SeaDNA - takes inspiration from the hagfish, which lives on the ocean floor and produces a defensive slime made of filaments with amazing properties. From these filaments we aim to make novel biopolymers that are biodegradable, lightweight, thinner than a human hair and potentially stronger than nylon, steel and even Kevlar. We plan to use E.coli cells to mass produce these polymers that will be very useful in health and medicine, food and manufacturing industries.

Synthetic Biology has the potential to generate novel biological materials as alternatives to petrochemical-derived products. The aim of SeaDNA is to develop a large-scale synthesis a novel biopolymer that has fascinating, and potentially very useful properties. This biopolymer is an intermediate filament produced by the Pacific hagfish (Eptatretus stoutii) as part of a defensive slime that it deploys to repel predators. The mechanical properties of the filaments as produced by hagfish are similar to those of spider dragline silk; making them a candidate for high-performance materials applications(Fudge et al., 2010). Isolation of filaments from hagfish is unlikely to be commercially viable. Recombinant production of these filaments is therefore highly desirable but has not been achieved to date. Here we have cloned the two hagfish intermediate filament genes into vectors to facilitate protein expression in E. coli. We have developed a protocol for the purification of the two synthetic proteins and for their assembly into filaments and threads. In parallel we have been working with the natural proteins isolated from hagfish slime have now begun to compare the properties of filaments derived from the natural and synthetic proteins to confirm that they have similar mechanical properties. Once this is confirmed, the synthesis must be scaled up to utilise the filaments in materials applications such as - replacement of nylon in durable textiles, eye sutures, replacement of ligaments and tendons, bandages, increase biocompatibility of implants by coatings, component in composite materials, replacement of Kevlar, etc. Another aim is to produce films, from which hydrophobic coatings may be made. Thus, further applications as antimicrobial hydrogels in - for example - contact lenses or, use as porous membranes to remove harmful materials from water, are a possibility. We have formed a company – Benthic Labs to develop the commercial potential of this novel biopolymer.

The Hagfish is a primitive fish found in the deepest benthic region of the ocean. Hagfish are often called “slime eels”. However, they are in the class Agnatha along with lampreys. This class of animals incorporates jawless fish. There are 76 species of Hagfish the world over and approximately 100 species in the class Agnatha. Hagfish are referred to as “living fossils” as they strongly resemble fossils of their 300 million year old ancestors. This does not mean that Hagfish have stopped evolving, rather that their body plan and strategy is still very successful today.

When the hagfish is agitated or preyed upon, it releases a substance from its glands (Gland Thread Cells: GTC) that are located along the side of its body(Downing et al., 1984). When this substance comes in contact with water, it creates a mucus gel that expands greatly in water. A video of the water absorbent properties of the slime can be seen here. This slime can clog the gills of the would-be attacker, causing them to retreat from the hagfish. This is an incredible defensive adaptation to ward off predators. An excellent video of this can be seen here on the National Geographic website.

Very little is known about Hagfish reproduction and only one of the 76 species has been successfully bred in captivity. Hagfish are scavengers and while jawless, they have two rows of tooth-like keratin structures, which they use to burrow into the carcasses of dead fish and whales resting on the seabed. They are known for eating their prey from the inside out. Hagfish often tie themselves in overhand knots while feeding to prevent them choking on their own slime. This results in them “sneezing” out the slime from their nostril. Hagfish are usually not eaten on account of their looks, their viscosity and unpleasant habits. One species, however, the inshore Hagfish, is valued in the Korean Peninsula. It is agitated to produce its slime that is used in a similar manner to egg whites in various forms of cookery. Tasty!

The slime or gel produced by hagfish is made primarily of two protein components(Fudge et al., 2005). The first component is mucin, which causes the rapid expansion and formation of gel with water. The second component is an intermediate filament, which gives the gel its mechanical properties. This project was primarily focused on production of the second component, the hagfish slime intermediate filaments.

