Phage display-derived alpaca nanobodies as potential therapeutics for Naja atra snake envenomation

ABSTRACT

Naja atra, the Chinese cobra, is a major cause of snake envenomation in Asia, causing hundreds of thousands of clinical incidents annually. The current treatment, horse serum-derived antivenom, has unpredictable side effects and presents manufacturing challenges. This study focused on developing new-generation snake venom antidotes by using microbial phage display technology to derive nanobodies from an alpaca immunized with attenuated N. atra venom. Following confirmation of the immune response in the alpaca, we amplified V_HH genes from isolated peripheral blood mononuclear cells and constructed a phage display V_HH library of 1.0 × 10⁷ transformants. After four rounds of biopanning, the enriched phages exhibited increased binding activity to N. atra venom. Four nanobody clones with high binding affinities were selected: aNAH1, aNAH6, aNAH7, and aNAH9. Specificity testing against venom from various snake species, including two Southeast Asian cobra species, revealed nanobodies specific to the genus Naja. An in vivo mouse venom neutralization assay demonstrated that all nanobodies prolonged mouse survival and aNAH6 protected 66.6% of the mice from the lethal dosage. These findings highlight the potential of phage display-derived nanobodies as valuable antidotes for N. atra venom, laying the groundwork for future applications in snakebite treatment.

IMPORTANCE

Chinese cobra venom bites present a formidable medical challenge, and current serum treatments face unresolved issues. Our research applied microbial phage display technology to obtain a new, effective, and cost-efficient treatment approach. Despite interest among scientists in utilizing this technology to screen alpaca antibodies against toxins, the available literature is limited. This study makes a significant contribution by introducing neutralizing antibodies that are specifically tailored to Chinese cobra venom. We provide a comprehensive and unbiased account of the antibody construction process, accompanied by thorough testing of various nanobodies and an assessment of cross-reactivity with diverse snake venoms. These nanobodies represent a promising avenue for targeted antivenom development that bridges microbiology and biotechnology to address critical health needs.

INTRODUCTION

Snake envenomation poses a significant global public health challenge and was designated a Category A Neglected Tropical Disease (NTD) by the World Health Organization (WHO) in 2017 (1). Although categorized as an NTD, snakebite continues to occur annually in regions beyond the tropics, including developed countries or urban centers (2). The worldwide incidence of new snakebite cases is approximately 2.7 million per year, resulting in 138,000 deaths. Notably, Asia bears the greatest burden, accounting for 70% of cases. Within Asia, Taiwan faces unique challenges, with approximately 1,000 reported cases of snake envenomation annually, of which Naja atra snakebite accounts for 2% to 14.7% (3, 4).

N. atra, commonly known as the Chinese cobra, belongs to the Elapidae family and is found predominantly in northern Laos, northern Vietnam, southeastern China, Hong Kong, and Taiwan. N. atra is a highly venomous member of the true cobras, and its venom can cause fatal harm to humans even in minute quantities. Clinical manifestations of N. atra envenomation include swelling, pain, muscle weakness, fever, gastrointestinal distress, wound infection, necrotizing soft tissue infection, and skin necrosis (5). The emergence of diverse syndromes can be attributed to the intricate composition of N. atra venom, a highly complex mixture of various toxic constituents (6, 7). These include three-finger neurotoxins (NTXs), three-finger cytotoxins (CTXs), phospholipase A2 (PLA2), cobra venom factor (CVF), nerve growth factor , cysteine-rich secretory proteins, serine proteases, metalloproteinases, and L-amino acid oxidases (8). The formidable toxic potency of these constituents imposes significant constraints on the development of therapeutic interventions targeting the diverse components of the venom (9).

Antivenom, which is typically derived from snake venom-hyperimmunized horse serum and consists mainly of immunoglobulin G (IgG), F(ab), and F(ab’)₂ antibody fragments, has been the primary treatment for snake envenomation for the past century (10). However, conventional antivenom is associated with unresolved problems, including serum sickness and the presence of notable amounts of non-toxin-neutralizing antibodies. Other challenges include limited peripheral penetration and the need to extend the elimination half-life of the antibody molecules. The manufacture of horse-derived antivenom requires advanced quality control, batch variability management, and complex production procedures that rely on two biological systems: the snake and the immunized animal (11). Therefore, the development of novel snake venom antidotes is now paramount.

In recent years, increasing scientific interest has focused on distinct immunoglobulin variants, particularly heavy chain antibodies (HCAbs), an IgG isotype found in camels, llamas, and alpacas. Unlike conventional antibodies (150 kDa), which consist of two pairs of heavy and light chains, HCAbs (90 kDa) lack light chains and the CH1 constant domain of the heavy chain. Since the publication reporting the first camelid-derived snake antivenom in 2006, several studies have explored the neutralizing potential of HCAbs against the venom of various snake species, including Echis ocellatus, Bothrops mattogrossensis, Bitis arietans, and Naja nigricollis, with some reports confirming their neutralizing ability (12–14).

While the neutralizing potential of camelid HCAbs makes them promising candidates for snake venom antidotes, they face some unresolved limitations similar to horse serum. Through artificial engineering, the variable domain of HCAbs can be selected, and these antibodies are often referred to as nanobodies (15 kDa), single-domain antibodies, or variable heavy domains derived from heavy chains (VHHs). Nanobodies typically consist of three complementarity-determining regions (CDRs), and the longer CDR1 and CDR3 regions play the main role in recognizing target molecules (15, 16). The advantages of nanobodies as snake venom antidotes include their compact size, which is smaller than that of conventional IgG, Fab, and scFv antibody fragments. Additionally, they possess tissue-penetrating abilities, low immunogenicity, high solubility, stability across diverse pH and temperature conditions, and the capacity to effectively recognize enzyme proteins (17–25). A small number of studies have investigated the development of nanobodies that target the venom of diverse snake species, such as Bothrops atrox, Bothrops jararacussu, Daboia russelii, and Naja kaouthia, with promising results in reducing venom toxicity (26–29). However, reports specifically addressing the development of nanobodies against N. atra venom targets are currently lacking.

This study focused on screening phage display-derived alpaca nanobodies for specific binding affinities for N. atra snake venom. The nanobodies were generated by immunizing an alpaca with crude snake venom as the antigen. We constructed the V_HH gene library, followed by screening and binding affinity assay. We also evaluated the cross-reactivity of these nanobodies against six venoms from prevalent snake species found in Taiwan and Southeast Asia. Moreover, we validated the snake venom-neutralizing efficacy of the nanobodies through mouse experiments. These findings underscore the potential of these nanobodies as prospective candidates for the future development of snake venom antidotes.

