Sensitive Immunofluorescent Detection of the PRAME Antigen Using a Practical Antibody Conjugation Approach

Bioconjugation of antibodies with various payloads has diverse applications across various fields, including drug delivery and targeted imaging techniques. Fluorescent immunoconjugates provide a promising tool for cancer diagnostics due to their high brightness, specificity, stability and target affinity. Fluorescent antibodies are widely used in flow cytometry for fast and sensitive identification and collection of cells expressing the target surface antigen. Nonetheless, current approaches to fluorescent labeling of antibodies most often use random modification, along with a few rather sophisticated site-specific techniques. The aim of our work was to develop a procedure for fluorescent labeling of immunoglobulin G via periodate oxidation of antibody glycans, followed by oxime ligation with fluorescent oxyamines. Here, we report a novel technique based on an in situ oxime ligation of ethoxyethylidene-protected aminooxy compounds with oxidized antibody glycans. The approach is suitable for easy modification of any immunoglobulin G, while ensuring that antigen-binding domains remain intact, thus revealing various possibilities for fluorescent probe design. The technique was used to label an antibody to PRAME, a cancer-testis protein overexpressed in a number of cancers. A 6H8 monoclonal antibody to the PRAME protein was directly modified with protected-oxyamine derivatives of fluorescein-type dyes (FAM, Alexa488, BDP-FL); the stoichiometry of the resulting conjugates was characterized spectroscopically. The immunofluorescent conjugates obtained were applied to the analysis of bone marrow samples from patients with oncohematological diseases and demonstrated high efficiency in flow cytometry quantification. The approach can be applied for the development of various immunofluorescent probes for detection of diagnostic and prognostic markers, which can be useful in anticancer therapy.

Oxime ligation, a "click" reaction of oxyamines (O-substituted hydroxylamines) with aldehydes and ketones in an aqueous solution at physiological pH, is widely used as a tool for bioconjugation [21]. The oxime linkage is stable enough [22] to ensure the integrity of such bioconjugates in various applications. Carbonyl groups suitable for oxime ligation can be introduced into the antibody molecule by periodate oxidation of the glycosylated site [5, 23,24], even one that has been genetically engineered [25]. Alternatively, the carbonyl group can be introduced into the antibody by enzymatic modification of the glycan part with a keto-sugar [26], or via site-specific incorporation of genetically encoded p-acetylphenylalanine into a peptide chain [27,28]. Whereas the two latter approaches require cumbersome enzymatic or genetic manipulations, periodate oxidation of the carbohydrate site can be easily applied to any full-size antibody. The glycosylation site of immunoglobulin G is located at the heavy chain, at a significant distance from the antigen binding domain, thus, the modification does not influence antibody affinity.
The obvious requirements for a fluorescent immunoconjugate (an antibody modified with fluorescent dyes) are high brightness, specificity, high chemical and spectral stability and the ability to preserve antibody affinity [5]. We aimed to develop a convenient procedure for labeling any immunoglobulin G via periodate oxidation followed by oxime ligation.
While some oxyamine derivatives of fluorescent dyes are commercially available, their high reactivity limits their storage time even in salt form, e.g., hydrochloride. Derivatives of luminescent dyes that are stable when stored with a masked oxyamine function and easily releasable before conjugation with carbonyl compounds are of potential interest. We aimed to develop fluorescent reagents with such properties, a procedure for their reaction with aldehydes, and to demonstrate the usefulness of the reagents for labeling of oxidized immunoglobulins. Here, we report the synthesis of fluorescent dye derivatives that contain an ethoxyethylidene-protected aminooxy function and their application to fluorescent labeling of antibodies.
To demonstrate the viability of the approach, we used an antibody to PRAME (preferentially expressed antigen in melanoma), a cancer-testis protein that is overexpressed in a number of cancers and is thought to be suitable as a diagnostic and prognostic marker in anticancer therapy [29][30][31][32][33][34][35][36][37][38][39][40][41]. Recently, we developed an approach for labeling an 6H8 monoclonal antibody to the PRAME protein with fluorescent dye Cy3 [42]. Here, we report the direct labeling of this antibody with protected-oxyamine derivatives of dyes with fluorescein-type emission.

