An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding

An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding

Accepted Manuscript An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding Sivadatch Chooduang, Wah...

3MB Sizes 6 Downloads 17 Views

Accepted Manuscript An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding Sivadatch Chooduang, Wahyu Surya, Jaume Torres, Panadda Boonserm PII:

S0003-9861(18)30515-0

DOI:

10.1016/j.abb.2018.10.006

Reference:

YABBI 7833

To appear in:

Archives of Biochemistry and Biophysics

Received Date: 2 July 2018 Revised Date:

9 October 2018

Accepted Date: 11 October 2018

Please cite this article as: S. Chooduang, W. Surya, J. Torres, P. Boonserm, An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding, Archives of Biochemistry and Biophysics (2018), doi: https://doi.org/10.1016/j.abb.2018.10.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT An aromatic cluster in Lysinibacillus sphaericus BinB involved in toxicity and proper in-membrane folding.

a

RI PT

Sivadatch Chooduanga, 1, Wahyu Suryab, Jaume Torresb and Panadda Boonserma,*

Institute of Molecular Biosciences, Mahidol University, Salaya, Phuttamonthon, Nakhon

Pathom, Thailand

School of Biological Sciences, Nanyang Technological University, Singapore

M AN U

SC

b

*

To whom correspondence should be addressed: Panadda Boonserm, Institute of Molecular

Biosciences, Mahidol University, Salaya, Phuttamonthon, Nakhon Pathom 73170, Thailand, Tel: +66 2441 9003; Fax: +66 2441 9906; E-mail: [email protected] Present address: Jetanin Institute for Assisted Reproduction, 5 Soi Chidlom, Ploenchit

TE D

1

AC C

EP

Road, Lumpinee, Pathumwan, Bangkok, Thailand 10330

1

ACCEPTED MANUSCRIPT ABSTRACT

The binary toxin from Lysinibacillus sphaericus has been successfully used for controlling mosquito-transmitted diseases. Based on structural alignments with other toxins,

RI PT

an aromatic cluster in the C-terminal domain of BinB (termed here BC) has been proposed to be important for toxicity. We tested this experimentally using BinB mutants bearing single mutations in this aromatic cluster. Consistent with the hypothesis, two of these mutations,

SC

F311A and F315A, were not toxic to Culex quinquefasciatus larvae and were unable to permeabilize liposomes or elicit ion channel activity, in contrast to wild-type BinB. Despite

M AN U

these effects, none of these mutations altered significantly the interaction between the activated forms of the two subunits in solution. These results indicate that these aromatic residues on the C-terminal domain of BinB are critical for toxin insertion in membranes. The latter can be by direct contact of these residues with the membrane surface, or by facilitating

TE D

the formation a membrane-inserting oligomer.

Keywords: Pore-forming toxin; Lysinibacillus sphaericus binary toxin; oligomerization; ion

AC C

EP

channel; BinB C-terminal domain

2

ACCEPTED MANUSCRIPT 1. INTRODUCTION

Mosquitoes are considered to be one of the most dangerous insects that spread several serious diseases in humans such as dengue, malaria, filariasis or japanese encephalitis [1, 2].

RI PT

However, the control of mosquitoes using synthetic pesticides cause adverse effects on humans and other organisms as well as the development of mosquito resistance [3-5]. In contrast, bioinsecticides have been developed and successfully integrated in the mosquito

SC

control programs worldwide since they are environmentally friendly, and exhibit high specificity to insect targets [6, 7].

M AN U

Lysinibacillus sphaericus (Ls) is a Gram-positive, spore-forming bacterium that produces insecticidal toxins active against mosquito larvae [8, 9]. L. sphaericus produces the binary toxin (Bin) which is composed of two proteins, BinA (42 kDa) and BinB (51 kDa). The protoxins, hereafter referred to as pro-A and pro-B, are synthesized during the sporulation phase

TE D

as crystalline inclusions [10]. Among the mosquito species, Culex sp. are the most susceptible to Bin toxin, followed by Anopheles sp. while Aedes sp. are Bin-refractory [11]. After ingestion by susceptible mosquito larvae, subunits pro-A and pro-B are processed by larval

EP

gut proteases, becoming activated forms of 40 kDa (act-A) and 45 kDa (act-B) [12, 13]. Although the precise molecular events following this activation are still under debate,

AC C

activated BinB (act-B) has been shown to specifically bind to a GPI-anchored maltase 1 receptor (Cpm1 in Culex sp. and Agm1 in Anopheles sp.). This receptor is present in brush border membranes of susceptible larval midgut [14-17] and directs the internalization of activated BinA (act-A) or act-A/act-B complex into susceptible insect cells [18, 19]. Maximum toxicity against mosquito larvae is achieved when both BinA and BinB components are present at equimolar amounts [20]. Using gel filtration, dynamic and static light scattering and sedimentation velocity, the precise stoichiometry of the two activated

3

ACCEPTED MANUSCRIPT subunits (act-A and act-B) in solution has been unequivocally shown to be heterodimers, whereas the protoxins form monomers [21]. The recent three-dimensional structures of pro-A, pro-B and act-B have been solved by X-ray crystallography [22, 23]. Both BinA and BinB share highly similar structures, with

RI PT

two functional N- and C-terminal domains. The N-terminal domain is globular, and has been proposed to be responsible for receptor recognition, based on its structural similarities with sugar-binding proteins or lectins. The C-terminal domain has an elongated shape, with two

SC

special structural features: the presence of a cluster of aromatic residues exposed on the surface, and a predominance of serine and threonine residues on the outer face of the β-

M AN U

sheets. These features are also found in other aerolysin type β pore-forming toxins (β-PFTs) [24]. Aerolysin has four domains, and domain 4 has an amphipathic β-barrel structure responsible for membrane insertion of the final heptameric complex [25, 26]. It is assumed that pro-aerolysin approaches the target cell as a water-soluble hydrophilic dimer which, after

TE D

being concentrated on the surface, binds to the glycosylphosphatidylinositol (GPI)-anchored protein receptor. Formation of the heptameric complex follows, resulting in exposure of the hydrophobic region of the toxin and thus membrane penetration [27]. Previously, BinB was

EP

shown to insert in model lipid bilayers and its C-terminal domain has been proposed to be

AC C

pore-forming region [28].

