Branched Linkers for Site-Specific Fluorescent Labeling of Antibodies

Fluorescent antibodies have proved to be an invaluable tool for molecular biology and diagnostics. They are routinely produced by modification of lysine residues, which leads to high heterogeneity. As such, their affinity may be compromised if the antigen-binding site is affected, the probability of which increases along with the degree of labeling. In this work, we propose a methodology for the synthesis of site-specific antibody-dye conjugates with a high degree of labeling. To this end, we synthesized two oxyamine-based branched triazide linkers and coupled them with a periodate-oxidized anti-PRAME antibody 6H8; two oxyamine-based linear monoazide linkers of similar structure were used as controls. The azide-labeled antibodies were subsequently conjugated with fluorescent dyes via SPAAC, a copper-free click reaction. Compared to their counterparts made with linear linkers, the branched conjugates possessed a higher degree of labeling. The utility of the methodology was demonstrated in the detection of the PRAME protein on the surface of the cell by flow cytometry.


Introduction
Fluorescently labeled antibodies are a powerful tool for various analytical and diagnostic approaches, such as flow cytometry and immunocytochemistry [1,2]. Most often, modified antibody synthesis is based on conjugation via maleimide-thiol or active esteramino group chemistry. Both approaches are not free of flaws: acylation of lysine residues may compromise the affinity if the antigen-binding site is modified, while the reduction of disulfide bonds followed by conjugate addition can impact the tertiary structure of the antibody [3,4]. In addition, these conjugation methods yield highly heterogeneous mixtures of products. For these reasons, site-specific modification of antibodies is regarded as preferable [5,6].
Most existing methodologies for site-specific antibody modification are rather intricate, relying on the introduction of non-natural amino acids [7], enzymatic remodeling of glycans [8], or proximity-induced chemistry [9]. In contrast, glycan modification via periodate oxidation followed by reacting the resulting aldehyde groups with nucleophiles constitutes a much simpler approach [10][11][12]. This methodology is at its best when coupled with oxime ligation, which yields stable conjugates without the need for C=N double bond reduction [13].

Synthesis of Azide Linkers
We used the previously developed pentaerythritol-based triazide 1 [14] as the starting material for creating new bifunctional branched cores (Scheme 1). The synthesis commenced with making the key compound 2. This mesylate was prepared from alcohol 1 using mesyl chloride and an appropriate base. Although in the present work, it was purified by column chromatography for characterization purposes, we found that its use in the next synthetic step does not require purification. Next, we obtained ethoxyethylideneprotected oxyamine 3 from the mesylate and sodium salt of ethyl-N-hydroxyacetimidate according to the method described earlier (Scheme 1A). This protected oxyamine core can be used as a branching reagent for antibody conjugates. In our case, it was employed to increase the loading of fluorescent dyes (Figure 1). The protecting group can be removed in situ, and ligation with carbonyl groups can instantly occur as we described earlier [15] by incubation with previously oxidized antibodies at pH 3-4.5. It is a convenient method for dyes unstable in the presence of a free oxyamino group, e.g., fluorescein. It should be noted, however, that prolonged incubation of antibodies at pH values of 3 or lower can lead to reduced affinity, even though the oxyamine deprotection occurs efficiently under these conditions. To this end, we found that the optimal pH value for in situ deprotection/oxime ligation is 4.0. At this pH, oxyamine deblocking and subsequent ligation occurs without significant damage to the integrity of the antibody. With a sufficient excess of the

