**6. Enigma and Future for CGA and Its Derived Peptides in the Critically Ill**

Based on available data, one first challenge is understanding the mechanisms and relevance of the multimerization of CGA and some of its derived peptides in acute stressing diseases. As shown in Figure 3, according to time from admission to the recovery of shock, we have studied by Western blotting analyses the plasma of septic shock patients with antibodies directed against VS-I. As indicated, we have observed time-dependent changes in the processing of VS-I-tagged multimers in the bloodstream during the first days of ICU stay. These data indicate time-dependent changes in the processing of molecules, which supports characterizing chromogranins (CGA, CGA-related peptides, and chromogranins B and C) as acute phase proteins. The multimers rapidly diminished and, in a second phase, progressively vanished from circulation at the time when the treatment by catecholamines was possible to stop. Whether multimerization is just a modality of transport for CGA from chromaffin cells to distant targets in tissues or a mechanism of protection of CGA from enzymatic processing within circulation in this setting remains unresolved and merits close understanding to avoid interrupting a physiological process intended to protect CGA from immediate endogenous processing. On the one hand, the pharmacological manipulation of multimers with low doses of therapeutic albumin enables the release of monomeric molecules if these molecules are already oxidized [16]. On the other hand, the release of CAT, which has no such disulfide bridge available, requires another pharmacological approach, possibly by limiting the upregulation of inducible enzymes responsible for the processing of CGA. In clinical settings with systemic inflammation, such intervention has been recommended for vascular iNO-synthase inhibition. Finally, there is another challenge in relation to ICU patients as far as health stress is concerned: would CGA predict the risk of re-admission to an ICU for an improving patient when he/she leaves the ICU for an intermediate care facility? This issue has never been reported to date but would be of significant interest in situations of overcrowded ICUs.

**Figure 3.** Typical Western blotting analyses of plasma samples to evaluate the time-dependent changes in extracellular processed VS-I (**top** panel), and simultaneous hemodynamic profile (**bottom** panel) during a human septic shock. Time (Day X) runs over 3 days from admission (Day 1, hour 10:00 pm (H22)) to the end of the third day (Day 3, 02.00 pm (H14) and is represented on the X-axis (top and bottom lines). Plasma samples harvested by 8 h, from admission until Day 3, were immediately centrifugated (4000 rounds/min at 4 ◦C), and SDS-PAGE electrophoresis followed by electrophoretic blotting with immunological detection of VS-I was performed in standard conditions. The anti-VS-I antibodies were a generous gift of Pr A. Corti, Milan, Italy. Periods of norepinephrine and dobutamine infusion are represented as bold arrows from admission until weaning. Please note that the processing of chromogranin A is notably changing around Day 2 at 2:00 (H14) when the patient's circulatory status no longer requires the two vasopressors. The global immunoblots are decreasing in intensity, and both small and large molecule multimerization is decaying (top panel). This phenomenon is contemporary to the decrease in doses of vasopressors (norepinephrine and dobutamine) and corresponds to circulatory failure recovery. Altogether, these data explain (i) why dosages of any of these proteins must be performed at a similar time window of the disease if a proper interpretation of their role is to be considered; (ii) that a pharmacological intervention must be scheduled at a moment when it can be efficient. Thus, our data on 4% albumin-impact on multimerization show that such an intervention must start as early as possible after admission, and it has no sense after the 5th day of disease onset [16]. Please remember that the apparent molecular weight (MW) of full-length CGA is around 70–75 kDa, and that of VS-I is approximately 18 kDa, which explains the immunoblotting of multimers on the top panel. Boxes (solid line, dashed line ... ) focus on processed molecules of interest, tagged with VS-I-antibodies: a careful identification must specify whether some multimers are not just large monomeric, full-length CGA molecules containing the VS-I domain.

