Aromatic Residues on the Side Surface of Cry4Ba-Domain II of Bacillus thuringiensis subsp. israelensis Function in Binding to Their Counterpart Residues on the Aedes aegypti Alkaline Phosphatase Receptor

Receptor binding is a prerequisite process to exert the mosquitocidal activity of the Cry4Ba toxin of Bacillus thuringiensis subsp. israelensis. The beta-sheet prism (domain II) and beta-sheet sandwich (domain III) of the Cry4Ba toxin have been implicated in receptor binding, albeit the precise binding mechanisms of these remain unclear. In this work, alanine scanning was used to determine the contribution to receptor binding of some aromatic and hydrophobic residues on the surface of domains II and III that are predicted to be responsible for binding to the Aedes aegypti membrane-bound alkaline phosphatase (Aa-mALP) receptor. Larvicidal activity assays against A. aegypti larvae revealed that aromatic residues (Trp327 on the β2 strand, Tyr347 on the β3–β4 loop, and Tyr359 on the β4 strand) of domain II were important to the toxicity of the Cry4Ba toxin. Quantitative binding assays using enzyme-linked immunosorbent assay (ELISA) showed similar decreasing trends in binding to the Aa-mALP receptor and in toxicity of the Cry4Ba mutants Trp327Ala, Tyr347Ala, and Tyr359Ala, suggesting that a possible function of these surface-exposed aromatic residues is receptor binding. In addition, binding assays of the Cry4Ba toxin to the mutants of the binding residues Gly513, Ser490, and Phe497 of the Aa-mALP receptor supported the binding function of Trp327, Tyr347, and Tyr359 of the Cry4Ba toxin, respectively. Altogether, our results showed for the first time that aromatic residues on a side surface of the Cry4Ba domain II function in receptor binding. This finding provides greater insight into the possible molecular mechanisms of the Cry4Ba toxin.


Introduction
Dengue infection, causing dengue fever, dengue hemorrhagic fever (DHF), and dengue shock syndrome (DSS), is a major global public health concern [1]. Dengue is a mosquitoborne Flavivirus infection that is primarily transmitted by Aedes aegypti [2]. Dengue is hyperendemic in tropical and subtropical regions and has been reported in more than 100 countries [2,3]. Globally, there were an estimated 390 million infections in 2010, of which 96 million cases had apparent clinical symptoms [4]. The geographical range of dengue is expected to expand into low-risk or dengue-free areas due to climate change, with approximately 6.1 billion people (60% of the world's population) predicted to be at risk of dengue infection in 2080 [5]. Currently, only one commercially available vaccine against and the Aa-mALP receptor (blue) prepared using VMD software [23]. Ball and stick models represent some potential binding residues (aromatic and hydrophobic side chains of domain II (Trp327, Tyr347, Ile358, and Tyr359) and domain III (Phe490 and Leu517) of the Cry4Ba toxin along with their counterparts on the Aa-mALP receptor).

Analysis of Possible Binding Residues between the Cry4Ba Toxin and the Aa-mALP Receptor
Analysis of the molecular docking interaction between the Cry4Ba toxin and the Aa-mALP receptor revealed that the Cry4Ba toxin used the surface residues of both domains II and III on a side of the molecule to bind to the Aa-mALP receptor (Figure 1). The docking result showed that the major domain II residues that function in binding to the Aa-mALP receptor are on a side surface rather than only as residues on the beta-hairpin loops located on the underside of the toxin. In our previous work, some binding residues from this docking result were studied and we found that Leu615 in the β22-β23 loop of the Cry4Ba domain III toxin was a crucial residue for Aa-mALP receptor binding [21]. In the present study, aromatic and hydrophobic residues of Cry4Ba domain II (Trp327, Tyr347, Ile358, and Tyr359) and domain III (Phe490 and Leu517) and their counterpart interacting residues (Gly513, Ser490, His500, Phe497, Leu53, and Leu52, respectively) on Aa-mALP ( Figure 1) predicted as responsible for binding were selected to study their relevance to binding and toxicity. The effects of the mutation of the selected Cry4Ba residues on the binding affinity to the Aa-mALP receptor were preliminarily determined using in silico mutation analysis using the BeAtMuSiC approach [24]. The BeAtMuSiC results revealed that all the alanine substitutions of the selected Cry4Ba residues decreased in stability of the Cry4Ba−Aa- and the Aa-mALP receptor (blue) prepared using VMD software [23]. Ball and stick models represent some potential binding residues (aromatic and hydrophobic side chains of domain II (Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 ) and domain III (Phe 490 and Leu 517 ) of the Cry4Ba toxin along with their counterparts on the Aa-mALP receptor).

