In this study we successfully synthesised and overexpressed four novel SELPs of similar size (55 kDa) but with variable content and ratios of silk to elastin i.e. 1:1 (SELP-109-A), 1:2 (SELP-1020-A, SELP-59-A) and 1:4 (SELP-520-A). Due to the novel elastin block (VPAVG) and the various silk to elastin ratios used processing of these polymers should give rise to a novel family of SELPs with unique biological and mechanical properties and thereby potentially extend the functionality of SELPs in general. Indeed, those polymers with a higher proportion of the elastin-like motif may be expected to be characterised by a reduced tensile strength, higher resilience, lower propensity for hydrogel formation, increased solubility and reduced crystallinity as compared to those with a higher proportion of silk (Dandu et al. 2009; Gustafson and Ghandehari 2010; Haider et al. 2005). The four polymers prepared in this study have been tailored to cover a large range of many of these properties and studies are currently underway to characterise these.
We examined the expression of these polymers with the currently most used method (Sambrook and Russell 2001) and found all proteins to be expressed intracellularly in a soluble form. In an attempt to simplify the production protocol we investigated the use of an auto-induction approach in which the inducing agent (in this case lactose) was added directly during media formulation. The best expression levels were obtained with TB+lac at 37°C when compared with LB+lac at the same temperature. However, despite the high expression levels, low cell densities (OD600 ≈ 2) were reached. This can be further optimised by adding low concentrations of glucose as previously demonstrated by Studier (Studier 2005), where the addition of glucose at low amounts (0.05% w/v) allowed cells to have an initial burst in growth. However, it should be noted that addition of glucose causes saturated cultures to become very acidic, limiting saturation density (Studier 2005). In this study we have chosen not to add glucose to medium composition, as our intention was to perform a preliminary study in the evaluation of protein expression levels where expression is induced in the early stages of growth. It is well known that the presence of glucose exerts catabolite repression, preventing the uptake and utilization of lactose, whilst glycerol does not interfere with lactose induction or the ability to metabolize this sugar for energy. In fact, addition of glycerol in auto-induction media was demonstrated to double the yield of target recombinant protein when compared to medium relying on lactose as the primary energy source. We have found that expression of the recombinant copolymers only occurs in the period between 9 to 15 hours of growth with no recombinant protein being produced earlier. This delay in the induction of protein expression is most likely attributed to the yeast extract used in the complex media formulation. The yeast extract supplies a variety of metabolites including complex carbohydrates that allow for fast initial growth rates. However, the yeast extract is also rich in amino acids that may prevent induction of target proteins by lactose in the early stages of growth. Indeed, in the absence of glucose, the presence of amino acids appears to inhibit induction until growth slows upon approach to saturation (Studier 2005).
One of the major factors affecting both bacterial cell growth and recombinant protein expression is the concentration of dissolved oxygen in the medium. It is commonly accepted, that high oxygen transfer rates lead to higher cell densities in bacterial cell cultures (Losen et al. 2004, Collins et al. unpublished). In shake flask fermentations, the OTR is influenced by several factors namely shaking frequency and shaking diameter, filling volume, flask shape and size, surface properties of the flask material and the physical-chemical properties of the liquid (Maier and Büchs 2001). Increasing the shaking frequency or lowering the filling volume leads to higher OTR due to liquid mixing and a high area of contact that allows efficient oxygen diffusion (Maier and Büchs 2001; McDaniel et al. 1965; McDaniel and Bailey 1969). In this work, we have performed a preliminary study on the effect of oxygen on the recombinant protein expression by evaluating the expression levels in shake flask cultures with different liquid to flask volume ratios. Cell density was shown to be inversely proportional to the filling volume without any deficiencies in protein expression (Figure 3) and thus suggesting that an optimization of oxygen levels could play a major factor in improving cell densities.
Auto-induction is more convenient than IPTG induction as it does not require any intermediate steps. The expression host is simply inoculated in the medium and let to grow without the need to monitor cell culture and add inducer at the appropriate time. In our study, induction of protein expression by lactose showed similar or even higher expression levels when compared with the conventional IPTG induction method. This work provides an insight into small/laboratory scale fermentations where auto-induction can provide major advantages as no extra culture handling is required. All copolymers were found to be successfully overexpressed in this auto-induction media, demonstrating its high potential as a facile method for production of recombinant SELPs in small/laboratory scale fermentations, but also for the production of several other proteins. In fact, in our laboratory, several other recombinant proteins were produced by this auto-induction method and purified for biotechnological purposes, namely, structural protein-based polymers, human growth factors, serine proteases, hydrolases, esterases, lipocalines, antimicrobial peptides and plant transcription factors (personal communication, Casal,
M.). In all the studies performed above we used purified lactose but other sources of lactose can be used. Waste pollutants derived from the dairy industry, like cheese whey, were previously used for the production of recombinant proteins with expression levels similar to those obtained with IPTG (Viitanen et al. 2003; De León-Rodríguez et al. 2006).