In BinB structure, the cluster of aromatic residues has been found in a region of about

30 amino acids with an alternate pattern of hydrophobic and hydrophilic residues, suggesting to constitute a transmembrane domain [22, 23]. In the present study, the role of aromatic amino acids in the putative transmembrane region of BinB has been characterized. Our results demonstrate that Phe-311 and Phe-315 of BinB are essential for larvicidal activity and interaction with membranes.

4

ACCEPTED MANUSCRIPT 2. MATERIALS AND METHODS 2.1. Site-directed mutagenesis To investigate the role of the cluster of aromatic residues located on the proposed TM region (Fig. 1B), five different mutants were made on pro-B: Y306A, Y314A, Y316A,

RI PT

F311A and F315A. Site-directed mutagenesis was performed following the PCR-based QuikChange mutagenesis protocol (Agilent Technologies). Mutagenic primers were designed to change the desired residues and either introduced or deleted restriction endonuclease

SC

recognition sites in order to distinguish the mutant plasmids from the wild type plasmids, see table 1. The PCR products were digested by DpnI and transformed into E. coli strain JM109

further confirmed by DNA sequencing.

M AN U

competent cells. The mutant plasmids were screened by restriction digestion analysis and

Primer

TE D

Table 1. Oligonucleotide primers for generation of BinB mutants.

5'-GATTTAAATATG GCCATTGGAGCAGATTTT-3'

Y306A-r

5'-ATCTGCTCCAATGGCCATATTTAAATC TTC-3'

Y314A-f

5'-T TTTGGCATGGCATTC TAT TTGAGATCTAG-3'

EP

Y306A-f

enzyme

Haelll

Bsml

5'-ATCTCAAATAGAATGCCATGCCAAAATCTG-3'

AC C

Y314A-r

Restriction

Sequence

Y316A-f

5'-CATGTATTTTGCT TTGCGATCTAGCGGATT-3'

Bglll

Y316A-r

5'-CCGCTAGATCGCAAAGCAAAATACATGCCA-3'

(Delete)

F311A-f

5'- GTATATTGGTGCAGATGCTGGCATGTA -3'

F311A-r

5'- CATGCCAGCATCTGCACCAATATACATATT -3'

F315A-f

5'- TTTGGCATGTATGCATATTTGAGATCTAGC -3'

Bsgl

Nsil F315A-r

5'- AGATCTCAAATATGCATACATGCCAAAATC -3'

5

ACCEPTED MANUSCRIPT * The letters f and r represent forward and reverse primers, respectively. Recognition sites either introduced or deleted in the primers for restriction endonuclease analysis are underlined and mutated nucleotides are shown in italic.

RI PT

2.2. Construction of truncated fragment

The C-terminal domain of BinB (BC), corresponding to residues I201 to Q448, was constructed by PCR using the following forward and reverse primers: 5′-TTT TAT GCT

SC

CAT ATG ATT CCT CAA TTA C-3′ and 5′-GCG CGG ATC CTC ACT GGT TAA TTT TAG GTA TTA ATT-3′ (restriction endonuclease sites are underlined) and cloned into a

automated DNA sequencing.

2.3. Expression of protein

M AN U

pET28b expression vector. The recombinant plasmid carrying the BC DNA was verified by

TE D

His-tagged proteins pro-A, pro-B (WT and mutants), and BC, were expressed in soluble form in E. coli BL21(DH3) pLysS after induction with 0.2 mM IPTG at 18 °C for 5 h following the protocols as described previously [29]. Harvested cell pellets were resuspended

EP

in lysis buffer (50 mM Tris-HCl, 200 mM NaCl, pH 8.0) and lysed with a sonicator. The cell

AC C

lysate was centrifuged at 15,000 g for 1 h and the supernatant was collected and filtered through a 0.2 µm syringe filter before purification.

2.4. Protein purification and overall secondary structure analysis A two-step purification was performed using affinity and size exclusion chromatography. A HiTrap™ Chelating HP (GE healthcare Life Science) column was precharged with 0.1 M NiSO4 and equilibrated with lysis buffer. The supernatant collected from the cell lysate (see above) was loaded into the column and washing was performed with

6

ACCEPTED MANUSCRIPT lysis buffer containing 25 mM imidazole and 50 mM imidazole. Bound (His)6-tagged protein was eluted with lysis buffer containing 100 mM imidazole. Protein-containing fractions were pooled and concentrated by ultrafiltration at 4 °C using a Centriprep column (30-kDa cutoff for pro-A or -B, and 10 kDa cutoff for BC).

RI PT

To obtain activated toxin, trypsin was added to a trypsin-to-toxin mass ratio of 1:10 and incubated at 37 °C for 2 h. The reaction was terminated by PMSF addition and the cleaved -activated-protein was purified by size-exclusion chromatography (Superdex 200 HR

SC

10/30 column, GE healthcare Life Science). This column was equilibrated with 50 mM TrisHCl pH 9.0 and 1 mM DTT. To observe the oligomerization of toxins in gel filtration, the

M AN U

same column was used but 50 mM NaCl was added to the buffer at 25 °C. The flow rate was 0.4 ml/min. Protein quality and quantity was estimated by SDS-PAGE and a Bradford assay, respectively. Overall secondary structure was determined using CD (Applied Photophysics),

TE D

and CD spectra were analysed using DichroWeb [30].

