Pesticidal fusion protein improvements

Abstract

Methods of increasing the biological activity of toxins. Methods of increasing the biological activity of pesticide toxins through the incorporation of pro-regions into nucleic acid constructs for the production of said toxins.

Claims

1. A method for increasing the pesticidal activity of a recombinant toxin, the method comprising: expressing a nucleic acid construct in vitro encoding a recombinant toxin fusion protein comprising the recombinant toxin linked to a carrier protein and a pro-region of a native toxin, the nucleic acid construct comprising: (i) a nucleic acid sequence encoding the recombinant toxin; (ii) a nucleic acid sequence encoding the pro-region of a native toxin; and (iii) a nucleic acid sequence encoding the carrier protein selected from snowdrop lectin (GNA), garlic lectin Allium sativum, pea lectin Pisum sativum(P-lec), peanut lectin Arachis hypogaea, or french bean lectin (PHA, phytohaemagglutinin).

2. The method of claim 1, wherein the recombinant toxin is an arthropod toxin.

3. The method of claim 1, wherein the recombinant toxin is ω-ACTX-Hv1a.

4. The method of claim 1, wherein the recombinant toxin is δ-amaurobitoxin-PI1a.

5. The method of claim 1, wherein the pro-region is rich in acidic amino acid residues.

6. The method of claim 1, wherein the pro-region comprises the amino acid sequence EDTRADLQGGEAAEKVFRR (SEQ ID NO: 1).

7. The method of claim 1, wherein the pro-region comprises the amino acid sequence ISYEEGKELFQKER (SEQ ID NO: 2).

8. The method of claim 1, wherein the pro-region alters or improves the folding of the recombinant toxin.

9. The method of claim 1, wherein the carrier protein is snowdrop lectin (GNA).

10. The method of claim 1, wherein the pro-region is a sequence between a signal peptide sequence and a mature protein N-terminus of a native toxin; optionally wherein the pro-region is identifiable by the steps of: comparing a sequence for a protein toxin isolated from its source with a sequence predicted by a gene encoding it; identifying N-terminal sequences not present in the isolated protein toxin sequence; identifying a signal peptide sequence in the N-terminal sequence not present in the isolated protein toxin sequence; and identifying the pro-region sequence between the signal peptide sequence and the isolated protein toxin sequence N-terminus.

11. A recombinant toxin fusion protein produced by the method of claim 1.

12. A pesticide composition comprising the recombinant toxin fusion protein produced by the method of claim 1.

13. The pesticide composition of claim 12, wherein the composition comprises the recombinant toxin fusion protein in an amount of between 0.001% and 99% by weight; between 0.5% and 98% by weight; or between 1.0% and 95% by weight.

14. A process for preparing a pesticide composition comprising an admixture of a pesticidal effective amount of the recombinant toxin fusion protein produced by the method of claim 1, with one or more suitable carriers, diluents, adjuvants, preservatives, dispersants, solvents, or emulsifying agents.

15. A molluscicide bait composition comprising the recombinant toxin fusion protein produced by the method of claim 1.

16. A method for treating a pest infection of a plant comprising applying a quantity of the recombinant toxin fusion protein produced by the method of claim 1 to the plant or its locus of growth.

17. A nucleic acid construct encoding the recombinant toxin fusion protein of claim 1.

18. The nucleic acid construct of claim 17, wherein the toxin gene sequence is adjacent to the pro-region sequence.

19. The nucleic acid construct of claim 17, wherein the nucleic acid construct is an expression construct.

20. The nucleic acid construct of claim 17, further comprising a sequence encoding an affinity tag.

21. A transgenic plant or progeny thereof having increased biological activity resulting from the expression of the recombinant toxin fusion protein encoded by the nucleic acid construct of claim 17.

Description

DETAILED DESCRIPTION

(1) The present invention will now be described with reference to the following non-limiting examples and figures, which show:

(2) FIG. 1: Schematic of gene structure of toxins containing pro-regions

(3) (A) Schematic of gene structure of toxin proteins containing pro-regions. (B) Sequence of spider toxin (SED ID NO: 19), Hv1a. Boxed amino acid sequence corresponds to pro-region.

(4) FIG. 2: Expression and purification of recombinant Hv1a and pro-Hv1a toxin

(5) (A) SDS-PAGE gel (20% acrylamide) analysis showing purification of recombinant Strep-tagged Hv1a toxin from culture supernatant. Lanes 1 & 2 are GNA standards (0.5 and 0.25 μg, respectively) and lanes 3 & 4 are peak fractions (10 μl) following elution from Streptactin column with 2.5 mM desthiobiotin. (B) Tris-Tricine gel (15% acrylamide) analysis of recombinant pro-Hv1a, lanes 1 & 2 are peak fractions (10 μl) following elution from a nickel affinity column with 0.2 M imidazole. Arrow depicts major protein product predicted to be pro-Hv1a from which the histidine tag has been cleaved. (C) Western blot analysis of sample in (B) using anti-His antibodies.

(6) FIG. 3: SDS-PAGE gel analysis of lyophilised samples of purified pro-Hv1a/GNA, Hv1a/GNA and MODHv1a/GNA

(7) SDS-PAGE analysis (17.5% acrylamide gel) of purified recombinant Hv1a containing fusion protein, gel stained for total proteins with Coomassie Blue. Loading as follows: Lane 1: pro-Hv1a/GNA; Lane 2: Hv1a/GNA; Lane 3: MODHv1a/GNA; Lanes 4-6: GNA standards of 1, 2 and 4 μg, respectively; Lanes 7-9: 12.5, 25 and 50 μg lyophilised purified pro-Hv1a/GNA (to enable quantification of fusion protein content).

(8) FIG. 4: Injection toxicity of recombinant pro-Hv1a, pro-Hv1a/GNA and Hv1a/GNA to Mamestra brassicae

(9) Percentage survival of 3rd-5.sup.th instar Mamestra brassicae larvae following the injection of different doses of recombinant pro-Hv1a, pro-Hv1a/GNA or Hv1a/GNA. (A) Percentage survival of 5.sup.th instar larvae following injection of various doses of pro-Hva1. (B) Percentage survival of 3.sup.rd-4.sup.th instar larvae following injection of various doses of pro-Hv1a/GNA (Dose A) or Hv1a/GNA (Dose B). (C) Percentage survival of 5.sup.th instar larvae following injection of various doses of pro-Hv1a/GNA (Dose A) and Hv1a/GNA (Dose B).

(10) FIG. 5: Ingestion toxicity of recombinant pro-Hv1a, Hv1a/GNA, pro-Hv1a/GNA and GNA to Mamestra brassicae

(11) Percentage survival of 3.sup.rd instar Mamestra brassicae larvae following ingestion of a single 2 μl droplet containing 20 μg of purified pro-Hv1a, Hv1a/GNA, pro-Hv1a/GNA or GNA. Control larvae were fed on a droplet containing no added protein (n=10 per treatment).

(12) FIG. 6: Ingestion toxicity of recombinant pro-Hv1a, pro-Hv1a/GNA, Hv1a/GNA or GNA to Acyrthosiphon pisum and Sitobion avenae

(13) Percentage survival of (A) Acyrthosiphon pisum (pea aphids) and (B) Sitobion avenae (cereal aphids) with artificial diets containing 0.05-0.75 mg/ml purified recombinant pro-Hv1a, pro-Hv1a/GNA, Hv1a/GNA or GNA.

(14) FIG. 7: Injection toxicity of recombinant pro-Hv1a/GNA and Hv1a/GNA to Deroceras reticulatum

(15) Percentage survival of Deroceras reticulatum (200 mg±40 mg) injected with 100, 50 or 25 μg of Hv1a/GNA or pro-Hv1a/GNA. n=18 for control treatment; n=10 for 100 μg dose; and n=8 for 50 and 25 μg doses.

(16) FIG. 8: Expression and purification of recombinant PI1a and Ao1bPro-PI1a toxin

(17) (A) PI1a toxin derived from a construct encoding the mature toxin sequence separated on “normal” SDS-PAGE; M indicates marker, loadings of PI1a are 5 and 10 μg. (B) PI1a toxin (5 μg) separated on SDS-PAGE after denaturation by 6 M urea. (C) Recombinant PI1a toxin derived from a construct containing the pro-region designated Ao1b on SDS-PAGE, loadings of Ao1bPro-PI1a are 2.5 μg. (D) Western blotting of purified Ao1bPro-PI1a (25, 50 & 100 ng) using anti-His antibodies.