Synthesis of Azide Linkers
We used the previously developed pentaerythritol-based triazide 1 [14] as the starting material for creating new bifunctional branched cores (Scheme 1). The synthesis commenced with making the key compound 2. This mesylate was prepared from alcohol 1 using mesyl chloride and an appropriate base. Although in the present work, it was purified by column chromatography for characterization purposes, we found that its use in the next synthetic step does not require purification. Next, we obtained ethoxyethylideneprotected oxyamine 3 from the mesylate and sodium salt of ethyl-N-hydroxyacetimidate according to the method described earlier (Scheme 1A). This protected oxyamine core can be used as a branching reagent for antibody conjugates. In our case, it was employed to increase the loading of fluorescent dyes (Figure 1). The protecting group can be removed in situ, and ligation with carbonyl groups can instantly occur as we described earlier [15] by incubation with previously oxidized antibodies at pH 3-4.5. It is a convenient method for dyes unstable in the presence of a free oxyamino group, e.g., fluorescein. It should be noted, however, that prolonged incubation of antibodies at pH values of 3 or lower can lead to reduced affinity, even though the oxyamine deprotection occurs efficiently under these conditions. To this end, we found that the optimal pH value for in situ deprotection/oxime ligation is 4.0. At this pH, oxyamine deblocking and subsequent ligation occurs without significant damage to the integrity of the antibody. With a sufficient excess of the dye (~100 equiv.) and an incubation time of about 1 h, it is possible to achieve an antibody labeling degree of 2-4.
On the other hand, many dyes can tolerate the presence of a protonated/free aminooxy group for some period of time. Should such dyes be used for labeling, incubation of the antibody at a more basic pH is possible. A suitable buffer for oxime ligation is the acetate buffer with a pH of 5.0. At this level of acidity, the periodate oxidation is at its best, and the side reaction of the formation of Schiff bases of the carbonyl groups with lysine residues is suppressed [16]. In contrast to the formation of Schiff bases, which proceeds better in a slightly alkaline environment, oxime ligation reactions are best performed under acidic conditions. Considering that, we also made the deprotected branched derivative 4 from 3 by incubation in methanolic HCl (Scheme 1A).
From mesylate 2 we obtained another branched bifunctional linker 6 with an amino group instead of the oxyamino group. For this purpose, the mesylate was converted to phthalimide 5, followed by deprotection with hydrazine (Scheme 1A). The bifunctional linker 6 itself may be of interest for different applications. In our case, it was condensed with the sulfo-Cy3 derivative 9 to yield the bifunctional dye-containing linker 10 (Scheme 1B).
The conversion of 7 to 9 was performed using the water-tolerant coupling reagent TSTU in a DMF-1,4-dioxane-H 2 O triple solvent system. This combination of solvents was chosen to improve the solubility of the metal salt 7. First, the monocondensation product 9 was obtained by desymmetrization (a mixture of mono and bis products was synthesized and separated by column chromatography). It was then reacted with reagent 6 to yield 10 (Scheme 1B). Acidic deprotection yielded the free oxyamine variant of the fluorescent linker 11 along with linkers 12 and 16, the latter being obtained from 7 (Scheme 1C).
Molecules 2023, 28, x FOR PEER REVIEW 3 of 18 dye (~100 equiv.) and an incubation time of about 1 h, it is possible to achieve an antibody labeling degree of 2-4. On the other hand, many dyes can tolerate the presence of a protonated/free aminooxy group for some period of time. Should such dyes be used for labeling, incubation of the antibody at a more basic pH is possible. A suitable buffer for oxime ligation is the acetate buffer with a pH of 5.0. At this level of acidity, the periodate oxidation is at its best, and the side reaction of the formation of Schiff bases of the carbonyl groups with lysine residues is suppressed [16]. In contrast to the formation of Schiff bases, which proceeds better in a slightly alkaline environment, oxime ligation reactions are best performed under acidic conditions. Considering that, we also made the deprotected branched derivative 4 from 3 by incubation in methanolic HCl (Scheme 1A).
From mesylate 2 we obtained another branched bifunctional linker 6 with an amino group instead of the oxyamino group. For this purpose, the mesylate was converted to phthalimide 5, followed by deprotection with hydrazine (Scheme 1A). The bifunctional linker 6 itself may be of interest for different applications. In our case, it was condensed with the sulfo-Cy3 derivative 9 to yield the bifunctional dye-containing linker 10 (Scheme 1B). The conversion of 7 to 9 was performed using the water-tolerant coupling reagent TSTU in a DMF-1,4-dioxane-H2O triple solvent system. This combination of solvents was chosen to improve the solubility of the metal salt 7. First, the monocondensation product 9 was obtained by desymmetrization (a mixture of mono and bis products was synthesized and separated by column chromatography). It was then reacted with reagent 6 to yield 10 (Scheme 1B). Acidic deprotection yielded the free oxyamine variant of the fluorescent linker 11 along with linkers 12 and 16, the latter being obtained from 7 (Scheme 1C).

