|
|
1Dept. of Genetics and Biochemistry and 2School of Materials Science and Engineering, Clemson University, Clemson, SC 29634, USA Phone: (864) 656 3060 Fax: (864) 656 6879 E-mail: aalbert@clemson.edu
Using spider silk and collagen as a model, we have investigated the role that various protein primary structural components play in fibre production. Spidroins, spider dragline silk protein components, are essentially characterised by an amino-acid repeat containing a glycine-rich motif (amorphous) followed by an alanine-rich motif (crystalline, putatively responsible for fibre strength). We have tested the importance of alanine runs in these proteins and the role of this motif in the mechanical properties of the resulting fibre. To test the importance of alanine-rich motifs in the spidroin-1 proteins, we engineered three types of spidroin-1-like genes containing sequence encoding for different amounts of alanine repeats in the protein (normal, low, and no alanine residues). We also have engineered three copolymer collagen-spidroin-1 genes using each of the three spidroin-1 synthetic genes. These copolymers were mimicked on the existing natural block collagen-silk-like protein copolymer found in the byssus thread of marine mussels. All of these constructs were introduced in yeast (Pichia pastoris) for protein production. We are currently purifying each of the recombinant proteins for structural analysis (CDspectroscopy).
Spider silk, spidroin, collagen, synthetic genes, Pichia pastoris
The golden orb weaver spider Nephila clavipes can manufacture seven types of silk of different compositions which exhibit different mechanical properties adapted for their specific use (e.g. web, cocoon, dragline). These silks are natural protein-based fibre polymers resulting from the self-assembly of many individual fibre protein molecules into a complex and stable network. Physical and chemical studies performed on spider dragline silk, as produced by the major ampullate glands of the spider, suggest that this silk exhibits unique properties that are superior to those displayed by the silkworm silk, and even to some manmade fibres. Dragline silk is a very tough fibre resulting from the combination of high extensibility and high strength [1]. Extensive information is available regarding N. clavipes' dragline silk composition, as is physical and mechanical data on the silk fibre.
Partial cDNAs sequences corresponding to the genes encoding the two protein components of the dragline silk were isolated [2 and 3]. The primary structure of these highly repetitive proteins, called Masp1 and Masp2 (major ampullate spidroin 1 and 2), basically consists of alanine motifs sandwiched between glycine-rich (Masp1) or proline containing glycine-rich regions (Masp2). Hypotheses concerning the secondary structures, as well as the role of alanine or glycine-rich regions were developed once the physical data (X ray) on dragline silk fibres and the actual protein sequences were available [4]. The alanine regions are supposed to adopt anti-parallel ß-sheet structures leading to the formation of crystals arranged in parallel to the fibre axis and ultimately responsible for the fibre strength [4].
Spider silk is an excellent model for investigating the relationship between the structure of the individual repetitive motifs present in the fibre proteins and their mechanical function in the resulting fibre. Understanding the rules governing this structure/function relationship is also the first step towards fibre protein engineering for the production of new designed protein-based biomaterials.
http://www.autexrj.org/No4-2003/0073.pdf
Many laboratories worldwide have focused their efforts on producing artificial silk produced in transgenic systems. Reports on the cloning of synthetic spidroin-like genes for protein production in bacteria (E. coli) or in yeast (P. pastoris) are numerous [5]. More recently, a spidroin cDNA was used for spidroin protein production in mammalian cells [6] and a synthetic spidroin-1-like gene was introduced in tobacco and potato for protein production, demonstrating that plants can also serve as bioreactors to produce engineered fibre proteins [7]. Only one group reported the spinning of silk-like proteins produced by engineered mammalian cells [6]. However, the artificially spun fibre did not display the same properties as the native silk.
To determine the secondary structure adopted by the alanine motifs present in the transgenically produced silk-like proteins and their putative role in the formation of crystals that would account for the fibre's strength, we have used N. clavipes’ major ampullate silk cDNA sequences (Masp1 and Masp2) as a model. We have explored the use of single-celled systems such as yeast (Pichia pastoris), and multicellular systems such as plants for fibre protein production. Through genetic engineering in yeast, we have generated several clones able to produce three customised spidroin-1like proteins containing variable amounts of alanine motifs (regular, less and none). In addition, inspired from the natural fibrous proteins forming the mussel byssal threads, we have also studied copolymeric proteins. To this end, we have engineered collagen-spidroin-1 copolymers by juxtaposing sequences encoding the helical part (GXY)n of a root-knot nematode cuticle collagen (Meloidogyne incognita) [8] to each of our customised synthetic spidroin-1 genes, and cloned them in yeast for protein production and characterisation.