Synthesis of Modifying Reagents
We found that the 1-Ethoxyethylidene group, used to protect O-alkylhydroxylamines, is completely stable under alkaline conditions and is easily removable under acidic conditions [43,44]. Our aim was to prepare derivatives of fluorescent dyes containing a protected oxyamine function. Bifunctional protected oxyamine-azide tetraethyleneglycol-based reagent 1 [42] was used as the starting compound. Selective azide reduction by Ph 3 P yielded bifunctional reagent 2, containing amine and a masked oxyamine function. Compound 2 was acylated by oxysuccinimide esters of fluorescent dyes-BODIPY FL (BDP-FL), 6-carboxyfluorescein (FAM), and Alexa 488 (AF488)-to yield protected derivatives 3-5 (Scheme 1). Attempts to remove the protection group in compounds 3-5 were unsuccessful due to considerable degradation of the dyes in the presence of the free oxyamine function. (AF488) -to yield protected derivatives 3-5 (Scheme 1). Attempts to remove the protection group in compounds 3-5 were unsuccessful due to considerable degradation of the dyes in the presence of the free oxyamine function.

Modification of Monoclonal Antibody 6H8
The presence of the PRAME protein was observed on the surface of K562 leukemic cells, which suggested the possibility of using PRAME as a target for immunotherapy. Since PRAME expression is not limited to leukemia, monitoring its expression in solid tumor cells will enable the development of new drugs for these diseases as well. In our study, the presence of the PRAME protein on the surface of melanoma cells was assessed by flow cytometry. According to the results obtained, this protein is active on the surface of mel P cells, which are metastatic melanoma. The intensity of the staining of mel P and K562 cells was comparable. Therefore, the PRAME protein is located on the surface of melanoma cells and can be used as a target for immunotherapy.
The modification of the 6H8 antibody was based on periodate oxidation, to generate aldehyde groups on the glycoside moieties of the immunoglobulin. The resulting oxidized antibodies were labeled by oxyamine dyes generated in situ from compounds 3-5 to yield fluorescent antibodies ( Figure 1).

Modification of Monoclonal Antibody 6H8
The presence of the PRAME protein was observed on the surface of K562 leukemic cells, which suggested the possibility of using PRAME as a target for immunotherapy. Since PRAME expression is not limited to leukemia, monitoring its expression in solid tumor cells will enable the development of new drugs for these diseases as well. In our study, the presence of the PRAME protein on the surface of melanoma cells was assessed by flow cytometry. According to the results obtained, this protein is active on the surface of mel P cells, which are metastatic melanoma. The intensity of the staining of mel P and K562 cells was comparable. Therefore, the PRAME protein is located on the surface of melanoma cells and can be used as a target for immunotherapy.
The modification of the 6H8 antibody was based on periodate oxidation, to generate aldehyde groups on the glycoside moieties of the immunoglobulin. The resulting oxidized antibodies were labeled by oxyamine dyes generated in situ from compounds 3-5 to yield fluorescent antibodies ( Figure 1).
(AF488) -to yield protected derivatives 3-5 (Scheme 1). Attempts to remove the protection group in compounds 3-5 were unsuccessful due to considerable degradation of the dyes in the presence of the free oxyamine function.

Modification of Monoclonal Antibody 6H8
The presence of the PRAME protein was observed on the surface of K562 leukemic cells, which suggested the possibility of using PRAME as a target for immunotherapy. Since PRAME expression is not limited to leukemia, monitoring its expression in solid tumor cells will enable the development of new drugs for these diseases as well. In our study, the presence of the PRAME protein on the surface of melanoma cells was assessed by flow cytometry. According to the results obtained, this protein is active on the surface of mel P cells, which are metastatic melanoma. The intensity of the staining of mel P and K562 cells was comparable. Therefore, the PRAME protein is located on the surface of melanoma cells and can be used as a target for immunotherapy.
The modification of the 6H8 antibody was based on periodate oxidation, to generate aldehyde groups on the glycoside moieties of the immunoglobulin. The resulting oxidized antibodies were labeled by oxyamine dyes generated in situ from compounds 3-5 to yield fluorescent antibodies ( Figure 1).  The absorption spectra of the labeled antibodies ( Figure 2A) contain the characteristic absorption bands of all of the dyes used for modification. The correlation of absorption at 280 nm and at the characteristic wavelengths allows for the determination of the dye load. Fluorescent conjugates with dye-antibody ratios of 4 (FAM), 2 (BDP-FL) and 2.5 (AF488) were obtained. The emission maxima ( Figure 2B) of the fluorescent 6H8 conjugates had the same wavelengths as those of the initial dyes FAM and AF488 (520 nm), and BDP-FL (509 nm). Thus, the method developed provides a load of 2-4 fluorescent dye per antibody molecule. Fluorescein-type dyes retained their photophysical properties under these conditions.

