The handle http://hdl.handle.net/1887/137749 holds various files of this Leiden University dissertation.
Author: Lahav-van der Gracht, A.M.F.
92
5
Exploration of Bioorthogonal Click Chemistry for the
visualization of MHC-I Epitope SIINFEKL on Cells
5.1 Introduction
Bioorthogonal click chemistry, which in previous chapters has been used for click to release (CtR), was originally designed for the selective ligation of two chemical groups in/on a biological sample.1 The approach, initially developed by Bertozzi and co-workers, relies on some key parameters for success: both molecules have to be mutually reactive, but should not react with other molecules within a cell.2 The product(s) of the reaction should also be stable, non-toxic in a cellular environment, and the reaction should be very specific and have a high reaction rate.3 Over the last two decades, multiple reactions have been developed4, such as the Cu(I)-catalyzed Huisgen cycloaddition (CCHC) between an azide and alkyne,5 and the inverse electron demand Diels-Alder (IEDDA) reaction,6 which will be used in this Chapter.
Previously, the use of the CCHC for the use of bioorthogonal groups in the detection of epitopes has been reported (Figure 1).5 This was only successful for the detection of exogenously loaded peptides in the MHC-I over-expressing cell line RMA-S11. The CCHC click on epitopes was further explored by the van den Bogaart-group to determine MHC-II loading levels.12 The main requirement for the use of bioorthogonal groups in the detection of epitopes is that the handles are small, stable, and the reaction proceeds with low background.5
93
Figure 1 | Bioorthogonal ligation strategy with CCHC. SIINFEPgL was successfully labeled inside
MHC-I complexes on RMA-S cells. Visualization was only possible with azide-CalFluor-488.5
In this Chapter the further exploration of CCHC click on epitopes is described, to allow for detection of lower levels of pMHCs on antigen presenting cells. Secondly, the IEDDA is also explored with the aim of live cell compatible labeling.13
CCHC Ligation
CCHC-reactions can be performed on either incorporated azide or alkyne handles with the opposing reaction partner. Both these handles are small in size, stable and biologically inert. 3,14 The azide handle has been successfully introduced in DNA15, proteins16 and lipids17 in cells for visualization by means of fluorophores covalently attached through click chemistry. Furthermore, it is very unlikely for an azide to change the properties of the molecules they are incorporated in, due to its low polarity and unlikeliness to form hydrogenbonds.3 Therefore, this molecule is very suitable for bioorthogonal labelling. The reaction kinetics of the CCHC reaction between an azide and an alkyne (k ≈ 10-100 M-1s-1)18 are very fast and therefore the CCHC reaction is quite often referred to as the model example of a click reaction.19
94
IEDDA ligation
The inverse-electron demand Diels-Alder reaction is the reaction between an electron-poor diene such as a 1,2,4,5-tetrazine and an electron rich, strained alkene.13 Its use as a bioorthogonal reaction was first reported by Fox and co-workers21, and Weissleder and co-workers.22 The major advantages of these reactions as bioorthogonal ligation reactions are that the reactions are fast and non-toxic, allowing live cell labeling.3 The reaction kinetics are accelerated by introducing ring strain into the alkenes, with rates as fast as 3*106 M-1s-1 having been reported for a dipyridinyltetrazine reacting with a cis-fused cyclopropane-transcyclooctene.23 The initial product of a transcyclooctene tetrazine (TCO/Tz) reaction is a 4,5-dihydropyridazine, this reacts further to form the final 1,4-dihydropyridazine.24
The main problem when using this reaction in vivo is that the most reactive tetrazines are unstable when exposed to a cellular environment for a long time, and these small tetrazines are cleared rapidly from the body.25 The fastest reaction kinetics in vivo published so far have been between a polymer-tetrazine and transcyclooctene-antibody construct, which proceeded with a rate of >6000 M-1s-1 in the mouse.26 The in vivo instability can be overcome by the use of more electron deficient tetrazines, as they are more stable in biological media. However, concomitant lower reaction rates have to be negated for these tetrazines, by using more reactive alkenes, such as trans-cyclooctenes (TCOs).25
Figure 2 | IEDDA ligation strategy for the reaction between trans-cyclooctene-SIINFEKL and tetrazine-BODIPY-FL.
95 that were developed for the work described in Chapter 2 were used, as well as cycloalkene- and cyclopropene-modified epitope peptides (Figure 3).
N H H N N H H N N H H N N H H2N OH O O O O O O O O HO NH2 O O OH R N H HO O O R=
mbTCO (2) cyclopropene (3) spirohexene (4) O NH H N O O H N O O NH2 O HO O N H H N O N H O O NH2 H N O N H O OH O H N O N H OH O SIINFEpgL (1)
Figure 3 | Three different handles introduced on the amine of the lysine for IEDDA ligation. The
handles used were alkyne, solubilized transcyclooctene (mbTCO), cyclopropene and spirohexene.
In this Chapter modified versions of the model peptide OVA257-264 (OT-I, SIINFEKL) in which the alkyne (1), mbTCO (2), cyclopropane (3) and spirohexene
(4) have been incorporated on the lysine, have been tested for their ligation
96
fluorophore-tetrazines/fluorophore-azides and the signal to noise ratio was determined by flow cytometry. Where relevant, the percentage of elimination was determined using a T-cell activation assay. The most promising tetrazine-fluorophores were analyzed with fluorescence microscopy in order to see if it is possible to visualize the labelling of epitopes inside MHC-I.
5.2 Results and Discussion
Generally, antigen presenting cells (APCs) present between 105 - 106 MHC-I complexes simultaneously on their cell surface, of which subsets are loaded with self-peptides,31 thus the number of potential ligation handle-containing epitopes to be presented in MHC-I is limited. This necessitates highly selective chemistry to obtain detectable signal. So far this has only worked when high concentrations of peptides (far higher than those required for T-cell loading) were used. Furthermore, fluorogenic fluorophores, such as CalFluor-488-azide (Figure 1) were previously required due to high background signals obtained in the negative controls.5 It was therefore first attempted to further optimize the CCHC ligation reaction to see whether this signal to noise could be improved upon.