Spidroin 1-like homopolymers (Figure 1)
A MaSp1 protein consensus repeat : (GGA) GQ (GGY) (GGL) (GGQ) GAGR (GGL) (GGQ) (GA)2 (A)3
Figure 1. Spidroin-1-like gene construction. A. Major ampullate spidroin 1 amino-acid consensus repeat [4]; B. Sets of oligonucleotides designed to span the region encoding this consensus repeat (sets B1, B2 and B3 are for the engineering of genes encoding for spidroin 1 with regular, less and no alanine motifs respectively); C. Annealing of each complementary oligonucleotides; D. Head to tail ligation/multimerisation to generate spidroin-1like genes (D1, D2 and D3 are genes encoding for spidroin-1-like proteins containing the regular amount of alanine motifs, a lesser amount and no alanine motifs respectively)
Sets of 5' phosphorylated complementary oligonucleotides spanning the region encoding the consensus repeat of N. clavipes spidroin 1 protein (Masp1) were synthesised (IDT technology). Three types of spidroin-1-like genes encoding for spidroin-1-like proteins containing normal amounts of alanine motifs, lesser and no alanine motifs were generated as described below. The different sets of complementary oligonucleotides were annealed at 65ºC for 5 hours, and were then slowly cooled to room temperature. In each case, the appropriate annealed reactions (Figure 1) were combined and ethanol precipitated. The recovered DNA pellets were subjected to a head to tail ligation performed overnight at 4ºC using T4 DNA ligase (Promega). The resulting multimerised genes were electrophoretically separated on 0.8% agarose gels, and the genes between 500 and 1000 bp were isolated by electroelution followed by ethanol precipitation [9]. Individual reactions were set up for 1 hour at 72ºC to simultaneously ‘fill’ and ‘A-tail’ each type of size-selected spidroin-1-like gene using Taq DNA polymerase (Fisher) and dNTPs (Promega). The tailed products were recovered by ethanol precipitation and ligated to a pGEMT-Easy vector (Promega) using a T4 DNA Ligase (Promega). The ligation reaction was used to transform competent E. coli strain JM109 (Promega). Clones containing plasmids with inserts between 500 and 600 bp were selected, and the plasmids DNAs were isolated using an alkaline lysis method [10]. The plasmid inserts were sequenced using an ABI PRISMTM Dye terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer) and M13 (forward and reverse) universal primers. The sequences were resolved on an ABI 373 Stretch sequencer (Applied Biosystems Inc.). Three clones each containing one of the spidroin-1-like candidate genes (normal, less or no alanine motif encoding genes) ranging between 500 and 600 bp were selected.
Collagen spidroin-1-like copolymers
All the PCR reactions were performed using a high-fidelity DNA polymerase (Pfu DNA polymerase, Promega) using specific primers designed to add appropriate restriction sites in the 5' and 3' of the genes to allow the cloning in the yeast expression vector. The collagen-like sequence used for the collagen-spidroin-1-like copolymer construction was obtained by performing a PCR reaction on the micol-2 collagen cDNA clone [8] using specific primers. The primers were specific to the region encoding for the helical part of this collagen, and were designed to add an Eco RI and a Bam HI sites in the 5' and 3' of the 423 bp amplified product respectively. The three types of spidroin-1-like genes (normal, less and no alanine motif encoding genes) were individually amplified by PCR using specific primers that allowed the addition of a Bam HI and a Sac II sites in the 5' and the 3' end of the gene respectively. All these spidroin-1-like and collagen PCR products were cloned in pGEMT-Easy (Promega) in E. coli strain JM109 (Promega), and the inserts of the candidate recombinant plasmids were sequenced. The 3 spidroin-1-like clones and the collagen clone containing the correct gene sequence flanked by the appropriate 5' and 3' restrictions sites were identified and selected.
The yeast expression vector pPICZaA (Invitrogen) was subjected to an Eco RI/Sac II restriction digestion, isolated by electroelution on agarose gel [9] and ethanol precipitated.
Homopolymer gene
The same PCR-based strategy mentioned above was used to add an Eco RI and a 3' Sac II restriction sites respectively in the 5' and the 3' of each of the three genes (normal, less and no alanine motif encoding genes) with specific primers. The PCR products were cloned in pGEMT-Easy (Promega) and used to transform competent E. coli strain JM109 (Promega). Recombinant plasmids containing the three types of spidroin-1-like genes were isolated using an alkaline lysis protocol [10] and their inserts were sequenced. Three clones, each containing one of the 3 spidroin-1-like gene types with the appropriate 5' and 3' restrictions sites, were identified and selected. The three plasmid candidates were individually subjected to a double restriction digestion with Eco RI/Sac II. Each of the released spidroin-1-like gene inserts was isolated separately by electroelution and ethanol precipitated. Three separate ligation reactions were set up between the Eco RI/Sac II pPICZaA vector and one of the Eco RI/Sac II spidroin-1-like genes mixed in a 1:3 molar end ratio respectively.