Synthesis of Antibody Conjugates
For the present work, we chose the anti-PRAME antibody 6H8 [17], which we had previously used in our studies [15,18]. The PRAME protein is expressed on the surface of melanoma and leukemia cells and is thus a promising target for diagnostic and therapeutic tools [19][20][21][22][23]. We set out to obtain antibody conjugates with linear and branched reagents (both dye-based and not) and then modify them with dyes using strain-promoted azide-alkyne cycloaddition (SPAAC) (Scheme 2). SPAAC was chosen, as it is less likely to compromise the antibody's affinity compared to copper-catalyzed azide-alkyne cycloaddition (CuAAC) [24].
previously used in our studies [15,18]. The PRAME protein is expressed on the surface of melanoma and leukemia cells and is thus a promising target for diagnostic and therapeutic tools [19][20][21][22][23]. We set out to obtain antibody conjugates with linear and branched reagents (both dye-based and not) and then modify them with dyes using strain-promoted azide-alkyne cycloaddition (SPAAC) (Scheme 2). SPAAC was chosen, as it is less likely to compromise the antibody's affinity compared to copper-catalyzed azide-alkyne cycloaddition (CuAAC) [24]. First, we obtained antibodies labeled with branched reagent 4 (Scheme 2A,B). For comparison, we also obtained an antibody labeled with linear linker 12. For that, we oxidized the antibodies with sodium periodate in an acetate buffer in a weakly acidic medium. Beforehand, we studied the effect of periodate concentration and oxidation time on the affinity of antibodies. The conclusion was that the most optimal is the use of 20 mM sodium periodate for 30 min in acetate buffer with a pH value of 5. Upon oxidation, the antibody was desalted and incubated with linear reagent 12 and branched reagent 4 in a weakly acidic acetate buffer for 1 h.
The resultant conjugates were purified and reacted with the fluorescein dibenzocyclooctyne (DBCO) derivative 21, which is suitable for effective click reaction with azides in biomolecules, to yield the linear and branched FAM conjugates 22 and 24. In a similar First, we obtained antibodies labeled with branched reagent 4 (Scheme 2A,B). For comparison, we also obtained an antibody labeled with linear linker 12. For that, we oxidized the antibodies with sodium periodate in an acetate buffer in a weakly acidic medium. Beforehand, we studied the effect of periodate concentration and oxidation time on the affinity of antibodies. The conclusion was that the most optimal is the use of 20 mM sodium periodate for 30 min in acetate buffer with a pH value of 5. Upon oxidation, the antibody was desalted and incubated with linear reagent 12 and branched reagent 4 in a weakly acidic acetate buffer for 1 h.
The resultant conjugates were purified and reacted with the fluorescein dibenzocyclooctyne (DBCO) derivative 21, which is suitable for effective click reaction with azides in biomolecules, to yield the linear and branched FAM conjugates 22 and 24. In a similar way, we obtained conjugates 25 and 28 with fluorescent linkers 16 and 11, respectively (Scheme 2C,D). They were then modified in a SPAAC click reaction with the sulfo-Cy5 DBCO dye 26. Due to the hydrophobicity of the DBCO groups, the final conjugates exhibited instability and were prone to aggregation in PBS. In order to overcome this problem, we used 0.01% Tween 80 in PBS as a stabilizer for the linear conjugates (22, 27) and 0.05% Tween 80 for the branched ones (24,29). In the presence of Tween 80, the conjugates were found to be stable for at least a week at +4 • C.