(18) FIG. 9: Characterisation of purified recombinant PI1a/GNA fusion proteins by SDS-PAGE

(19) (A) SDS-PAGE analysis of PI1a/GNA fusion protein (10 μg) and GNA standard (5 μg). (B) Deglycosylation of PI1a/GNA fusion protein using PNGase F (band indicated by open arrowhead), GNA standard (5 μg). (C) SDS-PAGE analysis of Ao1bPro-PI1a/GNA (1, 2 and 4 μg), GNA standard (5 μg). (D) SDS-PAGE analysis of Hv1aPro-PI1a/GNA (1, 2 and 4 μg), GNA standards (1, 2 and 4 μg).

(20) FIG. 10: Injection toxicity of recombinant PI1a and Ao1bPro-PI1a to Mamestra brassicae

(21) (A) Percentage survival of 5.sup.th instar Mamestra brassicae larvae following injection of different doses of purified recombinant PI1a. (B) Percentage survival of 5.sup.th instar Mamestra brassicae larvae following injection of different doses of purified recombinant PI1a (Dose A) or Ao1bPro-PI1a (Dose B).

(22) FIG. 11: Injection toxicity of recombinant PI1a/GNA, Ao1bPro-PI1a/GNA and Pro-HV1a-PI1a/GNA to Mamestra brassicae

(23) Percentage survival of 5.sup.th instar Mamestra brassicae larvae following injection of different doses of purified recombinant PI1a/GNA (A), Ao1b-ProPI1a/GNA (B) or Pro-Hv1a-PI1a/GNA (C). n=20 per treatment.

(24) FIG. 12: Ingestion toxicity of recombinant PI1a/GNA, Ao1bPro-PI1a/GNA and Hv1aPro-PI1a/GNA to Mamestra brassicae

(25) Percentage survival of 3.sup.rd instar Mamestra brassicae larvae following ingestion of a single 2 μl droplet containing 20 μg of purified PI1a/GNA (A), Ao1bPro-PI1a/GNA (B) or Hv1aPro-PI1a/GNA (C) fusion proteins. Controls in all cases were sucrose alone (no added protein); 30 μg of either PI1a toxin (mature or modified form) or GNA.

ABBREVIATIONS

(26) BB: Binding buffer

(27) ECL: Enhanced chemiluminescence

(28) HRP: Horseradish peroxidase

(29) PBS: Phosphate buffered saline

(30) SDS-PAGE: Sodium dodecyl sulphate polyacrylamide gel electrophoresis

(31) YPG: Yeast extract peptone glycerol

(32) Materials and Methods

(33) Cloning of Recombinant Hv1a, Pro-Hv1a and Pro-Hv1a/GNA Fusion Protein

(34) A synthetic gene encoding the mature Hv1a amino acid sequence was assembled using a series of overlapping oligonucleotides, with codon usage optimised for expression in yeast (Fitches et al., 2012). To create an expression construct coding for the mature Hv1a peptide the coding sequence was amplified by PCR using primers with PstI and SalI sites and purified from excised gel slices using a QiaQuick Gel Extraction Kit (Qiagen) as described in the manufacturer's protocol. The extracted DNA fragment was digested (PstI and SalI) and ligated into similarly digested yeast expression vector pGAPZαB (Invitrogen) that had been previously modified to contain a 5′ Strep tag in frame with the yeast α-factor pre-pro-sequence. The resulting plasmid was transformed into electrocompetent E. coli cells and selected clones were checked for the correct assembly of the construct by gel electrophoresis and DNA sequencing.

(35) The pro-Hv1a coding sequence was amplified by PCR using primers with Ps1I and XbaI sites (Forward: TACTGCAGCAGAAGATACTAGAGCT (SEQ ID NO: 3) and Reverse: ATTCTAGAATCACATCTCTTAAC (SEQ ID NO: 4). Gel extracted products were digested with Ps1I and XbaI and ligated into similarly digested yeast expression vector pGAPZaB. The resulting recombinant plasmid was transformed into E. coli and selected clones were checked for correct assembly of the construct by gel electrophoresis and DNA sequencing.

(36) To produce the pro-Hv1a/GNA construct, the pro-Hv1a coding sequence was amplified by PCR using primers with PstI and NatI sites (Forward: TACTGCAGCAGAAGATACTAGAGCT (SEQ ID NO: 3)and Reverse: ATGCGGCCGCATCACATCTCTTAAC (SEQ ID NO: 5)) and purified by gel electrophoresis as described above. Following restriction by PstI and NotI, the PCR product was ligated into a previously generated pGAPZaB plasmid containing the mature GNA coding sequence digested with the same enzymes. Selected clones containing the expression vector encoding the pro-Hv1a/GNA fusion protein were verified by DNA sequencing.

(37) The sequences of the Hv1a constructs are shown below:

(38) Native Hv1a:

(39) TABLE-US-00003 (SEQ ID NO: 6) SPTCIPSGQPCPYNENCCSQSCTFKENENGNTVKRCD

(40) Recombinant Hv1a (alpha factor signal sequence, Hv1a toxin, Strep tag green highlighted region, no pro-region, no carrier):

(41) TABLE-US-00004 (SEQ ID NO: 7) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAWSHPQFEKGL QSPTCIPSGQPCPYNENCCSQSCTFKENENGNTVKRCD

(42) Recombinant pro-Hv1a (alpha factor signal sequence, Hv1a toxin, pro-region, no carrier, (His).sub.6 tag):

(43) TABLE-US-00005 (SEQ ID NO: 8) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAEDTRADLQG GEAAEKVFRRSPTCIPSGQPCPYNENCCSQSCTFKENENGNTVKRCDALE QKLISEEDLNSAVDHHHHHH

(44) Recombinant Hv1a/GNA (alpha factor signal sequence, Hv1a toxin, no pro-region, carrier):

(45) TABLE-US-00006 (SEQ ID NO: 9) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFI NTTIASIAAKEEGVSLEKREAEAAASPTCI PSGQ PCPYN ENCCSQSCTFKEN ENGNTVKRCDAAADNIL YSGETLSTGEFLNYG SFVFI MQEDCNLVL YDVDKPIWATNT GGLSRSCFLSMQTDGNLVVYNPSN KPIWASNTGGQNGNYVCILQKDRNVVIYGTDRWATGVD

(46) Recombinant pro-Hv1a/GNA (alpha factor signal sequence, Hv1a toxin, pro-region, carrier, (His).sub.6 tag):

(47) TABLE-US-00007 (SEQ ID NO: 10) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNST 5 NNGLLFINTTIASIAAKEEGVSLEKREAEAAAEDTRADLQG GEAAEKVFRRSPTCIPSGQPC PYN ENCCSQSCTFKEN ENGNTVKRCDAAA DN IL YSG ETLSTGEFLNYGSFVFI MQEDCNLVL YDVDKPIWATNTGGLSR SCFLSMQTDGNLVVYNPSNKPIWASNTGGQNGNYVCILQKDRNVVIYGTD RWATGVDHHHHHH
Cloning of Recombinant PI1a and PI1a/GNA Fusion Proteins

(48) A double stranded DNA incorporating a sequence encoding the mature PI1a toxin (P83256), with codon usage optimised for yeast, was designed by the inventors and synthesised and supplied by ShineGene Molecular Biotech, Inc. (Shanghai 201109, China; www.synthesisgene.com) in the vector pUC57. Other oligonucleotides were supplied by Sigma Chemical Co.

(49) The PI1a coding sequence was transferred from pUC57 to the yeast expression vector pGAPZαB (Invitrogen) by digestion with PstI and XbaI, isolation of the coding sequence fragment by agarose gel electrophoresis, followed by ligation to pGAPZαB that had been digested with the same enzymes. DNA fragments were purified from excised gel slices using a QiaQuick Gel Extraction Kit (Qiagen). The resulting recombinant plasmid was cloned using standard protocols by transformation of electrocompetent E. coli cells. Selected clones were checked for correct assembly of the construct by DNA sequencing. To produce the modified construct for expression of PI1a two complementary synthetic oligonucleotides encoding the pro-region from U3-agatoxin-Ao1b (Q5Y4V7) were assembled and inserted into the PstI site of the PI1a expression construct. Correct assembly of the construct (ProAo1b-PI1a) was checked by DNA sequencing.