PRAME Detection by Flow Cytometry
Our preliminary experiments PRAME protein detection, based on phycoerythrin-labeled secondary antibodies to the 6H8 monoclonal antibody, showed that the intensity of the fluorescent signal in flow cytometry was low in comparison with control samples. This method was complicated by two-stage binding, which led to an unstable labeling construct on the cell surface. As a result, we opted for a direct labeling strategy, using a covalently labeled monoclonal antibody ( Figure 1) with several fluorophores.
The obtained fluorescent antibodies were applied to a quantitative analysis of PRAME expression in the model cell lines (K562 and AMO-1) using a flow cytometry approach ( Figure 3).  The emission maxima ( Figure 2B) of the fluorescent 6H8 conjugates had the same wavelengths as those of the initial dyes FAM and AF488 (520 nm), and BDP-FL (509 nm). Thus, the method developed provides a load of 2-4 fluorescent dye per antibody molecule. Fluorescein-type dyes retained their photophysical properties under these conditions.

PRAME Detection by Flow Cytometry
Our preliminary experiments PRAME protein detection, based on phycoerythrinlabeled secondary antibodies to the 6H8 monoclonal antibody, showed that the intensity of the fluorescent signal in flow cytometry was low in comparison with control samples. This method was complicated by two-stage binding, which led to an unstable labeling construct on the cell surface. As a result, we opted for a direct labeling strategy, using a covalently labeled monoclonal antibody ( Figure 1) with several fluorophores.
The obtained fluorescent antibodies were applied to a quantitative analysis of PRAME expression in the model cell lines (K562 and AMO-1) using a flow cytometry approach (Figure 3). tion at 280 nm and at the characteristic wavelengths allows for the determination of th dye load. Fluorescent conjugates with dye-antibody ratios of 4 (FAM), 2 (BDP-FL) and 2 (AF488) were obtained. The emission maxima ( Figure 2B) of the fluorescent 6H8 conjugates had the sam wavelengths as those of the initial dyes FAM and AF488 (520 nm), and BDP-FL (509 nm Thus, the method developed provides a load of 2-4 fluorescent dye per antibody mole cule. Fluorescein-type dyes retained their photophysical properties under these cond tions.

