Novel Bifunctional Affibody Molecules with Specific Binding to Both EBV LMP1 and LMP2 for Targeted Therapy of Nasopharyngeal Carcinoma

Antibodies are considered highly specific therapeutic agents in cancer medicines, and numerous formats have been developed. Among them, bispecific antibodies (BsAbs) have gained a lot of attention as a next-generation strategy for cancer therapy. However, poor tumor penetration is a major challenge because of their large size and thus contributes to suboptimal responses within cancer cells. On the other hand, affibody molecules are a new class of engineered affinity proteins and have achieved several promising results with their applications in molecular imaging diagnostics and targeted tumor therapy. In this study, an alternative format for bispecific molecules was constructed and investigated, named ZLMP110-277 and ZLMP277-110, that targets Epstein-Barr virus latent membrane protein 1 (LMP1) and latent membrane protein 2 (LMP2). Surface plasmon resonance (SPR), indirect immunofluorescence assay, co-immunoprecipitation, and near-infrared (NIR) imaging clearly demonstrated that ZLMP110-277 and ZLMP277-110 have good binding affinity and specificity for both LMP1 and LMP2 in vitro and in vivo. Moreover, ZLMP110-277 and ZLMP277-110, especially ZLMP277-110, significantly reduced the cell viability of C666-1 and CNE-2Z as compared to their monospecific counterparts. ZLMP110-277 and ZLMP277-110 could inhibit phosphorylation of proteins modulated by the MEK/ERK/p90RSK signaling pathway, ultimately leading to suppression of oncogene nuclear translocations. Furthermore, ZLMP110-277 and ZLMP277-110 showed significant antitumor efficacy in nasopharyngeal carcinoma-bearing nude mice. Overall, our results demonstrated that ZLMP110-277 and ZLMP277-110, especially ZLMP277-110, are promising novel prognostic indicators for molecular imaging and targeted tumor therapy of EBV-associated nasopharyngeal carcinoma.


Introduction
Epstein-Barr virus (EBV), a member of the herpesvirus family, has been reported to be associated with a number of malignancies, such as nasopharyngeal carcinoma (NPC), Burkitt's lymphoma (BL), Hodgkin's lymphoma, and gastric cancer [1]. In NPC, the latent membrane proteins LMP1 and LMP2 encoded by EBV are frequently detected and can play an important role in various biological processes, such as cell proliferation, cell migration, cell invasion, and apoptosis [2,3]. Moreover, studies have confirmed that both LMP1 and LMP2 can contribute to NPC pathogenesis [4,5], emphasizing the need to develop novel diagnostic and therapeutic approaches for EBV-associated NPC. Both LMP1 and LMP2 initiate and activate ERK-MAPK, JAK/STAT, NF-κB JNK/p38-SAPK, and PI3-K/Akt signal transduction pathways that affect undesirable phenotypic changes, including upregulation of c-Fos, c-Myc, and c-jun proto-oncogenes in NPC [6,7]. Consequently, monoclonal specificity in vitro and in vivo [10,39]. Based on these considerations, our group designed novel bifunctional affibody molecules for increased binding specificity and enhanced antitumor effects on NPC cells both in vitro and in vivo. We connected the LMP2A-N terminal affibody (Z LMP2AN 110) to the LMP1-C terminal (Z LMP1-C 277) using the (G4S)3 linker and further changed their positions to construct dual-affinity proteins, respectively, named Z LMP 110-277 and Z LMP 277-110. In this study, we constructed novel bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) with the potential for dual binding to the target proteins LMP1 and LMP2, antitumor effects on NPC cells, and in vivo evaluation of targeted therapy for NPC in xenograft tumor-bearing nude mice. The EBV Z LMP1-C 277 and EBV Z LMP2A-N 110 affibody molecules were selected using a phage display library that targets specific binders to the EBV LMP1-C (amino acids 186-387) and LMP2A-N (amino acids 1-119) proteins, respectively. The two monospecific affibody molecules (EBV Z LMP1-C 277 and EBV Z LMP2A-N 110) were connected by the linker (G4S) 3 between Z LMP1-C 277 and Z LMP2A-N 110 and His-tag to produce Z LMP 110-277 and Z LMP 277-110 bifunctional affibody molecules ( Figure 1A). The fusion linker provides spatial separation between two proteins while also improving biological activity and expanding expression yield [42]. Figure 1B shows a prediction of the three-dimensional structure of proteins using the SWISS-MODEL workspace ( Figure 1B). This indicates that the connection between the two may not change their respective spatial conformations through flexible peptides. molecules to form bifunctional molecules for dual targets and antitumor effects in mouse models, which inspired our group to design EBV LMP1-LMP2 bifunctional affibody molecules.
In our previous studies, we generated monospecific affibody molecules that target either LMP1 or LMP2 using phage display technology and explored the basis of their binding specificity in vitro and in vivo [10,39]. Based on these considerations, our group designed novel bifunctional affibody molecules for increased binding specificity and enhanced antitumor effects on NPC cells both in vitro and in vivo. We connected the LMP2A-N terminal affibody (ZLMP2AN110) to the LMP1-C terminal (ZLMP1-C277) using the (G4S)3 linker and further changed their positions to construct dual-affinity proteins, respectively, named ZLMP110-277 and ZLMP277-110. In this study, we constructed novel bifunctional affibody molecules (ZLMP110-277 and ZLMP277-110) with the potential for dual binding to the target proteins LMP1 and LMP2, antitumor effects on NPC cells, and in vivo evaluation of targeted therapy for NPC in xenograft tumor-bearing nude mice.