2.5. Mosquito-larvicidal activity assay

To test toxicity, 2nd instar Culex quinquefasciatus larvae were exposed to a mixture of

EP

pro-A and pro-B at a 1:1 molar ratio, which was serially diluted in two-fold steps. The protein

AC C

mixture was added into wells containing 10 larvae each. Subsequently, the larvae were left at room temperature for 2 days and mortality was recorded. LC50 was determined using Probit analysis [31].

2.6. Calcein release assay The calcein release assay used large unilamellar vesicles (LUVs) that were prepared from a mixture of phosphatidylcholine (PC) and phosphatidic acid (PA). The lipid mixture was evaporated under a nitrogen gas stream and resuspended with 200 µl of 60 mM calcein in

7

ACCEPTED MANUSCRIPT 50 mM carbonate buffer at pH 9.2. The mixture was subjected to five cycles of freezethawing, and was passed through a polycarbonate membrane. Free-calcein was removed by gel filtration using a HiTrapTM desalting column. The liposome concentration was determined by a phosphorus content assay. Toxins were added to LUVs at a molar protein-to-lipid ratio

RI PT

of 1:10 (0.125:1.25 µM). As a consequence of membrane perturbation, calcein was released and fluorescence intensity was monitored by a JASCO FP-6300 spectrofluorometer at the

2.7. Ion channel activity in planar lipid bilayers

SC

emission and excitation wavelengths of 520 and 485 nm, respectively.

M AN U

Ion channel activity was studied with Port-a-Patch (Nanion, Munich, Germany). Giant unilamellar vesicles (GUVs) composed of DPhPC/DOPS (molar ratio of 3:1) were prepared by electroformation using Vesicle Prep Pro, according to manufacturer’s instructions (Nanion, Germany). Planar lipid bilayers (PLB) were formed by spreading GUVs in 200 mM

TE D

NaCl, 10 mM HEPES, pH 4.0 (symmetric conditions) on the aperture of a 3-5 MΩ 10-NPC chip (Nanion, Germany) and applying slight suction until GΩ-seal was formed. Proteins were added in 50 µl of buffer solution at final concentration of 0.3 µM, and channel activity was

EP

recorded at -50 to +50 mV holding potential. Data analysis was performed with Clampfit

AC C

module of pClamp 9.2 (Axon Instruments/Molecular Devices, USA).

8

ACCEPTED MANUSCRIPT 3. RESULTS 3.1. Secondary structure of BinB mutants. To investigate the role of a cluster of aromatic residues located on the proposed ‘transmembrane’ region of BinB, single mutations were introduced, where the five aromatic

RI PT

residues Y306, Y314, Y316, F311 and F315, were replaced with alanine. The C-terminalaerolysin-like-domain of BinB (BC), encompassing residues I201 to Q448 (Fig. 1A, blue and yellow), was also expressed as a soluble form (data not shown) but it was not toxic when

SC

combined with pro-A, even when the final concentration increased to 1 µg/ml (see below), indicating that the missing N-terminal receptor binding domain is required for toxicity [32].

M AN U

Therefore, the five mutations indicated were only introduced in pro-B and not in the separated BC domain.

All BinB mutants were expressed in a soluble form and could be activated by trypsin, producing similar trypsin-resistant fragments of 45-kDa to the wild-type toxin [see [21] ].

TE D

None of these mutations altered significantly the overall secondary structure of the protein, as

AC C

EP

determined by circular dichroism (Fig. 1C).

9

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

Figure 1. The act-B structure [23]. (A) N-terminal (orange) and C-terminal (blue and yellow) domains of act-B; (B) the putative transmembrane (TM) region (yellow) and the cluster of aromatic residues (green); (C) Circular dichroism spectra of pro-B, and its five mutants (Y306A, F311A, Y314A, F315A, and Y316A). The measurement was performed at 0.1 mg/ml of toxin with a JASCO J-815 CD spectropolarimeter. Since all protein

TE D

concentrations are the same, the y-axis can be shown as not normalized units (mdeg).

EP

3.2. Oligomerization in solution by size-exclusion chromatography Given that both subunits of the Bin toxin have to work together for maximum

AC C

toxicity, the effect of BinB mutations on the act-A/B interaction was analysed by sizeexclusion chromatography. None of the BinB mutations affected this interaction; the two proteins (act-A and act-B mutants) coeluted at the same retention volume (13.9 ml) as the heterodimeric act-A/B wild-type mixture (data not shown) [21]. Therefore, none of these mutations affect significantly the oligomerizing behavior of the act-A/B heterodimer. Elution profiles of wild-type act-B and mutants Y314A and Y316A (Fig. 2) were consistent with a monomer and a minor (about 20% of the total) dimeric form (peak 1, 15.4 ml, 45 kDa and peak 2, 13.6 ml, 90 kDa). Even in fragment BC (monomer at 15.9 ml, 30 10

ACCEPTED MANUSCRIPT kDa), a very small peak at 14 ml (~69 kDa) was present, consistent with a dimeric form. However, this weak dimerization was completely abolished in mutants Y306A, F311A and F315A. Overall, this data shows that act-BinB has a small tendency to dimerize and this is prevented by some mutations in the aromatic cluster. The fact that dimerization was

RI PT

abolished by mutations Y306A, F311A and F315A suggests that these residues are at the

AC C

EP

TE D

M AN U

SC

interface in this homodimer.