(50) To produce a construct encoding the PI1a/GNA fusion protein, the mature PI1a coding sequence from a verified expression construct in pGAPZαB was excised by digestion with PstI and NotI and purified by agarose gel electrophoresis. A pGAPZαB plasmid containing the fusion protein construct Hv1a/GNA (Fitches et al., 2012) was digested with PstI and NotI to remove the Hv1a coding sequence and purified by agarose gel electrophoresis. The Hv1a coding region was subsequently replaced with PI1a by ligating the purified fragments and cloning the resulting recombinant plasmid. To produce the modified expression construct for PI1a/GNA containing the pro-region from U3-agatoxin-Ao1b, the PI1a/GNA expression construct was modified as described above; in addition, a pro-region from the pro-Hv1a toxin was also used in a further construct designated as Pro-Hv1a-PI1a. Selected clones containing the expression vector encoding the PI1a/GNA fusion proteins were verified by DNA sequencing. All DNA sequencing was carried out using Applied Biosystems ABI Prism 3730 automated DNA sequencers by DBS Genomics, School of Biological and Biomedical Sciences, Durham University, UK.

(51) Native PI1a:

(52) TABLE-US-00008 (SEQ ID NO: 11) GCLGEGEKCADWSGPSCCDGFYCSCRSMPYCRCRNNS

(53) Recombinant PI1a (alpha factor signal sequence, PI1a toxin, no pro-region, no carrier, (His).sub.6 tag):

(54) TABLE-US-00009 (SEQ ID NO: 12) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAGCLGEGEKC ADWSGPSCCDGFYCSCRSMPYCRCRNNSALEQKLISEEDLNSAVDHHHHH H

(55) Recombinant Ao1bPro-PI1a (alpha factor signal sequence, pro-region [Ao1b], PI1a toxin, no carrier, +(His).sub.6 tag):

(56) TABLE-US-00010 (SEQ ID NO: 13) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAISYEEGKEL FQKERGCLGEGEKCADWSGPSCCDGYCSCRSMPYCRCRNNSALEQKLISE EDLNSAVDHHHHHH

(57) Recombinant PI1a/GNA (alpha factor signal sequence, PI1a toxin, no pro-region, carrier, (His).sub.6 tag):

(58) TABLE-US-00011 (SEQ ID NO: 14) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAGCLGEGEKC ADWSGPSCCDGFYCSCRSMPYCRCRNNSAAADNIL YSGETLSTGEFLNYG SFVFIMQEDCNLVL YDVDKPIWATN TGGLSRSCFLSMQTDGNLVVYNPSN KPIWASNTGGQNGNYVCI LQKDRNVVIYGTDRWATGVDHHHHHH

(59) Recombinant Ao1bPro-PI1a/GNA (alpha factor signal sequence, pro-region [Ao1b], PI1a toxin, carrier, (His).sub.6 tag):

(60) TABLE-US-00012 (SEQ ID NO: 15) MRFPSI FTAVLFAASSALAAPVNTTTEDETAQI PAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAISYEEGKEL FQKERGCLGEGEKCADWSGPSCCDGCSCRSMPYCRCRNNSAAADNILYSG ETLSTGEFLNYGSFVFIMQEDCNLVL YDVDKPIWATNTGGLSRSCFLSMQ TDGNLVVYNPSNKPIWASNTGGQNGNYVCILQKDRNVVIYGTDRWATGVD HHHHHH

(61) Recombinant Hv1aPro-PI1a/GNA (alpha factor signal sequence, pro-region [Hv1a], PI1a toxin, carrier, (His).sub.6 tag):

(62) TABLE-US-00013 (SEQ ID NO: 16) MRFPSIFTAVLFAASSALAAPVNTTTEDETAQIPAEAVIGYSDLEGDFDV AVLPFSNSTNNGLLFINTTIASIAAKEEGVSLEKREAEAAAEDTRADLQG GEAAEKVFRRGCLGEGEKCADWSGPCCDGFYCSCRSMPYCRCRNNSAAAD NIL YSGETLSTGEFLNYGSFVFIMQEDCNL VL YDVDKPIWATNTGGLSRS CFLSMQTDGNLVVYNPSNKPIWASNTGGQNGNYVCILQKDRNVVIYGTDR WATGVDHHHHHH
Expression of Hv1a, Pro-Hv1a, Hv1a/GNA and Pro-Hv1a/GNA Fusion Proteins in Yeast

(63) pGAPZαB plasmids containing Hv1a, pro-Hv1a, Hv1a/GNA and pro-Hv1a/GNA sequences were amplified in E. coli, purified and linearised with BlnI (Takara). Linearised plasmids were transformed into Pichia pastoris strain SMD1168H (Invitrogen) using the EasyComp Transformation kit (Invitrogen) as described in the manufacturer's protocol. Transformed yeast clones were selected on YPG agar plates (1% yeast extract (w/v), 2% peptone (w/v), 4% glycerol (v/v), 1.5% agar (w/v)) containing zeocin (100 mg/ml). Selected clones were checked for expression by analysis of culture supernatants from small-scale shake flask cultures (10 ml) grown for 2-3 days in YPG-zeocin media at 30° C. Supernatant samples were separated by SDS-polyacrylamide gel electrophoresis; gels were blotted onto nitrocellulose and probed with anti-(His).sub.6 primary antibodies (BioRad) or anti-Strep antibodies, or for Hv1a/GNA and pro-Hv1a/GNA blots were probed with anti-GNA primary antibodies, followed by washing, probing with HRP-conjugated secondary antibodies (BioRad), and detection of bound antibodies by ECL, as described previously (Fitches et al., 2001; 2012).

(64) For protein production selected P. pastoris clones containing the integrated Hv1a, pro-Hv1a, Hv1a/GNA and pro-Hv1a/GNA expression cassettes were grown in either a 7.5 L BioFlo 110 bench-top fermenter (New Brunswick Scientific) or a 5 L Bio-Controlly ADI1010 bench-top fermenter (APPLIKON BIOTECHNOLOGY, Holland). YPG cultures (200 ml) of transformed P. pastoris were grown for 2-3 days at 30° C. with shaking (no zeocin antibiotics) before inoculating 2.5 L of sterile minimal media supplemented with PTM1 salts. Cultivation at 30° C., 30% dissolved oxygen, pH 4.5-5.0 with continuous agitation was carried out with a ramped glycerol feed (5-10 ml/h; total 1.25 l) over a period of 4 days. Culture supernatant was subsequently separated from cells by centrifugation (20 min, 8000 rpm; 4° C.), clarified by filtration through 2.7 μM and 0.7 μM glass fibre filters (GFD and GFF; Whatmann). For Hv1a only supernatant was adjusted to 50 mM phosphate buffer containing 0.3 M sodium chloride at pH 8.0 by adding 4× concentrated stock. Recombinant Hv1a was purified on streptactin columns (1 ml) with a flow rate of 0.5 ml/min. Columns were equilibrated in 50 mM phosphate buffer containing 0.3 M sodium chloride at pH 8.0. Strep-tagged Hv1a was eluted from columns using 2.5 mM desthiobiotin (in phosphate buffer; pH 8.0). For all other proteins supernatants were adjusted to 0.02 M sodium phosphate buffer, 0.4 M sodium chloride, pH 7.4 by adding 4× concentrated stock (4× Binding buffer [BB]). Recombinant pro-Hv1a Hv1a/GNA and pro-Hv1a/GNA were purified by nickel affinity chromatography on 5 ml HisTrap crude nickel columns (GE Healthcare) with a flow rate of 2 ml/min. After loading, the columns were washed with 1×BB (50 mM sodium phosphate; 0.4 M sodium chloride) and then with BB containing 0.025 M imidazole, and finally bound recombinant proteins were eluted with BB containing 0.2 M imidazole. In all cases eluted proteins were then checked for purity by SDS-PAGE, dialysed against deionised water using multiple changes to remove all small molecules, and freeze-dried. Concentrations of recombinant proteins were estimated by comparison to known amounts of GNA standards run on SDS-PAGE gels or by BCA analysis using a BCA™ Protein Assay Kit (Thermo Scientific).