Design and Construction of Bifunctional Affibody Molecules pET21a(+)/ZLMP110-277 and pET21a(+)/ZLMP277-110
The EBV ZLMP1-C277 and EBV ZLMP2A-N110 affibody molecules were selected using a phage display library that targets specific binders to the EBV LMP1-C (amino acids 186-387) and LMP2A-N (amino acids 1-119) proteins, respectively. The two monospecific affibody molecules (EBV ZLMP1-C277 and EBV ZLMP2A-N110) were connected by the linker (G4S)3 between ZLMP1-C277 and ZLMP2A-N110 and His-tag to produce ZLMP110-277 and ZLMP277-110 bifunctional affibody molecules ( Figure 1A). The fusion linker provides spatial separation between two proteins while also improving biological activity and expanding expression yield [42]. Figure 1B shows a prediction of the three-dimensional structure of proteins using the SWISS-MODEL workspace ( Figure 1B). This indicates that the connection between the two may not change their respective spatial conformations through flexible peptides.  DNA sequences encoding the bifunctional affibody molecules were cloned into pET21a(+) to generate the recombinant plasmids pET21a(+)/Z LMP 110-277 and pET21a(+)/Z LMP 277-110 ( Figure 1C). Then, the recombinant proteins were expressed in E. coli BL21 (DE3) after induction with 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG). Next, the His-tagged fusion proteins expressed were successfully purified by affinity chromatography using Ni-NTA resin and confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis ( Figure 1D). Western blotting further confirmed that fusion proteins were specifically recognized by the anti-His-tag mouse mAb ( Figure 1E). Moreover, SDS-PAGE analysis showed that the final products were obtained with high purity (95%), which can then be used for future investigations.