RESULTS

Humoral response to alpaca immunization

The immunization of alpaca involved a series of seven vaccinations with attenuated N. atra venom, administered at 2-week intervals (Fig. 1A). Before initiating the immunization, the crude N. atra venom was comprehensively analyzed using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, revealing a complex composition consisting mainly of proteins smaller than 20 kDa, with only a few falling within the range of 20 kDa to 150 kDa (Fig. 1B). The humoral response of alpaca sera to the multiple-venom proteins was evaluated via Western blotting under consistent conditions. Sera underwent a 1:5,000 dilution, encompassing samples collected from the preimmunization to the seventh immunization, followed by probing with anti-alpaca IgG. A discernible response was observed from the fourth immunization, with a robust signal ultimately detected upon testing serum from the seventh immunization. This finding indicated the elicitation of polyclonal-specific HCAbs, which recognize most of the venom proteins with molecular weights less than 20 kDa and even certain larger macromolecular proteins (Fig. 1B). An enzyme-linked immunosorbent assay (ELISA) was conducted to determine the binding titer. Serial dilutions ranging from 5,000× to 160,000× of the serum collected from the preimmunization to the seventh immunization were employed, followed by the incubation of anti-alpaca IgG. In addition, only background optical density was observed when testing the preimmunization serum, but the optical density reached approximately two after the third immunization and remained constant up to the seventh immunization (1:5,000 dilution) (Fig. 1C). These findings strongly indicate the development of a potent immune response to snake venom in the immunized alpaca, forming the basis for constructing a V_HH library from which to select high-affinity nanobodies against N. atra venom.

Fig 1.

Immunization and the humoral response in alpaca. (A) Schematic representation of the alpaca immunization process. (B) N. atra venom was separated via SDS-PAGE and visualized using Coomassie Blue staining (left panel). The humoral response to multiple venom proteins was monitored via Western blot analysis, utilizing a 1:5,000 dilution of sera collected from preimmunization to the 7th immunization (right panel). (C) Binding activity was further characterized by serial dilutions of serum (5,000× to 160,000×) on ELISA plates. The error bars indicate the standard deviation of the duplicate results. OD represents optical density.

Construction of the V_HH library

The construction of the V_HH library began with isolating peripheral blood mononuclear cells (PBMCs), followed by RNA purification and cDNA synthesis, as illustrated in the schematic diagram (Fig. 2A). These initial steps were the foundation for V_HH gene amplification through well-established microbiology and molecular biology techniques and protocols (30). Following the primary PCR, two predominant amplified products were identified, specifically representing the VHH-CH1-Hinge-CH2 region (approximately 1,000 bp) and the VHH-Hinge-CH2 region (approximately 800 bp), as indicated by arrows (Fig. 2B, left). The entire set of products of the primary PCR served as the template for the secondary PCR, resulting in targeted V_HH gene amplification (approximately 500 bp) (Fig. 2B, right). The secondary PCR product was subjected to restriction and ligation into the pComb3X phagemid vector, followed by transformation into Escherichia coli (E. coli) ER2738 through electroporation. The library was estimated to comprise approximately 1.0 × 10⁷ transformants, nearly 100% of which contained unique V_HH gene-coding sequences (data not shown). After VSCM13 helper phage infection of E. coli cells, the resulting V_HH library phages were employed in a biopanning process.

Fig 2.

Construction of a nanobody library from the peripheral blood mononuclear cells of an immunized alpaca. (A) Schematic diagram demonstrating the construction of the nanobody library. (B) (Left) Amplified products of the VHH-CH1-Hinge-CH2 region (approximately 1,000 bp) and the VHH-Hinge-CH2 region (approximately 800 bp) after the primary PCR are indicated with arrows. (Right) The amplified product of VHH (approximately 500 bp) after the secondary PCR is indicated by the arrow.

Selection of anti-N. atra venom nanobodies

The phage-displayed V_HH library was subjected to four rounds of biopanning to isolate specific anti-N. atra venom nanobodies. The titer of the eluted phages was approximately 5.2 × 10³ CFU after the first round of biopanning and progressively increased to 2.0 × 10⁶ CFU after the fourth round (Fig. 3A). The specificity of each biopanning round was further assessed through phage ELISA. Initially, no detectable binding activity to N. atra venom was observed using the original library phages; however, the binding response significantly exceeded an optical density of 1.8 after the fourth round of biopanning. Notably, the phages showed barely detectable cross-reactivity to BSA (Fig. 3B), which was consistent with the results obtained from alpaca serum. After confirming the specificity achieved through biopanning, the total phagemid DNA was purified after the fourth round of biopanning and transformed into protein-expressing E. coli TOP10F’. Thirteen clones were randomly selected for nanobody expression analysis. Eleven clones exhibited a noticeable increase in nanobody expression, with a dominant protein band at approximately 20 kDa observed via SDS-PAGE and Western blotting. However, clones aNAH6 and aNAH13 exhibited lower expression levels with smaller protein sizes, comparable to that of a previously isolated anti-SARS-CoV-2 N protein nanobody used as a positive control (Fig. 3C).

Fig 3.

Selection of nanobodies. (A) Monitoring of the eluted phage titer from each round of biopanning. (B) ELISA was used to assess the specific anti-N. atra venom binding activity of the amplified phages after each round of panning. Serum from the seventh immunization served as a positive control. Lib., representing the library phages before panning, was used as a negative control. The error bars indicate the standard deviation of the duplicate results. OD represents optical density. (C) SDS-PAGE of total lysates from 13 selected nanobody clones stained with Coomassie Blue dye (left panel). The total lysates were transferred for use in Western blot analysis and probed with anti-HA IgG to detect nanobody protein expression (right panel). Lane PC, a specific anti-SARS-CoV-2 N protein nanobody used as a positive control.