(65) Expression of PI1a and PI1a/GNA Fusion Proteins in Yeast

(66) pGAPZαB plasmids containing the PI1a and PI1a/GNA expression constructs were amplified in E. coli, purified and linearised with BlnI. Linearised plasmids were transformed into Pichia pastoris strain SMD1168H (Invitrogen) using the EasyComp Transformation kit (Invitrogen) as described in the manufacturer's protocol. Transformed yeast clones were plated and selected on YPG agar plates (1% yeast extract (w/v), 2% peptone (w/v), 4% glycerol (v/v), 1.5% agar (w/v)) containing zeocin (100 mg/ml). Selected clones (at least 10 for each construct) were checked for expression of recombinant proteins by analysis of culture supernatant from small-scale shake flask cultures grown for 2-3 days in YPG-zeocin media at 30° C. Samples of supernatant were separated by SDS-polyacrylamide gel electrophoresis; gels were blotted onto nitrocellulose and probed with anti-(His).sub.6 primary antibodies (BioRad) or anti-GNA primary antibodies, followed by washing, probing with HRP-conjugated secondary antibodies (BioRad), and detection of bound antibodies by ECL.

(67) Selected clones of P. pastoris containing the integrated PI1a and PI1a/GNA constructs were grown in a 5 L Bio-Controlly ADI1010 bench-top fermenter (Applikon Biotechnology, Holland). For fermentation, two 100 ml YPG cultures of P. pastoris containing toxin or fusion genes were grown for 2-3 days at 30° C. with shaking, prior to being used to inoculate 2.5 L of sterile minimal media supplemented with PTM1 salts. Cultivation at 30° C., 30% dissolved oxygen, pH 4.5-5.0 with continuous agitation was continued with a ramped glycerol feed (5-10 ml/h) over a period of 4 days. Culture supernatant was separated from cells by centrifugation (20 min at 5000 g), and adjusted to 0.02 M sodium phosphate buffer, 0.4 M sodium chloride, pH 7.4 by adding 5× concentrated stock. Recombinant proteins were purified by nickel affinity chromatography on 5 ml HisTrap crude nickel columns (GE Healthcare) with a flow rate of 2 ml/min. After loading, the columns were washed with 0.02 M sodium phosphate buffer, 0.4 M sodium chloride pH 7.4 and the bound proteins were eluted with 0.2 M imidazole in the same buffer. Eluted proteins were checked for purity by SDS-PAGE, dialysed against deionised water using multiple changes to remove all small molecules, and freeze-dried. Concentrations of recombinant proteins were estimated by comparison to known amounts of GNA standards run on SDS-PAGE gels or by BCA analysis using a BCA™ Protein Assay Kit (Thermo Scientific).

(68) Insect Bioassays

(69) 3.sup.rd-5.sup.th instar Mamestra brassicae larvae (approximately 30-55 mg in weight) were used for injection bioassays. Larvae were injected with varying doses of Hv1a, pro-Hv1a, pro-Hv1a/GNA, PI1a, Ao1bProPI1a, PI1a/GNA, Ao1bPro-PI1a/GNA, or Hv1aPro-PI1a/GNA (n=20 per dose) in 5 μl of PBS (phosphate buffered saline; 0.15 M NaCl, 0.015 M sodium-phosphate buffer, pH 7.2). Larvae for controls were injected with 5 μl 1×PBS. Paralysis and mortality was recorded 12-96 h after injection.

(70) Droplet-feeding assays were carried out to assess the oral activity of Hv1a/GNA, pro-Hv1a/GNA, PI1a, Ao1bProPI1a, PI1a/GNA, Ao1bPro-PI1a/GNA, or Hv1aPro-PI1a/GNA towards third to fifth instar larvae of M. brassicae. Larvae were starved for approximately 24 h before feeding in order to encourage droplet consumption. Larvae were fed with a 2 μl droplet containing 20 μg of the above fusion proteins, 30 μg of toxins or 30 μg of GNA, in 1×PBS solution containing 10% sucrose (w/w). Control larvae were fed on PBS/sucrose droplets containing no added protein. Treated larvae were placed on standard artificial diet after consumption of the droplet.

(71) The insecticidal effects of Hv1a/GNA and pro-Hv1a/GNA to Acyrthosiphon pisum (pea aphids) and Sitobion avenae (cereal aphids) was assayed by feeding 100 μl liquid artificial diet containing known concentrations of fusion proteins (Prosser and Douglas, 1992), using double parafilm sachets (diet droplet in the middle) to deliver diet to insects. The experiment used 1-2 day-old aphids and survival was assessed daily for six days.

(72) Injection Bioassays: Deroceras reticulatum (Mollusc Grey Field Slug)

(73) Hv1a/GNA and pro-Hv1a/GNA were tested for activity against adult slugs (Deroceras reticulatum) by injection into adult slugs (0.2-0.3 g). Slugs were chilled at 4° C. (for approximately 15 minutes) prior to injection of 25 μg, 50 μg or 100 μg of purified fusion proteins resuspended in 20 μl PBS. Mortality was assessed daily for 7 days.

(74) Statistical Analysis

(75) Survival data were analysed using Kaplan-Meier survival analysis, using Prism (v5) software. All other data analysis was carried out using Origin 8.5 graphing and data analysis software. ANOVA analysis (with Bonferroni-Dunn post-hoc tests) was carried out to determine any significant differences between treatments in the parameters measured.

(76) Results

(77) The present inventors have conducted experiments to investigate the effect of inclusion of a pro-region in an expression construct for a toxin on the biological activity of said toxin.

(78) Experiments Investigating the Toxicity of Hv1a

(79) Introduction

(80) To investigate the effect of inclusion of a pro-region on the toxicity of recombinant toxins, ω-Hexatoxin-Hv1a was used. ω-Hexatoxin-Hv1a is a toxin isolated from the funnel-web spider Hadroncyhe versuta. ω-Hexatoxin-Hv1a (or ω-ACTX-Hv1a) is a calcium channel antagonist and it has previously been shown that ω-ACTX-Hv1a can block invertebrate but not vertebrate calcium channels.

(81) Although it has been shown that ω-ACTX-Hv1a can be used on its own as a pesticide when applied topically to caterpillars (Khan et al., 2006), no further evidence for insecticidal activity of the peptide alone has been reported. In patent application PCT/GB2012/000287 the present inventors demonstrated that the toxicity of a recombinant toxin (ω-ACTX-Hv1a) expressed in Pichia pastoris could be enhanced by expressing the protein in fusion with the plant lectin GNA, which had previously been shown to cross the gut epithelium and deliver ‘passenger’ peptides from the gut to the circulatory system of invertebrate animals.

(82) To further investigate how the potency of toxins expressed in vitro might be improved, the present inventors analysed the DNA sequences of the genes encoding arthropod toxins. The arthropod toxins utilised in PCT/GB2012/000287 are small, cysteine-rich proteins belonging to several superfamilies of protein sequences (which include toxins from organisms other than arthropods). The encoding genes include two sequences that are not present in the final protein product; a predicted N-terminal signal peptide that is removed during translation and a predicted pro-region, between the signal peptide and the final sequence of the protein as isolated (see FIG. 1A).

(83) The present inventors investigated the effect of including the pro-region in the expression construct on the overall toxicity of the recombinant protein.

(84) In the first instance, ω-ACTX-Hv1a was used, as this toxin contains a predicted pro-region in its gene sequence (see FIG. 1B).