PRAME Detection by Flow Cytometry
Our preliminary experiments PRAME protein detection, based on phyco erythrin-labeled secondary antibodies to the 6H8 monoclonal antibody, showed that th intensity of the fluorescent signal in flow cytometry was low in comparison with contro samples. This method was complicated by two-stage binding, which led to an unstab labeling construct on the cell surface. As a result, we opted for a direct labeling strategy using a covalently labeled monoclonal antibody ( Figure 1) with several fluorophores.
The obtained fluorescent antibodies were applied to a quantitative analysis o PRAME expression in the model cell lines (K562 and AMO-1) using a flow cytometr approach (Figure 3).  We obtained data for the staining of K562 and AMO-1 cells with PRAME-recognizing antibodies. As a result of staining with antibodies labeled with fluorescent molecules, the cells of these lines acquired significantly higher fluorescence parameters compared to cells stained with an isotypic control ( Figure 4). We obtained data for the staining of K562 and AMO-1 cells with PRAME-recognizing antibodies. As a result of staining with antibodies labeled with fluorescent molecules, the cells of these lines acquired significantly higher fluorescence parameters compared to cells stained with an isotypic control (Figure 4). The PRAME expression level in the K562 line was found to be 1393 ± 83% relative to ABL (Abelson kinase gene). The PRAME expression level in the AMO-1 line was 414 ± 29% relative to ABL. Thus, the presence of the PRAME protein on the surface of K562 and AMO-1 cells was confirmed. It should be noted that the K562 line is widely used as a model for cells overexpressing PRAME. There exist a number of reports of high PRAME mRNA levels in these cells [45][46][47][48][49][50][51][52][53][54]. Moreover, the results of staining the K562 cell membrane with PRAME-recognizing antibodies have been previously investigated [32,55]. Our results are consistent with previous observations, indicating the occurrence of binding of PRAME-recognizing antibodies to the surface of a PRAME-expressing cell.
To the best of our knowledge, studies on PRAME activity in AMO-1 cells, a model of multiple myeloma, have not yet been conducted. We chose this line for research due to the fact that multiple myeloma is characterized by a high frequency of PRAME expression [56]. The activity of the PRAME gene and the intensity of staining of the AMO-1 membrane were lower compared to K562 cells (Figure 4). In general, this is not surprising, since the amount of mRNA of the PRAME gene, and of the content of the protein that it encodes within the cell is of a direct ratio [57]. Therefore, a lower PRAME RNA level in the cell results in a lower content of the mature protein.
The highest intensity of cell staining was observed when using antibodies labeled with BDP-FL and FAM dyes. A slightly lower intensity was observed when using the AF488 dye. Encouraged by these results, we used antibodies labeled with the BDP-FL dye to study the bone marrow of patients with oncohematological diseases.
Blast cells from patients with acute myeloid leukemia (AML) were also successfully stained with antibodies to PRAME (Table 1). The PRAME expression level in the K562 line was found to be 1393 ± 83% relative to ABL (Abelson kinase gene). The PRAME expression level in the AMO-1 line was 414 ± 29% relative to ABL. Thus, the presence of the PRAME protein on the surface of K562 and AMO-1 cells was confirmed. It should be noted that the K562 line is widely used as a model for cells overexpressing PRAME. There exist a number of reports of high PRAME mRNA levels in these cells [45][46][47][48][49][50][51][52][53][54]. Moreover, the results of staining the K562 cell membrane with PRAME-recognizing antibodies have been previously investigated [32,55]. Our results are consistent with previous observations, indicating the occurrence of binding of PRAME-recognizing antibodies to the surface of a PRAME-expressing cell.
To the best of our knowledge, studies on PRAME activity in AMO-1 cells, a model of multiple myeloma, have not yet been conducted. We chose this line for research due to the fact that multiple myeloma is characterized by a high frequency of PRAME expression [56]. The activity of the PRAME gene and the intensity of staining of the AMO-1 membrane were lower compared to K562 cells ( Figure 4). In general, this is not surprising, since the amount of mRNA of the PRAME gene, and of the content of the protein that it encodes within the cell is of a direct ratio [57]. Therefore, a lower PRAME RNA level in the cell results in a lower content of the mature protein.
The highest intensity of cell staining was observed when using antibodies labeled with BDP-FL and FAM dyes. A slightly lower intensity was observed when using the AF488 dye. Encouraged by these results, we used antibodies labeled with the BDP-FL dye to study the bone marrow of patients with oncohematological diseases.
Blast cells from patients with acute myeloid leukemia (AML) were also successfully stained with antibodies to PRAME (Table 1). There is a tendency towards a more intense staining of blast cells in patients with a higher level of PRAME expression. At the same time, the fluorescence of lymphocytes in the studied samples did not change. Furthermore, an increase in the fluorescence of blast cells did not occur in AML patients in remission, with blast cells of a normal immunophenotype.
The results obtained indicate the success of the modification of the 6H8 antibody with fluorescent labels. Antibodies can be used in the routine diagnostic practice of laboratories that perform immunophenotyping of bone marrow in patients with acute leukemia. Data on PRAME expression on the surface of leukemic cells can be useful for planning PRAMEspecific therapy.

Confocal Microscopy of K562 Cells Stained with AF488-Labeled 6H8 Monoclonal Antibody
To visualize the binding of the labeled 6H8 antibody with the PRAME antigen, a confocal microscopy of stained cells was performed ( Figure 5). The standard K562 cell line, which contains a surface form of PRAME protein, was selected for observation. Labeling was performed on live K562 cells prior to fixation using paraformaldehyde, to prevent permeabilization and the binding of the AF488-labeled antibody 6H8 with the intracellular form of the PRAME protein. The membrane structures and nucleus were stained with PKH26 and DAPI tracers, correspondingly. There is a tendency towards a more intense staining of blast cells in patients with a higher level of PRAME expression. At the same time, the fluorescence of lymphocytes in the studied samples did not change. Furthermore, an increase in the fluorescence of blast cells did not occur in AML patients in remission, with blast cells of a normal immunophenotype.
The results obtained indicate the success of the modification of the 6H8 antibody with fluorescent labels. Antibodies can be used in the routine diagnostic practice of laboratories that perform immunophenotyping of bone marrow in patients with acute leukemia. Data on PRAME expression on the surface of leukemic cells can be useful for planning PRAME-specific therapy.