Figure 2. Gel filtration of act-B, act-B mutants and BC. Elution profiles of act-B and its mutants. Insert: SDS-PAGE gel of peaks 1 and 2 for act-B showing a single band at 45 kDa. M is protein molecular weight markers.

3.3. Permeabilization of membranes - calcein release assay. To study how these pro-B mutations induce membrane disruption, a calcein release assay was performed where maximum release (100%) was achieved after addition of 11

ACCEPTED MANUSCRIPT detergent. Pro-A and pro-B individually affected calcein release to some extent, i.e., 50% and ~15% of the maximum, respectively (Fig. 3). The mixture pro-A and pro-B achieved >80% of the maximum release. This apparent additive result suggests that, although the protoxins

TE D

M AN U

SC

RI PT

do not interact in solution, they appear to do so in the presence of membranes.

EP

Figure 3. Permeabilizing activity on calcein-entrapped LUVs. Measurements were performed in three independent experiments, and each data sets were then averaged. Toxins:

AC C

A, pro-A; B, pro-B or its mutant; BC, BinB C-terminal domain. The error bar indicates two standard deviations. Bovine serum albumin (BSA) and solubilizing buffer (50 mM Tris-HCl, pH 9.0) were used as negative controls.

Pro-B mutants Y306A, Y314A and Y316A showed comparable or even higher (for Y316A) permeabilization than wild-type pro-B, but F311A and F315A completely abolished permeabilization (Fig. 3). These results can be explained if these mutations either prevent interaction with membranes or prevent formation of a pore for calcein after insertion. 12

ACCEPTED MANUSCRIPT When combined with pro-A, Y306A and Y314A led to some reduction relative to the mixture pro-A/B (45% and 50%, respectively), whereas Y316A, F311A and F315A abolished permeabilization (Fig. 3). Pro-B mutant Y316A on its own clearly disrupts membranes in the calcein release assay. When combined with pro-A (which also permeabilizes membranes),

RI PT

calcein release is abolished. This result is clearly incompatible with a simple disruption of a pro-A/B complex at the membrane surface or in the membrane domain, because both pro-A

SC

and mutant Pro-B Y316A would still be able to disrupt the membrane.

Although as we have shown pro-toxins are monomeric in solution, they may form

M AN U

oligomers of some kind at the membrane surface, once in contact with the membrane, or once inserted. This is very plausible from the observed additive effect of using both pro-toxins in the calcein release assay. If this complex is at the membrane surface, the presence of pro-B mutation Y316A may generate a sort of ‘inactive complex’ that would prevent insertion. We

TE D

speculate that only in the context of this putative pro-A/B heterodimer, but not for pro-B alone, this residue is in close apposition with the membrane, and Y316A acts in a similar fashion to mutations F311A and F315A. To summarize, the difference with the latter residues

EP

is that Y316A would require the interaction with pro-A to become important in the interaction with the membrane. In fact, according to the toxicity assay, Y316A is the third

AC C

most important mutation (after F311A and F315A).

Another possibility is that both pro-A and pro-B mutant insert and form a complex

inside the membrane that becomes unable to disrupt the lipid bilayer from within. This would be possible if for example this disruption involved interaction between heterodimers, in which this residue was relevant. At this moment, in the absence of structural data for any of these putative species, the mechanism can only be speculated.

13

ACCEPTED MANUSCRIPT Neither act-A or act-B elicited any calcein release (not shown), even when hybrid liposomes (LUVs with larvae gut crude-extracts) were used at the highest toxin concentration tested, i.e., 1.25 µM, consistent with previous reports [33]. However, the higher toxicity of the ‘activated’ forms has been reported before [9, 20, 34], and this toxicity probably requires

RI PT

interaction with membranes. The pores formed by these activated forms may be too small for calcein to pass through. Ions can pass through smaller diameter (~0.3 nm) channels, whereas calcein has a hydrated Stokes diameter of around 1.5 nm [35]. To test that the act-forms

M AN U

3.4. Ion channel activity measurements.

SC

interact with liposomal membranes, they were tested for ion channel activity.

Both activated forms were separately added into the external buffer reservoir. Under these experimental conditions, act-A and act-B displayed channel activity (Fig. 4A-B) with conductances of about 367 and 464 pS (Fig. 4D-E), respectively. Membrane instability

TE D

prevented data collection for the act-A/act-B mixture. In any case, these results demonstrate that although they do not permeabilize membranes to calcein, act-A and act-B indeed interact with membranes and form ion channels. In contrast, act-B mutants F311A and F315A

EP

showed no channel activity (data not shown), presumably because of poor or no interaction

AC C

with membranes. Thus, toxicity, calcein release and channel activity data support that mutations at Phe-311 and Phe-315 abolish toxicity because they interfere with proteinmembrane or intersubunit interaction. This data also suggests that pro-toxins and act-toxins interact differently with membranes. In the case of pro-toxins, the pores are larger, and the effect more disruptive, allowing the passage of calcein. For act-toxins, which are the physiologically relevant form, the interaction with membranes appears to be more subtle, producing smaller pores or channels and a less disruptive interaction with membranes.

14

ACCEPTED MANUSCRIPT BC showed the highest conductance, 878 pS (Fig. 4F), consistent with its high calcein release and a large disruption of the membrane. This high conductance may be more

AC C

EP

TE D

M AN U

SC

RI PT

indicative of large pores, rather than ion channels.

Figure 4. Channel activity of act-A, act-B and BC. (A) Current recorded under symmetric conditions after addition of act-A for 15 min. Only some representative voltages are shown; (B) same for act-B; (C) same for BC. The dotted line indicates the closed state. Vertical and horizontal bars represent measured current and time scale, respectively; (D-F) current-voltage relationships corresponding to (A-C), where the data was fitted to a straight line (n=2).