The hagfish slime intermediate filament, like other intermediate filaments is based on a coiled coil structure, which consists of a two alpha helical components wound around each other. Both the alpha and the gamma component are considered a type of keratin. The helixes are co-expressed and co-assembled in vitro, come together to form a coiled coil (which is to say that both helixes wrap around each other), in a 1:1 ratio to form heterodimers. As there is only one cysteine in the entire complex it is established that this does not occur to disulphide bonds forming, but probably ionic interactions and hydrogen bonding. 10nm intermediate filaments bundle together to form threads that in the case of the hagfish slime reach up to 60cm in length and 3 microns in diameter.

Interest in the hagfish slime intermediate filaments as novel biomaterials stems form the observation that when stretched the hagfish threads can undergo a conformational shift from alpha-helical to beta-strand, giving rise to beta-sheet crystals that resemble those found in spider silk (Fudge et al., 2003). Draw processing of native slime threads showed promising results and produced fibres with desirable mechanical properties (Fudge et al., 2010; Kim et al., 2013). Furthermore it was shown that the solubilized hagfish slime proteins could be used to cast thin films and to generate fibres, the mechanical properties of which could be altered depending on conditions used (Negishi et al., 2012). These studies from Fudge and co-workers highlight the potential of slime threads as a source of renewable and biodegradable high performance biomaterials, with properties similar to the much-hyped properties of spider silk. In addition, the hagfish proteins are in theory better candidates than spider silk proteins for recombinant production in bacteria due to their relatively small size and alpha helical structure (prior to stretching). Recombinant expression of these proteins has not been reported to date however.

The hagfish slime intermediate filament protein is a heterodimer composed of two subunits – termed alpha and gamma. Both subunits form alpha helices that then assemble together to form the coiled coil structure of the intermediate filament. Like other intermediate filaments, the assembled alpha and gamma subunits have non coiled-coil head and tail regions

The alpha subunit has a molecular weight of 66.6kDa and contains 643 amino acids. Some of the structural features of the alpha subunit are shown below.

The gamma subunit has a molecular weight of 62.7.kD and its sequence contains 603 amino acids. Some of the structural features of the gamma subunit are shown below.

E.coli are the bacteria of choice for the production of useful proteins, which may be difficult to harvest from nature and have been used for example in the production of human insulin. Here we cloned the hagfish genes coding for the alpha and gamma subunits of the slime intermediate filament into plasmids that will facilitate their production at high levels in E. coli.

The protein sequence of alpha and gamma subunit proteins were retrieved from the protein database. Rather than using the natural hagfish DNA sequence coding for these proteins we generated a DNA sequence by optimizing the codon usage of the DNA sequence to better match that found in highly expressed E. coli genes. This was performed using an online bioinformatics tool called OPTIMIZER (Puigbo et al., 2007). This means that the bases coding for the certain amino acids were changed. This makes no alterations to the actual protein itself, but increases the chance of the bacteria being able to translate the mRNA sequence efficiently into a protein. Double stranded DNA sequences corresponding to the codon optimized alpha and gamma genes were then ordered from Gen9.

Forward and reverse oligonucleotides containing the standard biobrick prefix and suffix were used to amplify the codon-optimized alpha and gamma genes by PCR. Following digestion with EcoRI and SpeI enzymes these DNA fragments were cloned into the pSB1-C3 vector that had been cut with the same enzymes. These genes have been supplied to the Registry with parts numbers BBa_K1518000 and BBa_K1518001.

To produce and purify the proteins that constitute hagfish slime we cloned the codon-optimized alpha and gamma genes in the pCDF Duet expression vector (Merck-Millipore). This is a low copy plasmid that is specifically designed to facilitate expression of multiple genes either from one plasmid or from separate plasmids maintained in the same cells. The pCDF Duet vector encodes a poly histidine tag that will be fused to the amino terminal end of each protein. The primary reason for this modification is to facilitate the rapid purification of the proteins by affinity chromatography. The His-tag allows the protein to be removed from all other proteins in the bacteria by use of a Ni-NTA pro-bond resin column (column chromatography).