Sequence classification of anti-N. atra venom nanobodies

The randomly selected nanobodies were classified into four groups based on sequence identity: aNAH1 (69.2%; 9/13), aNAH6 (15.4%; 2/13), aNAH7 (7.7%; 1/13), and aNAH9 (7.7%; 1/13), as determined through Sanger sequencing (Table 1). The amino acid sequences were then aligned with those encoded by immunoglobulin germline genes, which revealed sequence similarity to the IGHV3S53*01/IGHJ6*01 and IGHV3-3*01/IGHJ4*01 germline genes (Table 2). Further calculations were performed to determine the mutation rates in the framework regions (FRs) and CDRs. The mutation rates in each FR were as follows: FR1 (4% to 16%), FR2 (17.6% to 29.4%), FR3 (13.2% to 26.3%), and FR4 (9%). For the CDRs, the mutation rates were as follows: CDR1 (62.5% to 87.5%), CDR2 (50% to 57.1%), and CDR3 (88.5% to 88.9%). Importantly, the total mutation rate in FRs (14.3% to 19.8%) was lower than the total mutation rate in CDRs (76.2% to 79.2%) (Table 2).

TABLE 1.

Classification of 13 selected anti-N. atra venom nanobody clones according to DNA sequence

Groups	Selected anti-N. atra venom nanobody clones	Percentage (%)
1	aNAH1, aNAH2, aNAH3, aNAH4, aNAH5, aNAH8, aNAH10, aNAH11, aNAH12	69.2 (9/13)
2	aNAH6, aNAH13	15.4 (2/13)
3	aNAH7	7.7 (1/13)
4	aNAH9	7.7 (1/13)

TABLE 2.

Amino acid mutation rates in the framework region and complementarity-determining region of the nanobody

Clones	V-germline/J-germline	FR1 (%)	FR2 (%)	FR3 (%)	FR4 (%)	Total FRs (%)	CDR1 (%)	CDR2 (%)	CDR3 (%)	Total CDRs (%)
aNAH1	IGHV3S53*01/IGHJ6*01	4	17.6	26.3	9	16.5	87.5	57.1	88.9	79.2
aNAH6	IGHV3-3*01/ IGHJ4*01	8	29.4	13.2	9	14.3	62.5	50	88.5	76.2
aNAH7	IGHV3S53*01/IGHJ6*01	16	17.6	26.3	9	19.8	87.5	57.1	88.9	79.2
aNAH9	IGHV3S53*01/IGHJ6*01	12	17.6	26.3	9	18.7	87.5	57.1	88.9	79.2

Expression and purification of anti-N. atra venom nanobodies

Next, the nanobodies were produced on a larger scale and purified. SDS-PAGE was employed to analyze the initial results. Prominent bands, each with a molecular weight exceeding 20 kDa, were observed for the aNAH1, aNAH7, and aNAH9 clones, each at a concentration of approximately 5 mg/L. In contrast, aNAH6 exhibited a prominent band within the 15 to 20 kDa range, at an approximate concentration of 2.5 mg/L (Fig. 4A). Additionally, the nanobodies were fused with a human influenza hemagglutinin (HA) tag. Therefore, to confirm that these bands corresponded to nanobodies, anti-HA tag antibodies were employed. Subsequent Western blot analysis revealed sizes consistent with those shown in Fig. 4A, confirming that the purified products were recombinant nanobodies (Fig. 4B).

Fig 4.

Purification of the anti-N. atra venom nanobodies. (A) SDS-PAGE analysis showing the purified anti-N. atra venom nanobodies (approximately 20 kDa), stained with Coomassie Blue. (B) The purified nanobodies were subsequently probed with anti-HA IgG via Western blotting, to facilitate the identification of the nanobodies.

Sensitivity of anti-N. atra venom nanobodies

As snake venom is a mixture of multiple components, subsequent investigations were designed to evaluate the specificity of nanobodies for N. atra venom. First, snake venom proteins were separated according to their molecular weight and subjected to Western blot analysis. All four nanobodies recognized the proteins within the molecular weight range of approximately 8 to 17 kDa (Fig. 5A). To obtain a comprehensive assessment of their binding abilities, serial dilutions of the nanobodies were prepared, and the binding efficiency was measured through ELISA. The results indicated that aNAH1, aNAH6, aNAH7, and aNAH9 exhibited diminishing binding affinities at concentrations of 156.25 ng/mL, 10,000 ng/mL, 312.50 ng/mL, and 156.25 ng/mL, respectively (Fig. 5B). Furthermore, a competitive ELISA was conducted to quantify the binding affinities of the anti-N. atra venom nanobodies. A competition between serial dilutions of free-form venom, ranging from 200 to 0.2 µg/mL, and the immobilized N. atra venom against individual nanobodies was conducted. At the 50% inhibition of protein binding, the concentrations of N. atra venom were as follows: 1.65 µg/mL aNAH1, 1.07 µg/mL aNAH6, 1.41 µg/mL aNAH7, and 1.55 µg/mL aNAH9 (Fig. 5C). Subsequently, the dissociation constant (K_d value) of each nanobody was calculated through the conversion equation described in the Materials and Methods section. The K_d values for the anti-N. atra venom nanobodies ranged from 1.81 × 10⁻⁷ M to 6.71 × 10⁻⁸ M, with aNAH6 demonstrating the highest affinity of (9.30 ± 2.59) ×10⁻⁸ (Table 3).

Fig 5.

Characterization of the N. atra venom-specific nanobodies. (A) The sizes of multiple proteins in N. atra venom were determined through Coomassie Blue staining (left panel). The binding response was tracked via Western blot probing with individual anti-N. atra venom nanobodies (right panel). (B) ELISA was used to assess the binding activity of serially diluted (20,000 ng/mL to 0.04 ng/mL) anti-N. atra venom nanobodies to N. atra venom. Duplicate results are represented in a scatter plot overlaid with a four-parameter logistic curve. OD represents optical density. (C) Competitive ELISA of anti-N. atra venom nanobodies to estimate the K_d value. The purified nanobodies were preincubated with serially diluted venom proteins (200 to 0.2 µg/mL) before being applied to ELISA plates coated with venom proteins (details in text). The bound nanobodies, both with and without free-form venom proteins, were represented as B/B0 to illustrate the inhibitory percentage of binding specificity. Error bars represent the standard deviation of the duplicate results.

TABLE 3.

