Self-assembling membranes Results.html


Self-assembling membranes


The genes encoding protein domains were fused according to our BioBrick standard. The proteins APH1-p53, APH-p53, p53-APH, BCR-p53 and p53-BCR were expressed in Escherichia coli BL21(DE3) pLysS mostly in form of inclusion bodies (Figure 1, Figure 2, Figure 3).

Figure 1: Production and purification of APH1-p53. SDS-PAGE analysis of the cell lysate supernatant (lane 1), insoluble fraction (lane 2), and purified polypeptide (lane 3). The band of calculated size of the polypeptide APH1-p53, 8.44 kDa is visible and was confirmed by mass spectrometry.

Figure 2: Production of BCR-p53 and p53-BCR. SDS-PAGE analysis of BCR-p53 (lane 1) and p53_BCR (lane 2) insoluble cell lysate show that proteins were expressed in form of inclusion bodies.

Figure 3: Western blot of APH-p53. APH-p53 is predominantly expressed in form of inclusion bodies as confirmed by Western blot using anti-His-tag antibodies (lane 2) in contrast to soluble fraction of cell lysate (lane 1). The third line represents protein standard.

We used circular dichroism spectroscopy to check whether proteins possess the expected alpha-helical secondary structure and to determine their stability. Far-UV CD spectra of all tested proteins show two minima, at 208 nm and 222 nm, which is characteristic for alpha-helical secondary structure (Figure 4).

Figure 4: Far-UV CD spectrum of APH1-p53 confirmed predominant alpha-helical secondary structure. Similar spectra were obtained for other constructs.

Different length of APH-1 and APH domains has influence on protein stability (Gurnon, 2003). We measured stability of our fusion proteins (Figure 5) and in case of APH and APH-1 fusions with p53 showed that different stability of coiled-coil-forming domains results in different stabilities of fusion proteins, with APH-1-p53 being less stable than APH-p53. Constructs with BCR domain revealed similar stability as APH-p53 domain (data not shown).

Figure 5: Fraction of unfolded APH-p53 as a function of denaturant concentration. Far-UV CD spectra of APH-p53 (0.1 mg/ml) were measured in solution with different concentrations of GdnHCl

Dialysis of solubilized APH-p53 inclusion bodies already produced a membrane-like material. The scanning electron microscopy (SEM) image of this membrane showed a porous material (Figure 6). Several fissures visible on image are probably due to non-optimal preparation of the sample with vacuum drying.

Figure 6: SEM image of APH-p53 membrane-like material

The following protocol was used to prepare the membrane by self-assembly of fusion polypeptide (Figure 7). Polypeptide was dissolved in 9M LiBr and then refolded by dilution into 20mM HEPES so the final concentration of protein in suspension was 0.2 mg/ml. 500 µl of protein suspension was slowly deposited on a filter with 0.2 µm pores with liquid being constantly removed by gentle vacuum (Figure 7). After deposition of the assembled protein to the support membrane, 10 % of glutaraldehyde was added to covalently crosslink and stabilize the protein assembly. After one hour, the membrane was rinsed with MQ water. The step of deposition and crosslinking of the proteins was repeated if required.

Figure 7: Flow chart of the protocol for membrane preparation.

Figure 8: The filtration system. Details of filtration system are depicted in the circle.

Blue dextran solution (0.5 mg/ml), and M13 bacteriophages were filtered through the prepared membrane to check its permeability. Concentrations of blue dextran and bacteriophage titer in the flow-through were determined. When we compared the concentrations and titer before and after the filtration we obtained encouraging results. The protein membrane retained most of the blue dextran (Figure 9) and bacteriophages (Figure 10).

Figure 9: Filtration of blue dextran. Blue dextran with initial concentration of 0.5 mg/ml did not flow through the APH-p53 membrane as was determined by measuring of the filtrate absorbance at 625 nm.

Figure 10: Clearance of bacteriophages by APH-p53 membrane. A) Comparison of viral titer before and after filtration through APH-p53 membrane. B) Scheme shows that ordinary filter can not retain the virus, while the addition of polypeptide membrane above the filter support does.


Our experiments show that designed polypeptide materials can work as effective ultrafiltration membranes. There is an enormous range of variations in design and possible applications of such materials. The features of structure formed upon a self-assembly of such fusion proteins depend on:

  • structure (geometry, size) of associating non-coiled-coil domain (oligomerization-prone domain)
  • length of coiled-coil-forming segments and properties of coiled-coils especially at the residues at positions b, c and f that are exposed to the pore
  • length and flexibility of the linker between oligomerization-prone domain and a coiled-coil-forming segment

Our design is especially suited for the ultrafiltration purposes since the size of the pore could be defined and adjusted by variations in the length of the coiled-coil-forming segments and the physicochemical pore properties could be defined and adjusted by variations in the size of coiled-coil segment and especially by variations and modifications of residues at positions b, c and f. These residues could be positively or negatively charged, hydrophobic, hydrophilic, allowing for chemical modification (Asn, Gln, Cys), they could chelate metals acting as catalysts or other amino acid residues imparting specific interactions. The only limitation is that these amino acid residues or their modifications do not disable coiled-coil formation. Such filtration membranes would not only allow for separation of molecules based on their size but also on molecules’ properties. Artificial enzymes can be formed in pores by introduction of active sites (e.g. catalytic triads) into the desired geometry into the pores.

Locations of visitors to this page