Confocal Microscopy of K562 Cells Stained with AF488-Labeled 6H8 Monoclonal Antibody
To visualize the binding of the labeled 6H8 antibody with the PRAME antigen, a confocal microscopy of stained cells was performed ( Figure 5). The standard K562 cell line, which contains a surface form of PRAME protein, was selected for observation. Labeling was performed on live K562 cells prior to fixation using paraformaldehyde, to prevent permeabilization and the binding of the AF488-labeled antibody 6H8 with the intracellular form of the PRAME protein. The membrane structures and nucleus were stained with PKH26 and DAPI tracers, correspondingly. The incubation of K562 cells with the AF488-labeled 6H8 antibody led to staining of the PRAME protein on the surface of cells and in the cytoplasm. These results suggest The incubation of K562 cells with the AF488-labeled 6H8 antibody led to staining of the PRAME protein on the surface of cells and in the cytoplasm. These results suggest that the dye-modified antibody internalizes into the cell as a result of the endocytosis mechanism.

General Methods
Dimethylformamide (DMF) was dried over CaH 2 , then distilled under reduced pressure over P 2 O 5 and kept under argon in the dark. Tetrahydrofuran (THF), dioxane and diethyl ether were distilled over sodium-benzophenone under argon and kept in the dark under an inert atmosphere. Ethyl N-(11-azido-3,6,9-trioxaundecyloxy) acetimidate 1 was prepared as described [42]; NHS derivatives of 6-carboxyfluorescein, 6-carboxy-Alexa488, and BDP-FL were obtained from Lumiprobe. The 6H8 monoclonal antibody was obtained as described in a previous study [58]. All NMR data for the synthesized compounds were obtained in DMSO-d 6 solution at 30 • C. The 1D 1 H, 13 C, and 15 N NMR spectra were acquired using a Bruker AVANCE-III-600 spectrometer equipped with a room-temperature probe. MS spectra were acquired using Thermo Orbitrap Elite hybrid instrument using a Thermo Accela UPLC system, equipped with a Phenomenex Aeris XB-C8 widepore column (150 × 2.1 mm, 3.6 µm). Samples were eluted with a 5→85% gradient of MeCN in water, with 0.1% formic and 0.02% trifluoroacetic acids as the eluent additive. Detection was achieved by UV-VIS DAD and full scan MS (ESI+/−, 250-2000 au).

Antibody Staining
The modification of the 6H8 antibody was initiated by periodate oxidation in an acidic acetate buffer (pH 5) [23]. The reaction was carried out for 1 h at room temperature in the dark. The excess of sodium periodate was quenched with a glycerol solution and the oxidized antibody was desalted using a Sephadex G-50 column and an acidic acetate buffer with pH 3.6. The subsequent reaction with protected oxyamines 3-5 for 1 h produced fluorescently labeled antibodies. The reagents 3-5 have an unlimited storage time. The ethoxyethylydene protecting group was removed in situ in an acidic medium and instantly reacted with carbonyl groups of the oxidized antibody. The excess dye was removed using gel chromatography on Sephadex G-50 in a PBS buffer.

Cell Line Handling
Cultures of cell lines K562 (chronic myeloid leukemia) and AMO-1 (multiple myeloma) were incubated at +37 • C, in a CO 2 incubator on an RPMI-1640 medium containing 2 mM glutamine, 10 mM HEPES, 10% fetal bovine serum (FBS) and 40 µg/mL gentamicin sulfate. The cells were subcultured twice weekly in a new medium at a concentration of 100,000 cells per mL.

Bone Marrow Sample Collection
Bone marrow samples from patients with acute myeloid leukemia were provided by GeneTechnology, (Moscow, Russia). Each patient signed an informed medical consent form confirming their voluntary participation in the scientific research.