Dissociation constants of four representative anti-N. atra venom nanobodies

Clones	Concentration for 50% competition (μg/mL)	Kd value (M)
aNAH1	1.65	(1.41 ± 0.40) × 10−7
aNAH6	1.07	(9.30 ± 2.59) × 10 −8
aNAH7	1.41	(1.31 ± 0.28) × 10−7
aNAH9	1.55	(1.45 ± 0.30) × 10−7

Specificity of anti-N. atra venom nanobodies

The specificity testing assessed the potential cross-reactivity of the nanobodies with venom from six widespread snake species commonly found in Southeast Asia and Taiwan, namely, Bungarus multicinctus, Deinagkistrodon acutus, Daboia siamensis, N. atra, Protobothrops mucrosquamatus, and Trimeresurus stejnegeri. In the initial Western blot analysis, aNAH1, aNAH6, aNAH7, and aNAH9, which have molecular weights of approximately 10 kDa, specifically recognized N. atra venom. Importantly, the serum from the seventh immunization also showed no cross-reactivity (Fig. 6A). Subsequently, ELISA revealed exclusive recognition of N. atra venom by aNAH1, aNAH6, aNAH7, and aNAH9, which displayed strong binding with optical densities of 2. The serum exhibited a similar binding response to N. atra venom, with an optical density of 2, and minimal cross-reactivity with B. multicinctus venom (Fig. 6B). Further exploration of cross-reactivity included Western blotting with Southeast Asian cobra venoms, specifically N. kaouthia and Naja siamensis. The results showed that aNAH1, aNAH6, aNAH7, and aNAH9 did exhibit cross-reactivity with N. kaouthia and N. siamensis, with molecular weights of approximately 10 kDa. Notably, the serum from the seventh immunization also demonstrated cross-reactivity with N. kaouthia and N. siamensis (Fig. 7A). The four clones and the serum also all exhibited cross-reactivity with N. kaouthia and N. siamensis according to ELISA, with an optical density of 2 (Fig. 7B).

Fig 6.

Specificity analysis of N. atra venom-specific nanobodies. (A) Six snake venoms (from B. multicinctus, D. acutus, D. siamensis, N. atra, P. mucrosquamatus, and T. stejnegeri) were separated via SDS-PAGE and stained with Coomassie Blue (upper right). Subsequently, the six snake venoms were probed with each anti-N. atra venom nanobody and serum from the seventh immunization via western blotting. (B) ELISA of the anti-N. atra venom nanobody and serum from the seventh immunization to analyze cross-reactivity to the six snake venoms and BSA. The error bars indicate the standard deviation of the duplicate results. OD represents optical density.

Fig 7.

Cross-reactivity of N. atra venom-specific nanobodies to Southeast Asian cobra venom. (A) Three Southeast Asian snake venoms (N. kaouthia, N. siamensis, and N. atra) were separated via SDS-PAGE and stained with Coomassie Blue (upper right). Subsequently, the Southeast Asian snake venoms were probed with each anti-N. atra venom nanobody and serum from the seventh immunization for Western blot analysis. (B) ELISA of the anti-N. atra venom nanobody and serum from the seventh immunization to analyze cross-reactivity to the three Southeast Asian snake venoms and BSA. The error bars indicate the standard deviation of the duplicate results. OD represents optical density.

Venom neutralization assay of nanobodies

An in vivo mouse model was used to assess the effectiveness of the nanobodies in neutralizing N. atra venom. The absolute lethal dose (LD100) was first established through subcutaneous injections of different doses. The 10-µg group exhibited an 89% survival rate, with only one mouse dying within 3 hours and no apparent abnormalities. Conversely, the mice in the 20-µg, 30-µg, and 40-µg groups exhibited a 0% survival rate, with all the mice dying within 3 hours (Fig. 8A). Consequently, the LD100 for N. atra venom was determined to be 20 µg for subsequent neutralization assessments. The neutralization assay was divided into three parts: polyclonal IgG antibodies, monoclonal nanobodies, and a combination of multiple nanobodies. In the polyclonal testing, the IgG from the seventh immunization group achieved a 78% survival rate, which was significantly higher than those in the preimmunization IgG and PBS control groups but slightly lower than the 100% survival rate observed in the antivenin positive control group (Fig. 8B). Among the monoclonal nanobodies, aNAH1 displayed an 11% survival rate, aNAH6 and aNAH7 exhibited a 22% survival rate, and aNAH6 demonstrated a substantial 67% survival rate, whereas the unrelated nanobody had no protective effect (Fig. 8B). In the combination test, where equal amounts of four monoclonal nanobodies were combined and administered, the 4-mg group achieved a 44% survival rate, while the 1-mg group had a 0% survival rate; however, the survival time was markedly prolonged (Fig. 8B).

Fig 8.

Protective effect of the anti-N. atra venom nanobody in a mouse model. (A) Survival rate of mice (n = 9) after subcutaneous injection of various doses of N. atra venom proteins to determine the absolute lethal dose (LD100). (B) The survival rate of mice in each group after subcutaneous injection (n = 9). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

DISCUSSION

Snakebite is currently considered the most neglected NTD in humans, and the primary treatment involves the immediate injection of equine serum. However, ongoing concerns remain regarding the clinical effectiveness and safety of antivenoms (1, 31). Conventional antivenoms present challenges such as unsatisfactory treatment efficacy, a low proportion of toxin-neutralizing immunoglobulins, limited tissue penetration ability, and manufacturing complexities related to batch variability and the biological systems involved, which cannot be overlooked (11, 31). While previous studies have utilized various animals, such as rabbits, ducks, and chickens, to produce antivenom serum against N. atra venom, this research represents the first instance of using alpacas to generate N. atra venom antidotes (32–34). Adhering to conventional immunization methods involving attenuated snake venom with 0.125% glutaraldehyde, our approach aimed to minimize harm to alpacas while potentially impacting the generation of antibodies against native snake venom proteins (35). Although other studies have used nonattenuated snake venom in alpaca immunization and reported minimal side effects accompanying the identification of monoclonal nanobodies with comprehensive effects, our experiments utilizing attenuated snake venom also identified nanobodies with both potency and efficacy (29). The initial immunization involved a significantly lower dose (200 µg + Freund’s Complete Adjuvant) than the WHO-recommended dose of 1–4 mg for horse immunization, followed by seven subsequent immunizations (200 µg + Freund’s Incomplete Adjuvant), which was still considerably lower than the WHO-recommended dose of 5–10 mg for horse immunization (35). Following immunization, alpaca serum exhibited broad reactivity toward multiple molecules, from large to small (Fig. 1B). The response to smaller proteins was particularly notable, and an estimation based on molecular weight suggested the potential recognition of CTXs (<10 kDa), NTXs (approximately 14 kDa), and PLA2 (approximately 14 kDa) (6, 7). The alpaca displayed a robust immune response against N. atra snake venom with a titer of 160,000× (Fig. 1C), which was superior to the titer of 64,000× previously obtained using chickens (34). Our alpaca immunization protocol required three immunizations to elicit a significant response against N. atra snake venom, in contrast to the initial immunization response observed in llamas against N. kaouthia snake venom (29). Despite having a better immune response, alpacas are cheaper and easier to maintain than horses and sheep in many areas. These observations validate the utility of alpacas as hosts for snake venom immunization.