5.2.1 Evaluation of the CCHC click for visualization of pMHC-I
It was hypothesized that the high background signal obtained in the reported experiments was due to aspecific interactions of the fluorophores with the APC outer membranes. To first determine whether the reaction proceeded with high yield, it was assessed by fluorescent Native-PAGE analysis. To this end, the bioorthogonal epitopes SSIPgFARL (MHC-I immunodominant epitope from herpes simplex virus protein B), SIIPgFEKL and the wild type epitope SIINFEKL were refolded into the MHC-I complex.
pMHC-I complex production
97 glutathione). Addition of a specific high-affinity peptide, e.g. SIINFEKL, stabilizes the forming complex. The refolding mixture was purified using size exclusion chromatography, and the pure fractions were concentrated with an amicon-ultra 30k filter (Fisher Scientific, the Netherlands). The yield, differing between batches, was between 4-6 mg/ml and 0.5-1 ml. The purified complexes were analyzed on native-PAGE which showed single bands for all protein complexes tested (Figure 4c). The MHC-I complex containing SIINFEPgL runs lower on the native-PAGE than the complex containing SIINFEKL. This is observed for all complexes containing a propargylglycine in the peptide. The loss of one positive charge (propargylglycine instead of lysine) might be the cause of the difference in height on the gel.
Figure 4 | Production of H-2Kb and β2m in BL21 pLysS resulting in correctly folded pMHC-I.a)
SDS-PAGE analysis of protein expression in BL21 pLysS. After induction bands appear at 12 and 35 kDa, β2m and H-2Kb respectively. b) Western blot with specific antibodies against H-2Kb and
98
The refolded state of the pMHC-I complex was determined using circular dichroism (CD). Both pMHC-I complexes, containing either the wildtype epitope SIINFEKL or the substituted epitope SIINFEPgL, were analyzed using CD (Figure 5). Both spectra show similar bends and curves, resulting in similar percentages of β-sheets and α-helices corresponding well with literature.32 There is a difference between the minima values of the curves suggesting different structural stabilities for pMHC-SIINFEKL and pMHC-SIINFEPgL. However, when SIINFEKL dissociates the complex becomes unstable. In vivo, SIINFEKL has a relative fast kon (4.5-5.9 x 103 M-1s-1)33,34 and long residence time of 20-30
hours33,34 for binding to the H-2Kb. Nevertheless, as both complexes show similar structural elements, pre-dominantly β-sheet as described in the literature, it was concluded that these pMHC-I complexes were correctly folded.
Figure 5 | CD spectra of SIINFEKL or SIINFEPgL folded in MHC-I-Kb. The CD spectra of SIINFEKL
and SIINFEPgL show pre-dominantly β-sheet rich secondary structures. The values for the distribution of secondary structure elements are listed in the table on the right panel. The spectra were recorded at least as an average of 5 independent measurements. The distribution data were calculated using CDNN 2.1.
CCHC on purified pMHC-I complexes
99
Figure 6 | CCHC ligation on SSIPgFARL, SIINFEKL and SIIPgFEKL in MHCI complex. Ligation of
AF488- azide with CCHC on peptide-MHCI complexes visualized with fluorescence scanning (left) and coomassie staining (right) of native-PAGE gels. a) SSIPgFARL (barring an alkyne click handle) epitope in MHC-I complex (refolded H-2Kb-β2m with peptide obtained from bacterial culture,
purified after refolding) was clicked with AF488-azide and run on gel. b) SIINFEKL (Wt) and SIIPgFEKL (Pg4) epitopes in MHC-I (refolded complexes) were clicked with AF488-azide and run on gel.
Crystallization of pMHC-I complexes
H-2Kb has high affinity for the wildtype epitopes SIINFEKL and SSIEFARL.5 Substituting any of positions 1-7 of SIINFEKL for the non-natural amino acid propargylglycine led to complete abolishment of T-cell activation (Figure 7a).
Figure 7 | T-cell response towards propargylglycine containing epitopes. a) T-cell response
against OVA257-264 epitopes substituted with propargylglycine (Pg) for each of the positions in the
epitope shows only 20% response for position 8. b) HSV-GB498-505 epitopes with substitutions at
each of the positions shows a significant T-cell response of 60% for p2 and even slight T-cell responses for p5 and p8.
100
propargylglycine at different positions resulted in T-cell activation abolishment for positions 1, 3, 4, 6 and 7 (Figure 7b). Remarkable is the more than 60% T-cell response for p2 modification, this position is more accepting of change than expected. For position 4 and 7 it has been previously tested that any amino acid change on these positions does not interfere with MHC-I binding affinity5. Therefore, the loss in T-cell activation for both epitopes is solely caused by the TCR-pMHC-I interaction.
The T-cell hybridoma cell line B3Z,7 specific for SIINFEKL recognition, provides a good indication of T-cell response up to high picomolar/low nanomolar range of SIINFEKL. However, primary OT-I T-cells are sensitive towards low picomolar amounts of SIINFEKL.35 A proliferation assay was preformed staining primary OT-I T-cells with CFSE and incubating these cells for three days with SIINFEKL, SIINFEPgL or SIINFEKPg, as these cells can self-present peptides towards each other. From the proliferation profiles in Figure 8, it is clear that substituting position 7 from lysine to propargylglycine completely abolished T-cell activation while position 8 only reduced the T-cell response, confirming the B3Z results in Figure 7a.
Figure 8 | OT-I T-cell proliferation, activation with different modified epitopes. CFSE colored
OT-I T-cells were incubated for 3 days with 1nM of one of the peptides: SIINFEKL, SIINFEPgL and SIINFEKPg. SIINFEKL has proliferated the T-cells, having 6 distinct division peaks, whereas SIINFEKPg only shows 3 distinct division peaks and SIINFEPgL shows no division.
101
Figure 9 | Crystals of MHC-I-peptide complex containing SSIPgFARL. Crystallization conditions
were 0.1M PCTP pH 7.0 with 25% w/v PEG 1500, dropsize of 500nl with 70% protein solution and 30% buffer. The left picture is made with a contrast microscope and the right picture is made under UV light.