RNA Isolation and Real-Time PCR
The isolation of total RNA from cell lines and bone marrow materials of patients with acute myeloid leukemia was performed using the "RNA-extract" reagent set (GeneTechnology, Moscow, Russia), in accordance with the guidelines provided by the vendor. Subsequent synthesis of cDNA and the determination of expression levels of the PRAME gene and the Abelson kinase (ABL) housekeeping gene were performed using Oncoscreen-9Q and Oncoscreen-14Q kits (GeneTechnology, Moscow, Russia), respectively, according to the guidelines provided by the vendor. Real-time PCR was performed on a DT-Lite device (DNA-Technology, Moscow, Russia). The level of PRAME expression was assessed using a program provided by DNA-technology (Moscow, Russia).

Cells Staining and Flow Cytometry
K562 or AMO-1 cells (1 million), or 500 µL of patients' bone marrow, were used for flow cytometry. Both the K562 and AMO-1 cells were washed from the culture medium by centrifugation in a PBS solution for 5 min at 500 rpm. Bone marrow samples were subjected to hemolysis in 5 mL of 0.9% NH 4 Cl solution for 10 min to remove erythrocytes. Then, the cells were twice washed by centrifugation in PBS solution for 5 min at 500 rpm. The K562 and AMO-1 cells, as well as the patients' bone marrow cells, were resuspended in 1 mL of PBS, and 100 µL samples were used for staining with antibodies to PRAME and the isotypic control. The samples were treated with 0.1-0.5 µL of a 1 mg/mL solution of labeled antibodies. Additionally, bone marrow cells were stained with antibodies to antigens CD45, CD34, CD117, CD13, CD33, CD14, CD36, CD15, CD11c, CD41a, MPO, CD19, CD22, CD10, CD2, CD3, CD4, CD8, CD56 and CD15 (SonyBiotechnology, San Jose, CA, USA) to determine blast cell counts, as well as the number of normal lymphocytes and granulocytes. Incubation with antibodies was conducted for 15 min in the dark. After incubation, unbound antibodies were washed off and cells were analyzed on an ACEA NovoCyte flow cytometer (ACEA Biosciences, San Diego, CA, USA).
Cells of tumor lines and samples obtained from patients were stained with antibodies to the PRAME antigen, and the fluorescence of the cells was measured in comparison to the control stained with isotypic antibodies. In bone marrow samples from patients with acute leukemia, the degree of staining of blast cells and other cells with a normal immunophenotype was also assessed.

Cell Staining for Confocal Microscopy
K562 cells were transferred into a PBS-BSA buffer and incubated with AF488-labeled monoclonal antibody 6H8 at 37 • C. After 1 h, the cells were precipitated by centrifugation, resuspended in PBS, and incubated with 3.7% paraformaldehyde in PBS for 15 min at room temperature. The fixed cells were washed with PBS and stained for 20 min with DAPI, and for 10 min with a PKH26 tracker. Then, the staining medium was removed, and the cells were washed with PBS again. The coverslips were placed onto the slides with a drop of 10% Mowiol in 0.1 M Tris-HCl pH 8.5 and washed stained K562 cells. Slides were analyzed using an Eclipse TE2000 confocal microscope (Nikon, Tokyo, Japan) with 515/30 (blue), 590/50 (green), and red (650LP) filters sets.

Conclusions
This study aimed to contribute to the development of a facile method for fluorescent modification of antibodies with dyes containing a masked oxyamine group. After synthesis, these modification reagents can be stored for an unlimited time, thus allowing the rapid and convenient labeling of all types of immunoglobulin G. Our approach utilizes periodate oxidation as the first step, and can therefore potentially be applied to any glycosylated antibody.
The synthetic procedure was applied to a PRAME-specific monoclonal 6H8 antibodies, yielding a fluorescent immunoconjugate that was found to be effective for PRAME detection, both in model cell lines and in bone marrow samples of patients with acute leukemia. The constructed fluorescent probe allowed us to investigate PRAME expression in the AMO-1 cell line. Moreover, the PRAME-specific probe provided reliable data on bone marrow samples, thus, it can be considered as a potential and promising tool for prognostic and diagnostic purposes.