Analysis of the Binding Selectivity of Bifunctional Affibody Molecules to Cells Expressing LMP1 and LMP2
To measure the expression levels of LMP1 and LMP2 in NPC-positive cells (C666-1 and CNE-2Z) and NPC-negative cells (HNE-2), quantitative reverse transcription polymerase chain reaction (qRT-PCR) and Western blotting assays were performed. The results of qRT-PCR analysis showed that LMP1 and LMP2 were highly expressed in NPCpositive cell lines compared to an NPC-negative cell line ( Figure 3A,C). The qRT-PCR results were further confirmed by Western blotting assays ( Figure 3B,D). Next, we study the cellular binding of ZLMP110-277 and ZLMP277-110 to NPC-positive cells using an indirect immunofluorescence assay. Our results showed that bright green fluorescence appeared at the juxtamembrane area of C666-1 and CNE-2Z that were treated with bifunctional or monospecific affibody molecules, as well as higher fluorescence intensity in cells treated with ZLMP110-277 or ZLMP277-110 compared to ZLMP1-C277 and ZLMP2A-N110 ( Figure 3E,F). However, NPC-positive cells treated with SPA-Z scaffold (ZWT) and HNE-2 treated with bifunctional or monospecific affibody molecules did not exhibit any visible fluorescence   To measure the expression levels of LMP1 and LMP2 in NPC-positive cells (C666-1 and CNE-2Z) and NPC-negative cells (HNE-2), quantitative reverse transcription polymerase chain reaction (qRT-PCR) and Western blotting assays were performed. The results of qRT-PCR analysis showed that LMP1 and LMP2 were highly expressed in NPC-positive cell lines compared to an NPC-negative cell line ( Figure 3A,C). The qRT-PCR results were further confirmed by Western blotting assays ( Figure 3B,D). Next, we study the cellular binding of Z LMP 110-277 and Z LMP 277-110 to NPC-positive cells using an indirect immunofluorescence assay. Our results showed that bright green fluorescence appeared at the juxtamembrane area of C666-1 and CNE-2Z that were treated with bifunctional or monospecific affibody molecules, as well as higher fluorescence intensity in cells treated with Z LMP 110-277 or Z LMP 277-110 compared to Z LMP1-C 277 and Z LMP2A-N 110 ( Figure 3E,F). However, NPC-positive cells treated with SPA-Z scaffold (Z WT ) and HNE-2 treated with bifunctional or monospecific affibody molecules did not exhibit any visible fluorescence signal ( Figure 3G).  In addition, confocal immunofluorescence and co-immunoprecipitation (Co-IP) assays were performed to study intracellular protein-protein interactions. As shown in Figure 4A,B, bifunctional affibody molecules (green) can interact with both targets LMP1 and LMP2 (red) and colocalize (yellow). By comparison, monospecific affibody molecules can only interact with their individual target proteins, LMP1 or LMP2. SPA-Z scaffold (Z WT ) did not display any obvious fluorescence signals. The fluorescence intensity of C666-1 incubated with Z LMP 277-110 was higher than that incubated with Z LMP 110-277 (Supplementary Figure S2). Furthermore, Co-IP assays add more evidence that Z LMP 110-277 and Z LMP 277-110 interact with LMP1 and LMP2 proteins ( Figure 4E,F). Nonetheless, Z LMP1-C 277 and Z LMP2A-N 110 only interact with LMP1 or LMP2 ( Figure 4C,D). Taken together, these results suggest that Z LMP 110-277 and Z LMP 277-110 can recognize and bind simultaneously to both LMP1 and LMP2 proteins.  C666-1 cells were incubated with bifunctional or monospecific affibody molecules for three hours and analyzed by confocal microscopy. The rabbit anti-LMP1 or anti-LMP2 mAb and the mouse anti-His-tag mAb were used as primary antibodies. The goat anti-rabbit antibody conjugated with Cy3 (red) and the goat anti-mouse antibody conjugated with FITC (green) were used as secondary antibodies. The cell nuclei were stained by Hoechst3342 (blue) (400×). The merge images show the co- with bifunctional or monospecific affibody molecules for three hours and analyzed by confocal microscopy. The rabbit anti-LMP1 or anti-LMP2 mAb and the mouse anti-Histag mAb were used as primary antibodies. The goat anti-rabbit antibody conjugated with Cy3 (red) and the goat anti-mouse antibody conjugated with FITC (green) were used as secondary antibodies. The cell nuclei were stained by Hoechst3342 (blue) (400×). The merge images show the co-localization of bifunctional and monospecific affibody molecules with LMP1 and LMP2 (yellow). (E,F) Z LMP 110-277 and Z LMP 277-110 interact with both LMP1 and LMP2, and then an IP was performed with anti-LMP1 and anti-LMP2 antibodies. (C,D) Z LMP1-C 277 and Z LMP2A-N 110 complexed with LMP1 or LMP2 following an IP. Western blotting analysis was performed and incubated with rabbit anti-LMP1, anti-LMP2, or anti-His-tag mAb. IgG served as a negative control. Experiments were performed in triplicate.