From these purified pMHC-I complexes containing biorthogonal handles, crystals were obtained for each at different conditions. First growth of crystal was observed after 5 days, and these crystals were collected after 35 days. Crystals were on average thin (< 100 µm) and had a length of 200-300 µm. A typical obtained crystal for pMHC-I complex is shown in Figure 9. Four complexes were crystallized from which diffraction data could be collected, containing SIINFEKL, SIINFEPgL, SSIEFAPgL or SSIPgFARL in the binding groove. This data is solved by dr. Sven Hennig. Figure 10 shows the complex structure of SSIPgFARL-MHC-I, representative for all structures. Results show no significant changes between overall structures as the r.m.s.d. values are below 1.3 Å (Table 1).
Table 1 | Overview of structural overlay with SIINFEKL containing crystal set as reference.
r.m.s.d values were determined with overlay function in pymol of refined structures, kindly provided by dr. Sven Hennig.
Peptide sequences Cα-atoms chains A,C,P r.m.s.d. / Å Cα-atoms chains B,D,Q r.m.s.d. / Å
SIINFEKL reference n.a. 386 vs 385 1.01
SIINFEPgL 386 vs 385 1.27 387 vs 385 1.26
SSIEFAPgL 382 vs 385 1.22 382 vs 385 1.15
102
Figure 10 | Complex structure of MHC-I (H-2Kb (dark orange surface) - β2m (light orange
surface)) with bound SSIPgFARL peptide (blue sticks). A) Two perspectives resulting from a 90°
turn as indicated by the black arrows. B) Top view close-up of the peptide binding pocket. The
bound peptide is shown as blue sticks with its according 2Fo-Fc electron density map. N- and C-terminus of the peptide are labelled as well as all peptide amino acids including the non-natural propargylglycine (Pg).
103
Figure 11 | Superimposition of Cα-backbone trace of chains A, C and P of all peptide bound structures, shown as ribbon representation. A) Side- and top-view of the superimposed
structures shows no difference in the overall structural fold upon binding of all four different peptides. The backbone trace of the bound peptides is visible in the top view (black arrow). B) Close-up of the bound peptides with all peptide backbones shown as thick lines and all amino acid side chains shown as thin sticks. N- and C-termini are labelled.
CCHC on D1 dendritic cells
It was next explored, whether this CCHC reaction could be further optimized on the dendritic cell line D136. D1 cells were loaded with peptide for 1 hour in a flat-bottom wells plate, after which they were transferred to a v-bottom 96-well plate and washed with PBS. The cells were fixed with 0.2% PFA for 15 minutes at RT. Cells were washed 3x with PBS and incubated with CCHC mix containing 10 µM azide-AF488 (as the fluorogenic CF488 was not available) for 1 hour at 37°C. After click reaction the cells were washed 3x with PBA before flow cytometry analysis. Unfortunately, no conditions could be found (Table 2) that yielded significant signal-to-noise (Figure 12).
104
fixation improved the S/N for this experiment. However, fixation has no influence on background signal as AF488 levels for SIINFEKL loaded D1 cells stayed the same. It was also found (Figure 12b) that all obtained signals were independent of peptide nature or concentration. It was attempted to remove the excessive unreacted fluorophore after click reaction by performing a second ligation reaction with a solubilizing group (Figure 12c). In theory this would make the unclicked fluorophore more soluble and easier to remove by washing, however no improvement of signal was observed. Also different washing protocols did not improve the signal (Figure 12d). Based on these experiments the use of the CCHC on dendritic cell lines was abandoned.
Figure 12 | Flow cytometry analysis of CCHC ligation of AF488-azide on SIINFEKL or SIINFEPgL loaded D1 cells. D1 cells were loaded with peptide for 1 hour after which they were transferred
105 10 µM S/N = 0.87 PBS; BSA_5% 0.84 1 µM S/N = 0.76 PBS; PBS + 0.1% Triton 0.68 0.1 µM S/N = 0.80 PBS; PBS + 1% DMSO 0.73 Removing unclicked fluorophore PBS + 0.1% Triton; PBA_0.2% 0.86
Click mix S/N = 0.77 PBS + 0.1% Triton; BSA_5% 0.96
PEG3 linker S/N = 0.88 PBS + 0.1% Triton; PBS + 0.1% Triton
0.86
PEG1000 S/N = 0.95 PBS + 0.1% Triton; PBS + 1%
DMSO
0.92
5.2.2 Evaluation of the IEDDA click for visualization of pMHC-I
To investigate whether the IEDDA ligation reaction would give better results, a model epitope (SIINFEKL, OVA257-264, OT-I) was modified with three different strained alkene: mbTCO (2), cyclopropane (3) and spirohexene (4) (Figure 3).
Synthesis of 2 was performed by Mark de Geus as published6, the other two peptides (3 and 4) were synthesized by Alexi Sarris. mbTCO (2) was first selected
for assessing ligation in a recombinant folded pMHC-I complex. Folding of the peptide MHC-complex was performed the same as for the alkyne containing peptides and the obtained complex was analyzed by Native-PAGE (Figure 13). Ligation of a fluorophore-tetrazine Cy5-Tz (Figure 13) or BODIPY-tetrazine2 (Bp2) (Figure 14) to the complex prior to gel analysis showed a fluorescent
signal emerging at the correct height while no fluorescent signal was obtained for wild type pMHC-I complex.
106 N+ N O NH O N H N N NN N+ SO 3 -Cl -Cy5-Tz
Figure 13 | IEDDA ligation on SIINFEKL[mbTCO]L and SIINFEKL in MHCI complex. Ligation of
Cy5-tetrazine and BODIPY-tetrazine (Bp2) with IEDDA on peptide-MHCI complexes visualized with fluorescence scanning (left) and coomassie staining (right) of Native-PAGE gels. a) SIINFEK[mbTCO]L epitope in MHC-I complex (refolded H-2Kb-β2m with peptide obtained from
bacterial culture, purified after refolding) was clicked with Cy5- and BODIPY-tetrazine, after which samples were run on gel. b) SIINFEKL (Wt) and SIINFEK[mbTCO]L (mbTCO) epitopes in MHC-I (refolded complexes) were clicked with BODIPY-tetrazine and run on gel.