Copolymer gene
For the copolymer construction, Eco RI/Bam HI and Bam HI/Sac II double restriction digestions were performed respectively for the collagen clone and the three types of spidroin-1-like plasmid clones. For all constructs, the released spidroin-1-like or collagen gene inserts were individually isolated by electroelution and ethanol precipitated. Three separate ligation reactions were set up and each one contained a 1:3:3 molar end ratio of the prepared Eco RI/Sac II pPICZaA vector, the Eco RI/Bam HI collagen gene and one of the three Bam HI/Sac II spidroin-1-like gene respectively.
For all homopolymer and copolymer gene cloning, the different ligation reactions were used to transform the competent E. coli strain TOP10F' (Invitrogen). Candidate plasmids containing homopolymer or copolymer genes of the correct size were identified. These 6 plasmid clones were individually isolated and sequenced using vector specific primers: 5'AOX1, a Factor, and 3'AOX1 primers (Invitrogen).
Each of the 3 homopolymer or the 3 copolymer constructs were used separately to transform the Pichia pastoris strain GS115 (Invitrogen). Transformations and expression studies (using BMGH/BMMH media) were performed as suggested by the manufacturer (Easy selectTM Pichia Expression Kit, Invitrogen). Yeast transformants were obtained by plating the transformation reaction on a medium containing five times the suggested antibiotic concentration. Clones showing higher protein production levels were characterised as ‘super expressers’ and were selected for our study.
All protein extractions from yeast cells were performed according to the instructions specified by the manufacturer (Easy selectTM Pichia Expression Kit, Invitrogen). Proteins were purified by affinity chromatography (Talon Metal Affinity Resins, Clontech) or by precipitation with 30% ammonium sulphate.
All intracellular or secreted recombinant proteins were subjected to SDS-PAGE analysis of 4%/10% polyacrylamide gels followed by Coomassie blue staining or Western-blot analysis using antibodies directed against a myc epitope present in the fusion protein (ECLTM Western Blotting Analysis Systems, Amersham Pharmacia Biotech). Dot blot analysis of crude cellular protein extracts or correspondent culture medium were also used as a method to quickly screen and select yeast transformants producing the recombinant proteins. Each time, Western-blot analysis was used to confirm the presence, size and production level of each engineered protein.
Spidroin-1-like gene encoding for proteins containing a variable amount of alanine motifs (normal= ALA, less= LA, and none= NA) were generated using synthetic oligonucleotides that were designed to span the region encoding for the amino acid consensus repeat of the major ampullate spidroin-1 protein (Figure 1). The multimerisation process used to restore the repetitive nature of the gene can generate spidroin-1-like genes of sizes up to 6 kbp. However, for our study, we voluntarily selected spidroin-1-like genes ranging between 500 and 1000 bp for several reasons: 1) to be able to sequence the cloned genes and confirm that there were no mistakes in the sequences due to errors present in the synthetic oligonuclotides, which would result in the production of a truncated or modified protein; 2) to build gene cassettes containing sequences encoding for different fibre protein amino acid motifs that could later be used in different combinations for fibre protein engineering.
Collagen Spidroin 1-ALA Spidroin 1-LA Spidroin 1-NA
Figure 2. Copolymer gene engineering in yeast. A. Collagen, spidroin-1 ALA, spidroin-1-LA, and spidroin-1-NA cloned genes. B. PCR reaction with specific primers to add restriction sites and cloning in E. coli for sequencing.
C. Ligation of the collagen gene to each of the three spidroin-1-like genes to form the copolymer genes; D. Cloning of the copolymer genes in an expression vector in E. coli and sequencing (the pPICZaA vector allows the production of a recombinant protein that has a myc epitope and a polyhistidine tag in its carboxyl terminal for immunological detection and purification by affinity chromatography respectively. This vector also allows the secretion of the fusion proteins in the culture media facilitating the purification process); E. The recombinant vector carrying the copolymer genes was isolated and used to transform yeast; F. Selection of three types of yeast transformants, each of which produces and secretes one of the three types of collagen-spidroin-1 like proteins
Our goal was to determine the secondary structure adopted by the alanine motifs present in spidroins produced ‘in vivo’. we have engineered and cloned in yeast three types of spidroin-1-like proteins homogenous in size and only differing by the number of alanine motifs present in their sequence that would allow the study of the feasibility of a simple eukaryote such as yeast to manufacture, handle and secrete such fibre proteins. The three types of genes ranging between 500 and 600 bp were first cloned in E. coli for sequence verification, and then cloned in Pichia pastoris for production of three types of homopolymer proteins (spidroin-1-ALA, -LA, or -NA).
Copolymer collagen spidroin1-like genes (COALA, COLA, and CONA) were also generated using the sequences encoding the helical part of a cuticle collagen from a parasitic nematode in combination with each of our synthetic spidroin-1-like genes (Figure 2).