After the successful immunization of alpacas against N. atra snake venom, monoclonal nanobodies must be selected to optimize production processes and address challenges associated with serum purification batches. We were intrigued by the potential of nanobodies, as developing them as treatments for snake venom offers several advantages. Their smaller molecular weight (15 kDa) allows for more effective tissue penetration compared with conventional IgG, Fab, and scFv fragments, potentially shielding victims from the tissue-damaging effects of venom toxins (17). Furthermore, their smaller size contributes to lower immunogenicity when employed as therapeutic agents (21). Additionally, nanobodies frequently feature longer CDR3 loops, enabling them to target cryptic enzyme epitopes challenging for conventional antibodies to recognize (24, 25). With remarkable stability and solubility, they can be easily expressed and purified from E. coli expression systems, potentially enhancing cost effectiveness and efficiency (20, 22, 23). In our previous attempts, we utilized chicken scFv antibodies for screening N. atra venom antibodies; however, we only shortly prolonged the death of mice (34). Considering the high ratio of enzyme proteins and three-finger toxins in snake venom, we believe that nanobodies with longer CDR3 loops have a greater potential for selecting neutralizing antibodies. Therefore, we have shifted our focus toward this system to select effective venom-neutralizing nanobodies.

Conventional hybridoma screening methods present difficulties in generating stable hybridomas for camels, llamas and alpacas, leading to increased interest in phage display technology for antibody screening (36). This interest prompted us to employ the method proposed by Conrath et al. for constructing an alpaca V_HH gene library; we used the well-established phage display system for antibody screening (30, 37, 38). A two-step PCR strategy was employed to amplify the V_HH gene, with the primary PCR amplifying VHH-CH1-Hinge-CH2 and VHH-Hinge-CH2, followed by the subsequent secondary PCR for the specific amplification of the V_HH gene (Fig. 2A and B). In the nested PCR approach, it is advantageous to design the enzyme cutting site as SfiI in the second round to ensure a more specific final product (39). Subsequently, employing the pComb3X vector and VSCM13 phage system, our library size reached 1.0 × 10⁷ transformants, a range deemed sufficient for a V_HH library devoid of the heavy-light chain pairing complexities associated with scFv antibodies (39). The diversity of a library is crucial for generating high-quality antibodies, often measured by the size (40). Library size directly correlates with the likelihood of encountering unique and high-affinity antibodies against specific targets (40, 41). Previous investigations into V_HH phage libraries for animal toxin nanobody selection have consistently observed library sizes ranging from 10⁶ to 10⁸ (29, 42–45). The diversity is especially important when dealing with complex antigens such as animal toxins, where multiple epitopes exist during neutralizing (46, 47). Our post-immunized library, falling within the established range of 10⁶ to 10⁸, demonstrates that it possesses the necessary diversity and complexity required for successful antibody screening. Four rounds of biopanning were conducted using snake venom, a mixture of multiple components. Despite the complexity of the venom mixture, the increasing trend of the eluted phage titer and the phage ELISA results (Fig. 3A and B) indicated the amplification of nanobodies specific for N. atra snake venom after biopanning, which was consistent with our previous experience in antibody screening against various snake venoms (34, 48–52). These results crucially support the feasibility of establishing a V_HH library using alpaca PBMCs after immunization with snake venom.

Further analysis of 13 randomly selected clones from the fourth biopanning revealed four distinct groups, with a dominant clone in group 1 (aNAH1) constituting 69.2% and minor clones in groups 2 (aNAH6), 3 (aNAH7), and 4 (aNAH9), constituting 15.4%, 7.7%, and 7.7%, respectively (Table 1). BLAST comparisons of these clones indicated a close similarity to the germline sets IGHV3S53*01/IGHJ6*01 and IGHV3-3*01/IGHJ4*01, which was consistent with the findings of Bahrami-Dizicheh et al. that nanobodies derived from alpacas predominantly originate from the IGHV3S53 and IGHV3-3 germlines (15). Despite the absence of the three VL CDRs, the nanobodies can form a cavity with their three remaining CDRs, effectively balancing the importance of the three missing CDRs (53). Among our four nanobodies, CDR1 (62.5% to 87.5%), CDR2 (50% to 57.1%), and CDR3 (88.5% to 88.9%) exhibited high variability, as did the CDRs as a group (76.2% to 79.2%) (Table 2). This significant variability implies that these nanobodies were generated through VDJ recombination and exhibit specificity against N. atra snake venom. The binding mechanism of traditional antibodies is highly associated with the high variability of CDR3, whereas CDR1 and CDR2 have canonical structure constraints (54). In nanobodies with only heavy chains, studies have revealed a significant correlation between CDR3 structure and antigen binding, influenced by both the CDR3 sequence and the germline type (15, 55). Our findings consistently indicate a high degree of variation in the CDR3 region, indicating a strong association between antigen binding. Comparing animal toxin nanobody sequences revealed notable CDR3 region similarity for the same toxin (27, 29, 42, 43, 56, 57). These findings emphasize the importance of considering nanobody CDR3 in future snake venom antigen binding studies.

We conducted large-scale expression and purification of the four selected nanobodies for comprehensive characterization. Through a series of qualitative assays (Fig. 5A through C), we demonstrated that all four nanobodies specifically recognized N. atra snake venom. Moreover, we quantified the binding affinities of the nanobodies using competitive ELISA, obtaining K_d values ranging from 1.81 × 10⁻⁷ M to 6.71 × 10⁻⁸ M (Table 3), derived through a conversion formula based on optical density readings (58). The selected nanobodies demonstrate a binding affinity of approximately 10 nM, compared with other known selected animal toxin nanobodies, including scorpion toxins from Androctonus australis hector, Hemiscorpius lepturus, and Hottentotta saulcyi and snake venoms from Bothrops jararacussu and Naja kaouthia, ranging from 1 µM to 0.5 nM (27, 29, 42–45, 56). This substantiates the efficacy and binding affinity of our antibody within the specified range, validating its potential for therapeutic applications. However, it is crucial to acknowledge that the specific targets recognized by our four nanobodies remain unclear. Western blotting allowed us to determine the molecular weights, which suggested the hypothesis that these nanobodies may target three-finger toxin proteins, such as NTXs or CTXs, in accordance with the molecular weights observed in other studies (6).