In Vivo Tumor Target Efficacy of Z LMP 110-277 and Z LMP 277-110
Furthermore, the in vivo biodistribution and tumor-targeted ability of Z LMP 110-277 and Z LMP 277-110 were investigated in nude mice bearing NPC-positive cell lines (C666-1 and CNE-2Z) and NPC-negative cell lines (HNE-2) tumor xenografts. The nude mice bearing tumor cell xenografts were intravenously injected with Dylight-755-labeled bifunctional or monospecific affibody molecules of 100 µg in 150 µL phosphate buffered saline (PBS) and then observed with a near-infrared imaging system at different time points after tail vein injection. As shown in Supplementary Figure S3, Dylight-755-labeled bifunc-tional and monospecific affibody molecules were widely distributed throughout the body within 0.5 h after injection, excreted from the kidneys, and cleared from the body within 48 h. Then, we investigated the tumor-targeted ability of Dylight-755-labeled Z LMP 110-277, Dylight-755-labeled Z LMP 277-110, Dylight-755-labeled Z LMP1-C 277, and Dylight-755labeled Z LMP2A-N 110 in a tumor-bearing mouse model. In NPC-positive cell (C666-1 and CNE-2Z) xenograft models, we observed that a strong fluorescence signal was detected at the xenograft tumor site 1 h post-injection of Dylight-755-labeled bifunctional and monospecific affibody molecules, peaked at 4 h, and remained for 24 h (Figure 5A,B). Nevertheless, no tumor-specific fluorescence signal was detected post-injection of Dylight-755-labeled bifunctional and monospecific affibody molecules in the HNE-2 xenograft model ( Figure 5C). Moreover, the Dylight-755-labeled SPA-Z scaffold (Z WT ) affibody molecules did not show any tumor-specific signal in xenograft models. These results showed that Z LMP 110-277 and Z LMP 277-110 bifunctional affibody molecules could accumulate at tumor locations in NPC-bearing mice with high specificity in vivo.
Furthermore, the in vivo biodistribution and tumor-targeted ability of ZLMP110-2 and ZLMP277-110 were investigated in nude mice bearing NPC-positive cell lines (C66 and CNE-2Z) and NPC-negative cell lines (HNE-2) tumor xenografts. The nude mice be ing tumor cell xenografts were intravenously injected with Dylight-755-labeled bifu tional or monospecific affibody molecules of 100 µg in 150 µL phosphate buffered sal (PBS) and then observed with a near-infrared imaging system at different time points af tail vein injection. As shown in Supplementary Figure S3, Dylight-755-labeled bifunctio and monospecific affibody molecules were widely distributed throughout the bo within 0.5 h after injection, excreted from the kidneys, and cleared from the body with 48 h. Then, we investigated the tumor-targeted ability of Dylight-755-labeled ZLMP110-2 Dylight-755-labeled ZLMP277-110, Dylight-755-labeled ZLMP1-C277, and Dylight-755-labe ZLMP2A-N110 in a tumor-bearing mouse model. In NPC-positive cell (C666-1 and CNE-2 xenograft models, we observed that a strong fluorescence signal was detected at the xe ograft tumor site 1 h post-injection of Dylight-755-labeled bifunctional and monospec affibody molecules, peaked at 4 h, and remained for 24 h (Figure 5A,B). Nevertheless, tumor-specific fluorescence signal was detected post-injection of Dylight-755-labeled functional and monospecific affibody molecules in the HNE-2 xenograft model (Figu 5C). Moreover, the Dylight-755-labeled SPA-Z scaffold (ZWT) affibody molecules did n show any tumor-specific signal in xenograft models. These results showed that ZLMP1 277 and ZLMP277-110 bifunctional affibody molecules could accumulate at tumor locatio in NPC-bearing mice with high specificity in vivo.

In Vitro Efficacy of Bifunctional Affibody Molecules
Next, to determine the antitumor efficacy of Z LMP 110-277 and Z LMP 277-110, cell viability assays were performed using C666-1, CNE-2Z, and HNE-2 cell lines.  Figure 6A). In addition, colony formation assays were used to test the long-term effects of Z LMP 110-277 and Z LMP 277-110 on NPC-positive cell proliferation. As reported in Figure 6B