Degree of Labeling
The stoichiometry of the dye conjugates was determined by UV-Vis spectrophotometry using the molar absorbance coefficients listed in Table 1. The degree of modification of the conjugates 20 and 23 could not be measured by this method, so it was assessed indirectly by measuring the DOL of their fluorescein derivatives 22 and 24. Somewhat surprisingly, even though 20 and 23 were synthesized under virtually identical conditions, 23 appeared to have a substantially higher degree of modification, giving rise to conjugate 24 with a DOL of at least 21 (see Section 4 for details) as compared with its linear counterpart 22 (DOL = 3.7) (Figure 2A). way, we obtained conjugates 25 and 28 with fluorescent linkers 16 and 11, respectively (Scheme 2C,D). They were then modified in a SPAAC click reaction with the sulfo-Cy5 DBCO dye 26. Due to the hydrophobicity of the DBCO groups, the final conjugates exhibited instability and were prone to aggregation in PBS. In order to overcome this problem, we used 0.01% Tween 80 in PBS as a stabilizer for the linear conjugates (22, 27) and 0.05% Tween 80 for the branched ones (24,29). In the presence of Tween 80, the conjugates were found to be stable for at least a week at +4 °C.

Degree of Labeling
The stoichiometry of the dye conjugates was determined by UV-Vis spectrophotometry using the molar absorbance coefficients listed in Table 1. The degree of modification of the conjugates 20 and 23 could not be measured by this method, so it was assessed indirectly by measuring the DOL of their fluorescein derivatives 22 and 24. Somewhat surprisingly, even though 20 and 23 were synthesized under virtually identical conditions, 23 appeared to have a substantially higher degree of modification, giving rise to conjugate 24 with a DOL of at least 21 (see Section 4 for details) as compared with its linear counterpart 22 (DOL = 3.7) (Figure 2A).  In contrast, conjugates 25 and 28 synthesized under similar conditions have a comparable degree of modification ( Figure 2B) that can be measured directly, which suggests that they are better suited for the controlled introduction of azide moieties into antibodies. To obtain a reference for the 100% yield of the click reaction, we synthesized the low-molecular-weight sCy3-sCy5 conjugate 18 (Scheme 1D) and measured its sCy3/sCy5 absorbance ratio at their maxima ( Figure 2D). Surprisingly, it turned out that the value (0.91) was much higher than expected ([ε 548 sCy3+ε 548 sCy5]/ε 646 sCy5 = (1.62 × 10 5 + 1.20 × 10 4 ) M −1 cm −1 /(2.71 × 10 5 ) M −1 cm −1 = 0.64). This was used to evaluate the degree of modification of 25 with sCy5-DBCO, which proceeded with 82% yield, giving rise to conjugate 27 with a DOL of 3.0. Under analogous conditions, branched conjugate 28 yielded conjugate 29 that possessed a higher DOL of 7.4 as measured by sCy5 ( Figure 2C).

Study of Fluorescent Properties
In order to evaluate the fluorescent properties of the labeled antibodies, fluorescence emission spectra of the four final conjugates were recorded. While branched conjugate 24 demonstrated higher fluorescence intensity compared to 22 (Figure 3A), the approximately twofold increase was disproportionate to the more than fivefold difference in the DOL (21 vs. 3.7); the discrepancy can be ascribed to self-quenching of fluorescein due to fluorophores being in close proximity around the branched core. Contrarily, the difference in fluorescence intensity in conjugates 29 and 27 was completely proportional to the difference in the degree of labeling, indicating that no apparent self-quenching took place ( Figure 3B). This might be in part due to a longer linker connecting dye fragments to the core in 29 compared to 24 (Scheme 2). a DOL of 3.0. Under analogous conditions, branched conjugate 28 yielded conjugate 29 that possessed a higher DOL of 7.4 as measured by sCy5 ( Figure 2C).