15

ACCEPTED MANUSCRIPT 3.5 Effect of BinB mutants and truncated BinB on larvicidal activity. In general, all the pro-B mutants forming this tight cluster of aromatic residues show some reduction in toxicity when compared to the wild type (see Table 2), but the effect is not uniform. In two of the mutations, F311A and F315A, LC50 is ~100 and ~500 times higher

RI PT

than the pro-B wild type, whereas Y316A, and especially Y306A and Y314A show almost comparable toxicity to the wild type. However, since these Tyr residues are spatially close (Fig. 1B), the presence of the other two may be sufficient to retain toxicity.

SC

Table 2. Mosquito larvicidal activity of the mixture pro-A and pro-B against the 2nd-instar C. quinquefasciatus larvae.

pro-B WT pro-BY306A

pro-BY316A pro-BF311A

a

4

(2-6)

10

(7-13)

8

(5-11)

15

(12-18)

396

(311-502)

2,167

(1,289-5,097)

Not toxic

AC C

BC

EP

pro-BF315A

TE D

pro-BY314A

LC50 (ng/ml)a

M AN U

Toxin (+ pro-A)

LC50 was calculated from three independent experiments using the Probit analysis [31]. The

fiducial limit at 95% is shown in parentheses. 1 µg/ml final concentration of pro-A or pro-B alone and distilled water were used as negative control.

4. DISCUSSION The role of aromatic residues in membrane binding has been extensively studied [26, 36-41]. In equinatoxin II, a cluster of aromatic amino acids located in a loop was found to be

16

ACCEPTED MANUSCRIPT important for protein-membrane interaction, and mutations of these aromatic residues prevented binding of the toxin to the membrane [36, 42, 43]. In α-hemolysin, the indole ring of Trp-179 makes a contact with the quaternary ammonium head group of phosphocholine through cation–π interactions [40]. Crystallographic studies of δ-toxin from Clostridium

RI PT

perfringens in complex with glycerol demonstrate that Tyr-201 interacts with the glycerol backbone via hydrogen bonding [41]. In aerolysin and hemolytic lectin CEL-III, an aromatic belt is positioned to anchor the heptameric complex to the lipid head group–acyl chain

SC

boundary [26, 44].

To investigate the molecular mechanism of BinB, we used the structure of act-B [23]

M AN U

to guide the mutagenesis of the aerolysin-like C-terminal domain (BC) aromatic cluster. Phe mutations at the positions 311 and 315 of BinB abolished mosquito-larvicidal activity and prevented calcein release in the protoxin form and channel activity in the act-form, showing that these two residues are critical for the interaction of the toxin with membranes.

TE D

Our results also show that pro-A and pro-B on their own interact with membranes, as they release calcein upon interaction with liposomes. The effect is additive, showing that the two subunits interact in the membrane environment. Although calcein release was not

EP

observed for act-A and act-B, they also interact with membranes, as shown by the formation

AC C

of ion channels on artificial membranes. In contrast, act-B mutants F311A and F315A showed no channel activity, consistent with their lack of toxicity. Membrane instability prevented the analysis of the act-A/B mixture. Act-B inserted and produced a larger conductance than act-A, in contrast to previous results [33]. BC alone showed both the highest conductance and calcein release, supporting its important role of this domain for interaction with membranes. Previously we have reported that when the activated subunits of BinA and BinB are combined, they readily form a heterodimer in solution [21]. This dimerization, upon

17

ACCEPTED MANUSCRIPT interaction with a membrane or a receptor, may then trigger other conformational changes or further oligomerization, e.g., two heterodimers bound through a BinB interface. A reminiscent supramolecular assembly has been reported in staphylococcal γ-haemolysin and leucocidin pore-forming toxins, where their pore formation mechanism involves protein

RI PT

secretion as monomers, dimer formation on the cell surface, assembly of dimers to form octameric prepore, and finally a transmembrane β-barrel formation [45].

Despite the lack of correlation between calcein release and in vivo toxicity, e.g., BC

SC

showed a large release of calcein (>70%) yet had no effect on toxicity (Table 2), some information can be gained because the assay at least informs on the ability to disrupt

M AN U

membranes. Membrane disruption can simply be caused by non-native conformations [46] arising from mutations, e.g., Y316A, or truncations. However, the calcein release results are consistent with the relative toxicity (A > B) and with the respective membrane-binding roles of these two subunits [33]. They are also consistent with their known additive effect on

TE D

toxicity [34, 47], and with the fact that interaction of pro-A and pro-B with model membranes led to dramatic conformational changes, showing that they independently interact with membranes [28].

EP

Combined with channel activity, this data suggests that pro-toxins and act-toxins

AC C

interact differently with membranes. In the case of pro-toxins, the pores appear to be larger, and the effect more disruptive, allowing the passage of calcein. For act-toxins, which are the physiologically relevant form, the interaction with membranes may be more subtle, producing smaller pores or channels, and an insertion in membranes that is far less disruptive. These different behaviors may be caused by a conformational change after activation. In Cry1Ab, two distinct functional pre-pores have been reported to form after protoxin or protease-activated toxin binds to the cadherin receptor. These pre-pores actively induce pore formation and contribute to the insecticidal activity, although with different

18

ACCEPTED MANUSCRIPT characteristics [48]. In the present paper, we have shown that, although the protoxins do not interact in solution, they show additive effect in a calcein assay, suggesting that they interact in presence of membranes forming some membrane-disrupting complex. In contrast, the activated forms did not disrupt these membranes, although they form heterodimers in

experiments were performed in the absence of a receptor.