(85) Synthetic Gene Constructs, Expression and Purification of Recombinant Hv1a, Pro-Hv1a, Hv1a/GNA and Pro-Hv1a/GNA

(86) Recombinant Hv1a, pro-Hv1a, Hv1a/GNA and pro-Hv1a/GNA fusion protein constructs were synthesised based on the vector pGAPZαB, which possesses a strong constitutive promoter (GAPDH) to direct target gene expression. The pGAPZ vectors are integrating vectors in the expression host Pichia pastoris and selected transformants contain the expression construct integrated into the host genome. The constructs for expressing recombinant Hv1a and pro-Hv1a contained a predicted amino acid sequence corresponding to the published sequence for the toxin. For Hv1a, the mature peptide was cloned in frame with the yeast a-factor prepro-sequence and 8 amino acids encoding a Strep tag (i.e. WSHPQFEK (SEQ ID NO: 17) linked to the N-terminus of mature Hv1a sequence via a 3 amino acid linker ‘GLQ’). For Pro-Hv1a, the N-terminal pro-region was arranged in-frame C-terminal to a sequence encoding the yeast a-factor prepro-sequence, and the construct also contained sequences encoding the myc epitope and (His).sub.6 tag, supplied by the vector, at the C-terminus of the predicted product. For Hv1a/GNA, the mature toxin sequence was fused to the N-terminus of a coding sequence corresponding to residues 1-105 of mature snowdrop lectin (GNA) via a 3 amino acid (AAA) linker peptide. For pro-Hv1a/GNA the synthetic pro-Hv1a coding sequence was fused to the N-terminus of a coding sequence corresponding to residues 1-105 of mature snowdrop lectin (GNA) via a 3 amino acid (AAA) linker peptide. Both fusion protein constructs were also arranged in-frame with the yeast α-factor prepro-sequence, and C-terminal to a sequence encoding (His).sub.6 tag, supplied by the vector. The constructs were assembled by ligation of endonuclease digested DNA and were checked by DNA sequencing after cloning.

(87) Sequence verified clones, containing recombinant Hv1a, pro-Hv1a, Hv1a/GNA or pro-Hv1a/GNA were transformed into the P. pastoris protease-deficient strain SMD1168H and selected using the antibiotic zeocin. Selected clones were grown in shake flask cultures for 3-4 days at 30° C. and culture supernatants were analysed for the expression of recombinant proteins using western blotting. This enabled highly-expressing clones to be selected for production by bench-top fermentation. The majority of the analysed transformed yeast clones showed evidence of protein expression as judged by the presence of immunoreactive bands with the expected size on western blots (results not shown).

(88) Fermentation of the selected clones was carried out in bioreactors under controlled environmental conditions. The use of the pGAP alpha factor secretory signal that directs the secretion of expressed proteins out of the cells and into the growth media, enabled subsequent purification of recombinant proteins from fermented culture supernatants. Supernatants were obtained by centrifugation, clarified by filtration and recombinant proteins were subsequently purified by affinity chromatography (Streptactin for Hv1a and nickel affinity for pro-Hv1a and pro-Hv1a/GNA). Eluted peaks containing target proteins were desalted by dialysis and lyophilised. For yields of recombinant proteins, Hv1a was produced at approximately 5-10 mg/L culture supernatant; pro-Hv1a was produced at approximately 40 mg/L, as estimated by BCA quantification and Hv1a/GNA at approximately 40 mg/L and pro-Hv1a/GNA at approximately 21 mg/L, as estimated by semi-quantitative SDS-PAGE.

(89) As shown in FIG. 2A (lanes 3 and 4), 5′ Strep tagged mature purified Hv1a separated on SDS-PAGE gels as a single protein of approximately 6.5 kDa, comparable to the predicted molecular mass of 5.47 kDa (gels were not optimised for the separation of low molecular weight proteins). Purified recombinant pro-Hv1a was separated using Tris-Tricine SDS-PAGE gels and analysed by both staining for total protein and western blotting using anti-His antibodies (FIGS. 2B & C). The recombinant toxin pro-Hv1a gave a major protein band at approximately 4 kDa and further weaker bands in molecular mass range 6-16 kDa on Tris-Tricine gels. The dominant band of approximately 4 kDa is not immunoreactive with anti-(His).sub.6 antibodies (FIG. 2C) and the molecular mass is consistent with the predicted mass (4.06 kDa) following cleavage of the C-terminal tag region. Pull down of non-His tagged proteins along with His-tagged proteins from nickel affinity columns has previously been observed for other recombinant proteins recovered from Pichia supernatants. The 10 kDa protein that shows positive immunoreactivity with anti-(His).sub.6 antibodies corresponds to the predicted mass (9 kDa) for recombinant Hv1a including the pro-region, suggesting that cleavage of the pro-region is incomplete during processing by yeast cells. The 14 kDa band also immunoreactive with anti-(His).sub.6 antibodies may represent a dimeric form of Hv1a given that the predicted mass of Hv1a containing the C-terminal His region but no pro-region is 6.74 kDa.

(90) Lyophilised samples of purified pro-Hv1a/GNA, Hv1a/GNA and MODHv1a/GNA were analysed on SDS-PAGE gels (FIG. 3). Two bands of approximately 19 kDa and 14 kDa were observed for all fusion proteins. The predicted molecular mass for Hv1a/GNA is 16.27 kDa, and for pro-Hv1a/GNA without the pro-region but containing a (His).sub.6 tag is 16.95, both slightly less than the observed 19 kDa band. However, the identical separation of pro-Hv1a/GNA and Hv1a/GNA on SDS-PAGE gels suggests that the pro-region has been cleaved from pro-Hv1a/GNA during processing by P. pastoris cells. The predicted mass for intact MODHv1a/GNA is 17.09 kDa and correspondingly this protein runs as a slightly larger protein as compared to Hv1a/GNA and pro-Hv1a/GNA due to the presence of an additional histidine tag that is not present in Hv1a/GNA. In all cases the smaller 14 kDa band is immunoreactive with GNA antibodies (results not shown) and corresponds in size to GNA from which the Hv1a toxin has been cleaved. As observed previously for Hv1a/GNA (Fitches et al., 2012), the ratio of intact pro-Hv1a/GNA fusion protein to cleaved GNA was estimated as approx. 1:1 as judged by Coomassie blue staining on SDS-PAGE gels, whereas modification of the Hv1a sequence in MODHv1a/GNA results in a higher recovery of intact fusion protein (ratio intact:cleaved 2:1). Quantification of the Hv1a/GNA fusion proteins was based on comparative band intensity with GNA standards of known concentration as shown in FIG. 3.

(91) Injection Toxicity of Recombinant Hv1a, Pro-Hv1a, Hv1a/GNA and Pro-Hv1a/GNA Fusion Protein to Cabbage Moth (Mamestra brassicae) Larvae

(92) Injections of Hv1a at doses of up to 100 μg of the recombinant toxin did not result in any larval mortality with survival comparable to controls (n=40; survival >90%). This demonstrated that the expression of mature Hv1a peptide without an N-terminal pro-region resulted in the production of biologically inactive toxin, suggesting that the toxin was incorrectly processed and/or folded during synthesis by yeast cells. By contrast, injections of newly eclosed 3.sup.rd-4.sup.th (˜30-40 mg) and 5.sup.th instar (˜45-55 mg) M. brassicae larvae with either pro-Hv1a, Hv1a/GNA, MODHv1a/GNA or pro-Hv1a/GNA led to significant larval mortality.

(93) As shown in FIG. 4A the effects of recombinant pro-Hv1a after injection into 5.sup.th instar larvae were dose dependent. Injection doses of 10 μg and 20 μg pro-Hv1a/insect resulted in complete mortality 24 hours post injection and injection of 5 μg/insect resulted in 80% mortality 24 hours after injection. At the lowest dose of 1.25 μg/insect, all insects displayed flaccid paralysis and a temporary absence of feeding although some paralysed insects recovered and were able to resume feeding 2-3 hours after injection. From these assays, the estimated LD.sub.50 (lethal dose; 48 hours) for the recombinant pro-Hv1a was 25 μg/g insect. As summarised in Table 1 this is approximately 3-fold lower than that previously published for recombinant Hv1a toxin produced in E. coli, whose LD.sub.50 (72 hours) was ˜69 μg/g insect.