Cloning was achieved by digesting the pCDF-duet plasmid with two restriction enzymes; EcoRI and HindIII. The hagfish subunit DNA as synthesised was flanked by MfeI and HindIII restriction enzymes, allowing it to be compatible with the EcoRI and HindIII restriction enzyme sites in the pCDF-duet plasmid. Gel electrophoresis was performed on these samples and the DNA was subsequently extracted from the gel. A ligation reaction was then performed followed by a bacterial transformation into E. coli DH5α. Cells then cultured on a streptomycin plate. Colonies obtained were screened by colony PCR and positive colonies picked and cultured in liquid broth to prepare minipreps of plasmid DNA. Plasmid mini preps were performed individually to isolate the plasmids from the DH5α cultures. From this we obtained an isolated plasmid pCDF-Duet with an alpha subunit, and a plasmid pCDF-Duet with a gamma subunit. To verify that the isolated plasmids had the correct insert we performed a restriction digest. This DNA was also checked by sequencing of the inserts.

For protein expression the BL21 E. coli strain was used and expression induced using IPTG which binds to the lac operator to induce transcription. Following cell lysis purification of the intermediate filament alpha and gamma proteins was performed under a variety of conditions.

Transformation of the plasmids described above into E. coli BL21 cells was performed and cells plates on spectinomycin plates. An overnight culture from a single colony from both the alpha and gamma plates was grown and the next morning that culture volume was increased to 500ml with LB Broth (plus antibiotic). Gene expression was induced by the addition of 1mM IPTG at OD600 of 0.4 and for an incubation period of 3 hours. Induction of expression in log phase allows for robust transcription and translation of the target gene. Cultures were shaken at 200rpm for this period to provide sufficient aeration. Whole protein samples taken at 0 and 3 hours were analysed for protein expression by SDS PAGE. Expression of both proteins at the expected molecular weight was observed (see figure in next section).

Pelleted cells were lysed by sonication and a cleared lysate (supernatant) obtained by centrifugation. This was subjected to purification using a Ni-NTA column, washed extensively and bound proteins were eluted using imidazole. The purification fractions were analysed by SDS PAGE. Although strong expression of both protein was achieved these proteins were not obviously visible in the cleared lysates for either protein. No significant protein bands were seen in the elution fractions for either protein. This lead us to conclude that when expressed individually both proteins are insoluble and so are not present in the cleared lysate and cannot be purified.

This experiment was also carried out at 30 ̊C to aid in the correct structural formation of the proteins as they are being expressed. It is thought that lower temperatures help some protein expression as the cell is under a mild stress conditions causing increased synthesis of chaperone proteins to help protein folding. Also a reduced temperature lowers the rate of protein synthesis reducing the risk of this protein aggregating and not folding properly or killing the host cell. The protein remained insoluble at 30 ̊C. Co-expression of the alpha and gamma subunits was also attempted but did not yield significant amounts of soluble protein (data not shown).

Since the hagfish intermediate filament proteins were expressed but seemed to be insoluble we attempted to purify them under denaturing conditions. Following induction, cells were disrupted by sonication in lysis buffer containing 8M Urea. The His-Tagged proteins were then purified in the presence of urea using Ni-NTA probond resin. The results of the expression and purification were evaluated by SDS-PAGE for both the alpha and the gamma subunits. Both proteins were purified to apparent homogeneity using this approach.

We successfully both cloned and expressed both subunits of the hagfish thread intermediate filaments. Both subunits were insoluble, but could be purified under denaturing conditions. It is possible that with further optimization soluble expression of the proteins could be obtained. While soluble protein would be preferable, we decided to proceed with the protein purified under denaturing conditions to see if it could be assembled into filaments.

Working with the proteins that had been purified in 8M urea we developed a method to reconstitute the proteins and co-assemble them into filaments and threads that resemble those obtained from the natural proteins isolated from hagfish slime. Since we are interested in developing these proteins commercially as novel biomaterials we are not divulging the details of our proprietary co-assemble method.

Following isolation of the alpha and gamma slime intermediate filament proteins in urea we subjected them to our co-assemble protocol. Roughly equal amounts of alpha and gamma protein were used for co-assembly. As controls the co-assembly procedure was carried out for alpha and gamma subunits individually.

