When the hydrophobic (apolar) and polar moieties of elastomeric polypeptides are properly balanced, the polypeptides are soluble in water at lower temperatures but undergo folding and assembly transitions to increased order on raising the temperature. The temperatures, Tt, and heats, ΔHt, of these inverse temperature transitions are determined by differential scanning calorimetry for a series of elastomeric polypentapeptides: poly(VPAVG), poly(IPAVG), poly(VPGVG), poly(IPGVG), poly[0.5(VPGVG),0.5(IPGVG)] and poly[0.82(IPGVG),0.18(IPGEG)] where V = Val, P = Pro, A = Ala, G = Gly, I = lle and E = Glu.

On increasing the hydrophobicity as when replacing V(Val) by I(lle) which is the addition of one CH2 moiety per pentamer, the temperature of the transition is lowered by 15 to 20°C and the heat of the transition is increased by more than one kcal/mole, for the above examples, by more than a factor of two.

When differential scanning calorimetry thermograms are obtained on mixtures of poly(VPAVG) plus poly(IPAVG) or of poly(VPGVG) plus poly(IPGVG), it is found that the polypentapeptides self-separate, i.e., they de-mix, even though in the latter case the conformations have been shown to be essentially identical before and after their respective transitions.

When the polymer, poly[0.82(IPGVG),0.18(IPGEG)], is studied as a function of pH, increasing the degree of ionization is found to increase the temperature and to decrease the heat of the transition such that, with the correct balance of I with the variable E(GluCOO−), the values of Tt and ΔHt can be made to approach those of poly(VPGVG). Acid-base titration studies indicate that less than one Glu(COO−) in 200 residues can raise the value of Tt by 25°C and decrease ΔHt by 90%.

These and additional data are interpreted to mean that there exists an hierarchical hydrophobic folding, that the hierarchical hydrophobic folding can be modulated by changing the degree of ionization or by changes in a number of intensive variables, that changes in these intensive variables can be used to drive folding/unfolding-assembly/disassembly transitions under isothermal conditions, and that these unfolding/folding and disassembly/assembly transitions can be used to achieve signal transduction. This is called the ΔTt mechanism of free energy (signal) transduction.

Poly(γ-benzyl-L-glutamate) (PBLG) derivatized at its N-terminus with lipoic acid, a disulfide-containing moiety, self-assembles on gold from helicogenic solvents to give a thin film with the polypeptide α-helices orientation distribution different from the planar orientation in the unlabeled, physisorbed PBLG films (control) and Langmuir-Blodgett monolayers. The SA films were studied by angle-dependent XPS, reflection-absorption FTIR spectroscopy, and ellipsometry. The IR dichroic properties of the amide I and amide II bands were used to infer the orientational distribution of the helices in the self-assembled film and lead to two extreme pictures of the helix axis distribution function: (a) random (hemispherical distribution) and (b) perfect order with a tilt of 53° from the surface normal. Additional characterization is necessary to differentiate between these two distributions.

Proteins, due to their complex structures, possess infinite possibilities for assembling themselves as well as other biological and nonbiological materials into complex, multicomponent structures. Many examples from nature demonstrate the high degree of flexibility, strength and “intelligence” offered by protein-organized, self-assembling systems. However, several problems must be overcome before self-assembling protein-based products can become commercially feasible. One problem is the high cost of protein production. A second problem involves the question of producing nonidentical protein subunits in the proper stoichiometry for self-assembly. A third problem involves the stability of the protein components of larger structures.

We have been developing protein-producing bioreactors to address the problems associated with the commercial production of proteins, both as individual products as well as components of self assembling systems. The approach that we have taken is to clone the gene or cDNA that encodes the desired protein(s) and tailor the expression of the protein so that it can be produced at high levels under conditions where cell division is blocked. This assures that supplied nutrients go into product and not biomass, and at the same time provides conditions where cells can be immobilized without fracturing the support matrix during division. The blocking of cell growth also allows for the adjustment of protein stoichiometry by setting the ratios of cells that produce different subunits of complex structures.

To simplify downstream processing and purification, we have designed a system where nearly pure protein is secreted directly into the medium. This approach allows for automated downstream processing and separation of product from cells before it is degraded. The model system that we are using for the development of the protein-producing bioreactors utilizes genetically modified strains of Escherichia coli that secrete proteins which are usually found in the periplasmic space directly into the medium. The phosphatebinding protein serves as an ideal model protein for bioreactor development, since the regulatory elements that control its production are turned on by phosphate limitation, which also arrests cell division. We are presently determining if these regulatory and secretory elements can be used to direct the synthesis and secretion of heterologous proteins.