RI PT

solution. We see some conceptual similarities in the two systems, although in our case these

However, it is also important to note the experimental conditions for calcein release

SC

assay and BLM are different. The calcein release assay was performed at alkaline pH (~9.2) in presence of 50% of negative charge lipid (PA). This pH is justified because the Cpm1

M AN U

receptor, the recognition site of BinB, localizes on the epithelial cells of the gastric caecum and posterior stomach [49], and the physiological pH in gastric caecum and midgut of mosquito larvae has been shown to be around 10 [50, 51] or 8-9 [52]. The negative charge of the membranes has been shown to be required for the pro-toxin [33, 53]. Neither act-A or act-

TE D

B permeabilize membranes to calcein in these conditions. In contrast, experiments with planar bilayers were conducted at pH 4.0 in DPhPC and DOPS, which provided an approximately neutral charge. In these conditions, both act-A and act-B showed comparable

EP

channel activity in contrast to a previous report that used pH 7.0 [33]. The origin of these pH

AC C

dependencies is not clear and in any case channel activity may not be directly related to toxicity.

In summary, our results show that the aromatic residues, particularly residues Phe-311

and Phe-315, on the C-terminal domain of BinB is critical for proper toxin insertion and conformation in membranes.

Acknowledgments

19

ACCEPTED MANUSCRIPT This work was supported by the 60th Year Supreme Reign of His Majesty King Bhumibol Adulyadej and Postgraduate Student Exchange Program, Mahidol University, Thailand (to SC), Thailand Research Fund, Mahidol University and Synchrotron Light Research Institute (Public Organization) (BRG5980016) (to PB). W.S. is supported by

RI PT

Singapore Ministry of Education grant Tier 1 grant RG 51/13 (to J.T.).

References

R. Rattanarithikul, P. Panthusiri, Illustrated keys to the medically important mosquitos

SC

[1]

of Thailand, Southeast Asian J. Trop. Med. Public Health. 25 Suppl 1 (1994) 1-66. M.A. Tolle, Mosquito-borne Diseases, Curr. Probl. Pediatr. Health Care. 39 (2009)

97-140. [3]

R. Nauen, Insecticide resistance in disease vectors of public health importance, Pest

Manag. Sci. 63 (2007) 628-633.

J. Hemingway, N.J. Hawkes, L. McCarroll, H. Ranson, The molecular basis of

TE D

[4]

M AN U

[2]

insecticide resistance in mosquitoes, Insect Biochem. Mol. Biol. 34 (2004) 653-665. [5]

K.D. Chalegre, T.P. Romao, D.A. Tavares, E.M. Santos, L.M. Ferreira, C.M.

EP

Oliveira, O.P. de-Melo-Neto, M.H. Silva-Filha, Novel mutations associated with resistance to

AC C

Bacillus sphaericus in a polymorphic region of the Culex quinquefasciatus cqm1 gene, Appl. Environ. Microbiol. 78 (2012) 6321-6326. [6]

P.K. Mittal, Biolarvicides in vector control: challenges and prospects, J. Vector Borne

Dis. 40 (2003) 20-32. [7]

H.C. Chapman, Biological control of mosquito larvae, Annu. Rev. Entomol. 19

(1974) 33-59. [8]

I. Ahmed, A. Yokota, A. Yamazoe, T. Fujiwara, Proposal of Lysinibacillus

boronitolerans gen. nov. sp. nov., and transfer of Bacillus fusiformis to Lysinibacillus

20

ACCEPTED MANUSCRIPT fusiformis comb. nov. and Bacillus sphaericus to Lysinibacillus sphaericus comb. nov, Int. J. Syst. Evol. Micr. 57 (2007) 1117-1125. [9]

J.F. Charles, C. Nielson-LeRoux, A. Delecluse, Bacillus sphaericus toxins: molecular

biology and mode of action, Annu. Rev. Entomol. 41 (1996) 451-472. J.F. Charles, M.H. Silva-Filha, C. Nielsen-LeRoux, M.J. Humphreys, C. Berry,

RI PT

[10]

Binding of the 51- and 42-kDa individual components from the Bacillus sphaericus crystal toxin to mosquito larval midgut membranes from Culex and Anopheles sp. (Diptera:

[11]

SC

Culicidae), FEMS. Microbiol. Lett. 156 (1997) 153-159.

C. Berry, J. Hindley, A.F. Ehrhardt, T. Grounds, I. de Souza, E.W. Davidson, Genetic

M AN U

determinants of host ranges of Bacillus sphaericus mosquito larvicidal toxins, J. Bacteriol. 175 (1993) 510-518. [12]

A.H. Broadwell, P. Baumann, Proteolysis in the gut of mosquito larvae results in

further activation of the Bacillus sphaericus toxin, Appl. Environ. Microbiol. 53 (1987) 1333-

[13]

TE D

1337.

A.H. Broadwell, M.A. Clark, L. Baumann, P. Baumann, Construction by site-directed

mutagenesis of a 39-kilodalton mosquitocidal protein similar to the larva-processed toxin of

L.M. Ferreira, T.P. Romao, O.P. de-Melo-Neto, M.H. Silva-Filha, The orthologue to

AC C

[14]

EP

Bacillus sphaericus 2362, J. Bacteriol. 172 (1990) 4032-4036.

the Cpm1/Cqm1 receptor in Aedes aegypti is expressed as a midgut GPI-anchored alphaglucosidase, which does not bind to the insecticidal binary toxin, Insect Biochem. Mol. Biol. 40 (2010) 604-610. [15]

M.H. Silva-Filha, C. Nielsen-Leroux, J.F. Charles, Binding kinetics of Bacillus

sphaericus binary toxin to midgut brush-border membranes of Anopheles and Culex sp. mosquito larvae, Eur. J. Biochem. 247 (1997) 754-761.