Analysis of Possible Binding Residues between the Cry4Ba Toxin and the Aa-mALP Receptor
Analysis of the molecular docking interaction between the Cry4Ba toxin and the Aa-mALP receptor revealed that the Cry4Ba toxin used the surface residues of both domains II and III on a side of the molecule to bind to the Aa-mALP receptor (Figure 1). The docking result showed that the major domain II residues that function in binding to the Aa-mALP receptor are on a side surface rather than only as residues on the beta-hairpin loops located on the underside of the toxin. In our previous work, some binding residues from this docking result were studied and we found that Leu 615 in the β22-β23 loop of the Cry4Ba domain III toxin was a crucial residue for Aa-mALP receptor binding [21]. In the present study, aromatic and hydrophobic residues of Cry4Ba domain II (Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 ) and domain III (Phe 490 and Leu 517 ) and their counterpart interacting residues (Gly 513 , Ser 490 , His 500, Phe 497 , Leu 53 , and Leu 52 , respectively) on Aa-mALP ( Figure 1) predicted as responsible for binding were selected to study their relevance to binding and toxicity. The effects of the mutation of the selected Cry4Ba residues on the binding affinity to the Aa-mALP receptor were preliminarily determined using in silico mutation analysis using the BeAtMuSiC approach [24]. The BeAtMuSiC results revealed that all the alanine substitutions of the selected Cry4Ba residues decreased in stability of the Cry4Ba−Aa-mALP complex as ∆∆G bind > 0 values were obtained ( Figure 2).  Changes in binding free energy (ΔΔGbind) in the Cry4Ba-Aa-mALP complex following alanine substitution of the selected Cry4Ba residues analyzed using BeAtMuSiC approach [24], where ΔΔGbind > 0 means decreased binding affinity.

Expression and Purification of Cry4Ba and its Mutants
The selected aromatic and hydrophobic residues (Trp327, Tyr347, Ile358, and Tyr359 on domain II and Phe490 and Leu517 on domain III) of Cry4Ba which were predicted as binding residues were individually mutated to alanine using PCR-based site-directed mutagenesis. All mutants were successfully constructed. After 4 h induction with isopropyl-β-Dthiogalactopyranoside (IPTG), all mutant Cry4Ba toxins were expressed as inclusion proteins in E. coli JM109 for which the sizes on polyacrylamide gel were 130 kDa ( Figure 3A). The inclusion proteins of each mutant were purified from E. coli and solubilized in carbonate buffer, pH 9.2. Proteolytic activation of the solubilized protoxins by digestion with trypsin produced 65 kDa active fragments which were comparable to that of the Cry4Ba-R203Q toxin. The 65 kDa active toxins purified using a size-exclusion fast protein liquid chromatography (FPLC) system were shown in Figure 3B.  Changes in binding free energy (∆∆G bind ) in the Cry4Ba-Aa-mALP complex following alanine substitution of the selected Cry4Ba residues analyzed using BeAtMuSiC approach [24], where ∆∆G bind > 0 means decreased binding affinity.

Expression and Purification of Cry4Ba and Its Mutants
The selected aromatic and hydrophobic residues (Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 on domain II and Phe 490 and Leu 517 on domain III) of Cry4Ba which were predicted as binding residues were individually mutated to alanine using PCR-based site-directed mutagenesis. All mutants were successfully constructed. After 4 h induction with isopropylβ-D-thiogalactopyranoside (IPTG), all mutant Cry4Ba toxins were expressed as inclusion proteins in E. coli JM109 for which the sizes on polyacrylamide gel were 130 kDa ( Figure 3A). The inclusion proteins of each mutant were purified from E. coli and solubilized in carbonate buffer, pH 9.2. Proteolytic activation of the solubilized protoxins by digestion with trypsin produced 65 kDa active fragments which were comparable to that of the Cry4Ba-R203Q toxin. The 65 kDa active toxins purified using a size-exclusion fast protein liquid chromatography (FPLC) system were shown in Figure 3B.

