20ºC, BAL at 4ºC.We do not recommend storage longer than one month or repeated use of the same reagent stock.

DC isolation, peptide loading and -AsH probe labeling
Mature splenic DCs were isolated as previously described [35] and resuspended in complete RPMI.Peptides were added into 20 mM TCEP solution at twice the desired concentration for loading DCs.The solution was vortexed thoroughly and incubated at 37ºC for 30 minutes.Meanwhile, ReAsH stock solution was diluted to 100 uM in complete RPMI and added on DC suspension to obtain 1 uM nal ReAsH concentration.Cells were incubated at 37ºC for 30 minutes and washed twice with complete RPMI.ReAsH labeled DCs were mixed with equal volume of peptide-TCEP solution, incubated at 37ºC for 30 minutes, washed twice and resuspended in complete RPMI to proceed with FlAsH labeling.Flash was added to cell suspension to achieve 1 uM nal concentration, incubated at 37ºC for 30 minutes and washed twice with complete RPMI.To remove excess FlAsH and reduce non-speci c binding, cells were resuspended in 240 uM BAL diluted in HBSS with Ca ++ and Mg ++ , washed twice, then washed a third time with PBS and resuspended in PBS.

DC-T cell cocultures
Freshly isolated DCs were pulsed with peptide in complete RPMI and were incubated at 37 °C for 30 min followed by three washes with complete RPMI to remove unbound peptide.Pulsed DCs were then treated with -AsH as decribed above then seeded into at-bottom 96-well plates (Corning) at a density of 5 × 104 to 1 × 105 cells per well.T cells and iTreg cells were labeled with e450 or e670 (eBioscience) according to the manufacturer's protocol.Cells were cocultured at 37°C for for indicated periods, at a 1:10 ratio of DCs to T cells for proliferation assays and 1:1 for trogocytosis assays.

Trogocytosis assay
Trogocytosis assays were perfomed as described in [20].Brie y, freshly isolated splenic DCs were labeled with 4μM PKH26 (Sigma-Aldrich) for labeling of the cell membrane as described in Puaux et al.48.DCs were then pulsed with 10 μM peptide, treated with -AsH and were cocultured with T cells for up to 18 h.Cell conjugates were dissociated by washing cells with MACS buffer with 2 mM EDTA; cell suspensions were prepared for ow cytometry.

Flow cytometry
Cells were stained in FACS buffer containing uorochrome-conjugated antibodies for 30 minutes at 4°C, protected from light.Data acquisition was performed with BD LSRFortessa and BD LSR II cytometers (BD Biosciences).Data analysis was performed with FlowJo v10.8.1.

Adoptive transfer
Mice were anesthetized in an induction chamber with 1.5% iso urane USP (Baxter).Pulsed DCs were adoptively transferred through the right footpad in 50uL sterile PBS.T cells labelled with e450 or e670 (Thermo Fisher Scienti c) were injected intravenously in 100uL sterile PBS through the retro-orbital sinus.
Confocal microscopy 35mm dishes with a 14mm glass coverslip bottom (MatTek) were pre-treated with 10ug/ml bronectin (Sigma-Aldrich) in PBS for 1h at room temperature and washed twice.Splenic DCs were added onto the glass coverslip in complete RPMI and pulsed with 5-10 μM peptide at 37°C for 1 hour.Excess peptide was washed off with complete RPMI before addition of CD4 T cells.Co-cultures were imaged for 3-18h using a Leica SP-8 inverted microscope (Leica Microsystems) equipped with a full range of visible lasers, two hybrid detectors, three photomultiplier detectors, and a motorized stage.63× objective (Leica Microsystems) was used and microscope con guration was set up for 3D analysis (x, y, z) of the cellular layer.The following lasers were used: diode laser for 405 nm excitation; argon laser for 488 and 514 nm excitation; diode-pumped solid-state laser for 561 nm; and HeNe laser for 594 and 633 nm excitation.All lasers were tuned to minimal power (between 0.3% and 2%) to prevent photobleaching.z stacking of images of 10-12 µm were collected.Mosaic images of large cell culture areas (1 mm2) were generated by acquiring multiple z stacks using the Tile scan mode and were assembled into tiled images using LAS X v4.0 (Leica Microsystems).
Intravital two-photon laser-scanning microscopy of mouse popliteal lymph nodes Mice were anesthetized in 1.5% iso urane USP (Baxter) through a rodent nose cone followed by surgery to expose the right popliteal lymph node.Warm PBS was added to maintain lymph node moisture.Mouse body temperature was maintained with an infrared blanket (Braintree Scienti c) throughout the imaging process after which they euthanized through cervical dislocation under anesthesia.Mice were imaged on a glass stage with a two-photon laser-scanning microscopy setup with a Leica SP8 inverted confocal microscope with dual multi-photon lasers, Mai Tai and InSight DS (Spectra-Physics), L 25.0 waterimmersion objective, 0.95 NA (Leica Microsystems), and a 37 °C incubation chamber (NIH, Division of Scienti c Equipment and Instrumentation Services).The Mai Tai laser was tuned to 890 nm to excite e450and FlAsH; the InSight DS laser was tuned to 1,150 nm to excite DsRed/ PKH-26/ ReAsH and e670.For time-lapse imaging, a z stack consisting of 10-12 single planes (5 μm each over a total tissue depth of 50-60 μm) was acquired at 15 second intervals for a total period of 1-4 h.