21

ACCEPTED MANUSCRIPT [16]

O. Opota, J.F. Charles, S. Warot, D. Pauron, I. Darboux, Identification and

characterization of the receptor for the Bacillus sphaericus binary toxin in the malaria vector mosquito, Anopheles gambiae, Comp. Biochem. Physiol. 149 (2008) 419-427. [17]

I. Darboux, C. Nielsen-LeRoux, J.F. Charles, D. Pauron, The receptor of Bacillus

expression, Insect Biochem. Mol. Biol. 31 (2001) 981-990. [18]

RI PT

sphaericus binary toxin in Culex pipiens (Diptera: Culicidae) midgut: molecular cloning and

O. Opota, N.C. Gauthier, A. Doye, C. Berry, P. Gounon, E. Lemichez, D. Pauron,

SC

Bacillus sphaericus binary toxin elicits host cell autophagy as a response to intoxication, PLoS One. 6 (2011) e14682.

H. Lekakarn, B. Promdonkoy, P. Boonserm, Interaction of Lysinibacillus sphaericus

M AN U

[19]

binary toxin with mosquito larval gut cells: Binding and internalization, J. Invertebr. Pathol. 132 (2015) 125-131. [20]

L. Baumann, P. Baumann, Effects of components of the Bacillus sphaericus toxin on

TE D

mosquito larvae and mosquito-derived tissue culture-grown cells, Curr. Microbiol. 23 (1991) 51-57. [21]

W. Surya, S. Chooduang, Y.K. Choong, J. Torres, P. Boonserm, Binary toxin subunits

EP

of Lysinibacillus sphaericus are monomeric and form heterodimers after in vitro activation,

[22]

AC C

PLoS One. 11 (2016) e0158356.

J.P. Colletier, M.R. Sawaya, M. Gingery, J.A. Rodriguez, D. Cascio, A.S. Brewster,

T. Michels-Clark, R.H. Hice, N. Coquelle, S. Boutet, G.J. Williams, M. Messerschmidt, D.P. DePonte, R.G. Sierra, H. Laksmono, J.E. Koglin, M.S. Hunter, H.W. Park, M. Uervirojnangkoorn, D.K. Bideshi, A.T. Brunger, B.A. Federici, N.K. Sauter, D.S. Eisenberg, De novo phasing with X-ray laser reveals mosquito larvicide BinAB structure, Nature. 539 (2016) 43-47.

22

ACCEPTED MANUSCRIPT [23]

K. Srisucharitpanit, M. Yao, B. Promdonkoy, S. Chimnaronk, I. Tanaka, P.

Boonserm, Crystal structure of BinB: a receptor binding component of the binary toxin from Lysinibacillus sphaericus, Proteins. 82 (2014) 2703-2712. [24]

M.W. Parker, S.C. Feil, Pore-forming protein toxins: from structure to function, Prog.

[25]

RI PT

Biophys. Mol Biol. 88 (2005) 91-142.

M.W. Parker, J.T. Buckley, J.P.M. Postma, A.D. Tucker, K. Leonard, F. Pattus, D.

Tsernoglou, Structure of the aeromonas toxin proaerolysin in its water-soluble and

[26]

SC

membrane-channel states, Nature. 367 (1994) 292-295.

M.T. Degiacomi, I. Iacovache, L. Pernot, M. Chami, M. Kudryashev, H. Stahlberg,

M AN U

F.G. van der Goot, M. Dal Peraro, Molecular assembly of the aerolysin pore reveals a swirling membrane-insertion mechanism, Nat. Chem. Biol. 9 (2013) 623-629. [27]

V. Masignani, M. Pizza, R. Rappuoli, Bacterial toxins. The Prokaryotes: Human

Microbiol. 2013, pp. 499-554.

P. Boonserm, S. Moonsom, C. Boonchoy, B. Promdonkoy, K. Parthasarathy, J.

TE D

[28]

Torres, Association of the components of the binary toxin from Bacillus sphaericus in solution and with model lipid bilayers, Biochem. Biophys. Res. Commun. 342 (2006) 1273-

K. Srisucharitpanit, P. Inchana, A. Rungrod, B. Promdonkoy, P. Boonserm,

AC C

[29]

EP

1278.

Expression and purification of the active soluble form of Bacillus sphaericus binary toxin for structural analysis, Protein Expr. Purif. 82 (2012) 368-372. [30]

L. Whitmore, B.A. Wallace, DICHROWEB, an online server for protein secondary

structure analyses from circular dichroism spectroscopic data, Nucleic Acids Res. 32 (2004) W668-673. [31]

D. Finney, Probit analysis, Cambridge University Press. 3rd ed. (1971).

23

ACCEPTED MANUSCRIPT [32]

C. Tangsongcharoen, P. Boonserm, B. Promdonkoy, Functional characterization of

truncated fragments of Bacillus sphaericus binary toxin BinB, J. Invertebr. Pathol. 106 (2011) 230-235. [33]

J.L. Schwartz, L. Potvin, F. Coux, J.F. Charles, C. Berry, M.J. Humphreys, A.F.

RI PT

Jones, I. Bernhart, M. Dalla Serra, G. Menestrina, Permeabilization of model lipid membranes by Bacillus sphaericus mosquitocidal binary toxin and its individual components, J. Membr. Biol. 184 (2001) 171-183.

P. Baumann, M.A. Clark, L. Baumann, A.H. Broadwell, Bacillus sphaericus as a

SC

[34]

mosquito pathogen: properties of the organism and its toxins, Microbiol. Rev., 1991, pp. 425-

[35]

M AN U

436.

N. Yoshida, M. Tamura, M. Kinjo, Fluorescence Correlation Spectroscopy: A new

tool for probing the microenvironment of the internal space of organelles, Single Mol. 1 (2000) 279-283.