The specificity was thoroughly assessed using venom from several species of snakes with high clinical relevance, including four that are commonly found throughout Asia and two that are common in Taiwan (59). These six species were selected based not only on their frequency of occurrence and their geographical distribution but also on their uniqueness to be divided into six groups according to proteomic analysis of their snake venom composition (60). In cross-reactivity experiments involving the venom of these six major snake species (Fig. 6A and B), we were surprised to observe that the monoclonal nanobodies recognized only the venom of N. atra. A comparative analysis with prior investigations revealed that N. atra, within the spectrum of six major snake species, contains species-specific proteins, including CVF, NTXs, and CTXs (60). Evaluating the molecular weights of these proteins in relation to the proposed target antigen hypothesis mentioned above suggested that the recognition specificity of the four nanobodies is likely directed toward NTXs or CTXs. Importantly, however, the alpaca serum did show minor cross-reactivity with B. multicinctus. Notably, both N. atra and B. multicinctus belong to the Elapidae family and are known for venom predominantly comprising PLA2 and 3-finger toxins (60). This cross-reactivity did not occur from other snakes with venom containing PLA2, which suggested potential immunological homogeneity among three-finger toxins.

The neutralization assay showed that the polyclonal IgG from the alpaca did not provide complete protection, potentially owing to the use of a lower immunization dose than that used in traditional horse immunization protocols. Nonetheless, even with minimal immunization, we were still able to select monoclonal nanobodies with neutralizing abilities, and similar findings in previous studies support the idea that strategic immunization approaches can yield nanobodies that are effective against rare or low-venom snakes (61). Within the monoclonal nanobody part of the test, survival rates ranged from 11% to 67%. Although we referred to an experiment on snake venom neutralization in mice performed by WHO, evaluating the corresponding antigen and its role in snake venom responses is crucial for monoclonal nanobody neutralization experiments. Mouse experiments involving N. atra venom present challenges, because this venom usually causes tissue necrosis instead of typical neurotoxicity (62). Our monoclonal nanobodies aNAH1, aNAH7, and aNAH9 might bind to NTXs, but they do not effectively block the binding epitopes between NTXs and the acetylcholine receptor. However, we cannot exclude the possibility that the antigen recognized by aNAH1, aNAH7, or aNAH9 is instead CTXs, which would explain poor snake venom neutralization in this mouse model. On the other hand, surprisingly, aNAH6 showed effective snake venom neutralization. Currently, our team is identifying the antigen targeted by aNAH6. We are also conducting experiments to elucidate the role of this antigen in the snake venom reaction and to determine whether this antigen can lead to mouse death.

In summary, we successfully identified and isolated high-affinity alpaca nanobodies targeting N. atra venom from the immune library of 1.0 × 10⁷ transformants, which is large enough to ensure a diverse pool for effective nanobody selection. Following biopanning, we selected four individual nanobody clones, namely, aNAH1, aNAH6, aNAH7, and aNAH9, which exhibited remarkable binding affinities, including one K_d value as low as 10 nM. The cross-reactivity assessments showed that these nanobody clones exhibited exclusive recognition of Naja spp. venom, indicating the recognition of a unique antigen within this snake genus. While the aNAH6 clone demonstrated a noteworthy 66.6% survival rate in the neutralization assay, additional investigations are needed: precise identification of the targeted antigen, determination of the mechanisms involved in venom neutralization, in-depth examination of the tissue penetration ability, and evaluation of the therapeutic efficacy. These comprehensive analyses will provide valuable insights into the potential therapeutic applications of these nanobody clones against N. atra envenomation.

MATERIALS AND METHODS

Snake venom collection

All the crude snake venom samples (from B. multicinctus, D. acutus, D. siamensis, N. atra, P. mucrosquamatus, T. stejnegeri, N. kaouthia, and N. siamensis) were collected following WHO guidelines and ethical standards. Approval for the venom collection was obtained from the Forestry Bureau of Taiwan (approval code: 1120028288), the Taipei City Government (approval code: 1116014994), and the New Taipei City Government (approval code: 1113374426). These approvals collectively ensured our compliance with legal and ethical standards for data usage. Live snakes were captured from various regions of Taiwan, and specimens were collected from snakes through venom ejection, adhering to established ethical practices. The venom was then lyophilized and stored at −80°C until use. The stock venom was diluted to a concentration of 10 mg/mL in PBS and sterilized using a 0.22-µm membrane filter before use. Subsequent dilutions, under experimental requirements, were carried out within the same buffer system.

Alpaca immunization

All of the alpaca experiments followed established methodologies and were conducted following approved ethical standards (63). Approval for the experiments was obtained in advance from the Taipei Medical University Institutional Animal Care and Use Committee (ethical approval code: LAC-2021-0218). A 1-year-old male alpaca (Lama pacos) was used for the immunization. Before immunization, N. atra venom (200 µg) was attenuated with 0.125% glutaraldehyde (Sigma-Aldrich, St. Louis, MO, USA) at room temperature (RT) for 1 hour. The attenuated venom was then emulsified with 50% complete Freund’s adjuvant (Sigma-Aldrich, St. Louis, MO, USA) and subcutaneously injected into five different areas on the flank of the alpaca. The subsequent immunizations (second to seventh) involved subcutaneous injections of a mixture of 200 µg of attenuated N. atra venom and 50% incomplete Freund’s adjuvant (Sigma-Aldrich, St. Louis, MO, USA) at 2-week intervals. Blood samples were collected from the jugular vein 2 weeks after each immunization to measure the immune titer.

V_HH library construction

The phage display V_HH library was constructed according to established methodologies (30, 37, 38). Two weeks after the final immunization, 100 mL of blood was collected, and PBMCs were isolated through Ficoll-Paque Plus density gradient centrifugation (Cytiva Sweden AB, Uppsala, Sweden). Total RNA was extracted from 10⁸ PBMCs suspended in TRIzol reagent (Life Technologies, Carlsbad, CA, USA), and cDNA was synthesized from 20 µg of total RNA using HiScript I reverse transcriptase (Bionovas, Toronto, Ontario, Canada) following the manufacturer’s protocol.