Downregulation of the MEK/ERK/p90RSK Signal Transduction Pathway by Bifunctional Affibody Molecules in NPC Cells
The expression of LMP1 and LMP2 contributes to tumor cell proliferation, survival, motility, and invasion [2,3] and is mediated by several signal transduction pathways. Raf-MEK-ERK activates the 90 kDa ribosomal S6 kinase (p90RSK), resulting in increased transcription factors [43]. To explore the effects of bifunctional affibody molecules on the MEK/ERK/p90RSK pathway, Western blotting was performed. As shown in Figure 7A,B, there is a decrease in the level of phospho-Raf-1 (Ser338) in a concentration-and time-dependent manner after treatment with Z LMP 277-110 compared to the control groups mock and SPA-Z scaffold (Z WT ). As stated by the above result, 10 µM was selected for further study of the downstream targets. Our results showed decreased expression levels of phospho-MEK1/2 (Ser217/Ser221) , phospho-ERK1/2 (Thr202/Thr204) , phospho-p90RSK (Ser380) , and the transcription factors (c-Fos and c-Myc) after treatment with Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, and Z LMP2A-N 110 in NPC-positive cell lines ( Figure 7C,D). Moreover, bifunctional affibody molecules had more pronounced effects on the MEK/ERK/p90RSK signaling pathway than monospecific affibody molecules ( Figure 7E,F). Figure 7G shows a schematic illustration of bifunctional affibody molecule downregulation of the MEK/ERK/p90RSK pathway in NPC cells. Taken together, our results further confirmed that bifunctional affibody molecules significantly inhibit the proliferation and progression of NPC cells.
tion factors (c-Fos and c-Myc) after treatment with ZLMP110-277, ZLMP277-110, ZLMP1-C277, and ZLMP2A-N110 in NPC-positive cell lines ( Figure 7C,D). Moreover, bifunctional affibody molecules had more pronounced effects on the MEK/ERK/p90RSK signaling pathway than monospecific affibody molecules ( Figure 7E,F). Figure 7G shows a schematic illustration of bifunctional affibody molecule downregulation of the MEK/ERK/p90RSK pathway in NPC cells. Taken together, our results further confirmed that bifunctional affibody molecules significantly inhibit the proliferation and progression of NPC cells.

In Vivo Therapeutic Efficacy of Bifunctional Affibody Molecules
In vivo therapeutic efficacy of Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, and Z LMP2AN 110 was evaluated in C666-1, CNE-2Z, and HNE-2-bearing nude mice by measuring tumor growth in mice. Tumors grew much faster in groups treated with PBS and SPA-Z scaffold (Z WT ) than Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, Z LMP2AN 110, and cisplatin ( Figure 8A-D). More importantly, treatment with bifunctional affibody molecules strongly decreased tumor growth in NPC-bearing nude mice after 15 days of treatment compared to monospecific affibody molecules. Nevertheless, Z LMP 110-277 and Z LMP 277-110 did not exhibit any inhibitory effects on HNE-2 tumor growth ( Figure 8E,F). Altogether, these results support the idea that bifunctional affibody molecules Z LMP 110-277 and Z LMP 277-110, especially Z LMP 277-110, are promising anti-cancer agents for EBV-associated NPC.