A. Drechsler, C. Potrich, J.K. Sabo, M. Frisanco, G. Guella, M. Dalla Serra, G.

TE D

[36]

Anderluh, F. Separovic, R.S. Norton, Structure and activity of the N-terminal region of the eukaryotic cytolysin equinatoxin II, Biochemistry. 45 (2006) 1818-1828. Q. Hong, I. Gutierrez-Aguirre, A. Barlic, P. Malovrh, K. Kristan, Z. Podlesek, P.

EP

[37]

AC C

Macek, D. Turk, J.M. Gonzalez-Manas, J.H. Lakey, G. Anderluh, Two-step membrane binding by Equinatoxin II, a pore-forming toxin from the sea anemone, involves an exposed aromatic cluster and a flexible helix, J. Biol. Chem. 277 (2002) 41916-41924. [38]

J.M. Mancheno, H. Tateno, D. Sher, I.J. Goldstein, Laetiporus sulphureus lectin and

aerolysin protein family, Adv. Exp. Med. Biol. 677 (2010) 67-80. [39]

C.G. Savva, S.P. Fernandes da Costa, M. Bokori-Brown, C.E. Naylor, A.R. Cole,

D.S. Moss, R.W. Titball, A.K. Basak, Molecular architecture and functional analysis of NetB, a pore-forming toxin from Clostridium perfringens, J. Biol. Chem. 288 (2013) 3512-3522.

24

ACCEPTED MANUSCRIPT [40]

S. Galdiero, E. Gouaux, High resolution crystallographic studies of alpha-hemolysin-

phospholipid complexes define heptamer-lipid head group interactions: implication for understanding protein-lipid interactions, Protein Sci. 13 (2004) 1503-1511. [41]

J. Huyet, C.E. Naylor, C.G. Savva, M. Gibert, M.R. Popoff, A.K. Basak, Structural

RI PT

insights into delta toxin pore formation, PLoS One. 8 (2013) e66673. [42]

P. Malovrh, G. Viero, M.D. Serra, Z. Podlesek, J.H. Lakey, P. Maček, G. Menestrina,

G. Anderluh, A novel mechanism of pore formation: membrane penetration by the N-

SC

terminal amphiphatic region of equinatoxin, J. Biol. Chem. 278 (2003) 22678-22685. [43]

K. Kristan, G. Viero, P. Macek, M. Dalla Serra, G. Anderluh, The equinatoxin N-

M AN U

terminus is transferred across planar lipid membranes and helps to stabilize the transmembrane pore, FEBS J. 274 (2007) 539-550. [44]

H. Unno, S. Goda, T. Hatakeyama, Hemolytic lectin CEL-III Hheptamerizes via a

large structural transition from α-Helices to a β-barrel during the transmembrane pore

TE D

formation process, J. Biol. Chem. 289 (2014) 12805-12812. [45]

D. Yamashita, T. Sugawara, M. Takeshita, J. Kaneko, Y. Kamio, I. Tanaka, Y.

Tanaka, M. Yao, Molecular basis of transmembrane beta-barrel formation of staphylococcal

EP

pore-forming toxins, Nat Commun. 5 (2014) 4897. S.W. Gan, W. Surya, A. Vararattanavech, J. Torres, Two different conformations in

AC C

[46]

hepatitis C virus p7 protein account for proton transport and dye release, PLoS One. 9 (2014) e78494. [47]

K. Singkhamanan, B. Promdonkoy, T. Srikhirin, P. Boonserm, Amino acid residues in

the N-terminal region of the BinB subunit of Lysinibacillus sphaericus binary toxin play a critical role during receptor binding and membrane insertion, J. Invertebr. Pathol. 114 (2013) 65-70.

25

ACCEPTED MANUSCRIPT [48]

I. Gomez, J. Sanchez, C. Munoz-Garay, V. Matus, S.S. Gill, M. Soberon, A. Bravo,

Bacillus thuringiensis Cry1A toxins are versatile proteins with multiple modes of action: two distinct pre-pores are involved in toxicity, Biochem. J. 459 (2014) 383-396. [49]

E.W. Davidson, Binding of the Bacillus sphaericus (Eubacteriales: Bacillaceae) toxin

Entomol. 25 (1988) 151-157. [50]

R.H. Dadd, Alkalinity within the midgut of mosquito larvae with alkaline-active

SC

digestive enzymes, J. Insect Physiol. 21 (1975) 1847-1853. [51]

RI PT

to midgut cells of mosquito (Diptera: Culicidae) larvae: relationship to host range, J. Med.

P.J. Linser, K.E. Smith, T.J. Seron, M.N. Oviedo, Carbonic anhydrases and anion

[52]

M AN U

transport in mosquito midgut pH regulation, J. Exp. Biol. 212 (2009) 1662-1671. D.Y. Boudko, L.L. Moroz, P.J. Linser, J.R. Trimarchi, P.J. Smith, W.R. Harvey, In

situ analysis of pH gradients in mosquito larvae using non-invasive, self-referencing, pHsensitive microelectrodes, J. Exp. Biol. 204 (2001) 691-699.

T. Kunthic, B. Promdonkoy, T. Srikhirin, P. Boonserm, Essential role of tryptophan

TE D

[53]

AC C

EP

residues in toxicity of binary toxin from Bacillus sphaericus, BMB Rep. 44 (2011) 674-679.

26

ACCEPTED MANUSCRIPT Highlights •

Single mutations of aromatic residues, F311A and F315A, of BinB resulted in the loss of larvicidal activity against Culex quinquefasciatus larvae. Both F311 and F315 aromatic residues on the C-terminal domain of BinB are critical

EP

TE D

M AN U

SC

RI PT

for toxin insertion in membranes.

AC C