For each homopolymer and copolymer clone, up to 50 independent yeast transformants were analysed for protein production. Yeast cellular extracts and culture media were subjected to Dot blot analysis, and the positive results were confirmed by Western blot analysis to verify the molecular weight of the proteins and determine the level of protein production. Six yeast clones producing the spidroin-1-ALA, LA, and NA and the copolymer collagen spidroin-1-like COALA, COLA, and CONA at higher levels (the ‘super expresser’) were identified and selected for our study (Figure 3).
1 2 3 4 5 1 2 1 2
Figure 3. Western blot analysis showing the intracellular spidroin-1 homopolymer and copolymer protein production. A Lane 1 to 3; Molecular weight marker, positive control for intracellular expression (b-Galactosidase, 119 KDa), negative control (protein extract from untransformed P. pastoris) respectively; Lanes 4 and 5 are the spidroin-1LA homopolymer and copolymer respectively. B Lanes 1 and 2: spidroin-1-NA homopolymer and copolymer respectively. C Lanes 1 and 2: Spidroin-1-ALA homopolymer and copolymer respectively. The predominant spidroin-1-like homopolymer proteins detected correspond to dimer forms and their sizes are 39.4, 36.6 and 33.1KDa for spidroin-1-LA, -NA and -ALA respectively (fine arrows panels A, B and C). The copolymer proteins detected also correspond to dimer forms of the collagen-spidroin-1-like proteins and their sizes are 59.2, 60.9, and 57.4 KDa for COLA, CONA, and COALA respectively (bold arrows in panels A, B and C)
Expression studies were conducted to determine the best time of protein production and secretion for each of the homopolymers (ALA, LA, and NA) and copolymers (COALA, COLA, and CONA). All the different spidroin-1-like proteins and corresponding copolymers are produced intracellularly by the different yeast clones. Secreted proteins have been detected by both Dot blot and Western blot analysis for the copolymers COALA and COLA (Figure 4).
T0 T2 T4 T6 T 8 T10
T0 T2 T4 T6 T 8 T10 1 2 3 4 5 6 7 8 9 10 11
Figure 4. Dot blot and Western blot analysis showing the intracellular and secreted copolymer COALA protein. Panels A and B show Dot blot analysis of the intracellular protein fraction and culture media (secreted fraction) respectively. Panel C shows the Western blot analysis of the secreted proteins (same samples as in B; Lanes 1 to
11: T0 to T10). Sampling of both cells and media (without cells) from the cultures was carried out at regular time intervals after the initial start of induction (T0 = start of induction, T1 to T10 are collections from 4 hours and up to 88 hours after induction)
For the homopolymer protein production, although Dot blot analysis suggested that the homopolymer proteins spidroin-1-LA and -ALA were not only being highly produced in the cell but were also being secreted, Western blot analysis did not confirm these results. The major form of the homopolymer protein detected in the cell is the dimer, suggesting that these proteins have a tendency to aggregate or that they might be crosslinked, since the dimer forms are stable through treatment with SDS and heat. However, the solubility of these proteins in the cell extract buffer and the culture media might be different, due to the difference in salt concentration and pH conditions of the two buffers. It is possible that the secreted proteins are precipitating out of the culture media thus preventing them from entering the polyacrylamide gel while the intracellular proteins are maintained in solution. This would provide a possible explanation of why Western blot analysis on secreted spidroin-1-like proteins gives no evidence of the protein’s presence although Dot blot analysis was capable of detecting their presence. Our results also suggest that the addition of a collagen sequence seems to enhance the overall solubility of these proteins. In a general manner, the proteins containing no alanine motifs (homopolymer spidroin-1-NA and copolymer CONA) seem less stable than the alanine motif containing engineered homopolymer and copolymer proteins. The fact that we have observed mostly dimeric forms of homopolymer and copolymer proteins could be justified in two ways. The first would be gene recombination due to the repetitive nature of the spidroin sequences that would result in a protein of different size. However, our results (not shown) on the screening of first-generation clones show that all of the clones producing the proteins displayed the same profile in Western blot analysis. All the proteins produced by the various gene constructs were present predominantly as dimers. Since the odds of any recombination phenomena happening simultaneously in all the individual clones and giving rise to the same ‘recombined’ genes are very slim, the most plausible explanation would then be an increase in protein size due to intermolecular crosslinking. To eliminate the first possibility (gene recombination), we are currently conducting Southern blot experiments to determine the number of synthetic genes inserted in the yeast clone genomes and their size.
We are currently purifying these spidroin-1-like homopolymer and copolymer proteins, focusing our efforts on optimising their purification process using metal ion affinity chromatography in combination with heat treatment, salt concentration and pH modifications.
We thank the National Textile Center for the funding of this research and Dr R. S. Hussey (University of Athens, GA) for providing the mi-col-2 cDNA clone.
bravenet.com