NH O O +H 2N NH2 OH O N N N N F3C O -O N+BN -F F NH O N N N N N+BN -F F NH O N N N N N+BN -F F NH O N N N N N N N N+BN -F F NH O N N N N Bp1 Bp2 Bp3 Bp4 N+BN -F F H N O N N N N N+BN -F F H N O N N N N N+BN -F F H N O N N N N N N N+BN -F F H N O N N N N N Bp6 Bp7 Bp8 Bp5 N N N N HN O N H O N -O3S N B -N+ F F Rh7 Bp7s NN N N N N OH Tz1
Figure 14 | The different tetrazines used in this Chapter. The BODIPY fluorophores are numbered
107 not shift its average fluorescence. For 4 + Bp7 compared to SIINFEKL + Bp7, the
corresponding dotplots and histograms of the total and Bp+ population are given in Figure 15e/f. The mean intensities of only the positive populations were also analyzed and confirmed a high shift in signal to noise of these cells (Figure 15g/h). This was not due to the accumulation of fluorophore in dead/dying cells, as a live/dead exclusion had been performed in the gating strategy using size of the cells determined with forward scatter (fsc) and side scatter (ssc) (Figure 15e). However, what the reason is, remains to be elucidated.
108
Figure 15 |Flow cytometry analysis of BODIPY-tetrazines clicked to spirohexene, cyclopropene or mbTCO handles on SIINFEKL. a) Ligation of 2 µM BODIPY 1-8 to preloaded peptide D1 cells
109 After the initial promising results with Bp7, a more soluble version (Bp7s) was also tested at different concentrations of fluorophore and compared to Bp7 (Figure 16). Ligation to SIINFEKL loaded D1 cells gave low background signal 46 ± 3 GMFI for Bp7s and 130 ± 7 GMFI for Bp7 at 0.5 µM (Figure 16). This solubilized version of Bp7 (Bp7s) has even lower background than Bp7, but still did not yield a shift of the bulk population of cells (Figure 16b), even a lower percentage of cells was Bp+ compared to Bp7, for spirohexene at 0.5 µM Bp 13 ± 0.5 % vs 23 ± 0.5 % respectively.
Figure 16 |Flow cytometry analysis of Bp7 and Bp7s clicked to spirohexene, cyclopropene or mbTCO handles on SIINFEKL. Both BODIPYs were incubated at 0.2, 0.5 and 2 µM concentration
with peptide loaded D1 cells. a) The geometric mean fluorescence intensity of all cells. b) Percentage Bp+ cells.
110
All Bodipys show T-cell response, meaning that after the overnight incubation tautomerization and elimination has taken place. Bp7 shows the least amount of elimination resulting in 30% T-cell response compared to SIINFEKL.
Figure 17 |T-cell uncaging assay of Bp1-8, Bp7s Rh7. D1 cells were preincubated with 100nM
mbTCO-SIINFEKL or SIINFEKL and treated with 2µM BODIPY or tetrazine. Data is normalized towards SIINFEKL samples. Tz1, as good uncaging tetrazine, is taken as an extra control, N=1, n=3, error bars represent SD.
A concentration dependence, when the amount of available handle is altered, is expected. To check this, D1 cells were treated with different concentrations of the SIINFEKL or SIINFEKL-spiro epitope (4) for 2.5 hours, after which the cells
111
Figure 18 | Flow cytometry analysis of Bp7/Bp7s on spirohexene (4) at different peptide concentrations. Peptides were incubated for 2.5 hours (except for 1 sample which was incubated
for 1.5 hours as indicated in the bar chart) with D1 cells, after which they were treated with 0.2µM Bp7s or Bp7. a) The geometric mean fluorescence intensity of all cells. b) Percentage of Bp+ cells.
BODIPY-112
tetrazine 7. d) Dot plot overlay of dual stained samples (10µM peptide) containing H-2Kb-PE and
BODIPY-tetrazine 7s.
5.2.3 Visualization of pMHC-I by IEDDA with fluorescence microscopy
In order to obtain visual information on the high-IEDDA population, it was attempted to visualize the epitope inside MHC-I on cells. The three handles were tested in a fluorescence microscopy experiment with Bp7s as the fluorophore. Instead of D1 cells DC2.4 cells39 were used as these cells are more adherent and adhere more to the plate. DC2.4 cells were incubated with 10 µM peptide for 3 hours. After washing the cells were treated with 2µM Bp7s for 30 minutes. The cells were washed and left in PBS in a 96-well plate for fluorescence evaluation using the EVOS FL fluorescence microscope. This microscope has as maximal magnification of 60x, which for optimization purposes is suitable. Similar to the flow cytometry results, cyclopropene did not yield a detectable signal. MbTCO showed some cells with a detectable but low fluorescence. Spirohexene-treated cells showed a small population of cells of high fluorescence intensity (Figure 19), which – due to their size – are likely aggregates.
Figure 19 | Microscopy overlay pictures of IEDDA clicked peptides on DC2.4 cells. DC2.4 cells were
incubated with 10µM peptide for 3 hours after which cells were washed and treated for 30 minutes with 2µM Bp7s. The overlay picture shows the GFP and Trans channel taken with EVOS FL at 60x objective.
5.3 Conclusion
113 not deliver any signal in this system. The mbTCO handle, though very effective for uncaging, is too labile with these tetrazine-fluorophores as it eliminates from the epitope, thus eliminating this handle for visualization purposes. Cyclopropene and spirohexene are both very stable handles which result in a stable covalent product. Unfortunately, in this system cyclopropene-SIINFEKL does not yield any signal. The spirohexene-modified peptides precipitate under culture conditions.
114
5.4 Experimental section
Cell culture
D1
D1 cells were cultured in D1 medium (IMDM, 10% fetal calf serum, 2mM GlutaMAX, 100 IU/ml Penicillin, 50 μg/ml Streptomycin, 50 μM 2-mercaptoethanol, 30% R1 medium36). The D1 cells were sub-cultured after 3 or 4 days, depending on the visible
health of the cells, number of cells and the experimental requirements. Generally 1E5/ml to 2E5/ml were seeded in a fresh culture dish. The culture conditions were 5% CO2 at 37°C.