Expression and Purification of Cry4Ba and its Mutants
The selected aromatic and hydrophobic residues (Trp327, Tyr347, Ile358, and Tyr359 o domain II and Phe490 and Leu517 on domain III) of Cry4Ba which were predicted as bindin residues were individually mutated to alanine using PCR-based site-directed mutagene sis. All mutants were successfully constructed. After 4 h induction with isopropyl-β-D thiogalactopyranoside (IPTG), all mutant Cry4Ba toxins were expressed as inclusion pro teins in E. coli JM109 for which the sizes on polyacrylamide gel were 130 kDa ( Figure 3A The inclusion proteins of each mutant were purified from E. coli and solubilized in ca bonate buffer, pH 9.2. Proteolytic activation of the solubilized protoxins by digestion wit trypsin produced 65 kDa active fragments which were comparable to that of the Cry4Ba R203Q toxin. The 65 kDa active toxins purified using a size-exclusion fast protein liqui chromatography (FPLC) system were shown in Figure 3B.

Larvicidal Activity of Cry4Ba and Its Mutants
The larvicidal activity of Cry4Ba and its mutants were assayed against 2-day-old A. aegypti larvae. Larval mortality was recorded after 24 h feeding A. aegypti larvae with E. coli cells expressing Cry4Ba-R203Q or its mutants. Bioassays revealed that the mutants W327A, Y347A, and Y359A of domain II exhibited significant decreases in toxicity (to 40-70%) (p < 0.003, Student's t-test), with Y359A showing the highest toxicity reduction to approximately 40%, while the I358A, F490A, and L517A mutants retained their larvicidal activities (>80%) which were comparable to that of the Cry4Ba-R203Q toxin ( Figure 4). pUC12 vector (nc, a negative control) and (B) FPLC-purified 65 kDa trypsin-treated p solubilized protoxins from (A). M represents the molecular mass standards.

Larvicidal Activity of Cry4Ba and its Mutants
The larvicidal activity of Cry4Ba and its mutants were assayed against 2-d aegypti larvae. Larval mortality was recorded after 24 h feeding A. aegypti larv coli cells expressing Cry4Ba-R203Q or its mutants. Bioassays revealed that th W327A, Y347A, and Y359A of domain II exhibited significant decreases in toxic 70%) (p < 0.003, Student's t-test), with Y359A showing the highest toxicity red approximately 40%, while the I358A, F490A, and L517A mutants retained their activities (>80%) which were comparable to that of the Cry4Ba-R203Q toxin ( Fig   Figure 4. Mosquitocidal activity of E. coli (~10 8 cells/mL) expressing 130 kDa protoxins R203Q or its domain II (DII) mutants (W327A, Y347A, I358A, and Y359A) and domai mutants (F490A and L517A) against A. aegypti larvae. Cells harboring the pUC12 vector as a negative control. Error bars indicate standard error of the mean from three independ iments.

Expression and Purification of Aa-mALP and its Mutants
The amino acid residues Leu52, Leu53, Ser490, Phe497, His500, and Gly513 of with predicted binding with Leu517, Phe490, Tyr347, Tyr359, Ile358, and Trp327 of th toxin, respectively, were individually mutated to alanine. All of the mutants wer fully generated and expressed as inclusion proteins in E. coli BL21 for which th polyacrylamide gel were 54 kDa ( Figure 5A). The purified inclusion proteins w bilized in PBS buffer containing 8 M urea, then purified and refolded in a nicke riacetic acid (Ni-NTA) affinity column. After desalting and buffer exchange, the proteins were obtained in carbonate buffer (pH 9.2) with a size of 54 kDa on the g 5B). To confirm that the purified Aa-mALP and its mutants were refolded as form, the alkaline phosphatase activity was analyzed based on dot blotting and with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT genic substrate (Supplementary Figure S1).