(94) As shown in FIGS. 4B & C injections of 3-4.sup.th and 5.sup.th instar M. brassicae larvae with pro-Hv1a/GNA demonstrated increased toxicity as compared to Hv1a/GNA. For example injections of 10 μg Hv1a/GNA into 5.sup.th instar larvae resulted in 50% mortality after 72 hours as compared to 100% mortality recorded 24 hours after injection of the same dose of pro-Hv1a/GNA. Significant larval mortality (75%) was observed at a pro-Hv1a/GNA dose of 2.5 μg/insect whereas injections of 5 or 10 μg Hv1a/GNA did not result in any significant levels of mortality for 5.sup.th instar larvae. Similar results were observed following the injection of smaller 3-4.sup.th instar larvae (FIG. 4B). As shown in Table 1 LD.sub.50 values estimated for pro-Hv1a/GNA were some 10-fold lower (4.6 μg/g insect) as compared to an LD.sub.50 of 55 μg/g insect for Hv1a/GNA. This demonstrates that the addition of the pro-region to the Hv1a/GNA construct results in the production of fusion protein that is significantly more toxic as compared to protein derived from a construct encoding the mature toxin sequence fused to GNA. More surprisingly, the LD.sub.50 value of 4.6 μg/g insect calculated for pro-Hv1a/GNA is some 5-fold lower than 25 μg/g insect estimated for pro-Hv1a. This shows that linkage of the pro-Hv1a to GNA results in a protein with higher biological activity than either pro-Hv1a alone, or Hv1a/GNA without the pro-region. The LD.sub.50 value of 4.6 μg/g insect calculated for pro-Hv1a/GNA is also over 2-fold lower than the literature value for native Hv1a (as opposed to recombinant Hv1a; see table 1). This literature value represents the LD.sub.50 value in Heliothis, a species of the same insect order as M. brassicae. We hypothesise that fusing GNA to pro-Hv1a further acts to facilitate correct processing and folding of the Hv1a toxin by P. pastoris cells, allowing for further increases in biological activity to be attained. Injections of GNA alone at up to 40 μg/insect do not result in mortality of M. brassicae larvae.

(95) TABLE-US-00014 TABLE 1 Toxicity of recombinant toxins and fusion proteins in injection bioassays with Mamestra brassicae larvae Hv1a Pro- Hv1a (E. (literature) Hv1a Pro-Hv1a Hv1a/GNA Hv1a/GNA coli) LD.sub.50 12 μg/g >1000   25 μg/g 55 μg/g 4.6 μg/g 69 μg/g (Heliothis μg/g (48 h) (72 h) (48 h) (72 h) sp.)* <25 μg/g (72 h) *Data not available for M. brassicae
Ingestion Toxicity of Recombinant Hv1a, Pro-Hv1a and Pro-Hv1a/GNA Fusion Protein to Cabbage Moth (M. brassicae) Larvae

(96) The oral activity of proHv1a/GNA and Hv1a/GNA was assessed by feeding 2 μl droplets containing 20 μg of fusion protein to newly eclosed third instar M. brassicae larvae. Control treatments were 20 μg of either GNA or pro-Hv1a, in addition to a no-added protein control group. As shown in FIG. 5 and Table 2, significant effects were observed only for larvae fed on pro-Hv1a/GNA, with 90% mortality recorded 5 days after the ingestion of a single droplet of fusion protein. By contrast, mortality was only 30% for the Hv1a/GNA fusion protein, only slightly greater than the 20% and 15% mortality observed for GNA and pro-Hv1a treatments, respectively. Similar results were observed in assays where a single dose of 20 μg of pro-Hv1a/GNA fusion protein was found to cause 30% mortality of fifth instar larvae over 4 days, whereas no mortality was observed for larvae fed on either 20 μg of Hv1a/GNA, or pro-Hv1a (results not shown).

(97) TABLE-US-00015 TABLE 2 Toxicity of recombinant toxins and fusion proteins in oral feeding assays with Mamestra brassicae larvae Pro- Hv1a Pro-Hv1a Hv1a/GNA Hv1a/GNA (E. coli) Percentage 85% (5 d) 70% (5 d) 10% (5 d) 100% (5 d) Survival 600 μg/g 500 μg/g 500 μg/g 180 μg/g
Ingestion Toxicity of Recombinant Hv1a, Pro-Hv1a and Pro-Hv1a/GNA Fusion Protein to Pea (A. pisum) and Cereal (Sitobion avenae) Aphids

(98) Recombinant pro-Hv1a protein, pro-Hv1a/GNA and Hv1a/GNA were tested for oral activity against pea and cereal aphids by incorporation into artificial diet at concentrations of 0.125 mg-0.75 mg/ml (125-750 ppm). As observed for lepidopteran larvae, purified pro-Hv1a/GNA was found to be significantly more toxic than Hv1a/GNA to both aphid species (FIGS. 6A & B). Pro-Hv1a/GNA at 750 ppm caused 100% mortality of pea aphids after 3 days, whereas the same dose of Hv1a/GNA resulted in only 50% mortality after 8 days of feeding. At a lower dose of 500 ppm, mortality after 8 days of feeding was 100% for pea aphids fed on pro-Hv1a/GNA as compared to 20% for Hv1a/GNA.

(99) Pro-Hv1a/GNA was also found to be significantly more toxic than Hv1a/GNA to cereal aphids. As shown in FIG. 6B, 100% mortality was recorded for cereal aphids fed on diets containing 250 ppm of pro-Hv1a/GNA for 7 days as compared to 60% for Hv1a/GNA fed aphids. Cereal aphids appear to be more susceptible to pro-Hv1a/GNA than pea aphids as significant levels of mortality were observed at levels as low as 125 ppm pro-Hv1a/GNA (80% mortality after 2 days of feeding) whereas no mortality was recorded for pea aphids fed on the same dose of fusion protein.

(100) Injection Toxicity of Recombinant MODHv1a and Pro-Hv1a/GNA Fusion Protein to Grey Field Slugs (Deroceras reticulatum)

(101) MODHv1a/GNA and Pro-Hv1a/GNA were tested for activity against slugs (D. reticulatum) by injection into adult slugs (˜0.2 g). MODHv1a/GNA corresponds to the modified form of Hv1a/GNA, where a single amino acid change at the C-terminus of Hv1a has been shown to improve expression of intact fusion protein but has equivalent toxicity to Hv1a/GNA. Slugs were chilled at 4° C. (for ˜15 minutes) prior to injection of 25, 50 or 100 μg of purified Hv1a/GNA resuspended in 20 μl PBS. Mortality was assessed daily for 7 days. FIG. 7 shows dose dependent mortality observed for both treatments. Mortality was significantly greater for pro-Hv1a/GNA as compared to MODHv1a/GNA for all doses injected (P<0.05; Mantel Cox tests). For example, 100% mortality was recorded 3 days after injection of 50 μg of pro-Hv1a/GNA as compared to 10% mortality observed 5 days after injection of 50 μg of MODHv1a/GNA.

(102) Experiments Investigating the Toxicity of PI1a

(103) Introduction

(104) Results obtained for the Hva1/GNA fusion protein were extended by taking a toxin protein whose gene sequence did not include a predicted pro-region and incorporating a pro-region into the expression construct based on similar sequences in the global protein database. The toxin δ-amaurobitoxin-PI1a from the spider Pireneitega luctuosa was utilised.

(105) Expression and Purification of Recombinant PI1a and PI1a/GNA

(106) Expression constructs for the production of recombinant proteins in Pichia pastoris were based on the vector pGAPZαB, which contains a strong constitutive promoter used to direct expression of the recombinant protein and which is designed to integrate into the host genome at the GAPDH locus, giving stable transformants. Expression constructs for the production of recombinant PI1a contained a synthetic coding sequence corresponding to the published amino acid sequence for the toxin designated PI1a, arranged in-frame C-terminal to a sequence encoding the yeast α-factor prepro-sequence. Constructs containing toxin pro-regions had these inserted between the yeast α-factor prepro-sequence and the PI1a toxin sequence. The pro-regions used were taken from the closely related toxin U3-agatoxin-Ao1b from the spider Agelena orientalis (a cDNA sequence including the pro-region is not available for PI1a), designated Ao1bPro-PI1a, and from the pro-region for the Hv1a atracotoxin, as previously described, designated Hv1aPro-PI1a. The expression constructs also contained C-terminal sequences encoding the myc epitope and (His).sub.6 tag, supplied by the vector.