DC2.4
DC2.4 cells39 were cultured in DC2.4 medium (IMDM, 10% fetal calf serum, 2 mM
GlutaMAX, 100 IU/ml Penicillin, 50 µg/ml streptomycin, 10 mM HEPES, 1 mM pyruvate, 1x non-essential amino acids (NEAA)). The DC2.4 cells were sub-cultured every 2 or 3 days, depending on the visible health of the cells, number of cells and their confluence and the experimental requirements. General passage is 1:10 to 1:20. The culture conditions were 5% CO2 at 37°C.
B3Z
The B3Z T-cell hybridomas7 were cultured in B3Z medium (IMDM, 2mM glutamax, 10%
fetal calf serum, 100 IU/ml penicillin, 50 µg/ml streptomycin) with 0.25 mM 2-mercaptoethanol added fresh for new sub-culturing. Cells were passaged three times a week, 2.5E4 cells/ml in a new flask with a life percentage of at least 90%. Cells were maintained to stay below 1E6/ml. The culture conditions were 5% CO2 at 37°C.
T-cell activity assay in vitro B3Z
115 transported and kept for a week in enriched cages before they were sacrificed using cervical dislocation. To follow T-cell proliferation in vivo, OT-I/CD45.1 T-cells were labeled with the intracellular fluorescent dye CFSE (Molecular Probes). After a FACS analysis to assess the CFSE labeling efficiency, the OT-I cells were seeded in a 96-well microtiter plate (50,000 cells/well in 50 µl). Peptides (1nM) were added and incubated for 1 hour at 37°C. 100 µl of complete media was added and cells were left for 3 days. After incubation cells were washed with PBA and incubated with anti-CD8α-APC antibody and anti-CD45.1-V450 antibody to gate for live CD8+/CD45.1+ T cells.
Proliferation was assessed by flow cytometric analysis of the CFSE dilution on a LSR-II flow cytometer (BD biosciences). Data was analyzed using the FlowJo software 10.2 (Miltenyi Biosciences).
Synthesis of SIINFEKL, SIINFEPgL, SSIPgFARL and SSIEFAPgL
Peptides were synthesized using standard Fmoc solid phase peptide synthesis starting from Fmoc-Leu-Wang resin (0.64 mmol/g). The resin was removed with a cleavage cocktail (92.5% trifluoroacetic acid (TFA), 5% triisopropylsilate(TIPS) and 2.5% H2O) left
for at least 1 hour, while shaking. The protecting groups were removed by drying the peptide with N2 gas and treatment with the same TFA solution for at least 3 hours.
Finally, peptides were precipitated with diethyl ether, by adding the peptide in cleavage cocktail dropwise until the TFA concentration was 10-20% in 50 mL total volume and storing it for 30 minutes at -20˚C. The peptide was centrifuged 10 minutes at 4000 rpm at 4˚C and washed two times with diethyl ether. An LCMS sample was taken and diluted in 200 µl of H2O: tert-butanol:acetonitrile (1:1:1 v/v). The sample was analyzed using a
linear gradient of ACN (10-90%) in TFA (0.1%) for 10 minutes on the Agilent Technologies 6120 Quadrupole LC/MS. Peptides were purified using High Performances Liquid Chromatography (Prep column Gemini C18 110A 150x21.20 5μm) using 15 to 45% gradient (A: 0.1% TFA in MilliQ H2O, B: ACN). LC-MS measurements were done on
an API 3000 Alltech 3300 with a Grace Vydac 214TP 4.6 mm x 50 mm C4 column.
Production of MHC-I (H-2Kb-β2m) complexed with modified peptides: SIINFEKL,
SIINFEPgL, SSIEFAPgL and SSIPgFARL
Expression and harvesting
This method was adapted from Toebes et al. (2009).40 The extracellular domains of
116
The C-terminus of H-2Kb was decorated with the biotinylation sequence
GLNDIFEAQKIEWHE which is accepted by the BirA biotin ligase.41 The vector used for
both subunits was pET3a. Induction of protein expression was started upon addition of IPTG (0.4 mM) at 30°C. Optimal production was obtained after 20 hours. Cells were harvested and lysed with 10 ml lysis buffer (50 mM Tris HCl pH 8.0; 25% (w/v) sucrose; 1 mM EDTA; 1mM PMSF; 2 mM DTT) and 4 mg lysozyme was added per 5 g of bacterial pellet and incubated for 30 minutes on ice. After incubation MgCl2 (10 mM final
concentration), MnCl2 (1 mM final concentration) and benzonase (1/10000 dilution)
were added and this was incubated on ice for 30 minutes. The solution was sonicated (10% amplitude 5 seconds on, 10 seconds off for a total of 5 minutes on, kept on ice) in 20 ml detergent buffer (0.2 M NaCl; 1% (w/v) sodium deoxycholate monohydrate; 1% Ipegal; 20 mM Tris Cl pH 7.5; 2 mM EDTA; 0.02 mM PMSF; 0.04 mM DTT) per 5 g of bacterial pellet. After incubation on ice for at least 30 minutes, the lysate was centrifugated 20 minutes at 14 000 g, 4°C. The supernatant was discarded and the pellet was resuspended in wash solution (0.5% (v/v) Triton X-100; 100 mM NaCl; 1 mM EDTA; 2 mM DTT) by drawing it through an 18-G needle. After washing the pellet three times, the final pellet, mainly containing the inclusion bodies, was stored at -20°C.
Refolding and purification
H-2Kb and β2M were dissolved in denaturing buffer (8 M urea; 100 mM Tris HCl pH 8).