Expression and Purification of Aa-mALP and Its Mutants
The amino acid residues Leu 52 , Leu 53 , Ser 490 , Phe 497 , His 500 , and Gly 513 of Aa-mALP with predicted binding with Leu 517 , Phe 490 , Tyr 347 , Tyr 359 , Ile 358 , and Trp 327 of the Cry4Ba toxin, respectively, were individually mutated to alanine. All of the mutants were successfully generated and expressed as inclusion proteins in E. coli BL21 for which the sizes on polyacrylamide gel were 54 kDa ( Figure 5A). The purified inclusion proteins were solubilized in PBS buffer containing 8 M urea, then purified and refolded in a nickelnitrilotriacetic acid (Ni-NTA) affinity column. After desalting and buffer exchange, the Aa-mALP proteins were obtained in carbonate buffer (pH 9.2) with a size of 54 kDa on the gel ( Figure 5B). To confirm that the purified Aa-mALP and its mutants were refolded as the native form, the alkaline phosphatase activity was analyzed based on dot blotting and detection with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT) chromogenic substrate (Supplementary Figure S1).

Quantitative Analysis of Cry4Ba Toxin-Aa-mALP Receptor Interactions
The relevance on receptor binding of Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 on domain II and Phe 490 and Leu 517 on domain III of Cry4Ba was analyzed using ELISA. The quantitative binding studies revealed that mutation of aromatic residues on Cry4Ba domain II (W327A, Y347A, and Y359A) affected binding to the immobilized Aa-mALP receptor. The binding activities of these aromatic residues were significantly lower than for the Cry4Ba-R203Q toxin (p < 0.02, Student's t-test) ( Figure 6A). In contrast, the mutants I358A, F490A, and L517A did not affect binding to the Aa-mALP receptor, as shown in ( Figure 6A).

Quantitative Analysis of Cry4Ba Toxin-Aa-mALP Receptor Interactions
The relevance on receptor binding of Trp327, Tyr347, Ile358, and Tyr359 on domain II and Phe490 and Leu517 on domain III of Cry4Ba was analyzed using ELISA. The quantitative binding studies revealed that mutation of aromatic residues on Cry4Ba domain II (W327A, Y347A, and Y359A) affected binding to the immobilized Aa-mALP receptor. The binding activities of these aromatic residues were significantly lower than for the Cry4Ba-R203Q toxin (p < 0.02, Student's t-test) ( Figure 6A). In contrast, the mutants I358A, F490A, and L517A did not affect binding to the Aa-mALP receptor, as shown in ( Figure 6A).
Further characterization was performed by mutation of the Aa-mALP residues predicted to interact with the selected Cry4Ba residues and the effects of mutation on toxinreceptor interactions were studied. The residues Gly513, Ser490, His500, Phe497, Leu53, and Lue52 of Aa-mALP which interacted with Trp327, Tyr347, Ile358, Tyr359, Phe490, and Leu517 of the Cry4Ba toxin, respectively, were individually substituted with alanine and determined for interaction with the Cry4Ba-R203Q toxin using ELISA. The results revealed that the mutants S490A, H500A, F497A, L53A, and L52A affected binding with the toxin as significant decreases in their binding values were observed in the range 30-50% (p < 0.006, Student's t-test), with L53A having the highest decrease of approximately 30%, while the G513A mutant retained its binding to the toxin that was comparable to that of the wildtype Aa-mALP receptor ( Figure 6B).

Discussion
Receptor binding is a prerequisite step to exert the larvicidal activity of Cry toxins Further characterization was performed by mutation of the Aa-mALP residues predicted to interact with the selected Cry4Ba residues and the effects of mutation on toxinreceptor interactions were studied. The residues Gly 513 , Ser 490 , His 500 , Phe 497 , Leu 53 , and Lue 52 of Aa-mALP which interacted with Trp 327 , Tyr 347 , Ile 358 , Tyr 359 , Phe 490 , and Leu 517 of the Cry4Ba toxin, respectively, were individually substituted with alanine and determined for interaction with the Cry4Ba-R203Q toxin using ELISA. The results revealed that the mutants S490A, H500A, F497A, L53A, and L52A affected binding with the toxin as significant decreases in their binding values were observed in the range 30-50% (p < 0.006, Student's t-test), with L53A having the highest decrease of approximately 30%, while the G513A mutant retained its binding to the toxin that was comparable to that of the wild-type Aa-mALP receptor ( Figure 6B).

