**3. Fine-Tuned Albumin Infusion Modifies CGA-Derived Peptides Multimers In Vitro and Impacts on Nosocomial Infection Occurrence**

Care-related infections are a matter of worry in the ICU: they increase the costs of treatment and length of stay and kill, on average, 10 to 15% of the patients. Finally, acute stress and anti-microbial treatments are also responsible for the emergence of transmittableresistant microbes. Despite hygiene, antibiotics stewardship, and precautions, these infections will occur in many patients with systemic inflammatory conditions at the acute phase of a disease. Oxidative stress induces damage to proteins within the circulation and beyond. The mechanism of oxidative stress undoubtedly affects many proteins, including those belonging to innate defense. In several patients, we recorded multimers of interest at the acute phase of the disease, including multimers of granins (see Figures 2 and 3). These multimers persist longer in the circulation of those patients with the highest and longest rate of infusion of norepinephrine for shock. Human serum albumin (HSA) displays properties for the care of critically ill patients. Among others, HSA provides an opportunity to restore in vitro the native status of proteins, as reported previously by our group [15]. However, this effect has never been reported in ICU patients that are prone to develop either colonization or infection. We decided to perform a pilot study on critically ill patients at risk of nosocomial infection to test whether therapeutic non-oxidized HSA would prevent such infections [16]. This study included a biochemical analysis of the interactions of the CGA-derived peptide VS-I, for which the anti-microbial activity is related to its non-oxidative state [15]. The results were that: (i) in vivo, therapeutic HSA significantly lessens both colonization and infection occurrences in patients with shock; (ii) this was possible provided therapeutic HSA is prescribed as a continuous low dose infusion of 4% HSA. We showed, in addition, that, in vitro, both natural and recombinant VS-I develop biochemical interactions with several natural and synthetic isoforms of albumin (HSA, bovine serum albumin, therapeutic HSA) via the hydrophobic domain of VS-I17-40 (which includes the disulfide bridge C17-C38). This allows the oxidation of VS-I to be reversed,

rendering it more efficient as an anti-microbial protein even in tissues where the pH decreases at 6 in microcirculation during shock. We deduced that the rate of therapeutic HSA infusion is essential in vivo when seeking to restore the physiologic activities of defense. A prospective multicenter open-label randomized trial confirmed this data in septic shock patients for which continuously infused 4% therapeutic HSA over the first week of shock decreased nosocomial infection by two-thirds when compared with the intermittent infusion of similar doses of 20% HSA [17]. These results are noteworthy because they explain the discrepancies existing in meta-analyses on the benefit of therapeutic HSA: many studies postulate the lack of efficiency of therapeutic albumin in septic conditions, whereas others have found significant improvements in restricted populations [18,19]. In fact, protocols of infusion and amounts of therapeutic albumin differ from one study to another, and so do the isoforms of albumin tested, which explains why physicians do not achieve the goal of defense reversion. Our final proposition is that therapeutic albumin should be: (i) chosen as 4% albumin with a high potential of antioxidant activity [20]; (ii) infused continuously at a rate of 10–12 mL/kg/24 h over 5 days to limit the risk of care-related infections.

**Figure 2.** Gel filtration HPLC of purified and oxidized VS-I samples to evaluate the impact of different concentrations of therapeutic human serum albumin (HSA) on multimers of VS-I. For the four graphs, the X-axis corresponds to the elution time (expressed in min), which is linearly related to the molecular weight of the components included in the peaks. The Y-axis corresponds to absorbance expressed in milliUnits of absorbance. Whether in vitro or in vivo, the oxidation of VS-I leads to multimerization, as shown on the first chromatogram (upper graph, numbered 1). The monomeric form of the peptide corresponds to peak E, while peaks A-D correspond to VS-I multimers. When adding fresh, non-oxidized therapeutic HSA at a molar albumin/VS-I ratio (R) from 1/10 to 1/105 (as shown on chromatograms 2–4), the release of the monomeric VS-I increases (see changes in the amplitude of peak E). This counterintuitive result explains the possible release of monomers of VS-I with the restoration of its anti-microbial properties [15].