For a 5 g pellet 15 ml was used. A refolding reaction of 50 ml was set up which stirred in a dark, cold room for a total of 4 days. To 50 ml refolding buffer (5 ml 1 M Tris HCl pH 8; 4.2g L-arginine monohydrochloride; 2.5 ml 100 mM reduced glutathione; 0.5 ml 50 mM oxidized glutathione; 0.2 ml 0.5 M EDTA; 5 μl protease inhibitor; 41.8 ml H2O) 27 μM (final concentration) of conditional ligand (SIINFEKL, SIINFEPgL, SSIEFAPgL, SSIPgFARL) was added and 1 mM PMSF on the first day. A final concentration of 6 μM β2M was added on three consecutive days, so 2 μM final concentration on each day. Secondly, a final concentration of 3 μM H-2Kb was dropwise added on three consecutive
117 for 10 minutes. The membrane was incubated with 4 mL blocking buffer with primary antibody, anti MHC-I H-2kb biotin (mouse, 1:2000) or anti β2m (goat, 1:1000) for 90
minutes rotating at RT. Washed three times with 15 mL TBST for 10 minutes. The membrane was incubated with secondary antibody, donkey-anti-goat IgG-HRP (1:4000) for β2m or goat-anti-mouse-HRP (1:4000) for H-2Kb for 60 minutes. Washed three times
with 15 mL TBST and once with 15mL TBS for 10 minutes. The blot was moved to a dark box, 10 mL luminol, 100 μL enhancer and 3 μL H2O2 were added. This was incubated for
a couple of minutes and pictures were made with the Bio-Rad gel-doc system.
Circular Dichroism
SIINFEKL-MHC-I (stock concentration 5.4 mg/ml) and SIINFEPgL-MHC-I (stock concentration 3.5 mg/ml) proteins (stored in PBS, 150 mM NaCl and 5 % v/v glycerol) were diluted to a protein concentration of 0.35 mg/ml. The protein samples were measured on a Jasco J815 CD spectrometer (Easton, MD). The spectra were recorded with a minimum of 5 repetitions and averaged upon baseline subtraction. The percentages of secondary structural elements were calculated with CDNN 2.1 software comparing 33 reference spectra from the database.
Crystallization of MHC-I (H-2Kb-β2m) complexed with various epitopes
Crystals were obtained with sitting-drop vapor diffusion technique with a drop size of 500 nL containing either 50/50 protein solution and buffer solution or 70/30. For each peptide a new batch of refolded protein with the peptide was prepared resulting in different crystallization conditions for each peptide. The conditions were for SIINFEKL 0.1 M Bis-Tris propane pH 6.5 with 0.2 M sodium citrate tribasic dehydrate and 20% w/b PEG 3350; SIINFEPgL (B11) 0.1 M MES pH6.0 with 0.2 M Calcium chloride dehydrate and 20% w/v PEG 6000 (C11) 0.1M HEPES pH 7.0 with 0.2 M calcium chloride dehydrate and 20% w/v PEG 6000; SSIEFAPgL (B11) 0.1 M MES pH6.0 with 0.2 M Calcium chloride dehydrate and 20% w/v PEG 6000 (C10) 0.1 M HEPES pH 7.0 with 0.2 M magnesium chloride dehydrate and 20% w/v PEG 6000; SSIPgFARL 0.1 M PCTP pH 7.0 with 25% w/v PEG 1500.
Native-PAGE: CCHC reaction on pure MHC-I + peptide
CCHC reactions (1mM CuSO4, 100 µM Sodium Ascorbate, 5 mM THTPA, 3 mM Tris
118
minutes. 1,5 mm 6% Native-PAGE gels were made by first polymerizing the running gel (4mL acrylamide 30% 37:1, 15,78mL 0.375M Tris-HCl pH=8.8, 200µL 10% APS, 20µL TEMED) followed by polymerizing the stacking gel on top (1,34mL acrylamide 30% 37:1, 8,55mL 0.375M Tris-HCl pH=8,8, 100µL 10% APS, 10µL TEMED). Samples were loaded and run for 15 minutes at 150V followed by 45 minutes at 175V. Picture was taken with the ChemiDocTM MP, using the AlexaFluor488 and Cy3 channel. The gel was stained with
Coomassie Blue for 30 minutes, de-stained with demi water over night and imaged with the ChemiDocTM MP, using the Coomassie Blue channel.
CCHC click reaction on D1 cells
100,000 cells per well, in a flat-bottom 96-well plate, were seeded in 50µL and incubated with the epitope (12.5µL of 500µM/50µM/5µM) for 1 hour at 37°C, 5% CO2.
The cells were transferred to a v-bottom 96-well plate and fixed in 25µL of 0.2% PFA in PBS for 15 minutes at room temperature. Cells were washed three times with PBS (137 mM NaCl, 2.7 mM KCl, 10mM Na2HPO4, 1.8 mM KH2PO4 in H2O), each time
centrifugated for 3 minutes at 1500 rpm. 25µL of the CCHC click mix (1mM CuSO4,
10mM sodium-ascorbate, 1mM tris(3-hydroxypropyltriazolylmethyl)amine (THPTA), 10mM amino-guanidine, 10µM AF488-azide in PBS) was added. This was incubated for 1 hour at 37°C. Cells were washed three times with PBA, each time centrifugated for 3 minutes at 1500 rpm. Cell were taken up in 100µL of FACS-buffer (PBS/EDTA) for flow cytometry analysis. Flow cytometry analysis was performed on the Guava EasyCyte Flow Cytometer by Merck.
Native-PAGE: IEDDA reaction on pure MHC-I + peptide
IEDDA reactions (1 µM Bp2 or Cy5-Tz, 5 µg pMHC-I in PBS) was incubated at RT for 5 minutes. 1,5mm 6% Native-PAGE gels were made by first polymerizing the running gel (4mL acrylamide 30% 37:1, 15,78mL 0.375M Tris-HCl pH=8.8, 200µL 10% APS, 20µL TEMED) followed by polymerizing the stacking gel on top (1,34mL acrylamide 30% 37:1, 8,55mL 0.375M Tris-HCl pH=8,8, 100µL 10% APS, 10µL TEMED). Samples were loaded and run for 15 minutes at 150V followed by 45 minutes at 175V. Picture was taken with the ChemiDocTM MP, using the AlexaFluor488 and Cy3 channel. The gel was stained with
Coomassie Blue for 30 minutes, de-stained with demi water over night and imaged with the ChemiDocTM MP, using the Coomassie Blue channel.
