Nanobody affinity improvement: Directed evolution of the anti-ochratoxin A single domain antibody

Highlights

• Homology modeling and molecular docking for affinity maturation of the anti-ochratoxin A single domain antibody in vitro.

• Phage-displayed two-site saturated mutation single domain antibody libraries were constructed.

• Mutant single domain antibodies with improved affinity to ochratoxin A were obtained.

Abstract

The characteristics of single domain and ease of gene manipulation of the single domain antibody (sdAb) make it suitable for affinity maturation in vitro. Since the affinity of antibodies can influence the immunoassays' sensitivity, a nanobody (Nb), the anti-ochratoxin A sdAb (AOA-sdAb), was herein selected as the model antibody to explore feasible approach for improving its affinity. Homology modeling and molecular docking were used to analyze the interaction between OTA and the AOA-sdAb. After alanine scanning verification, Gly53, Met79, Ser102, and Leu149 were determined as the key amino acids of the AOA-sdAb. Two site-directed saturated mutation libraries were constructed by two-site mutation against those four key amino acids. After biopanning and identification, a mutant Nb-G53Q&S102D was obtained with a half maximal inhibition concentration (IC₅₀) of 0.29 ng/mL and a K_D value of 52 nM, which is 1.4-fold and 1.36-fold lower than that of the original sdAb, respectively. The computer simulation analysis indicated that the hydrogen bond, hydrophobic interaction, and side chain steric hindrance of amino acid residues are critical for the binding affinity of the AOA-sdAb. Overall, the techniques shown in this study are effective ways at ‘identifying residues involved in antigen binding’ that can be altered by site-directed mutation.

Introduction

Due to the advantages of speediness, low-cost, operability, sensitivity, and selectivity, immunoassays have been widely used as a tool for detection of contaminants in food and environment [1,2]. There is an urgent need for sensitive immunoassays to minimize consumer exposure to those hazardous materials, and thus improving the detection sensitivity has been a research hotspot of immunoassays. As far as the principle of immunoassays is concerned, there are mainly two research directions to improve detection sensitivity. Developing various signal amplification systems for the interaction of antigen and antibody is a common and effective strategy for sensitivity improvement, such as fluorescence substrate signal amplification and chemical modification signal amplification [3,4]. Another one is antibody affinity maturation for improving the affinity of the antibody for antigen. The affinity improvement of traditional antibodies generated by immunizing animals is restricted by immune tolerance. Due to their complex structures, these antibodies are also not applicable to affinity maturation in vitro. In contrast, the advantage of the ease of gene manipulation makes the genetically engineered antibodies very suitable for antibody maturation in vitro for higher affinity [5,6].

With the advances in genetic engineering and protein engineering, various kinds of genetically engineered antibodies have been developed, such as single-chain variable fragment, antigen-binding fragment, and single-domain antibody (sdAb) [[7], [8], [9]]. The sdAb, also known as nanobody (Nb), is highly attractive for its nanoscale size, good solubility in water, and high tolerance to the harsh environment [9]. The sdAb is generally derived from the variable domain of the heavy-chain antibody in Camelidae and of the immunoglobulin new antigen receptor in Chondrichthyes, which is named VHH and VNAR, respectively [10,11]. Specific sdAbs could be selected from three kinds of phage-displayed sdAb libraries, including immune library, naïve library, and semisynthetic/synthetic library [5]. The immune library is generally used for screening sdAbs with high affinity, and numerous sdAbs have been developed for food and environment analysis [[12], [13], [14]]. Moreover, the characteristic of single domain of sdAbs is propitious to gene modification and different engineered sdAbs have been reported, such as sdAbs with improved affinity by antibody maturation in vitro, sdAbs fused with enzyme, and sdAbs tagged with peptide [[15], [16], [17], [18]]. For instance of sdAbs affinity maturation in vitro, Jiao and co-workers constructed a saturation mutagenesis sdAb library by using the computer simulation and screened a mutant sdAb 2-C9 with improved affinity and enhanced broad-spectrum activity [15]. Qiu et al. obtained a mutant sdAb N-28-T102Y by directed modification that showed a 3.2-fold higher affinity than its parent sdAb N-28 and retained 55% activity after 5 min of heat treatment at 90 °C [16].

Many strategies have been tried to induce antibody affinity maturation in vitro, including error-prone PCR, DNA shuffling, chain shuffling, UV or chemical mutagenesis [[19], [20], [21], [22]]. However, the randomness of mutation makes the selection time-consuming and laborious which limits those methods' wide application in antibody affinity maturation in vitro. It has been reported that the combination of computational simulation and high-throughput screening techniques is efficient to obtain ligand-binding proteins with higher affinity [23,24]. The accuracy of the three-dimensional structure of proteins is critical for computational simulation of the interaction of protein and ligand. There are several techniques for obtaining the three-dimensional structure of proteins, such as the X-ray crystallography, nuclear magnetic resonance spectroscopy, and cryo-electron microscopy [25]. Nevertheless, these methods are not broadly used owing to the high cost, stringent experimental conditions, and time-consuming and labor-intensive procedures [26]. By contrast, protein homology modeling is a simple and effective method and could provide structural models for a given protein with an unknown structure [27]. Typically, the basic local alignment search tool (BLAST) was used to search for homologous sequences, followed by modeling through servers, such as SWISS-MODEL, 3D-Jury, I-TASSER and MODELLER [[27], [28], [29], [30], [31]]. The SWISS-MODEL, an automated server, has structural models for 2.2 million unique sequences for selection of the most suitable templates and provides a quality estimation of the resulting model [27,32]. Moreover, the SWISS-MODEL provides a comprehensive quality assessment that combines the Qualitative Model Energy Analysis (QMEAN) estimate with the global quality estimation (GMQE) score, which helps to obtain a more accurate model [27,33]. Furthermore, a combinatorial strategy involving homology modeling and molecular docking could be used to predict the protein function and analyze the intermolecular interaction of antibody and antigen, which will contribute to improving the affinity of recombinant antibodies by antibody maturation in vitro [34].

In our previous study, sdAbs against the mycotoxin ochratoxin A (OTA) have been selected from an alpaca immune library, which shows good specificity and affinity with OTA [12,35]. Moreover, various sensitive immunoassays based on the AOA-sdAb named as Nb28 have been established for OTA with different signal amplification modes [4,18,[35], [36], [37], [38], [39], [40], [41], [42]]. In the present work, we try to improve the affinity of Nb28 for OTA by antibody maturation in vitro for higher sensitivity in immunoassay. Homology modeling and molecular docking were performed to find the key binding sites of Nb28 which were confirmed by alanine scanning. Based on the key amino acids, two different strategies of site-directed saturation mutation were performed and discussed. Then the mutants with improved affinity selected from the phage-displayed mutation library were characterized and compared with the parent sdAb Nb28. To our knowledge, few studies have been reported on in vitro affinity maturation of sdAbs for small molecules from immune libraries.