Discussion
In recent years, researchers have focused on new therapy approaches to enhance the efficacy of drugs on tumor cells as compared to conventional cancer treatments. Current clinical practice utilizes mAbs extensively as treatment options for several cancer types, such as ovarian cancer, breast cancer, B-cell non-Hodgkin's lymphoma, etc. [44][45][46]. Moreover, blinatumomab, an anti-CD19/CD3 bispecific antibody, showed great therapeutic efficacy with a good safety profile in lymphoid leukemia patients [47]. However, there are many pitfalls associated with mAbs and bispecific antibodies, such as limited depth of tumor penetration, potential mechanisms of resistance, and high manufacturing costs. Therefore, new strategies for the development of molecularly targeted tumor therapy are urgently needed.
Since affibody molecules were introduced three decades ago as antibody mimetics, a number of them have been generated and characterized. More recently, an affibody molecule against IL-17 (ABY-035) has entered clinical study and been demonstrated to be safe and tolerable [29]. So far, numerous bioengineered protein probes (scFv, Fv, and Fab) with smaller size, shorter circulation time, and deep tumor penetration have been used in a variety of medical applications, including blocking protein-protein interaction, targeted delivery of payload, inhibitory effect of peptide aggregation, and molecular imaging diagnosis [48,49]. In this study, we generated two novel bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) capable of simultaneously binding to LMP1 and LMP2, which display oncogenic properties and induce phenotypic alteration in epithelial cells. Then, we successfully produced these bifunctional affibody molecules using a prokaryotic expression system and evaluated their binding affinity and specificity to LMP1 and LMP2. The bifunctional affibody molecules showed binding to both LMP1 and LMP2 as compared to their monospecific counterparts. These results were similar to previous published studies from our laboratory [50,51]. In addition, immunofluorescence co-localization studies demonstrated that bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) could bind specifically to NPC-positive cell lines (C666-1 and CNE-2Z), as well as colocalize with both LMP1 and LMP2 proteins. Moreover, co-immunoprecipitation provided more evidence of the protein-protein interactions of bifunctional affibody molecules with LMP1 and LMP2 proteins. A study with radiolabeled affibody Z HER2:342 demonstrated fast tumor uptake in xenograft models and provided a higher imaging-specific contrast in HER2positive SKOV-3 xenografts than the HER2 scFv antibody fragment [35]. By contrast, we also noticed similar data that Dylight-755-labeled bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) could rapidly accumulate in tumor locations in NPCpositive xenograft models. Our results suggest that Z LMP 110-277 and Z LMP 277-110 have good binding affinity and specificity to LMP1 and LMP2 both in vitro and in vivo and also solidify the potential of using Z LMP 110-277 and Z LMP 277-110 as molecular imaging probes.
NPC is a tumor of the head and neck that is frequently caused by the Epstein-Barr virus [1]. Despite recent advances in mAbs, radiotherapy, and surgery, overall patient cure is achieved in less than 50% of cases [8]. In addition, distant metastasis is the leading cause of treatment failure in NPC, with a median survival of 10 months or less [52]. Concurrent cisplatin chemoradiotherapy is the standard of care for locally advanced NPC [53], but it is associated with significant toxicity in NPC patients [54]. Therefore, new treatment agents that have the potential to improve NPC patient outcomes and that show a reduced toxicity profile are desperately needed. There are a variety of potential mechanisms for the use of mAbs in the treatment of cancer. For example, antibodies may target a specific portion of tumor development, such as growth factor receptors or receptor-ligand interactions, to kill tumor cells using death-receptor-mediated pathways [55,56]. Moreover, antibodies are capable of inducing complement-mediated cytotoxicity (CDC) or antibody-dependent cellular cytotoxicity (ADCC) [57]. Nonetheless, some of these antibodies were sometimes unable to stimulate antibody-mediated immune responses, including CDC and ADCC, which are important for destroying malignant tumor cells. In order to overcome this problem, fragment variable (Fv), single-chain Fv antibody fragments (scFvs), fragment antigen binding (Fab), and affibody molecules have been successfully used in cancer therapy and targeted drug delivery [58] and have been confirmed to target tumor-specific antigens with therapeutic effects. Among others, affibody molecules can be engineered to form dimers and trimers by varying the length of their peptide linkers, which can effectively improve binding affinity to target proteins [51,59,60]. In vitro cell viability, colony formation, and flow cytometry assays indicated that treatment with bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) inhibited the growth of NPC cell lines (C666-1 and CNE-2Z) and were non-cytotoxic toward the NPC-negative cell line (HNE-2). Importantly, the Z LMP 110-277 and Z LMP 277-110, especially Z LMP 277-110, were more potent than either of the monospecific affibody controls (Z LMP1-C 277 and Z LMP2A-N 110) in NPC-positive cells. Furthermore, in vivo evaluation of antitumor therapeutic efficacy showed that the antitumor activities of Z LMP 110-277 and Z LMP 277-110 were higher than those of Z LMP1-C 277 and Z LMP2A-N 110 monospecific affibody controls. Although cisplatin's antitumor efficacy was higher, mice treated with cisplatin showed significant weight loss compared to Z LMP 110-277 and Z LMP 277-110. Our results confirmed that Z LMP 110-277 and Z LMP 277-110, especially Z LMP 277-110, significantly decreased NPC cell proliferation both in vitro and in vivo.
LMP1 and LMP2 are key EBV-encoded oncoproteins that activate numerous cell signaling pathways, such as ERK-MAPK, JNK/AP1, P13K, and NF-κB, influencing cell proliferation and other cellular responses [6,7]. Therefore, identifying the LMP1 and LMP2 signaling proteins that are involved in the underlying pathological mechanisms of EBV-associated NPC is essential for successful drug discovery and therapeutic targets. A potent small molecule inhibitor to disrupt LMP1 and LMP2 oncogenic pathways and the origin of tumor-initiating cells could potentially be a novel and independent prognostic modality and therapeutic target for patients with EBV-associated NPC. In the present study, Z In summary, we constructed novel bifunctional affibody molecules (Z LMP 110-277 and Z LMP 277-110) and evaluated their binding affinity and specificity towards LMP1 and LMP2 using surface plasmon resonance, indirect immunofluorescence assays, co-immunoprecipitation assays, and near-infrared imaging both in vitro and in vivo. Moreover, in vitro and in vivo studies showed that Z LMP 110-277 and Z LMP 277-110, especially Z LMP 277-110, showed significantly stronger inhibitory actions than Z LMP1-C 277 and Z LMP2A-N 110 in the proliferation of NPC cells. Moreover, Z LMP 110-277 and Z LMP 277-110 could inhibit phosphorylation of proteins modulated by the MEK/ERK/p90RSK signaling pathway, ultimately leading to suppression of oncogene nuclear translocations. Overall, our results showed that Z LMP 110-277 and Z LMP 277-110, especially Z LMP 277-110, are promising novel prognostic indicators for molecular imaging and targeted treatment of EBV-associated NPC.