Study of Fluorescent Properties
In order to evaluate the fluorescent properties of the labeled antibodies, fluorescence emission spectra of the four final conjugates were recorded. While branched conjugate 24 demonstrated higher fluorescence intensity compared to 22 (Figure 3A), the approximately twofold increase was disproportionate to the more than fivefold difference in the DOL (21 vs. 3.7); the discrepancy can be ascribed to self-quenching of fluorescein due to fluorophores being in close proximity around the branched core. Contrarily, the difference in fluorescence intensity in conjugates 29 and 27 was completely proportional to the difference in the degree of labeling, indicating that no apparent self-quenching took place ( Figure 3B). This might be in part due to a longer linker connecting dye fragments to the core in 29 compared to 24 (Scheme 2).
For additional proof of the structure of conjugates 27 and 29, we studied the FRET effect in the sCy3-sCy5 pair ( Figure 3C,D). Energy transfer was observed in both conjugates; curiously, it was found to be more efficient in 27 compared to 29 despite the lower number of sCy5 fragments per one sCy3.

Antibody Conjugate Affinity Determination
We studied the effect of sodium periodate concentration and temperature on the affinity of the modified 6H8 antibody. For this purpose, 6H8 was oxidized at two different periodate concentrations, 5 mM and 20 mM, at room temperature and at 0 °C ( Figure 4A). The buffer chosen was 20 mM acetate with pH 5.0 and 150 mM NaCl. In all cases, the reaction proceeded for 30 min. After which, excess periodate was quenched with a 20% solution of glycerol in water. The antibodies were purified by gel filtration and then incubated with semicarbazide for 60 min in an acetate buffer at pH 5.0. The modified antibodies were then analyzed by ELISA. It was found that in the case of antibody 6H8, the periodate concentration in the range of 5-20 mM had no significant effect on the affinity. Similarly, the temperature had no effect, and in all cases, the antibody retained its affinity. Based on the data obtained, we chose 20 mM sodium periodate concentration, room temperature, incubation for 30 min, and an acetate buffer with pH 5.0 to obtain the fluorescent conjugates. The affinity of the four final conjugates, also determined by ELISA, was shown to be largely uncompromised ( Figure 4B). For additional proof of the structure of conjugates 27 and 29, we studied the FRET effect in the sCy3-sCy5 pair ( Figure 3C,D). Energy transfer was observed in both conjugates; curiously, it was found to be more efficient in 27 compared to 29 despite the lower number of sCy5 fragments per one sCy3.

Antibody Conjugate Affinity Determination
We studied the effect of sodium periodate concentration and temperature on the affinity of the modified 6H8 antibody. For this purpose, 6H8 was oxidized at two different periodate concentrations, 5 mM and 20 mM, at room temperature and at 0 • C ( Figure 4A). The buffer chosen was 20 mM acetate with pH 5.0 and 150 mM NaCl. In all cases, the reaction proceeded for 30 min. After which, excess periodate was quenched with a 20% solution of glycerol in water. The antibodies were purified by gel filtration and then incubated with semicarbazide for 60 min in an acetate buffer at pH 5.0. The modified antibodies were then analyzed by ELISA. It was found that in the case of antibody 6H8, the periodate concentration in the range of 5-20 mM had no significant effect on the affinity. Similarly, the temperature had no effect, and in all cases, the antibody retained its affinity. Based on the data obtained, we chose 20 mM sodium periodate concentration, room temperature, incubation for 30 min, and an acetate buffer with pH 5.0 to obtain the fluorescent conjugates. The affinity of the four final conjugates, also determined by ELISA, was shown to be largely uncompromised ( Figure 4B). finity of the modified 6H8 antibody. For this purpose, 6H8 was oxidized at two different periodate concentrations, 5 mM and 20 mM, at room temperature and at 0 °C ( Figure 4A). The buffer chosen was 20 mM acetate with pH 5.0 and 150 mM NaCl. In all cases, the reaction proceeded for 30 min. After which, excess periodate was quenched with a 20% solution of glycerol in water. The antibodies were purified by gel filtration and then incubated with semicarbazide for 60 min in an acetate buffer at pH 5.0. The modified antibodies were then analyzed by ELISA. It was found that in the case of antibody 6H8, the periodate concentration in the range of 5-20 mM had no significant effect on the affinity. Similarly, the temperature had no effect, and in all cases, the antibody retained its affinity. Based on the data obtained, we chose 20 mM sodium periodate concentration, room temperature, incubation for 30 min, and an acetate buffer with pH 5.0 to obtain the fluorescent conjugates. The affinity of the four final conjugates, also determined by ELISA, was shown to be largely uncompromised ( Figure 4B).