(107) Three expression constructs were created for the production of recombinant PI1a/GNA fusion protein and all contained the mature PI1a coding sequence fused to the N-terminus of a coding sequence corresponding to residues 1-105 of mature snowdrop lectin (GNA) via a 3 amino acid linker peptide; again, the fusion proteins were arranged in-frame C-terminal to the α-factor prepro-sequence, and N-terminal to a sequence encoding the (His).sub.6 tag, supplied by the vector. Modified fusion protein constructs also contained the pro-regions of Ao1b and Hv1a as described above, inserted between the yeast α-factor prepro-sequence and the mature coding sequence of PI1a; they were designated Ao1bPro-PI1a/GNA and Hv1aPro-PI1a/GNA. The constructs were assembled by restriction-ligation and were checked by DNA sequencing after cloning.

(108) Verified clones of expression constructs were transformed into the protease-deficient P. pastoris strain SMD1168H, using antibiotic (zeocin) selection for transformants. Approximately 50 resistant colonies were obtained for each expression construct. Culture supernatant from selected clones grown in shake-flask cultures was analysed for production of recombinant proteins by western blotting, to allow selection of clones producing the highest levels of PI1a and PI1a/GNA fusion proteins. Screening of large numbers of transformed yeast clones was not necessary, since most clones were expressing recombinant proteins, as judged by the presence of immunoreactive bands of the expected size on western blots of culture supernatants (results not shown).

(109) For each construct, the best-expressing clone of those screened in small-scale cultures was selected for large-scale protein production by bench top fermentation. Culture supernatants were purified by nickel affinity chromatography and eluted peaks were desalted by dialysis and lyophilized. Yields of recombinant proteins were comparable to other fusion proteins prior to optimisation; PI1a was produced at approximately 26 mg/L, Ao1bPro-PI1a at approximately 32 mg/L, PI1a/GNA at approximately 21 mg/L, Ao1bPro-PI1a/GNA at approximately 32 mg/L and Hv1aPro-PI1a at approximately 13 mg/L as estimated by semi-quantitative analysis.

(110) Purified recombinant PI1a toxins were analysed by SDS-PAGE and western blotting (FIG. 8). The recombinant toxin PI1a produced by the construct without the added pro-region ran as a closely spaced double band at an indicated molecular weight of approximately 18 kDa on SDS-PAGE gels (FIG. 8A); both bands were immunoreactive with anti-(His).sub.6 antibodies. The predicted molecular weight of recombinant PI1a, including the tag sequences is 6.87 kDa. The double band of toxin was reproducible with different gels, samples and use of reducing agents prior to electrophoresis, but was considered to be an artefact of the gel system, possibly as a result of poor binding of SDS to the polypeptide. When the same samples were pre-treated with 6 M urea, PI1a gave a single band at an indicated molecular weight of 14 kDa (FIG. 8B); the shift in mobility is indicative of gel artefacts, and the single band indicates homogeneity of the product. Further analysis on urea-containing gels gave single bands for PI1a, with indicated molecular weights of ˜11 kDa without blocking cysteine residues and ˜9 kDa after treatment with iodoacetamide to block cysteine residues; these results are diagnostic of incorrect molecular weights under “normal” conditions due to residual secondary structure and interactions between cysteine residues prior to or during electrophoresis. N-terminal sequencing verified incomplete processing of the Kex2 pGAPZαB cleavage site resulting in an additional glutamic acid and alanine residue at the N-terminus and a predicted product mass of 7.07 kDa. Interestingly, PI1a produced by the modified construct incorporating the pro-region (Ao1bPro-PI1a) ran as a closely spaced double band at ˜9 kDa under “normal” SDS-PAGE conditions, with some evidence of a diffuse band at higher molecular weight (FIG. 8C). The predicted molecular mass of the peptide including the additional pro-region is 8.6 kDa. N-terminal sequencing confirmed that the pro-region was present in the protein product and that cleavage had occurred between alanine and the primary residue of the Ao1b pro-region isoleucine giving a predicted molecular mass of 8.46 kDa.

(111) The “normal” PI1a/GNA fusion protein (i.e. derived from a construct that did not contain an additional pro-region) separated on SDS-PAGE gels as two major proteins of indicated sizes of 18 and 21 kDa (FIG. 9A). The 18 kDa protein, immunoreactive with anti-GNA antibodies (results not shown) corresponded in mass to that predicted for recombinant PI1a/GNA (17.1 kDa). The 21 kDa protein was also immunoreactive with anti-GNA antibodies, and had an identical N-terminal sequence to the 18 kDa band. Treatment with the deglycosylating enzyme PNGase F, which cleaves carbohydrate side chains attached to Asn residues through N-glycosidic bonds, removed this band, while the intensity of the “correct” band for the PI1a/GNA fusion protein increased as a result of the treatment (FIG. 9B). This result suggests that the extra band is due to “core” glycosylation of the fusion protein by P. pastoris during synthesis and secretion. GNA contains no potential N-glycosylation sites, but the PI1a toxin sequence contains a potential N-glycosylation site (N-X-S/T) at Asn-35. The N-terminal sequence of the single band was determined as E-A-A-A-G- (SEQ ID NO: 18), as expected for the fusion protein after removal of the yeast α-factor prepro-region during translation and secretion from P. pastoris. In addition a small amount of a band at an indicated molecular weight similar to recombinant GNA (12.7 kDa), which was immunoreactive to anti-GNA antibodies (results not shown), suggesting a small amount of cleavage of the fusion protein into its components was occurring during production and purification. The ratio of intact PI1a/GNA fusion protein to cleaved GNA was estimated as ˜30:1 as judged by Coomassie blue staining on SDS-PAGE gels.

(112) Both of the PI1a/GNA fusion proteins derived from constructs containing additional pro-region sequences (i.e. Ao1bPro-PI1a/GNA and Hv1aPro-PI1a/GNA) separated on SDS-PAGE gels as two major staining bands of approximately 17 and 21 kDa (FIGS. 9C & D). The smaller 17 kDa protein corresponds in mass to that predicted for Ao1bPro-PI1a/GNA and Hv1aPro-PI1a/GNA (16.94 kDa) following removal of the pro-region, suggesting that in both cases the pro-region is removed during processing by yeast cells. The larger 21 kDa protein band is most likely to represent glycosylated protein, as was observed for PI1a/GNA. Ao1bPro-PI1a/GNA and Hv1aPro-PI1a/GNA both expressed as 100% intact fusion protein with no evidence by SDS-PAGE and western blotting for cleavage between the toxin and GNA sequences.

(113) Injection Toxicity of Recombinant PI1a and Ao1bPro-PI1a Protein to Cabbage Moth (M. brassicae) Larvae

(114) Newly eclosed 5.sup.th instar larvae (˜45-55 mg in weight) of M. brassicae were injected with purified recombinant proteins to assess and compare in vivo activity of the toxins and fusion proteins. FIG. 10A shows survival of larvae following injection with PI1a and FIG. 10B shows survival following injection of comparable doses of PI1a and Ao1bPro-PI1a. Larvae injected with PI1a toxin all displayed flaccid paralysis within 1-2 hours (little mobility and almost a complete absence of feeding). Most mortality was observed within the first 24 hours. After a period of paralysis, some insects showed progressive recovery and were able to recommence feeding. The effects of PI1a were dose dependent, with mortality after 24 hours ranging from 75% at 20 μg toxin/insect to 20% at 1.25 μg toxin/insect. Even at high doses of toxin, complete mortality after 72 hours was not observed. From these assays, the estimated LD.sub.50 (48 hours) for the recombinant PI1a was 4.1 μg/insect or 82 μg/g insect, based on an average larval weight of 50 mg (see Table 3). The literature LD.sub.50 value for native PI1a (as opposed to recombinant PI1a) is 9.5 μg/g. This literature value represents the LD.sub.50 value in Spodoptera, a species of the same insect order as M. brassicae.