Template-Based Protein Structure Prediction
The three-dimensional models of the fusion proteins were constructed using the SWISS-MODEL work space, and the biophysical properties were predicted by submitting the protein (amino acid) sequence of bifunctional affibody molecules on the ExPASY server (https://swissmodel.expasy.org/interactive, accessed on 15 November 2022).

Biosensor Analysis
A Biacore T200 (GE Healthcare, Uppsala, Sweden) was used to evaluate the binding affinity of bifunctional affibody molecules and their interactions with LMP1 and LMP2. In our laboratory, we have previously stored pET21a/EBV LMP1-C and pET21a/EBV LMP2A-N, transformed into E. coli strain BL21-competent cells for soluble expression. The EBV LMP1-C and LMP2A-N were immobilized on the surface of the sensor chip CM5 (GE Healthcare) as ligands according to the method described previously [38]. After that, serial dilutions of samples were made, and the analytes were injected over the sensor chip surface to monitor protein-protein interaction. Binding curves were fitted to a 1:1 Langmuir model using BIAcore T200 evaluation 3.0.2 software.

LMP1 and LMP2 Expression Level Analysis in NPC Cells
All cells were cultured, and total ribonucleic acid (RNA) was extracted from C666-1, CNE-2Z, and HNE-2Z cells using the TRIzol reagent Takara from Biomedical Technology Co., Ltd. (Beijing, China). RNA is reverse transcribed into complementary deoxyribonucleic acid (cDNA), followed by two sets of primers specifically designed for each target gene and the Power SYBR Green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). The results were analyzed using QuantStudio Real-Time PCR software (Life Technologies, Carlsbad, CA, USA).
Furthermore, Western blotting was performed to confirm LMP1 and LMP2 expression levels in NPC cells. Briefly, cells were cultured and harvested, washed with ice-cold PBS, and lysed directly in lysis buffer. After that, equal amounts of proteins are separated by SDS-PAGE on 15% gels and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, Burlington, MA, USA). Membranes were blocked in 5% skim milk in PBS with 0.5% Tween 20 (PBST) for 2 h, incubated with primary antibodies, anti-LMP1 (Abcam 136633) or rabbit anti-LMP2A-NCD prepared in-house, overnight at 4 • C, and detected with a fluorescently tagged secondary antibody. Protein bands were visualized using the Western Blotting Imaging System (Clinx, Shanghai, China), and protein expression was quantified by ImageJ 1.33 software (National Institutes of Health). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control.

Indirect Immunofluorescence Assay
An immunofluorescence assay was used to analyze the binding specificity of bifunctional affibody molecules to NPC cells. CNE-2Z, C666-1, and HNE-2 cells were grown on cover slips using cell culture dishes (1 × 10 5 cells/well) for 24 h. Then, cells were treated with 100 µg/mL of affibody molecules (Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, and Z LMP2A-N 110) or Z WT affibody (negative control) and incubated for 3 h at 37 • C. After washing with PBS, cells were fixed with 4% paraformaldehyde and permeabilized using 0.3% Triton X-100 at room temperature. The cells on cover slips were then blocked with blocking buffer for 1 h, followed by incubation with the primary antibody (mouse anti-His-tag mAb) overnight at 4 • C. Afterwards, cells are stained with FITC-conjugated goat anti-mouse IgG secondary antibodies at 37 • C for 1 h. The cell nuclei were stained with propidium iodide (PI) dye, and cells were imaged using a confocal fluorescence microscope (Nikon C1-i, Tokyo, Japan).
To further prove the specificity of bifunctional affibody molecules to LMP1 and LMP2 proteins, colocalization analyses using double immunofluorescence were performed. The procedure is similar to that given above.