Discussion
Receptor binding is a prerequisite step to exert the larvicidal activity of Cry toxins [25]. Domains II and III of Cry toxins have been reported to function in binding to receptors on epithelial cells of the larval midgut (for review, see [26]). For the Cry4Ba toxin, several amino acid residues on β-hairpin loops, especially those located on the lower part of domain II such as Thr 386 , Ser 387 , Ser 388 , Pro 389 , Ser 390 , and Asn 391 on the β6-β7 loop, Glu 417 on the β8-β9 loop, and Tyr 455 and Asn 456 on the β10-β11 loop, have been reported to be important for binding and toxicity [16,22]. Our previous study [21], using molecular docking, revealed a possible Cry4Ba toxin-Aa-mALP receptor interaction in which surface residues on the side of the domain II, such as residues on β2 and β4 and residues on the β-hairpin loops of domains II and III rather than only the residues on the β-hairpin loops at the bottom of domain II, are responsible for binding to the Aa-mALP receptor (Figure 1). According to a previous study, three fundamental hotspot residues (tryptophane, tyrosine, and arginine) were reported to contribute more significantly to binding affinity in protein-protein interfaces compared with others (for review, see [27]). In addition, hydrophobic residues were reported to be important for protein-protein interactions [21,28]. In the present work, some aromatic and hydrophobic residues on the surface of Cry4Ba domain II (Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 ) and domain III (Phe 490 and Leu 517 ) that were predicted as binding residues were therefore selected to study their relevance to the binding and toxicity of the Cry4Ba toxin. In silico analysis of the change in binding free energy (∆∆G bind ) of a protein-protein complex upon mutation of the amino acid residue to alanine using the BeAtMuSiC approach revealed moderate destabilization change (∆∆G bind~1 .5-2.0 kcal/mol), suggesting that these residues may be essential for binding to the Aa-mALP receptor. The ∆∆G bind values analyzed using BeAtMuSiC have been reported to be more than 0.5 kcal/mol of alanine substitution of residues crucial for bioactivity [21,29].
PCR-based site-specific mutagenesis was performed to substitute Trp 327 , Tyr 347 , Ile 358 , Tyr 359 , Phe 490 , and Leu 517 of Cry4Ba residues to alanine. All of the Cry4Ba mutant proteins were heterologously expressed in E. coli at similar levels compared to the Cry4Ba-R203Q protein and their tryptic digestion patterns were comparable with that of the Cry4Ba-R203Q toxin, implying that the mutations did not affect folding of the mutant proteins. Protease sensitivity change was reported to be caused by structural alteration by mutation of Cry toxins [30]. Larvicidal activity assays against A. aegypti showed a significant decrease in toxicity of aromatic residue mutations on Cry4Ba domain II (W327A, Y347A, Y359A), suggesting that these aromatic residues, especially Tyr 359 which had a more adverse effect on toxicity, may be related to receptor binding. The involvement in receptor binding of the aromatic residues Trp 327 , Tyr 347 , and Tyr 359 of Cry4Ba domain II was supported by Cry4Ba-Aa-mALP binding assays using ELISA that showed a decrease in binding of W327A, Y347A, and Y359A to the Aa-mALP receptor which correlated with their larvicidal activities. Notably, not all of the predicted binding residues were involved in the binding and toxicity of the toxin. However, the decreases in both the receptor binding and toxicity of the Cry4Ba mutants W327A, Y347A, and Y359A implied an important role of these aromatic residues.