(115) Recombinant PI1a produced from the modified expression construct, including the pro-region from U3-agatoxin-Ao1b, showed similar toxic effects to PI1a, but was consistently more effective at lower doses than PI1a produced from the construct without this additional sequence (FIG. 10B). Again, the major effects of Ao1bPro-PI1a on mortality were observed during the first 24 hours, with mortality ranging from 80% at 10 μg toxin/insect to 30% at 1.25 μg toxin/insect. In these assays, there was a trend for mortality caused by toxin produced by the modified construct to continue to increase to 72 hours and the highest dose of toxin (10 μg toxin/insect) caused 100% mortality at 72 hours. Assays carried out at the same time with PI1a produced by the unmodified construct gave similar results to the previous assay and direct comparison between the two samples in the same assay showed that differences between PI1a produced by the unmodified and modified constructs were statistically significant when identical dose survival curves were analysed (FIG. 10B). The estimated LD.sub.50 (48 hours) for recombinant PI1a produced from the modified construct (Ao1bPro-PI1a) was ˜1.0 μg/insect, or 21 μg/g insect based on a mean larval weight of 50 mg; this is equivalent to an increase in toxicity of μ4-fold (see Table 3).

(116) TABLE-US-00016 TABLE 3 Toxicity of recombinant toxins and fusion proteins in injection bioassays with Mamestra brassicae larvae Hv1aPro- Pl1a Ao1bPro- Pl1a/ Ao1bPro- Pl1a/ (literature) Pl1a Pl1a GNA Pl1a/GNA GNA LD.sub.50 9.5 μg/g 82 21 μg/g 11 μg/g 7.6 μg/g <5 μg/g (Spodoptera μg/g (28 μg/g (19 μg/g (<12 μg/g sp.)* fusion) fusion ) fusion) *data not available for M. brassicae

(117) The PI1a/GNA fusion protein also caused paralysis and mortality when injected into M. brassicae larvae and was significantly more effective than toxin alone (FIG. 11A). When larvae were injected with 1.25-10 μg fusion protein/insect (equivalent to 0.50-4.0 μg PI1a/insect, since the molecular weight of recombinant PI1a is ˜0.404 of that of the PI1a/GNA fusion protein), significant mortality was observed at all doses, and complete mortality after 24 hours was observed at the highest dose (FIG. 11A). As with PI1a, most mortality was observed within the first 24 hours of the assay and effects were dose dependent, ranging from 100% mortality at 10 μg fusion protein/insect to 35% mortality at 1.25 μg fusion protein/insect. Mortality at this lowest dose of fusion protein increased to 65% after 72 h whereas mortality from injection of 1.25 μg toxin alone did not change from 20% between 24 and 72 hours. From these assays, the estimated LD.sub.50 (48 hours) for the recombinant PI1a/GNA fusion protein was 1.4 μg/insect, or 28 μg/g insect, based on a mean larval weight of 50 mg. The LD.sub.50 is equivalent to 0.56 μg PI1a toxin per insect, making the PI1a/GNA fusion protein ˜7 times as active as the toxin produced by the unmodified construct, and ˜2 times as active as the toxin produced by the modified Ao1b-Pro-PI1a construct, on a molar basis. A similar ratio is obtained by using mortality figures at 72 hours. Direct comparisons of mortality produced by identical doses of toxin and fusion proteins show that the three treatments are different from each other, and from control, at p<0.0001 (ANOVA). In all these assays, no mortality of control-injected insects was observed over 72 hours.

(118) As observed with PI1a toxin, addition of the Ao1b pro-region to the PI1a/GNA fusion protein expression construct resulted in a protein product with enhanced biological activity (FIG. 11B). The fusion protein product derived from this construct had an estimated LD.sub.50 (48 hours) of 0.94 μg/insect, with increased mortalities at all doses except the highest. Addition of an alternative pro-region from the pro-HV1a toxin, to the PI1a/GNA expression construct also enhanced the biological activity of the resulting fusion protein (FIG. 11C); this protein had an estimated LD.sub.50 (48 hours) of <0.6 μg/insect, although overall mortality values were similar to the pro-Ao1b-PI1a/GNA fusion protein. These data suggest that a two-fold increase in toxicity can be obtained by including pro-regions in the expression constructs for P1a/GNA.

(119) Ingestion Toxicity of Recombinant PI1a/GNA, Ao1bPro-PI1a/GNA and Pro-HvlaPI1a/GNA Proteins to Cabbage Moth (M. brassicae) Larvae

(120) A similar increase in toxicity of fusion proteins derived from expression constructs including pro-regions to that observed in injection assays was also observed in droplet feeding assays with 3rd stadium M. brassicae larvae (FIG. 12). Following ingestion of a single 2 μl droplet containing 20 μg of fusion protein, mortality after 5 days was 40% for PI1a/GNA; 50% for Ao1bPro-PI1a/GNA and 70% for Hv1aPro-PI1a/GNA (data summarised in Table 4). Minimal reductions in survival (0-20%) were observed for control treatments where larvae were fed on 30 μg toxin or GNA and survival curves for controls were significantly different to fusion protein treatments. This provides further evidence that the addition of pro-regions to the PI1a/GNA construct results in increased biological activity. As for injection studies the use of the Hv1a pro-region was seen to result in the greatest enhancement of toxicity over the non-modified PI1a/GNA fusion protein.

(121) TABLE-US-00017 TABLE 4 Toxicity of recombinant toxins and fusion proteins in oral feeding assays with Mamestra brassicae larvae Pl1a (recom- Ao1bPro- Ao1bPro- Hv1aPro- binant) Hv1a Pl1a/GNA Pl1a/GNA Pl1a/GNA Percentage 90% (5 d) 80% (5 d) 60% (5 d) 50% (5 d) 30% (5 d) Survival 400 μg/g 400 μg/g 500 μg/g 500 μg/g 500 μg/g

REFERENCES

(122) Becker & Guarente High-efficiency transformation of yeast by electroporation. Methods Enzymol. (1991) 194:182-187. Berent S. L., Mahmoudi M., Torczynscki R. M., Bragg P. W., Bollon A. P. Comparison of oligonucleotide and long DNA fragments as probes in DNA and RNA dot, southern, northern, colony and plaque hybridizations. Biotechniques. (1985) 3:208-220 Douglas, A. E. and Prosser, W. A. Synthesis of the essential amino acid tryptophan in the pea aphid (Acyrthosiphon pisum) symbiosis. Journal of Insect Physiology (1992) 38: 565-568. Fitches, E. C., Pyati, P., King, G. F. and Gatehouse, J. A. Fusion to snowdrop lectin magnifies the oral activity of insecticidal ω-Hexatoxin-Hv1a peptide by enabling its delivery to the central nervous system. PLoS One (2012) 7:e39389 Fitches, E., Woodhouse S. D. Edwards, J. P., Gatehouse, J. A. In vitro and in vivo binding of snowdrop (Galanthus nivalis agglutinin; GNA) and jackbean (Canavalia ensiformis; Con A) lectins within tomato moth (Lacanobia oleracea) larvae; mechanisms of insecticidal action. Journal of Insect Physiology (2001) 47:777-787. Fletcher J. I., Smith R., O'Donoghue S. I., Nilges M., Connor M., Howden M. E. H., Christie M. J., King G. F. The structure of a novel insecticidal neurotoxin, omega-atracotoxin-HV1, from the venom of an Australian funnel web spider. Nature Structural Biology (1997) 4:559-566. Kaas, Q., Westermann, J. C. and Craik, D. J. Conopeptide characterization and classifications: an analysis using ConoServer. Toxicon (2010) 55:1491-1509 Khan, S. A., Zafar, Y., Briddon, R. W., Malik, K. A. and Mukhtar, Z. Spider venom toxin protects plants from insect attack. Trangenic Research (2006) 15:349-357 Nielsen, H., Engelbrecht, J., Brunak, S. and von Heijne, G. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering (1997) 10:1-6 Sambrook & Russell, Molecular Cloning: A Laboratory Manual: 3.sup.rd edition Sherman et al (1986) Methods in Yeast Genetics, A Laboratory Manual. Cold Spring Harbour, NY. Southern. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol (1975) 98:503-517 Tedford, Sollod, Maggio and King. Australian funnel-web spiders; master insecticide chemists. Toxicon (2004) 43:601-618 Windley, Herzig, Dziemborowicz, Hardy, King and Nicholson. Spider-Venom Peptides as Bioinsecticides. Toxins (2012) 4:191-227 Wong, E. S., Hardy, M. C., Wood, D., Bailey, T. and King, G. F. SVM-based prediction of propeptide cleavage sites in spider toxins identifies toxin innovation in an Australian tarantula. PLoS One (2013) 8(7):e66279