All four newly developed copolymers were expressed at high levels using the optimised fermentation conditions. Treating the soluble crude lysate with an acidification treatment allowed removal of most of the E. coli endogenous proteins. This methodology was previously employed for the purification of a recombinant major ampullate spidroin I (MaSp1) protein from the spider Nephila clavipes (Xia et al. 2010). The crude lysate of MaSp1 producing cells was adjusted to pH 4 and led to precipitation of host cell proteins whereas the recombinant protein remained in the supernatant after centrifugation. However, the supernatant from the acidic treatment was still enriched with host cell proteins. In our study, we demonstrated that acidification at pH 3.0-3.5 is much more efficient for removal of E. coli endogenous proteins than pH 4. In fact, upon acidification of the SELP-1020-A crude lysate, a highly pure polymer fraction was obtained in just one step. Minor contaminants remained for SELP-520-A and SELP-59-A and these were removed by a simple ammonium sulphate precipitation step, followed by resuspension in cold water and dialysis against water.
SELPs are structural proteins lacking complex folding mechanisms and spontaneously self-assemble into β-sheets by hydrogen-bonding mediated processes. The stability of SELPs at low pHs is well known with many SELP copolymers being extensively processed into materials in formic acid solutions without any reported negative effects (Nagarajan et al. 2007; Qiu et al. 2009,2010). Furthermore, it was previously demonstrated that pH had no effect on the equilibrium swelling ratio of SELP hydrogels between pH 2.0 and pH 12.0, indicating an absence of pH sensitivity (Dinerman et al. 2002; Haider et al. 2005; Dandu et al. 2009). However, the substitution of an amino acid in the elastin-like block by a glutamic acid generated a pH stimuli-sensitive polymer (Nagarsekar et al. 2003, 2002). Here, we took advantage of this known acidic pH stability to separate the novel SELP polymers from the unstable contaminants. The acidic pH stability was further supported by comparing the FTIR spectra between pure lyophilized copolymer samples with and without the acidic treatment.
Concerning SELP-109-A, purification was not carried out as the polymer precipitated during cell lysis, possibly due to the high content of silk blocks that promote gelation (Megeed et al. 2002; Cappello et al. 1998). With the appropriate composition, SELPs undergo an irreversible sol–gel transition through crystallization of the silk-like blocks. This spontaneous transition is influenced by the number of silk-like blocks as well as by environmental conditions like temperature (Haider et al. 2004). As the formation of hydrogen-bonds is the major driving force behind gelation, copolymers with larger silk-like blocks will provide more contact points for potential inter-chain junctions (Haider et al. 2005; Cresce et al. 2008). On the other hand, the inclusion of larger elastin-like blocks will increase the spacing between the silk-like blocks, increasing its flexibility and aqueous solubility (Haider et al. 2005). In fact, when analysed by solution viscometry and under the same conditions, the rate of gelation was found to reflect the increase in silk-like block content (Cappello et al. 1998). In the case of SELP-109-A and due to the large silk-like and small interrupting elastin-like blocks, gelation occurs rapidly. A protocol for purification of SELP-109-A will be further developed, possibly involving the use of chaotropic agents in the lysis buffer in order to inhibit hydrogen bonding. However it is beyond the purpose of this work where our main focus is to report the genetic construction and especially, a facile method for recombinant SELP expression and purification. For the remaining copolymers, precipitation with ammonium sulphate revealed to be an easy method for completion of SELP purification, allowing for the obtaining of highly pure polymer fractions with relatively low concentrations of ammonium sulphate. The volumetric productivities after lyophilisation were 185, 151 and 198 mg/L for SELP-1020-A, SELP-520-A and SELP-59-A, respectively, in comparison to the previously described 25–30 mg/L purified with IMAC (Dandu et al. 2009; Haider et al. 2005; Nagarsekar et al. 2002).
In an attempt to improve and simplify the purification process further we investigated whether the pH 3.5 treatment may also allow for cell disruption and hence enable both protein release and purification in one-simple step. Our analysis indicated that the treatment was indeed just as efficient at cell disruption, as monitored by SDS-PAGE, as was sonication. Up to 0.8 g of cell pellets (wet weight) per ml of resuspension buffer, which was the maximum cell concentration that could be easily resuspended, were found to be efficiently disrupted by incubation at pH 3.5 with, in addition, the removal of the majority of contaminating proteins being also achieved. Indeed a comparative investigation of cell viability following treatment further supported the suitability of the approach for cell disruption as it was found to be even more effective in reducing cell viability than sonication. Further purification and protein concentration can be obtained by precipitation with 20% ammonium sulphate with subsequent resuspension and dialysis against water followed by lyophilisation allowing for the obtention of highly pure, stable polymers. Hence we have developed a simple non-chromatographic method that allows for both cell disruption and the obtention of highly pure protein with reduced cost. This facile method is suited for proteins expressed in the soluble form and stable at low pH as in the case of the structural proteins hereby exemplified.
SELP copolymers purified by the method developed in this study did not show any cytotoxicity in mouse myoblast cell line C2C12 and even presented a slight increase in cell viability [see Additional file 2]. This suggests that these novel SELPs can be processed into materials and explored for biomedical applications; which is currently under study at our group. Although this work is focused on lab-scale fermentations, the purification method has already been validated in a high-density fermentation experiment with a 500 L fermenter, showing full recovery of the recombinant copolymer with high purity.