IEDDA click reaction on D1 cells
100,000 D1 cells, dissolved in D1 culturing medium and 30 % R1 supplement, were seeded per well on an uncoated flat bottom 96-well plate and incubated with 10µM peptide for 24 hours with conditions 5% CO2 at 37℃. After the removal of the medium
the cells were incubated with one of the tetrazine-fluorophores at indicated concentration for 3 hours at 37℃ and 5% CO2 in the dark. All samples were transferred
119
Antibody staining
After preincubation with peptides, dendritic cells were treated with antibodies for 20 minutes on ice with a dilution of either anti-H-2Kb (1:200) or 25D1.16-APC (1:500). Cells
were washed three times with PBA (PBS + 0.5% BSA + 0.02% NaN3), each time centrifugated for 3 minutes at 1500 rpm. The cell pellet was taken up in 100µL of PBA and analyzed with the Guava EasyCyte Flow Cytometer developed by Merck.
Fluorescence microscopy
100.000 DC2.4 cells per well, in a flat-bottom 96-well plate, were seeded in 50µL and incubated with the epitope (12,5µL of 20µM/10µM) for 1 hour at 37°C, 5% CO2. Cells
were washed two times with PBS, each time centrifugated for 3 minutes at 1500 rpm. 25µL of the TCO click mix (10µM/1µM BODIPY-FL in PBS) was added. This was incubated for 30 minutes to 1 hour at 37°C. Cells were washed one time with PBA, each time centrifugated for 3 minutes at 1500 rpm. Cell were taken up in 50µL of PBS for fluorescence microscopy analysis. Fluorescence microscopy was performed in 50µL of PBS. Images were taken with 60x objective of the Evos TM ® FL Auto 2 Imaging System
by ThermoFisher Scientific.
5.5 References
(1) Devaraj, N. K. The Future of Bioorthogonal Chemistry. (2018) ACS Cent. Sci. 4, 952–959. (2) Tian, Y.; Lin, Q. Fitness Factors for Bioorthogonal Chemical Probes. (2019) ACS Chem. Biol. 14, 2489–2496.
(3) McKay, C. S.; Finn, M. G. Click Chemistry in Complex Mixtures: Bioorthogonal Bioconjugation. (2014) Chem. Biol. 21, 1075–1101.
(4) Lang, K.; Chin, J. W. Bioorthogonal Reactions for Labeling Proteins. (2014) ACS Chem. Biol. 9, 16– 20.
120
(6) van der Gracht, A. M. F.; de Geus, M. A. R.; Camps, M. G. M.; Ruckwardt, T. J.; Sarris, A. J. C.; Bremmers, J.; Maurits, E.; Pawlak, J. B.; Posthoorn, M. M.; Bonger, K. M.; Filippov, D. V.; Overkleeft, H. S.; Robillard, M. S.; Ossendorp, F.; van Kasteren, S. I. Chemical Control over T-Cell Activation in Vivo Using Deprotection of Trans-Cyclooctene-Modified Epitopes. (2018) ACS Chem. Biol. 13, 1569–1576.
(7) Karttunen, J.; Shastri, N. Measurement of Ligand-Induced Activation in Single Viable T Cells Using the LacZ Reporter Gene. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 3972–3976.
(8) Porgador, A.; Yewdell, J. W.; Deng, Y.; Bennink, J. R.; Germain, R. N. Localization, Quantitation, and In Situ Detection of Specific Peptide–MHC Class I Complexes Using a Monoclonal Antibody. (1997) Immunity
6, 715–726.
(9) Bozzacco, L.; Yu, H.; A. Zebroski, H.; Dengjel, J.; Deng, H.; Mojsov, S.; Steinman, R. M. Mass Spectrometry Analysis and Quantitation of Peptides Presented on the MHC II Molecules of Mouse Spleen Dendritic Cells. (2011) J. Proteome Res. 10, 5016–5030.
(10) Araman, C.; Pieper-Pournara, L.; Leeuwen, T. van; Kampstra, A. S. B.; Bakkum, T.; Marqvorsen, M. H. S.; Nascimento, C. R.; Groenewold, G. J. M.; Wulp, W. van der; Camps, M. G. M.; Overkleeft, H. S.; Ossendorp, F. A.; Toes, R. E. M.; van Kasteren, S. I. Bioorthogonal Antigens Allow The Study Of Intracellular Processing And Presentation Of Post-Translationally Modified Antigens. (2018) bioRxiv Prepr. 1–19. (11) Ljunggren, H. -G; Öhlén, C.; Höglund, P.; Franksson, L.; Kärre, K. The RMA-S Lymphoma Mutant; Consequences of a Peptide Loading Defect on Immunological Recognition and Graft Rejection. (1991) Int. J.
Cancer 47, 38–44.
(12) Baranov, M. V.; Bianchi, F.; Schirmacher, A.; van Aart, M. A. C.; Maassen, S.; Muntjewerff, E. M.; Dingjan, I.; ter Beest, M.; Verdoes, M.; Keyser, S. G. L.; Bertozzi, C. R.; Diederichsen, U.; van den Bogaart, G. The Phosphoinositide Kinase PIKfyve Promotes Cathepsin-S-Mediated Major Histocompatibility Complex Class II Antigen Presentation. (2019) iScience 11, 160–177.
(13) Oliveira, B. L.; Guo, Z.; Bernardes, G. J. L. Inverse Electron Demand Diels–Alder Reactions in Chemical Biology. (2017) Chem. Soc. Rev. 46, 4895–4950.
(14) Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.; Chang, P. V.; Miller, I. A.; Lo, A.; Codelli, J. A.; Bertozzi, C. R. Copper-Free Click Chemistry for Dynamic in Vivo Imaging. (2007) Proc. Natl. Acad. Sci. U. S.
A. 104, 16793–16797.
(15) El-Sagheer, A. H.; Brown, T. Click Chemistry with DNA. (2010) Chem. Soc. Rev. 39, 1388.
(16) Beatty, K. E.; Liu, J. C.; Xie, F.; Dieterich, D. C.; Schuman, E. M.; Wang, Q.; Tirrell, D. A. Fluorescence Visualization of Newly Synthesized Proteins in Mammalian Cells. (2006) Angew. Chemie Int. Ed. 45, 7364– 7367.