Even the Cry4Ba mutants I358A, F490A, and L517A retained their toxicity and binding to the Aa-mALP receptor, while their counterpart Aa-mALP mutant residues H500A, L53A, and L52A showed significant decreases in binding to the toxin. This may have been due to the binding of His 500 , Leu 53 , and Lue 52 of Aa-mALP residues to more than one binding residue of the Cry4Ba toxin; therefore, a single mutation of these residues provided more effect on binding activities. For example, the Cry4Ba-F490A mutant retained binding to the wild-type Aa-mALP receptor, while its counterpart Aa-mALP-L53A mutant showed a large decrease in binding to the toxin that may have been due to Lue 53 bonding not only to Phe 490 but also to Arg 52 and Glu 522 of the toxin.
Tryptophane is one of three top-range hot-spot residues in protein-protein interfaces; the Trp/Ala mutation was reported to create a large cavity due to its large size and aromatic nature [31]. The presence of conservative Trp on the protein surface was reported as indicating a highly possible binding site [31]. Mutation W327A may affect Cry4Ba toxin-Aa-mALP receptor interaction and hence toxicity by disruption of the hydrophobic interaction of the aromatic side chain of Trp 327 on the β2 strand of the Cry4Ba toxin and the backbone of Gly 513 of the Aa-ALP receptor ( Figure 7A). Mutation by Gly513Ala of the Aa-ALP receptor that did not affect the binding of the toxin may support the binding of Trp 327 to the backbone of Gly 513 . Tyrosine was one of the enriched amino acids acting as a hot-spot residue in protein-protein interactions that primarily contributed hydrogen-bonding with polar and charge residues or interaction with backbone atoms and sidechain carbon [32,33]. Mutation Y347A eliminated the hydrogen-bonding between Tyr 347 on the β3-β4 loop of the Cry4Ba toxin and Ser 490 of the Aa-mALP receptor ( Figure 7B), which may have been a cause of decreased receptor binding and hence toxicity of the toxin. The binding interaction Tyr 347 -Ser 490 was also supported by the decreased binding of the Aa-mALP-S490A mutant to the Cry4Ba toxin. Mutation Y359A may affect the binding and toxicity of the Cry4Ba toxin by interrupting the hydrophobic interaction between Tyr 359 on the β4 strand of the Cry4Ba toxin and Phe 497 of the Aa-mALP receptor ( Figure 7C). The binding interaction of Tyr 359 -Phe 497 was also supported by the decreased binding of the Aa-mALP-F497A mutant to the Cry4Ba toxin.
Aromatic amino acids, especially Tyr, have been reported to play an important role in receptor binding which is a prerequisite for the toxicity of many toxins, such as the cytolethal distending toxin (Cdt) of Aggregatibacter actinomycetemcomitans [34], the alpha toxin (AT) of Clostridium septicum [35], and the binary toxin (Bin) of Bacillus sphaericus [36]. It is possible that the aromatic residues Trp 327 , Tyr 347 , and Tyr 359 on the side surface of Cry4Badomain II are required for binding with the Aa-mALP receptor to exhibit larvicidal activity.
Previous reports mainly focused on the exploration of amino acid residues on βhairpin loops located on the lower part of the Cry4Ba domain II function in receptorbinding [16,22], whereas the present study investigated the binding function of residues on β strands and β-hairpin loops on the side of the Cry4Ba toxin. We clearly showed that the aromatic residues Trp 327 on the β2 strand, Tyr 347 on the β3-β4 loop, and Tyr 359 on the β4 strand of the Cry4Ba-domain II were involved in binding to the Aa-mALP receptor and hence, toxicity. The finding in this work would be helpful to understand the receptor-binding mechanism of the Cry4Ba toxin.