Flow Cytometry
The four fluorescent conjugates were used to detect PRAME-expressing cells (lines THP-1, K562, MelP, and WI-38 with PRAME overexpression [25]) in flow cytometry experiments. The median fluorescence intensity of the stained cells was determined, and the unstained cells were used as controls (Figures 5-8). The data is summarized in Table 2.
Molecules 2023, 28, x FOR PEER REVIEW 9 of 18 The four fluorescent conjugates were used to detect PRAME-expressing cells (lines THP-1, K562, MelP, and WI-38 with PRAME overexpression [25]) in flow cytometry experiments. The median fluorescence intensity of the stained cells was determined, and the unstained cells were used as controls (Figures 5-8). The data is summarized in Table 2.           For fluorescein-based conjugates 22 and 24, fluorescence was registered in the FITC channel. For lines K-562, THP-1, and MelP, the difference between the stained cells and the control was marginal, and sufficient fluorescence was only observed for PRAME-overexpressing WI-38 ( Figure 7A). The branched conjugate 24 was found to perform worse than its linear counterpart 22 (Table 2), which we believe is due to a combination of factors: For fluorescein-based conjugates 22 and 24, fluorescence was registered in the FITC channel. For lines K-562, THP-1, and MelP, the difference between the stained cells and the control was marginal, and sufficient fluorescence was only observed for PRAMEoverexpressing WI-38 ( Figure 7A). The branched conjugate 24 was found to perform worse than its linear counterpart 22 (Table 2), which we believe is due to a combination of factors: (1) self-quenching of fluorescein (as described in Section 2.4) and (2) lower affinity of 24 compared to 22 ( Figure 3B). It should also be noted that for high-DOL conjugates, the mean DOL of the cell-bound fraction can be lower than that of the free conjugate, as affinity is generally inversely correlated with the DOL [12]; should this be the case, the fluorescence of stained cells becomes lower than expected.
In the case of conjugates 27 and 29, sufficient fluorescence was registered for all cell lines in the APC channel. Contrary to what was observed for fluorescein-labeled antibodies, the branched conjugate 29 consistently performed better than 27, giving rise to cell fluorescence intensity that was 1.6 to 3 times higher ( Table 2). In this case, the higher DOL was successfully translated into increased cell fluorescence thanks to the lack of self-quenching in 29 and the small difference in the affinity of 27 and 29.

