Ultrafiltration Fractionation of Bovine Hemoglobin Hydrolysates: Prediction of Separation Performances for Optimal Enrichment in Antimicrobial Peptide.

Hydrolysis of bovine hemoglobin (bHb), the main constituent of bovine cruor by-product, releases a natural antimicrobial peptide (NKT) which could present a major interest for food safety. To enrich this, tangential ultrafiltration can be implemented, but ultrafiltration conditions are mainly empirically established. In this context, the application of a simulation method for predicting the NKT yield and enrichment was investigated. Ultrafiltration performances were studied for decolored bHb hydrolysates at different degrees of hydrolysis (DH; 3%, 5%, 10% and 18%) and colored hydrolysates (3% and 5% DH) with 1 and 3 kg·mol-1 regenerated cellulose membranes. The simulation method helped to identify the most promising hydrolysate (in terms of NKT enrichment, yield and productivity) as the 3% DH colored hydrolysate, and UF conditions (volumetric reduction factor of 5 and 3 with 1 and 3 kg·mol-1 membrane, respectively) for higher antimicrobial recovery. A maximal enrichment factor of about 29 and NKT purity of 70% in permeate were observed. The results showed that the antimicrobial activity was in relation with the process selectivity and NKT purity. Finally, this reliable method, applied for predicting the ultrafiltration performances to enrich peptides of interest, is part of a global approach to rationally valorize protein resources from various by-products.

The influence of rough lipopolysaccharide structure on molecular interactions with mammalian antimicrobial peptides.

The influence of Escherichia coli rough lipopolysaccharide chemotype on the membrane activity of the mammalian antimicrobial peptides (AMPs) human cathelicidin (LL37) and bovine lactoferricin (LFb) was studied on bilayers using solid state (2)H NMR (ssNMR) and on monolayers using the subphase injection technique, Brewster angle microscopy (BAM) and neutron reflectivity (NR). The two AMPs were selected because of their differing biological activities. Chain-deuterated dipalmitoylphosphatidylcholine (d62-DPPC) was added to the LPS samples, to highlight alterations in the system properties caused by the presence of the different LPS chemotypes and upon AMP challenge. Both LPS chemotypes showed a temperature dependent influence on the packing of the DPPC molecules, with a fluidizing effect exerted below the DPPC phase transition temperature (Tm), and an ordering effect observed above the Tm. The magnitude of these effects was influenced by LPS structure; the shorter Rc LPS promoted more ordered lipid packing compared to the longer Ra LPS. These differential ordering effects in turn influenced the penetrative activity of the two peptides, as the perturbation induced by both AMPs to Ra LPS-containing models was greater than that observed in those containing Rc LPS. The NR data suggests that in addition to penetrating into the monolayers, both LL37 and LFb formed a non-interacting layer below the LPS/DPPC monolayer. The overall activity of LL37, which showed a deeper penetration into the model membranes, was more marked than that of LFb, which appeared to localise at the interfacial region, thus providing evidence for the molecular origins of their different biological activities.

Size-exclusion HPLC as a sensitive and calibrationless method for complex peptide mixtures quantification.

This work describes an original methodology to quantify complex peptide mixtures by size-exclusion high-performance liquid chromatography (SE-HPLC). The methodology was first tested on simulated elutions of peptide mixtures. For this set of experiments, a good estimation of the total peptide concentration was observed (error less than 10 %). Then 30 fractions obtained by ultrafiltration of hydrolysates from two different sources were titrated by Kjeldahl or BCA analysis and analysed by SE-HPLC for an experimental validation of the methodology. Very good matchs between methods were obtained. The linear working range depends on the hydrolysate but is generally between 0.2 and 4gL(-1) (i.e. between 10 and 200µg). Moreover, the presence of organic solvents or salts in samples does not impact the accuracy of the methodology contrary to common quantification methods. Hence, the findings of this study show that total concentration of complex peptide mixture can be efficiently determinate by the proposed methodology using simple SE-HPLC analysis.