Conclusions
In conclusion, we identified that the aromatic residues Trp327 on the β2 strand, Tyr347 on the β3-β4 loop, and Tyr359 on the β4 strand of the Cry4Ba toxin played essential roles in binding to the Aa-mALP receptor. Our work highlighted the importance of the side surface residues of the Cry4Ba molecule on receptor binding which provides greater insight into the possible mechanism of the Cry4Ba toxin. Further protein engineering could

Conclusions
In conclusion, we identified that the aromatic residues Trp 327 on the β2 strand, Tyr 347 on the β3-β4 loop, and Tyr 359 on the β4 strand of the Cry4Ba toxin played essential roles in binding to the Aa-mALP receptor. Our work highlighted the importance of the side surface residues of the Cry4Ba molecule on receptor binding which provides greater insight into the possible mechanism of the Cry4Ba toxin. Further protein engineering could be applied to achieve a more potent toxin to control the mosquito vectors of dengue viruses.

Materials
Mutagenic primers designed to a specific mutation at selected residues of the Cry4Ba toxin and the Aa-mALP receptor were purchased from Macrogen (Seoul, Republic of Korea). pfu DNA polymerase, restriction enzymes, and isopropyl-β-D-thiogalactopyranoside (IPTG) were purchased from Vivantis Technologies (Selangor, Malaysia). Nickel-nitrilotriacetic acid (Ni-NTA) affinity columns and tolylsulfonyl phenylalanyl chloromethyl ketone (TCPK)treated trypsin were purchased from Thermo Fisher Scientific (Rockford, IL, USA). The rabbit anti Cry4Ba antibody was kindly provided by the Bacterial Toxin Research Innovation Cluster (BRIC), Institute of Molecular Biosciences, Mahidol University, Thailand. The 96-well maxi-binding immunoplates were purchased from SPL Life Science (Gyeonggi, Republic of Korea). Horse radish peroxidase (HRP)-conjugated goat anti-rabbit IgG and the 3,3,5,5-tetramethylbenzidine (TMB) substrate were purchased from Cell Signaling Technology (Beverly, CA, USA). Alkaline phosphatase chromogenic substrate 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT) was purchased from Sigma-Aldrich (St Louis, MO, USA). PD-10 desalting and 10 kDa Vivaspin concentrator columns were purchased from Cytiva (Uppsala, Sweden). Bradford's reagent was purchased from HiMedia Laboratories (Maharashtra, India). Aedes aegypti eggs were purchased from the Department of Medical Sciences, Ministry of Public Health of Thailand (Bangkok, Thailand). Unless otherwise indicated, all other reagents were of analytical grade.

In Silico Binding Analysis of Cry4Ba Toxin-Aa-mALP Receptor Interaction
The in silico binding interaction between the active form of the Cry4Ba toxin (PDB:1W99) and the homology modeled Aa-mALP using ClusPro 2.0 molecular docking software was obtained from our previous work [21]. Interacting residues were analyzed using LIGPLOT 1.4.4 software [37]; then, aromatic and hydrophobic residues on the surface of Cry4Ba domains II and III predicted as being responsible for binding were selected for further site-specific mutagenesis. A possible effect of alanine substitution of possible binding residues of Cry4Ba on toxin-receptor interactions was preliminarily analyzed by computing the binding affinity changes (∆∆G bind ) using the in silico mutation analysis BeAtMuSiC approach [24]. The BeAtMuSiC software calculated the binding free energy based on statistical potentials, where the mutation is destabilizing if ∆∆G bind > 0, while when ∆∆G bind < 0, the respective mutation is stabilizing [38].

Construction of Cry4Ba Mutant Plasmids
Six aromatic and hydrophobic residues (Trp 327 , Tyr 347 , Ile 358 , and Tyr 359 on domain II, and Phe 490 and Leu 517 on domain III of the Cry4Ba toxin) predicted as potential binding residues were selected and individually substituted with alanine (alanine scanning) using the Quick Change site-directed mutagenesis procedure developed by Stratagene (La Jolla, MA, USA). Mutagenic primers were designed based on the nucleotide sequence of the cry4Ba gene (NCBI accession number X07423) to substitute each selected residue with alanine (Supplementary Table S1). A p4Ba-R203Q plasmid, encoding 130 kDa Cry4Ba-R203Q (in which one trypsin-sensitive residue (Arg 203 ) was mutated to Gln to produce a 65 kDa activated toxin for easy purification as described elsewhere [39]), was used as a template for polymerase chain reaction (PCR)-based site-directed mutagenesis by the activity of pfu DNA polymerase. Mutagenized plasmids were transformed into E. coli strain JM109. Mutant clones were primarily identified based on digestion with restriction endonuclease and subsequently confirmed using DNA sequencing from the commercial services of First BASE Laboratories (Selangor, Malaysia).

Construction of Aa-mALP Mutant Plasmids
Six potential binding residues of the Aa-mALP receptor (Leu 52 , Leu 53 , Ser 490 , Phe 497 , His 500 , and Gly 513 ) with predicted binding with Leu 517 , Phe 490 , Tyr 347 , Tyr 359 , Ile 358 , and Trp 327 of the Cry4Ba toxin, respectively, were individually substituted with alanine using the method described above. Mutagenized primers (Supplementary Table S1) were designed according to the nucleotide sequence of the alkaline phosphatase gene from the A. aegypti midgut (NCBI accession number GQ395622). A pET-Aa-mALP plasmid encoding C-terminal His-tagged Aa-mALP under the control of the T7 promoter [20] was used as a template for site-directed mutagenesis. Mutagenized plasmids were transformed into E. coli strain BL21(DE3). Mutant clones were identified as described above.