The following specific primers were listed in Table 4, referred from the previous studies (37, 38). The primary PCR amplification for amplifying VHH-CH1-Hinge-CH2 and VHH-Hinge-CH2 was using 1 µL of cDNA products mixed with 60 pmol of CALL01/CALL02 primer each, containing 50 µL of Taq DNA Polymerase 2× Master Mix RED (Ampliqon A/S, Odense M, Denmark), with the remaining volume made up using water, resulting in a final volume of 100 µL. The primary PCR was carried out under the following conditions: 94°C for 5 min, 30 cycles of 94°C for 15 s, 56°C for 15 s, 72°C for 2 min, and a final extension at 72°C for 10 min. The PCR products were subsequently subjected to gel electrophoresis using a 1% agarose gel in TAE buffer. The size of the products was estimated through the 100 bp Plus DNA Ladder (Bionovas, Toronto, Ontario, Canada). The primary PCR products were then purified from a 1% agarose gel using the Gel Advanced Gel Extraction Miniprep System (Viogene, Sunnyvale, CA, USA). The secondary PCR was performed to selectively amplify V_HH genes. The ALPSC-F1/ALPSC-R2 primers included the SfiI cutting sites for subsequent cloning into the phagemid vector. Using 100 ng of the primary PCR product mixed with 60 pmol of ALPSC-F1/ALPSC-R2 primer each, containing 50 µL of Taq DNA Polymerase 2 × Master Mix RED, with the remaining volume made up using water, resulting in a final volume of 100 µL. The second PCR was carried out under the following conditions: 94°C for 5 min, 25 cycles of 94°C for 15 s, 56°C for 15 s, 72°C for 2 min, and a final incubation step at 72°C for 10 min. The 500bp fragments were then purified as discussed above in preparation for restriction enzyme digestion.

TABLE 4.

The forward and reverse primers used in primary and secondary PCR of V_HH library construction

	Primer name		Sequence (5′ → 3′)	Reference
First PCR	CALL001	Forward	GTC CTG GCT GCT CTT CTA CAA GG	(37)
	CALL002	Reverse	GGT ACG TGC TGT TGA ACT GTT CC	(37)
Second PCR	ALPSC-F1	Forward	GAG GAG GAG GAG GAG GTG GCC CAG GCG GCC SAG KTG CAG CTC GTG GAG TCN GG	(38)
	ALPSC-R2	Reverse	GAG GAG GAG GAG GAG GAG CTG GCC GGC CTG GCC ACT AGT GGG GTC TTC GCT GTG GTG CG	(38)

^a The underlined parts indicate the SfiI restriction enzyme cutting site.

The V_HH genes were subsequently digested with the SfiI restriction enzyme (New England Biolabs, Ipswich, MA, USA) and inserted into the pComb3X vector. pComb3X is a phagemid that has both bacteriophage and plasmid properties and it is widely used in phage display technology (64). For V_HH genes, 5 µg of purified PCR product was mixed with 60 units SfiI/μg DNA (20 U/µL), 20 µL of 10× rCutSmart Buffer (New England Biolabs, Ipswich, MA, USA), and water to a final volume of 200 µL. The vector preparation was carried out using 20 µg of pComb3X, 6 units SfiI/μg DNA (20 U/µL), 20 µL 10× rCutSmart Buffer, and water to a final volume of 200 µL. All digests were incubated at 50°C for 5 hours and purified as discussed above. The V_HH genes were ligated into the pComb3X vector with 70 ng of purified V_HH genes and 140 ng of purified vector pComb3X, containing 10 µL (1 U/µL) T4 DNA ligase (Life Technologies, Carlsbad, CA, USA) and 40 µL of 5× T4 DNA Ligase Buffer (Life Technologies, Carlsbad, CA, USA) in a final volume of 200 µL. The ligation product was then transformed into E. coli ER2738 via MicroPulser electroporation (Bio-Rad, Hercules, CA, USA), after which a small portion was plated on LB agar to determine the library size. The remaining culture was infected with VCSM13 helper phage and amplified overnight in Super Broth (3% tryptone, 2% yeast extract, and 1% MOPS), followed by the precipitation of recombinant phages using 4% PEG-8000 (Avantor, Radnor, PA, USA) with 3% NaCl. The precipitated phages were suspended in PBS at a yield of approximately 10¹² PFU/mL for the biopanning procedure.

Biopanning

The selection of N. atra venom-specific nanobodies involved four rounds of biopanning on ELISA plates. First, 2.5 µg of N. atra venom was immobilized on the bottom of the well and incubated at 4°C overnight. The wells were then blocked with 3% BSA in PBS at 37°C for 1 hour. Approximately 10¹¹ PFU of the recombinant phages were added and allowed to incubate at 37°C for 2 hours. Subsequently, the wells were washed 10 times with PBS containing 0.05% Tween 20. Each wash involved vigorous pipetting 30 times to remove the unbound phages. The bound phages were eluted using 0.1 M glycine-HCl (pH 2.2) and neutralized with 2 M Tris base buffer. The eluted phages were immediately used to infect E. coli ER2738 for 10 min, and a small portion was plated on LB agar to determine the eluted phage titers (CFU). The remaining infected cells were cultured overnight at 37°C and the amplified phages were collected for the next round of biopanning.

Expression and purification of nanobodies

The total phagemid DNA purified from the amplified culture after the fourth round of biopanning was transformed into E. coli TOP10F’ and plated on LB agar. Thirteen individual clones were randomly selected for subsequent expression analysis. Each clone was then cultured overnight and diluted at a ratio of 1:100 in Super Broth supplemented with 20 mM MgCl₂ and 50 µg/mL ampicillin (Avantor, Radnor, PA, USA). After 8 hours of culture at 37°C, protein expression was induced by adding 1 mM isopropyl-β-D-1-thiogalactopyranoside (Bionovas, Toronto, Ontario, Canada) for 16 hours. The bacterial cells were centrifuged, and the pellet was resuspended in His-binding buffer solution (6 M urea, 20 mM imidazole, 500 mM NaCl, and 20 mM sodium phosphate, pH 7.4) before sonication to achieve lysis. The nanobodies were then purified using Ni²⁺ Sepharose resin (Cytiva Sweden AB, Uppsala, Sweden) following the manufacturer’s protocol.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting

In the SDS-PAGE and Western blotting experiments, snake venoms (B. multicinctus, D. acutus, D. siamensis, N. atra, P. mucrosquamatus, T. stejnegeri, N. kaouthia, and N. siamensis), total lysate containing nanobodies, and purified nanobodies were separated via SDS-PAGE. The separated proteins were stained with Coomassie Blue and transferred to polyvinylidene fluoride membranes. The membranes were then blocked with 5% skim milk in PBS. For the humoral response assay, alpaca serum collected from preimmunization to the seventh immunization was diluted at 1:5,000 for immunoprobing. In the expression and purification assays, anti-HA IgG (Proteintech, Rosemont, IL, USA) was diluted at 1:5,000 for immunoprobing. For the binding activity and cross-reactivity assays of the nanobodies, the purified nanobody from each clone (10 µg/mL) was used. After an hour of probing at 37°C, the membranes were washed with PBST and incubated with the following secondary antibodies: HRP-conjugated goat anti-alpaca IgG (Jackson ImmunoResearch, West Grove, PA, USA) diluted 1:10,000 for the humoral response assay, HRP-conjugated rabbit anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA) diluted 1:10,000 for expression and purification assay, and mouse anti-HA IgG diluted 1:5,000 followed by HRP-conjugated rabbit anti-mouse IgG diluted 1:10,000 for the nanobody assay. Membrane visualization was achieved with 3,3ʹ-diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA).

Enzyme-linked immunosorbent assay

In the ELISA experiments, the 96-well half-area ELISA plate (Corning Incorporated, Corning, NY, USA) was coated with snake venoms (B. multicinctus, D. acutus, D. siamensis, N. atra, P. mucrosquamatus, T. stejnegeri, N. kaouthia, and N. siamensis) and BSA, each at a concentration of 10 µg/mL in 25 µL, and incubated at 4°C overnight. After incubation, the wells were blocked with 150 µL of 5% skim milk in PBS at 37°C for 1 hour. For the humoral response assay, alpaca serum collected from preimmunization to the seventh immunization was serially diluted (ranging from 5,000× to 160,000×) for immunoprobing in 25 µL. In phage ELISA, the amplified phages (10¹¹ to 10¹² PFU) obtained after each round of biopanning were applied 25 µL. For the binding activity and cross-reactivity assays of the nanobodies, purified nanobody from each clone (10 µg/mL) was used in 25 µL. After 1 hour of probing at 37°C, the wells were washed six times with 150 µL PBST and incubated with 25 µL of the following secondary antibodies: HRP-conjugated goat anti-alpaca IgG (Jackson ImmunoResearch, West Grove, PA, USA) diluted 1:10,000 for the humoral response assay, HRP-conjugated mouse anti-M13 IgG (Sino Biological, Beijing, China) diluted 1:5,000 for phage ELISA, and mouse anti-HA IgG diluted 1:5,000 followed by HRP-conjugated rabbit anti-mouse IgG diluted 1:10,000 for the nanobody assay. After the final six washes, binding activities were assessed using 25 µL of 3,3′,5,5′-tetramethylbenzidine substrate (BD OptEIA; BD Pharmingen, San Diego, CA, USA) per well and stopped by 12.5 µL of 1N HCL. The optical density readings at 450 nm were obtained to evaluate the binding interactions.

Competitive enzyme-linked immunosorbent assay and dissociation constants

In the competitive ELISA, wells coated with N. atra venom were blocked, and a twofold dilution series of N. atra venom (ranging from 200 to 0.2 µg/mL) was prepared. The different dilutions were mixed with an equal volume of each individual nanobody (10 µg/mL) at 37°C for 1 hour. The unrelated nanobody (anti-SARS-CoV-2 N protein nanobody) was used as a negative control. These mixtures were then added to wells coated with immobilized N. atra venom and incubated at 37°C for 1 hour. The subsequent steps were carried out as previously described. The K_d value was calculated using the following equation: B0/(B0 – B) = 1 + K_d/A0, as reported by Friguet et al. (58). In this equation, B0 represents the optical density reading in the absence of the free form of N. atra venom proteins, B represents the optical density reading in the presence of the free form of N. atra venom proteins, and A0 represents the concentration of the N. atra venom proteins.

Neutralization assay in mice

The neutralization activity of the anti-N. atra venom nanobody in mice was assessed following the WHO protocol in accordance with approved ethical standards (65, 66). Approval for the experiments was obtained from the Taipei Medical University Institutional Animal Care and Use Committee before initiation (ethical approval codes: LAC-2019-0473 and LAC-2020-0339). First, a pilot assay was conducted to determine the LD100 of N. atra venom, defined as the lowest dosage that caused 100% fatality in mice. Four groups of IRC mice (n = 9/group; male; 12–14 g) were subcutaneously inoculated with 10 µg, 20 µg, 30 µg, or 40 µg of N. atra venom diluted in 200 µL of PBS and incubated at 37°C for 1 hour. The survival rate of the mice was monitored for 48 hours after envenomation. The LD100 was determined to be 20 µg, and this concentration was used for subsequent neutralization assessments. For the neutralization assay, 11 samples were prepared by incubating the LD100 venom with various treatments, including PBS, an unrelated nanobody (anti-SARS-CoV-2 N protein nanobody) (4 mg), horse-derived antivenom (4 mg) (branch no: 60-06-0010; manufacturer: Taiwan CDC), alpaca IgG from the preimmunization (4 mg), alpaca IgG from the seventh immunization (4 mg), purified aNAH1 (4 mg), purified aNAH6 (4 mg), purified aNAH7 (4 mg), purified aNAH9 (4 mg), and two different concentrations (1 mg or 4 mg) of a mixture of the four purified anti-N. atra venom nanobodies. These mixtures were incubated at 37°C for 1 hour and subcutaneously injected into ICR mice (n = 9/group; male; 12–14 g). The survival rate of the treated mice was monitored for 48 hours after inoculation.

Statistical analysis

All ELISA data were obtained from independently duplicated experiments and are presented as the mean ± standard deviation. The graphical results were generated with GraphPad Prism 6 software (GraphPad Software, Boston, USA). Four-parameter logistic analysis was performed and used in the presence of the binding activity of nanobody titration. Kaplan-Meier survival analysis was carried out, and the log-rank test was used for statistical analysis of the nanobody neutralization assay. The statistical analyses were performed using GraphPad Prism 6 software (GraphPad Software, Boston, USA) with a significance threshold of P < 0.05.