Immunoprecipitation
Briefly, C666-1 cells received treatment with bifunctional or monospecific affibody molecules (100 µg/mL) and were incubated at 37 • C for 3 h. Cells were washed in icecold PBS and then lysed with RIPA buffer containing protease inhibitors. Then, anti-LMP1 (Abcam 136633) or anti-LMP2 (Abcam, Clone15F9) antibodies were combined with disuccinimidyl suberate bound to protein A/G plus agarose. Afterwards, the pellets were resuspended in SDS sample buffer and analyzed by Western blotting.

In Vivo Tumor Imaging
Near-infrared (NIR) imaging was used to investigate the distribution and tumorspecific targeting ability of bifunctional affibody molecules in nude mice. C666-1, CNE-2Z, and HNE-2 cells (1 × 10 7 cells/100 µL PBS, respectively) were subcutaneously injected into the right forearm of nude mice (n = 5 per group). The tumor-bearing mice were used for imaging when tumor volume reached 300-500 mm 3

Flow Cytometry Assay
Cell cycle alteration was detected using flow cytometry. C666-1, CNE-2Z, and HNE-2 cells were harvested after treatment for 24 h with Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, or Z LMP2A-N 110 (100 µg/mL) and then fixed in 70% ethanol overnight at 4 • C. Fixed cells were washed in PBS and stained with propidium iodide (PI) (50 µg/mL PI, 100 µg/mL Rnase) for 30 min. Distributions of the cell cycle were determined using a FACS Calibur flow cytometer (BD, Biosciences, San Jose, CA, USA), and the resulting data were analyzed using ModFit LT version 3.0 software.

Western Blotting Assay for Cell Signaling Pathway Proteins
C666-1 and CNE-2Z cells were seeded at (1 × 10 5 cells/well) in 6-well plates and incubated for 36 h with Z LMP 110-277, Z LMP 277-110, Z LMP1-C 277, Z LMP2A-N 110, or SPA-Z scaffold (Z WT ) (10 µM). After that, the cells were detached and lysed in lysis buffer (RIPA buffer, Beyotime, Beijing, China) containing protease and phosphatase inhibitors. Then, proteins were separated by 15% SDS-PAGE and transferred to a PVDF membrane. The membranes were briefly blocked with 5% skim milk and incubated with a primary antibody (Supplementary Table S1) at 4 • C overnight. After washing, the membranes were incubated with secondary antibodies and then imaged using the Western blotting imaging system. GAPDH served as a reference gene.

In Vivo Antitumor Efficacy
C666-1, CNE-2Z, and HNE-2 tumor-bearing mice were divided randomly into 5 groups (n = 5 per group) and treated by tail vein injection every three days for 30 days with the following 0.1 mL: (I) PBS; (II) SPA-Z scaffold (Z WT ) 100 nmol/kg; (III) Cisplatin 5 mg/kg (used as a positive control); (IV) Z LMP 110-277 100 nmol/kg; (V) Z LMP 277-110 100 nmol/kg; (VI) Z LMP1-C 277 100 nmol/kg; and (VII) Z LMP2A-N 110 100 nmol/kg. The therapeutic efficacy of the different regimens was monitored by daily measurements of tumor sizes and body weight. At day 30, all mice were sacrificed, and the tumors were surgically removed carefully. The tumors and major organs were collected and frozen at −80 • C for further use.

Statistical Analysis
The data were expressed as mean ± standard deviation (SD). Statistical analysis of the results was performed with a Student's t-test, and a probability (p) value < 0.05 was considered statistically significant. All the graphs were performed using GraphPad Prism software.

Institutional Review Board Statement:
This study was carried out in strict compliance with the guidelines for Experimental Animals established by the State Science and Technology Commission. All animal experiments were performed in accordance with protocols approved by the Ethics Committee of Wenzhou Medical University (The permit license number: wydw2021-0352).

Data Availability Statement:
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.