General Methods
Dimethylformamide (DMF) was purified by distillation with benzene/water followed by vacuum distillation over CaH 2 and stored over 3 Å molecular sieves. Dichloromethane (DCM) and methanol (MeOH) were purified by distillation; dimethyl sulfoxide (DMSO) was purified by vacuum distillation and stored over 3 Å molecular sieves. Compounds 7, 21, and 26 were from Lumiprobe. Reactions were monitored by thin-layer chromatography (TLC) carried out on 0.20 mm Merck silica gel plates (60 F 254 ) using UV light as a visualizing agent and an alkaline aqueous solution of KMnO 4 and heat as developing agents. Merck Silica gel 60 was used for flash column chromatography. Size exclusion chromatography was performed using Sephadex G10 and G50 (Pharmacia Fine Chemicals). UV-Vis absorbance spectra were recorded on a Cary 100 UV-Visible spectrophotometer. Fluorescence spectra were recorded on a Per-kinElmer LS55 luminescence spectrometer. NMR spectra were recorded on Bruker 700 and 800 MHz instruments and calibrated using residual solvent as an internal reference (for CDCl 3 : 1 H, δ 7.26 ppm and 13 C, δ 77.16 ppm; for DMSO-d 6 : 1 H, δ 2.50 ppm and 13 C, δ 39.52 ppm) unless otherwise noted. The following abbreviations were used to explain the multiplicities: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, app = apparent, and br = broad. High-resolution mass spectra (HRMS) were recorded on a Thermo Scientific Orbitrap Exactive mass spectrometer using ESI (Electrospray ionization).

General Procedure B
A solution of a DBCO dye derivative in DMSO was added to a solution of a 6H8 conjugate in 200 µL, PBST (pH 7.4, 0.01% or 0.05% (v/v) Tween 80). The mixture was agitated on an orbital shaker for 2 h at 23 • C, after which the resultant conjugate was purified on a Sephadex G-50 column equilibrated in PBST (pH 7.4, 0.01% or 0.05% (v/v) Tween 80). The yield and degree of labeling of the conjugate were determined spectrophotometrically.

Conjugate 24
Conjugate 23 (0.053 mg, 0.35 nmol, 1.0 equiv) was treated with reagent 21 (11.1 mM in DMSO, 14.5 µL, 161 nmol, 460 equiv) according to general procedure B (buffer for purification: 0.05% PBST). Due to the high degree of labeling of conjugate 24, it is impossible to determine its precise quantity and degree of labeling. In assumption that the reaction was quantitative, it yielded 0.053 mg (100%) of 24 with an average degree of labeling of 21.

Flow Cytometry
Cell lines THP-1, K562, MelP, and WI-38 with PRAME overexpression were used for staining with labeled antibodies. Before staining, the cells were washed from the culture medium in 1% BSA PBS solution. After washing, the cells were diluted to a concentration of approximately 500,000 per mL. Staining was performed in samples of 100 µL of cell suspension containing 1 µg of antibody conjugate. Incubation was performed for 15 min in the dark. After that, unbound antibodies were washed away with 1% BSA solution in PBS. Flow cytometry was performed on an ACEA NovoCyte instrument; fluorescence was registered in the FITC and APC channels. Table 3 shows the most appropriate channels for each of the dyes.

ELISA Assay
A high protein-binding capacity 96-well ELISA plate was coated with recombinant PRAME protein at a concentration of 100 ng in 100 µL of PBS per well. The plate wells were triple washed with PBST, after which, the 6H8 antibody and its derivatives (100 µL per well in PBS with 1% BSA) were added in duplicate at different concentrations. Following incubation on an orbital shaker for 1 h at 23 • C and three PBST washes, antimouse Fcspecific HRP-labeled antibodies (100 µL per well in PBS with 1% BSA) were added to the wells. After incubation on an orbital shaker for 1 h at 23 • C and triple washing with PBST, a solution of OPD (100 µL per well) was added. The reaction was stopped by adding 10% sulfuric acid, and the optical density (OD) was measured at 490 nm on a Packard SpectraCount BS10000 (PerkinElmer, Waltham, MA, USA) microplate reader.

Conclusions
We have developed a methodology that allows for producing fluorescent antibodies with a higher degree of labeling. It relies on the periodate oxidation of IgG glycans followed by oxime ligation with branched linkers and fluorescent labeling by SPAAC. Antibody affinity was found to be largely unaffected in the course of the synthesis. The utility of the approach was demonstrated by the detection of the PRAME protein in cell membranes by flow cytometry. The results of this work can be used as the basis for the synthesis of branched antibody-based conjugates, such as fluorescent probes and antibody-drug conjugates with an increased drug load.