Expression and Purification of Cry4Ba and Its Mutants
The 65 kDa active Cry4Ba toxins were prepared as described previously [40]. In brief, 130 kDa Cry4Ba-R203Q and its mutants protoxins were overexpressed as cytoplasmic inclusions in E. coli JM109 upon induction with 0.1 mM IPTG for 4 h. After solubilization in carbonate buffer (50 mM Na 2 CO 3 /NaHCO 3 (pH 9.2)) for 1 h, the protoxins were activated by digestion with TCPK-treated trypsin (1:20, w/w) for 16 h into 65 kDa active toxins. The 65 kDa trypsin-activated toxins were purified using a size exclusion fast protein liquid chromatography system on a Superdex-200 HR column from Amersham-Pharmacia Biotech (Piscataway, NJ, USA) as described previously [22] and then concentrated using a Vivaspin concentrator column (10 kDa MWCO). The purified proteins were determined for their concentrations based on Bradford assay and analyzed using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).

Expression and Purification of Aa-mALP and Its Mutants
The 54 kDa Aa-mALP and its mutants were prepared as described previously [20]. In brief, upon induction with 0.1 mM IPTG for 4 h, the His-tag-fused Aa-ALP proteins were overexpressed as inclusion in E. coli BL21. Inclusions were solubilized for 1 h at 25 • C in phosphate-buffered saline (PBS) containing 8 M urea, pH 7.5. Purification and refolding were performed in a Ni-NTA affinity column using gradients of decreasing urea concentrations and elution with PBS containing 500 mM imidazole, pH 7.5. Buffer exchange into a carbonate buffer (pH 9.2) was performed on a PD-10 desalting column according to the manufacturer's protocol. The concentrations of purified Aa-mALP and its mutant proteins were determined using Bradford assay and then analyzed by SDS-PAGE.

Mosquito Larvicidal Activity Assays
The mosquitocidal activity of the Cry4Ba-R203Q and its mutants was investigated against two-day-old A. aegypti larvae. The assays were carried out in 48-well cell culture plates at 30 • C. Each well contained 10 larvae in 1 mL of cell suspension (10 8 cells of E. coli expressing Cry4Ba-R203Q toxin or its mutants in ddH 2 O) and a total of 100 larvae were used to assay for each toxin. E. coli containing a pUC19 plasmid without the insecticidal toxin gene was used as a negative control. Three independent replications were performed for each treatment.

Cry4Ba Toxin-Aa-mALP Binding Assays Using ELISA
Binding interactions between the Cry4Ba toxin and the Aa-mALP receptor were analyzed using ELISA, as described previously with some modifications [21]. The purified Aa-ALP receptor or its mutants were coated (2.5 µg in 200 µL of 50 mM carbonate buffer, pH 9.2) to 96-well maxi-binding immunoplates at 4 • C for 4 h. Then, after washing five times with PBS containing 0.02% Tween 20 (PBS-T), pH 7.4, and blocking with 5% (w/v) skimmed milk in PBS-T, pH 7.4 for 1 h, the coated wells were incubated with 100 nM of each purified Cry4Ba-R203Q, its mutants, or a negative control ligand-bovine serum albumin at 37 • C for 2 h. The Cry4Ba toxin proteins bound to the immobilized Aa-mALP receptor were detected based on incubation with rabbit anti-Cry4Ba antibodies (1:10,000 dilution) for 1 h, followed with HRP-conjugated goat anti-rabbit IgG (1:5000) for 1 h. The TMB substrate was added to the wells and incubated for 10 min. The reaction was stopped by the addition of 1 N HCl and absorbance at 450 nm was measured using a Multiskan ELISA plate reader (Thermo Fisher Scientific, Rockford, IL, USA).

Statistical Analysis
Each experiment was performed at least three independent times. Statistics were calculated using the Student's t-test to analyze significant differences between the wildtype and each mutant. A p value < 0.05 was considered statistically significant.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxins15020114/s1, Figure S1: Alkaline phosphatase activities of purified Aa-mALP wild-type and its mutants. Table S1: Mutagenized primers for alanine substitution of selected amino acid residues on domains II and III of the Cry4Ba toxin and the Aa-mALP receptor.