Molecular, interfacial and foaming properties of pulse proteins

Pulses are important protein sources in the current protein transition, but much of the behavior of pulse proteins in food systems (e.g. foam) is still unknown. This study compared the similarities and differences of the proteins from three common pulses: lentil, faba bean, and chickpea, at the molecular, mesoscopic and functionality levels. For each source, the proteins were extracted with a simple one-pot alkaline extraction, resulting in protein extracts with high protein content (67–85%) rich in globulins (68–81%) and albumins (19–32%). All pulse protein fractions were predominantly in a native state with high solubilities (83–91%), high denaturation enthalpies (8.9–10.1 J/(g protein)) and low aggregation levels. Lentil protein showed higher absolute value of the ζ-potential (−18.5 ± 1.2 mV) than faba bean (−13.2 ± 0.8 mV) and chickpea protein (−12.4 ± 1.2 mV). The behavior of all pulse proteins at the air-water interface including adsorption kinetics, interfacial dilatational rheology and interfacial microstructure was investigated and linked to their foaming properties. The pulse protein with the highest vicilin and convicilin content (39.5% in lentil protein) showed the shortest adsorption lag time (300 ms), and demonstrated the highest foam overrun (290 ± 17%). All three pulse protein fractions formed interfaces with disordered solid-like behavior and consisted of heterogeneous network structures. These structures were mostly comprised of intact globulin proteins for lentil and faba bean proteins, while mixtures of intact globulins and unfolded proteins/albumins were observed for chickpea protein. The interfacial stiffness, interfacial density and ζ-potential of these pulse proteins showed a high positive correlation with their foam stabilization properties. Lentil protein has the highest values of those parameters and showed the longest foam half-life time of 115 (± 22) min, while faba bean and chickpea proteins showed comparably lower values of those parameters and shared similarly shorter foam half-life times of 46 (± 19) min and 38 (± 6) min, respectively. These findings can aid to provide clearer directions in screening pulse proteins for desirable performance in aerated food products.