A second issue concerns the possible use of CGA-derived peptides with defense properties in helping to keep implantable medical devices immune to infectious attacks once implanted. Such a development would not only be of interest to surgeons who implant several categories of prostheses but would also represent a significant breakthrough for critically ill patients. Although these proteins are endogenous peptides with few detectable side effects at nano- or micro-molar concentrations, convenient therapeutic use in humans has never been found. No data exist on in vivo consequences of the infusion of VS-I to test vasoactivity in humans. However, some experimental ex vivo studies have described the effect of VS-I on human vessels and in experimental animals [31–34]. Although this peptide has proven capable of ex vivo fungicidal and anti-bacterial activity even in multidrugresistant microbes, it has never been tested in vivo as an anti-microbial drug. No ethical issues have been reported as explaining such a situation. Still, one can reasonably imagine how large the amount of VS-I required to produce persistent physiological effects on humans would be. One additional explanation is that the efficiency and the cost/effectiveness ratio of the drug used as a single anti-infectious agent would be questionable. Nonetheless, it does not seem unreasonable to test this approach to prevent colonization by superbugs in critical patients as they frequently display certain forms of immune suppression [35] or even as local adjuvant treatment. Duration of the peptide availability for long implantation is another limit for such use. Our group has recently proposed that implantable devices can be coated whenever possible to lessen the risk of care-related infections [27,28]. We have succeeded in incorporating CAT—and its active core CTL—by linking them to materials through a spacer, which is cleavable by enzymes from bacterial strains prone to colonizing intravascular prostheses. This opens the perspective of a new defense strategy for fighting infection of implants: the innovative component of the strategy is that the availability of the AMP is longer and diminishes only if the microbe is present with the required enzyme for the release of the coated peptide. For example, we successfully used the endo-protease Glu-C produced by S. aureus, although it had been previously shown that hyaluronidase from both S. aureus and yeast also works [27,28,30]. To make the strategy even more efficient, we have also tested D-stereoisomers of some CGA-derived peptides: these isomers proved very stable against enzymatic proteolysis. The preliminary results encourage further investigation. First, the dimeric form of CTL linked by three polyethylene glycols substantially enhances the anti-bacterial activity against S. aureus. In contrast, dimerization was not required to ensure better destruction of C. albicans. Second, the D- CTL peptide displays interesting, enhanced activity against some Gram-negative superbugs relevant as far as ICU patients' infections are considered. Third, the non-toxic peptide DOPA5T- CTL can be employed as a "self-killing strategy" regarding S. aureus having certain protease activities; in addition, once released, the anti-microbial CGA-derived peptides still boost local immunity in dendritic cells and CD14 cells as well in tissues where a high concentration of microbes would justify a sustained release of the host defense peptide. These results indicate that medical use is foreseeable with some synthetic peptides such as CTL or VS-I in the near future, provided that technical and scientific progress enables the perfect impregnation of the implant.

A third issue arose recently: COVID became a significant threat in the hospital as far as its critical forms are concerned. De Lorenzo et al. asserted that CGA concentrations could predict death [36]. They showed that dying COVID patients demonstrated higher CGA levels on admission than survivors. Indeed, in our differently designed study of COVID patients admitted for acute respiratory failure and hypoxemia, admission plasma CGA concentrations instead predicted the occurrence of morbidity rather than mortality, which was better forecast by the CAT/CGA ratio [14]. Our study suggested that the stressing challenge of COVID was probably not hypoxemia itself—in line with its effect in vitro—since matched ICU control patients without hypoxemia did display levels of CGA similar to those of hypoxic COVID patients [37]. Because standard inflammation parameters did also not correlate with either CGA or CAT, we also examined the possibility that long-lasting circulating concentrations of CAT could interfere with systemic inflammation. According

to a previous study by our group, such concentrations physiologically boost, through a cell-penetrating-peptide mechanism, the release of pro-inflammatory molecules by PMNs, molecules which are currently recognized as biomarkers of severity in COVID [23]. In addition to this mechanism of inflammation, circulating CAT will also be available to engage in a molecular receptor-linked action on nAChR. There are strong scientific arguments to prove that CAT is a non-competitive inhibitor of this receptor, which provides a new and better comprehension of imbalanced neural regulation of innate immunity in critically ill patients [38]. This activity of CAT will be of major significance regarding all critically ill patients in explaining morbidity linked to the failure to arrive at a proper balance between pro- and anti-inflammatory pathways after life-threatening stress. Indeed, the action of CAT on nAChR provides an explanation not only for the occurrence of care-related infection but also for long-lasting multiple organ failure-associated myopathy [39].

Finally, there are other relevant issues regarding critically ill patients in connection with the in vivo processing of CGA. Among the difficult matters to explore is the involvement of CGA and its derivatives in acute cardiac diseases, whereas its involvement in chronic heart diseases has already been investigated [40–45]. Acute cardiac failure is far from exceptional in acute-onset conditions such as shock. No compelling reason exists to exclude a possible role for full-length CGA or selected CGA-derived peptides in transient cardiomyopathy or some severe arrhythmias.