(17) Neef, A. B.; Schultz, C. Selective Fluorescence Labeling of Lipids in Living Cells. (2009) Angew.
Chem. Int. Ed. 48, 1498–1500.
(18) Presolski, S. I.; Hong, V.; Cho, S.; Finn, M. G. Tailored Ligand Acceleration of the Cu-Catalyzed Azide−Alkyne Cycloaddition Reaction: Practical and Mechanistic Implications. (2010) J. Am. Chem. Soc. 132, 14570–14576.
121
Pretargeted Live Cell Imaging Tetrazine-Based Cycloadditions : Application to Pretargeted Live Cell Imaging. (2008) Bioconjug. Chem. 19, 2297–2299.
(23) Darko, A.; Wallace, S.; Dmitrenko, O.; Machovina, M. M.; Mehl, R. A.; Chin, J. W.; Fox, J. M. Conformationally Strained Trans-Cyclooctene with Improved Stability and Excellent Reactivity in Tetrazine Ligation. (2014) Chem. Sci. 5, 3770–3776.
(24) Selvaraj, R.; Fox, J. M. Trans-Cyclooctene-a Stable, Voracious Dienophile for Bioorthogonal Labeling. (2013) Curr. Opin. Chem. Biol. 17, 753–760.
(25) Karver, M. R.; Weissleder, R.; Hilderbrand, S. A. Synthesis and Evaluation of a Series of 1,2,4,5-Tetrazines for Bioorthogonal Conjugation. (2011) Bioconjug. Chem. 22, 2263–2270.
(26) Devaraj, N. K.; Thurber, G. M.; Keliher, E. J.; Marinelli, B.; Weissleder, R. Reactive Polymer Enables Efficient in Vivo Bioorthogonal Chemistry. (2012) Proc. Natl. Acad. Sci. U. S. A. 109, 4762–4767.
(27) Hoffmann, J. E.; Plass, T.; Nikic, I.; Aramburu, I. V.; Koehler, C.; Gillandt, H.; Lemke, E. A.; Schultz, C. Highly Stable Trans-Cyclooctene Amino Acids for Live-Cell Labeling. (2015) Chem. Eur. J. 21, 12266–12270. (28) Yang, J.; Šečkutė, J.; Cole, C. M.; Devaraj, N. K. Live-Cell Imaging of Cyclopropene Tags with Fluorogenic Tetrazine Cycloadditions. (2012) Angew. Chem. Int. Ed. 51, 7476–7479.
(29) Cole, C. M.; Yang, J.; Šečkutė, J.; Devaraj, N. K. Fluorescent Live-Cell Imaging of Metabolically Incorporated Unnatural Cyclopropene-Mannosamine Derivatives. (2013) ChemBioChem 14, 205–208. (30) Rossin, R.; Van Duijnhoven, S. M. J.; Ten Hoeve, W.; Janssen, H. M.; Kleijn, L. H. J.; Hoeben, F. J. M.; Versteegen, R. M.; Robillard, M. S. Triggered Drug Release from an Antibody-Drug Conjugate Using Fast “Click-to-Release” Chemistry in Mice. (2016) Bioconjug. Chem. 27, 1697–1706.
(31) Yewdell, J. W.; Reits, E.; Neefjes, J. Making Sense of Mass Destruction: Quantitating MHC Class I Antigen Presentation. (2003) Nat. Rev. Immunol. 3, 952–961.
(32) Fadel, T. R.; Li, N.; Shah, S.; Look, M.; Pfefferle, L. D.; Haller, G. L.; Justesen, S.; Wilson, C. J.; Fahmy, T. M. Adsorption of Multimeric T Cell Antigens on Carbon Nanotubes: Effect on Protein Structure and Antigen-Specific T Cell Stimulation. (2013) Small 9, 666–672.
(33) Eisen, H. N.; Hou, X. H.; Shen, C.; Wang, K.; Tanguturi, V. K.; Smith, C.; Kozyrytska, K.; Nambiar, L.; McKinley, C. A.; Chen, J.; Cohen, R. J. Promiscuous Binding of Extracellular Peptides to Cell Surface Class I MHC Protein. (2012) Proc. Natl. Acad. Sci. U. S. A. 109, 4580–4585.
122
(35) Rosette, C.; Werlen, G.; Daniels, M. A.; Holman, P. O.; Alam, S. M.; Travers, P. J.; Gascoigne, N. R. J.; Palmer, E.; Jameson, S. C. The Impact of Duration versus Extent of TCR Occupancy on T Cell Activation: A Revision of the Kinetic Proofreading Model. (2001) Immunity 15, 59–70.
(36) Winzler, C.; Rovere, P.; Rescigno, M.; Granucci, F.; Penna, G.; Adorini, L.; Zimmermann, V. S.; Davoust, J.; Ricciardi-Castagnoli, P. Maturation Stages of Mouse Dendritic Cells in Growth Factor-Dependent Long-Term Cultures. (1997) J. Exp. Med. 185, 317–328.
(37) Devaraj, N. K.; Hilderbrand, S.; Upadhyay, R.; Mazitschek, R.; Weissleder, R. Bioorthogonal Turn-on Probes for Imaging Small Molecules inside Living Cells. (2010) Angew. Chem. Int. Ed. 49, 2869–2872. (38) Hughes, L. D.; Rawle, R. J.; Boxer, S. G. Choose Your Label Wisely: Water-Soluble Fluorophores Often Interact with Lipid Bilayers. (2014) PLoS One 9, e87649.
(39) Shen, Z.; Reznikoff, G.; Dranoff, G.; Rock, K. L. Cloned Dendritic Cells Can Present Exogenous Antigens on Both MHC Class I and Class II Molecules. (1997) J. Immunol. 158, 2723–2730.
(40) Toebes, M.; Rodenko, B.; Ovaa, H.; Schumacher, T. N. M. Generation of Peptide MHC Class I Monomers and Multimers through Ligand Exchange. (2009) Curr. Protoc. Immunol. No. SUPPL. 87, 1–20. (41) Fairhead, M.; Howarth, M. Site-Specific Biotinylation of Purified Proteins Using BirA. In Site-Specific