It was found that filaments and threads were formed when both the alpha and gamma subunits were present (the threads 1µm to 2µm in diameter). As expected under control conditions, in which only one subunit was present, little to no filamentous precipitate was formed. This suggests that the subunits are indeed forming heterodimers with then copolymerising into filaments and threads. The filamentous structure of the assembled structures could be imaged readily by bright field microscopy.

These results take us one step closer to making a polymer that is biodegradable, strong, elastic and lightweight by using synthetic biology. The future of this project involves characterising these proteins further, improving the solubility and optimizing their expression patterns. We hope that our work will open the opportunities to explore new polymer designs that will be superior to present day alternatives.

Experiments were performed on a sample of natural hagfish slime in the lab. This was to gain insight into how the proteins work, to prove methods read about in scientific papers would work and to have a framework to compare the synthetically produced protein with the natural protein.

We sourced a sample of the slime from a scientist who went out on a fishing ship that frequently catch wild hagfish accidently. He agitated a wild hagfish and collected the natural slime in a jar with some salt water.

Once the slime arrived in the lab it was divided in half, one half being washed in sodium citrate to strip off mucins associated with the slime, then washed with 20mM DTT to wash off any remaining mucins. This left behind intermediate filaments, which were then freeze dried for future use. The freeze-dried products were stored at -20°C to prevent degradation of the filaments.

When working with the intermediate filaments they were dissolved in formic acid solution for one minute and centrifuged at 3200 RPM for one minute, leaving the desired protein in solution. It was attempted to use Negishi’s method to dropcast the intermediate filaments according to her paper, however this method worked intermittently (Negishi et al., 2012). Spin coating was used eventually to produce thin films: intermediate filaments were dissolved in formic acid and dropped on to a silicon chip with a thin layer of MgCl2 to produce 150nm thick films.

This was repeated without using MgCl2 to produce thin films with roughness of 10nm with a 6nm variation in height, and a thickness of 10nm. Dissolving in pure formic led to an inaccurate protein dope concentration and failure for some gels to precipitate. These experiments were performed by placing the silicon chip on to a spin coater at 3200rpm with a 5s acceleration time and a 25s spin time.

We have also performed some prelimminary experiments to examine the reported antimicrobial properties of hagfish slime proteins. These data are provided as an appendix in the Notebook menu.

With these experiments completed we will now be in a better position to work with the synthetically produced intermediate filaments. The availability of natural filaments will facilitate comparative work with synthetic filaments at a later date.

Click to download the SeaDNA protocol.

Click to download the SeaDNA Hagfish Slime Antimicrobial Properties document.

• Downing, S.W., R.H. Spitzer, E.A. Koch, and W.L. Salo. 1984. The hagfish slime gland thread cell. I. A unique cellular system for the study of intermediate filaments and intermediate filament-microtubule interactions. J Cell Biol. 98:653-69.
• Fudge, D.S., K.H. Gardner, V.T. Forsyth, C. Riekel, and J.M. Gosline. 2003. The mechanical properties of hydrated intermediate filaments: insights from hagfish slime threads. Biophys J. 85:2015-27.
• Fudge, D.S., S. Hillis, N. Levy, and J.M. Gosline. 2010. Hagfish slime threads as a biomimetic model for high performance protein fibres. Bioinspir Biomim. 5:035002.
• Fudge, D.S., N. Levy, S. Chiu, and J.M. Gosline. 2005. Composition, morphology and mechanics of hagfish slime. J Exp Biol. 208:4613-25.
• Kim, B.S., K.E. Park, W.H. Park, and J. Lee. 2013. Fabrication of nanofibrous scaffold using a PLA and hagfish thread keratin composite; its effect on cell adherence, growth, and osteoblast differentiation. Biomed Mater. 8:045006.
• Negishi, A., C.L. Armstrong, L. Kreplak, M.C. Rheinstadter, L.T. Lim, T.E. Gillis, and D.S. Fudge. 2012. The production of fibers and films from solubilized hagfish slime thread proteins. Biomacromolecules. 13:3475-82.
• Puigbo, P., E. Guzman, A. Romeu, and S. Garcia-Vallve. 2007. OPTIMIZER: a web server for optimizing the codon usage of DNA sequences. Nucleic Acids Res. 35:W126-31.