Two-photon laser-scanning microscopy of live lymph node sections
Live lymph node sections were imaged ex vivo by two-photon microscopy.To maintain tissue architecture, popliteal lymph nodes were immediately transferred to PBS with 1% BSA and kept on ice immediately following harvest.Residual fat and connective tissue were removed using a Leica MZ6 StereoZoom microscope (Leica Microsystems).Lymph nodes were suspended in 38 °C agarose in DMEM and kept over ice.Following complete agarose solidi cation, complete RPMI was added and blocks were cut containing one lymph node each.Lymph node blocks were cut into 200 µm sections with a Leica VT1000 S vibrating blade microtome (Leica Microsystems) at speed 5 in ice-cold PBS and frozen at -80 °C.Prior to imaging, complete RPMI medium was added to tissue sections and incubated at 37 °C for 2 h.Sections were held down with tissue anchors (Warner Instruments) in 35mm dishes with a 14mm glass coverslip bottom (MatTek) and were imaged with a Leica SP8 inverted ve-channel confocal microscope equipped with an environmental chamber (NIH, Division of Scienti c Equipment and Instrumentation Services) and a motorized stage.The microscope con guration was set up for four-dimensional analysis (x, y, z, t) of cell segregation and migration within tissue sections.A diode laser for 405 nm, an argon laser for 488 and 514 nm, a diode-pumped solid-state laser for 561 nm, and a HeNe laser for 594 and 633 nm excitation wavelengths were tuned to minimal power (between 0.3% and 2%).z-stack of images of 10-25 µm were collected.Mosaic images of whole lymph nodes were generated by acquiring multiple z stacks using a motorized stage to cover the whole lymph node area and assembled into a tiled image with the LAS X software.For time-lapse analysis of cell migration, tiled z stacks were collected over time (1-4 h).

Image analysis
Analysis of confocal and intravital two-photon images was performed using Imaris v9.2.1 and v10.0.0 (Bitplane) as previously described [35].In brief, e450 and e670 signals were utilized for reconstructing the 3D structure of T cells and iTreg cells as surface objects.To reconstruct the 3D structure of DCs, overlapping DsRed, PKH-26, and ReAsH signals were employed.Imaris v10.0.0's built-in algorithms were used to calculate 3D volumes, track velocities, surface-to-surface colocalization parameters, and FlAsH Although both T cell types trogocytosed double-pulsed DC cell membrane, FlAsH signal was picked up only by OTII T cells, suggesting that FlAsH tracks pMHCII mobility with a high precision (Fig 5c).Furthermore, we visualized dynamic CD4 + T cell-DC contact and pMHCII mobility in vivo at 18 h after the transfer.DCs were labeled with ReAsH and pulsed with OVA CACA and/or LCMV GP (61-80) and treated with FlAsH.Recipient mice were injected with unpulsed, OVA CACA single or OVA CACA -GP (61-80) double pulsed DCs via footpad along with a 1:1 mixture of labeled naïve OTII and SMARTA T cells i.v.via retroorbital sinus.Popliteal lymph node microscopy con rmed that OTII and SMARTA T cells had stable contacts with double-pulsed DCs (Fig 5d, Supplementary video 2).Similar to Eα-speci c T cells, SMARTA also failed to pick up FlAsH signal despite showing signs of antigen-speci c contact such as increased volume, reduced sphericity and decreased track speed (Fig 5 e-f).

(
green) intensities of cells.Statistical analysis Statistical analysis was conducted on Prism v9.5.1 (GraphPad).Statistical tests and relevant P values for each gure are indicated in the gure or gure legend.Declarations Author Contributions B.A. conceived the project; B.A. and M.A. designed the experiments and wrote the manuscript; M.A., J.A.S., D.W., R.K., and O.K. performed the experiments; B.A. and J.K. analyzed the data; E.M.S., J.A.S. and, D.W. edited the manuscript.

Figures
Figures

Figure 1 FlAsH
Figure 1

Figure 3 T
Figure 3

Figure 5 FlAsH
Figure 5

Figure 6 Nuanced 5 CD45. 1 +
Figure 6 Proliferation and activation status of OTII T cells were not altered by the presence or placement of the tag (Fig 2a-d).Subsequently, DCs were pulsed with OVA CACA vs. native peptide, treated with FlAsH and cocultured with uorescently labeled OTII T cells for live confocal microscopy imaging.Within the rst hours of coculture, OTII T cells established stable contact with OVA CACA -pulsed DCs and FlAsH-labeled peptide gradually concentrated at the contact site (Fig 2e).At 24 hours after treatment, OTII T cells in the peptide-treated groups exhibited increased volume and a shift towards a less spherical shape, indicative of T cell blasting (Fig 2f).To delineate whether FlAsH labeling reports free vs. MHCII-bound peptide in cells, we initially assessed how surface MHCII expression of cells correlates with the OVA CACA -FlAsH signal.C57BL/6 splenocytes were incubated with OVA-Biotin or OVA CACA and peptide distributions among the T cells, B cells and DCs were quanti ed by ow cytometry using streptavidin or FlAsH (Fig 4a).Both streptavidin and FlAsH signals correlated signi cantly with MHCII expression, providing the strongest signal in DCs, where CD4 +